This page and those pages it links to it will discuss government projects and
how our politicians are foolishly wasting hundreds of millions - even billions - of
taxpayers' hard-earned dollars on exorbitantly expensive IT projects that are
either already failed or are almost guaranteed to fail. How bad is the situation?
In California, the total amount wasted over the last 10 years on foolish IT
projects probably exceeds the currently projected (as of May 2010) state budget
shortfall of $20 Billion by quite a bit. The waste per year on exorbitantly
overpriced IT projects has increased to the point of several billion dollars. Most
of these projects were put into place with lobbying and are held in place -
against all rationality and dramatic budget shortages - by lobbying.
In many of the discussions, I am going to name names. If any of the politicians
involved don't like it, they're certainly welcome to publicly challenge me on what
I'm going to say here. Challenge me publicly that I can't write - at minimal cost
and in a short time - the software I'm describing and have it work well. Let's
have a competition. The voters (particularly in California where the state
government is dead flat broke and throwing away money at breathtaking speed)
need to hear what is going on with this incredible waste of their money so they
can make intelligent decisions in future elections.
And I may be in California, but this foolishness spans a number of states,
including Colorado, Indiana, Maryland, New York, and Texas for starters.
England probably had the most monumentally foolish IT project of all.
To get a handle on just how much of your hard-earned tax dollars are being
literally thrown away on doomed state IT projects, you might want to start by
reading two files linked to this page. Each file holds an entire series of articles
written for the Daily Recorder, Sacramento, California. I have also written a
number of freelance columns for the Daily Recorder on computer technology for
courts and lawyers. The two series, which were written about a failed project
called CalWIN and its spinoff, CBMS (Colorado Benefits Management System),
are:
The CalWIN project has been the subject of a number of TV news discussions
and the CBMS project, which is a spinoff of CalWIN, has been the subject of
hundreds of news media discussions in Colorado where it sparked lawsuits,
appointment of a special oversight person (a retired dentist turned politician
who didn't know beans about computers), reimbursement requests from the
federal government for erroneous payouts due to software miscalculations, and
a number of legislative discussions. The system is famous from coast to coast
as an example of a poor IT project and has even been the subject of business
graduate school classes at the University of Denver on everything not to do
when putting an IT project together. There has been a stipulated settlement
order in place since December 2007 in a lawsuit (Davis v. Henneberry) brought
against the state, but the current vendor, Deloitte Consulting, has been unable
to fix much of anything and the problems continue unabated. This idiotic IT
project has cost Colorado taxpayers around a quarter billion dollars as well as
causing enormous difficulties for their welfare system. The parent California
version? Pretty close to a billion.
Plenty of people in California had an opportunity to call a halt to the CalWIN
(parent) version. In Sacramento County, the County Board of Supervisors had
clear warning that something was drastically wrong. When the system went live
in their county, it had already gone live in neighboring Placer County, sparking
newspaper articles about the disastrous results. After the Sacramento County
Board of Supervisors heard from one worker after another, and were besieged
at their public meetings by welfare workers begging for the county to stop the
pain, I offered to help. But they weren't interested. Specifically, also, the head of
their Department of Human Assistance, Bruce Wagstaff, spoke with me. But he
wasn't interested in getting help, either. Note carefully that these folks poured
additional tens of millions of dollars into this foolish project, all the time refusing
offers of help. And now (2009 & 2010) their county is crying that it's broke. At
one point, one of my newspaper technology column readers intervened and
arranged a meeting with the newest elected supervisor, Jimmy Yee, because
he assured me Yee was honest and would be interested in fixing things. Yee
wasn't interested, either. TV news clips about CalWIN and its failures continue
from time to time, your tax money keeps getting poured into this bottomless pit,
and no politicians are willing to do anything about it.
The foolishness associated with such projects for creating exorbitantly
expensive poor quality software to do simple processing of unsophisticated
welfare data is not, however, limited to California, Texas, and Colorado. I
encourage anyone who reads these materials to pick up the phone and call the
people involved. Demand that they stop this incredible waste of your tax money.
One upcoming IT project in California is SOMS (Strategic Offender
Management System) for the California Department Of Corrections &
Rehabilitation. Look at http://www.cdcr.ca.
gov/News/2008_Press_Releases/June_19.html to see another IT project
disaster in the making. This one may well bleed the California taxpayers of at
least $100 Million. As a matter of fact, I heard in 2009 it was probably going to
cost about $250 Million. Here we go again! Prison inmate data, I assure you, is
not complex.
The California prison system, in combination with law enforcement agencies,
might actually be a logical starting place for a demonstration project for what
could be done to link up agencies with inexpensive, powerful, accurate, easy to
use software. I understand the computers on the medical side of the prison
system aren't even compatible with the system that tracks the inmates in
general and neither is compatible with computers at law enforcement agencies
such as sheriff's departments or police departments. A revamping of the entire
system is obviously in order and it wouldn't have to cost a mint. But then, of
course, there are the RFP's written by people who know absolutely nothing
about writing modern software, the lobbying, the kickbacks, etc.
And another huge money pit into which California's Administrative Office Of The
Courts is just aching to throw your tax money is a project called the California
Court Case Management System (CCMS) now being handled by a company
with a well-established track record of failures. (They're currently sitting on top
of Colorado's CBMS mess which is about to explode again into the news.) If it
weren't for the failed economy, and the current lack of state money, the
managers and backers of this crazy project would be pouring more than $1.75
BILLION into this junk at the rate of hundreds of millions per year. If you're
thinking that's a typo I assure you it's not. That's BILLION not million.
Temporarily they want to pour a mere $100 Million/year or so (what one person
who should know better by the use of simple common sense publicly termed
mere "life support") into this nonsensical project to "maintain" it while waiting for
the budget to recover in a few years so they can then follow that up with at least
another billion dollars of taxpayer funds. The administrators don't understand
the technology, have relied on the wrong people for advice, and are now stuck
with a monumental lemon. They are having difficulty coming to grips with their
situation, which is now being dictated more by the psychology of human
behavior than by technology.
Recently, with the announcements about impending closures of many courts in
Los Angeles, the Mayor of Los Angeles practically begged the court system not
to close courts. And there have been layoffs at other courts, resulting in long
lines of citizens waiting for service. Sacramento County has had long wait times
at the filing counter. One TV newspiece showed a lady who regularly files
pleadings as a legal runner sitting in a camping chair as she waits hours to get
to the counter at Sacramento County Superior Court. That has occurred also in
Alameda County. TV reports about San Diego have shown lines extending out
to the street. This is caused by the clerks being unable to work quickly with the
awful CCMS software. Absolutely none of this is necessary.
Want to see just how bad the results from such poorly done IT projects can get?
Contact Ginger Sylvester at the Sacramento Superior Court and ask for a copy
of the April 2010 CCMS briefing slides.
And if this waste of your taxpayer money just plain ticks you off, take note that I
offered to create case management software with stunning capabilities for the
California superior courts and do the first county for FREE as a demonstration
and still didn't have one taker. My wife told me to just withdraw that offer so it's
off the table now. I'll still do it, but not for free, although prototypes for people to
try out hands-on would still be free. The court leadership in one county told me
they couldn't accept help because their hands were tied. But the first county
court to take me up on the offer to create good case management software will
get it for peanuts compared to the $1 Million/judge price tag projected for
CCMS.
Perhaps it will take the utter collapse of some county court's old mainframe
(COBOL-based) system before anyone in charge of the funding will be willing to
admit that CCMS is a huge mistake and be willing to accept help. If that
happens, my number is (530) 306-6370. I can't guarantee the recovery of lost
data (although I would certainly try to recover it) but creating a system to work
for the future could be done on an emergency basis in probably a week to 10
days. Hooking up other counties into a true system with sharing of data could go
just as fast per county. And it's sometimes possible to recover current data even
if you can't get all the ancient software designed to run the files containing that
data to work.
Apparently we need the court system equivalent of a huge trainwreck with dead
bodies littered around before the people managing CCMS are willing to look at
better solutions. Please take note that I believe CCMS has been backed by
Ronald George, the Chief Justice of the California Supreme Court, only
because he has been misled by various people around him upon whom he has
relied for technology advice. Justice Ming Chin, head of the Court Technology
Advisory Committee (CTAC), is in the same position. He has been misled.
Some of these supposed "experts" - who apparently have been handing out
advice to the "suits" or "robes" - ought to know that they don't know how to write
modern software and should thus know they are not in a position to give advice
on the subject to anyone in charge, or to anyone holding the purse strings, or to
anyone whose backing they seek. As a prime example, Sheila Calabro, the
Director of the Southern Regional Office of the AOC, who was placed directly in
charge of the CCMS project, should know better than to believe she has any
knowledge of how modern software to run court case data management should
be written. I suspect the top leadership of the judicial branch has been given
bad advice by quite a number of people although I don't know exactly who all
those people are. Misleading the Chief Justice of a state supreme court to
obtain backing from such an official is a serious matter, at least in my opinion.
Those higher up the line than Ms. Calabro, including the Legislature and
ultimately the taxpayers, should force a termination of the CCMS (or at least a
time out) and just spend the approximately $10 Million or less in legitimate
vendor cost that it would take to create a system of electronic superior court
case records and a system to manage them. That is certainly a heckuva lot
better than continuing to spend over $8 Million/month on CCMS, with no end in
sight. This situation is a classic example of where we should follow the old
adage and not throw good money after bad.
The CCMS software is so bad that one of the pilot counties, Los Angeles,
essentially trashed it and now only uses it in some small court for minor matters.
And also please look at the justice system software pages where more
boondoggles specifically linked to the justice system will be discussed.
WHY??? For God's sake why???
So why are such government IT projects draining state coffers dry and forcing
hard-working taxpayers to pony up tax money so it can literally be thrown away?
Well, first there are several common misconceptions about the creation of such
government software and systems and these misconceptions foster such
monumentally foolish wastes of public monies.
But first, before we get into the technology issues, let's take a look at the
psychology at work in these boondoggles. These "multi-hundreds of millions"
projects are all essentially "con" jobs. As P. T. Barnum once put it, "There's a
sucker born every day." We are all subject to being conned, no matter how
smart we are. Sometimes it's much easier to con someone than in the average
situation because the "connor" takes advantage of a "connee" who is really
eager and is totally ignorant about the subject matter. If the potential pigeon has
already decided that "something must be done," then the setting is complete.
Pigeon drop schemes, Ponzi schemes, and the like all feed off of the desire to
get rich quick. We all have something we would really like to have
accomplished. The con artist takes advantage of the existing desire. If someone
wants some result really badly and is highly ignorant of the underlying subject
matter, then you have all the makings of a really good con. People want to
believe. Religious investment schemes are often such good cons that when
investors are presented with the facts, they go into heavy denial. "He was such
a nice man. He prayed with us every day." They can't believe they were taken.
And politicians have a big problem in this area because they have to figure out
a way to tell their constituents it wasn't their fault after they just spent huge
amounts of taxpayer money on something that was really simple and
inexpensive to do.
When confronted with the possible reality, human beings often go into denial.
Politicians and bureaucrats won't look at alternatives to what they have wasted
money on because that might result in the reality of the situation becoming
public. Then they would have to explain how they foolishly fell for a con job and
spent the taxpayers' (aka voters) money. So they look for another charlatan with
a new set of glowing promises instead of putting a stop to what they started.
And I have personally seen examples of this in the IT arena when simple
solutions were right in front of the politicians.
Politicians and government bureaucrats are often easily taken when it comes to
IT projects. Few know much of anything about computers. Fewer still know
anything about writing modern software. Thus it becomes fairly easy to talk
government bureaucrats into spending hundreds of millions of dollars and years
of time on something that could, in reality, be done in a few weeks or months for
just a few million dollars, with that price including full deployment, education of
thousands of users, and distribution of software, as well as the initial
development. Faced with the reality that something that has been in progress
for ten years could have been created in ten weeks, any politicians who were
the least bit involved will duck and hide rather than confront the situation. So it
often takes a new generation of leaders who can point back at those they
replaced and point out how utterly foolish they were.
Voters need to consider this in upcoming elections.
So what are the key ingredients to a really good con of government officials on
IT projects? Well, first you impress them with how you're going to fulfill lots of
supposed requirements. Get the people you're going to fleece in on creating the
method of picking their pockets. Have them set out detailed requirements. If
they can't seem to do it, give them some help. Have them set themselves up to
be taken. The best formula is to have someone in government who thinks they
know how to set out detailed specifications for an IT project ("I wrote some
mainframe software 25 years ago so I know how to do this"), then urge them to
get going on it. I saw that process in action on one project that has cost the
California taxpayers nearly a billion dollars and people involved in it are still in
denial. And it still hits the evening TV news from time to time as people note
how the software makes tons of mistakes handling simple calculations. And the
project expanded to a second state and the folks in the second state are also in
denial. And that state is on their fourth "savior" and still plowing through tons of
money. It is one of the most marvelous con jobs I have heard of. It's so good it
belongs in a college psychology textbook. So far it's only hit the college
classrooms (at DU) as a classic example of everything not to do when
designing an IT project.
When you're setting up "requirements," don't just describe what the software
should do in simple, straightforward terms like "software that will enable a
particular group of workers to do their jobs faster, more accurately, and easier."
That will never do for a good con job. Set down thousands of detailed "technical
requirements" to make it all sound really complicated. If you can create a list of
such "requirements" and then put it on paper, and the stack is at least two feet
high, you have the makings of a superb con. One that would make Bernie
Madoff envious. And, to take a line or borrow a word or two from linguistics
masters such as J. Clark Kelso, Deloitte Consulting, EDS, and others involved
in or being taken by these schemes in California, use key, important sounding
words like "strategic" and "system" and "statewide" and "integrated". Skillfully
blend words like those together and you have the makings of a con that can hit
nine figures easily.
For something that could have been designed in a couple of weeks and would
actually have worked.
Now, back to the technology involved in government IT projects and common
misconceptions that lead to these boondoggles.
If the project is going to serve many end-users, that necessarily makes the
software to handle it complex (and thus expensive), right? Wrong. Dead Wrong.
Same thing for projects that are going to handle lots of records. Same for
projects that are going to handle lots of money. Same thing for projects that are
going to handle "important" data instead of "unimportant" data. And the fact that
it's government data instead of data to track cookie sales for a national
organization of children? "But Richard, this is court data."
Doesn't mean a damn thing.
Well, if none of these commonly touted factors (vulture IT vendors love to
emphasize them to naive bureaucrats and politicians) have any direct
correlation to the complexity (and thus the legitimate design cost) of the
software required, what is important? It's the complexity of the functions
required to be performed by the software. Want to see complex? Look at
SuperSingles-n-Doubles.
Software that has to process complex functions for 500 people and that will be
handled by one person can be incredibly more difficult to design than software
to handle 1,000,000 records that will be handled by 60,000 workers. There is no
necessary correlation between the number of people or records involved and
the legitimate design cost of the software.
"But Mr. Politician, look here. This software has to be worth at least $300 for
each install, right? Surely you understand that value. We charge by the number
of butts in the seats. So, let's see now. $300 times 60,000 is $18,000,000 and
you can understand how $18 Million worth of software could take $100 Million to
design. Right? And you want to underwrite our design costs, of course, right?
And then, of course, we have to contribute something to your re-election
campaign, and your political party could use a couple bucks, and you wouldn't
really mind if we made a little profit, would you? So you can understand now the
$400 Million price tag, right?"
So if it's more complicated to design software to handle data for certain types of
complex bowling events (and it definitely is) than it is to design software to
handle court data or welfare data or unemployment insurance data, then how
can the prices you typically hear connected with government IT projects to
handle such simple data make any sense?
Well, they don't. And how can it make any sense for RCP Software to only
charge $79.95 or $99.95 for software that required more design work than
software to handle many government operations would require?
First of all in the software world, there is the factor of economies of scale. A
major software manufacturer that spends $100 Million on developing a
particular application but sells 20 Million copies to completely separate
purchasers, simply spreads the cost out and makes a profit on everything
above $5 + distribution and marketing costs per copy. Microsoft Office,
Microsoft Windows, etc. don't cost you a million bucks a copy. So I can still
make a buck off bowling software, even charging very low prices. And besides,
I'm not out to make a killing on this bowling software and I am definitely not
trying to gouge bowlers or league secretaries or others in the bowling world. I'm
trying to help them and I have noticed they definitely appreciate that. I have
received numerous compliments and thanks when it comes to the bowling
software. And besides, it's fun to write this stuff.
By way of contrast, there are usually fewer copies of government software that
actually need to be distributed than what the vulture vendors tell naive
government officials. Vulture vendors love to bamboozle ignorant government
purchasers with confusing license provisions, and put emphasis on the large
number of users, etc.
But the reality is that a large government office can often take distribution of a
few copies or even a single copy of specialized software and work from it. But
that's of course how an honest vendor would do it. So it's theoretically
necessary to charge more per copy but that still doesn't need to be very much.
Note carefully that many of the copies can be absolutely identical.
But that inability to spread cost among millions of independent users still
doesn't mean the total project has to cost hundreds of millions of dollars.
Overall costs should correlate much more strongly to design complexities rather
than to the number of users, copies, etc. Design complexity in turn correlates
strongly to the amount of programming time required. Let's look at that situation
a little more closely.
Many governmental IT projects priced at figures like $500 Million or even $1
Billion or more could be done for $10 Million or less. The end-users just need
current computers outfitted in a standard manner. The legitimate "design" cost
might be only $2 Million. Distribution, installation, and education costs, with
some left over for vendor profit would consume the other $8 Million. Ongoing
maintenance costs could be minimal, particularly if the software works
flawlessly.
One additional observation of the high pricing of many government projects is in
order here. The number of users doesn't logically relate to design costs at all
but, in the case of poor software that has lots of bugs, maintenance cost does
relate to the number of users. If the software is well written, there is no
maintenance cost problem. The difference between 2 calls in a month and 15
doesn't really mean much. But with many of the vulture vendors lobbying and
getting government contracts, they might have 6,000 calls for help in a month.
Outfits like Deloitte Consulting that provide extremely poor software have to
field tens of thousands of cries for help. Thus the irony is that the worse the
software is the more it costs.
Recently (end of April 2010) I had the privilege of sitting down with some nice
folks who work in a particular government agency in California. They are in a
certain amount of distress over a "system" that they have had to work with and
which doesn't work worth a damn. They have stated their distress publicly. I
heard their pain and talked to them. I think they got the message that I could
quickly create what they need in the way of data management software, and that
it would work like a Swiss watch compared to what they are suffering with now,
and that I could do it for a tiny fraction of what has already been spent. But they
are timid about doing anything about the situation. They feel their hands are
tied. I'm not going to name the local agency or people involved or where they
are located because they were all very nice and polite. But I hope they will
rethink what they told me about being unable to do anything really constructive
about their situation.
Interestingly, while I was sitting there, one of the attendees told the local IT head
honcho about problems she was having that day with the system. The IT head
honcho responded with an answer that I knew instantly was, to a 99% certainty
level or better, not the primary source of the problem.
So the people at this agency know they are immersed in a disaster but won't
speak up to anyone in government who could do something about it. This, dear
taxpayers, is what is wrong with California government. No one who could
influence anyone to get something constructive done, save money, and make
government work better speaks up. The state-level agency involved here is on a
dead certain path to a huge trainwreck. The state government is dead flat
broke. The taxpayers cannot afford to finance any more of such foolishness. The
project involved is grounded on colossal ignorance of modern technology. I
don't mean a little bit of ignorance, I mean absolutely breathtaking, stunning,
wall-to-wall ignorance. I would be willing to bet that not one single person
involved in the entire project (and that is hundreds and hundreds of people) has
any substantial knowledge of how to write modern software.
The local chapter of the agency involved has, probably in reliance on
assurances from local IT personnel, tentatively planned to do a workaround that
would incorporate the bug-laden software with which they are already working.
The risk involved in such an undertaking is HUGE. And I told these folks that.
The base software on which they will be relying apparently has thousands if not
tens of thousands of coding errors. And the local folks know it. I mean they have
to know it because they told the public all about their experiences, describing
problem after problem after problem. The agency involved deals with important
data and if it goes down, the impact on society can be devastating. And they
have already had several shutdowns. The tentative solution planned by the local
chapter of the agency is absolutely devoid of any common sense.
If this local agency has a catastrophe and they finally call in desperation and just
agree to do what I tell them to do, will I help them? Yes, I'll take the high road
and help them to the extent possible. The agency's task is too important to allow
them to fail. But if they need a panic bailout, it won't be free. And under no
circumstances will I spend any time whatsoever debating with them or any of
their IT people how to rescue them from disaster. I just hope they have the
common sense to make backup copies of all data they can save. And,
considering the road down which they are contemplating traveling, there is
absolutely no guarantee that anyone - myself included - can bail them out as to
their current data if they encounter certain types of smashups.
Among the ideas this local agency is apparently seriously considering to work
around their problems is the idea of buying nearly $300,000 worth of server
hardware to host the bug laden, ancient-type software that has already been
written. To describe this plan as spectacularly foolish barely does it justice. It is
the equivalent of buying a titanium-platinum container to hold yesterday's
garbage. Garbage in a high-priced container is still garbage. A server bought
over the phone from Dell or HP for less than $5,000, combined with modern
bug-free software, could do a far superior job and at far less total cost. For
$10,000 you could get a server from Dell with spectacular capabilities.
It is such a shame that people who ought (by use of simple common sense) to
know better enter into such foolish ventures. The evidence of the total
worthlessness of the software they have been dealing with is all around them.
They are immersed in that fact every day. It is a shame that apparently the only
way to really have that hit home is to have a total collapse of the system and
then have no help from the software manufacturer available.
Before venturing into the briar patch they are contemplating entering, I would
suggest to them that they have their CTO participate in a little test of her skills.
Since she is supremely confident that she can handle the situation in front of
her, she should be willing to demonstrate that to the folks in the suits, right? It is
a test to determine if the person understands and can detect dependent coding
errors. This problem is obviously present in the software they are presently
using, based on the plain English descriptions I heard by non-techie users of
the system of various phenomena they encountered.
Of course the problem with finding and fixing dependent coding errors is that
they cause random problems. The problems occur sometimes but not at other
times, depending on the combination of code being relied upon by the particular
action being taken by the end user. The mathematical possibilities for where the
parts of the problem could be located are staggering in some of these
situations. We are talking exponential numbers of possibilities. It can be so bad
that an army of Einsteins would have trouble finding the problem(s). And
although you may have some general idea of where the problem is located, you
don't even necessarily know how many such errors are present, what type they
are, how they interact with each other, or whether they are even visually
obvious. You can be looking right at the problem code and not realize
something is wrong. I have personally encountered such a situation and, if it
were not for the fact that I thoroughly tested every single sub-area before
moving on when writing the application involved, I doubt I would ever have
figured it out.
Fortunately, I knew better than to tempt fate because I don't have a 400 IQ, so I
routinely tested every single layer before moving on. Up popped a bug. I knew it
had to be something I had done very recently. So I looked at that code.
Carefully. Over and over. Then I did elimination tests. Everything seemed
correct. Then finally, bingo! I found the small rubber raft floating somewhere in
the Pacific Ocean. But even when I looked at the code string that I had narrowed
the problem down to, I couldn't see the problem. After I knew for sure where it
was located, I performed some alternative type invocations of code in the same
dependent situation. And then it hit me. Then I figured out how to fix it.
Want to get a taste of how tough the problem can be? Look at this sequence.
615, 616, 617, 618, 619, 620, 621, 622, 623, 624
There's a problem with that sequence. Don't see it? I know damn well there is,
at least in a certain situation, because I ran into it. One of the numbers in that
sequence is non-usable. Now imagine that you are dealing with the problem of
locating the problem because you were so confident of your abilities that you
wrote layer after layer of code without testing each small group and now all you
know is that the problem likely lies between 14 and 1,216. Maybe. Not really
sure. I mean, there's only an outside chance it's somewhere else. And gosh, if I
see the problem, I'll recognize it instantly. Right?
But hey, you lobbied the right politicians, and the government (read "taxpayers")
is paying you by the hour to find the problem, so what's the big deal, right? I
only found the problem in the situation described above because I knew the
problem had to lie in this short sequence. Elimination testing with alternative
invocations located the problem and indicated the fix.
The code that these folks would be dependent on likely has many, many errors
or problems, perhaps ten thousand or more. And many could be of the type I
personally encountered. Trying to fix such problems could take eons. And since
you don't know in advance if the problems you might encounter are going to be
crippling, you don't know whether you will find yourself up against a wall,
absolutely needing to find and fix the problem. And to top it all off, these folks
don't even have guaranteed access to the source code.
One can only hope these folks will come to their senses in time. If not, and if one
of those random, critical bugs hits them smack between the eyeballs, I
guarantee you will read about it in the newspaper, see it on TV, and be asking
how in the world could such supposedly intelligent people have been so foolish
as to tempt fate like this.
So have other attempts been made to get elected officials in California off of the
track of just throwing money away on what should be simple and inexpensive IT
projects? You bet. Quite a while ago I contacted my two supposed
representatives to the California Legislature and offered to help. I explained to
them about the potential savings reaching into the billions if foolish IT projects
were scrapped in favor of better approaches that would cost only a tiny fraction
of what was currently being spent. That would be Senator Dave Cox and
Assemblyman Ted Gaines.
Neither was the slightest bit interested. I learned later from an article in the
Sacramento Bee, and information the newspaper had collected and then posted
online, that both of my representatives had been lobbied by organizations
whose members are employees of IT vendors that have been ripping off the
state government right and left on these crazy IT projects. Free computers.
Boxes of chocolates. Is that the basis on which these two officials - both of
whom loudly proclaim they are out to cut the cost of government and shave
down taxes - are supposed to be making decisions? As a taxpayer who has to
foot the bill for all this, what do you think?
The May 2010 CIO Report On CCMS
In early May 2010, an e-mail came around from the State Bar. There was a link
to more information on CCMS and that included the CIO's report to the
Legislature on CCMS. Before taking the contents of that report as being a
serious technical analysis of what is wrong, the California Legislature needs to
take a deep breath and remind themselves of certain critically important facts.
First of these is that the position of CIO is a political appointment. Neither the
present CIO, nor her predecessors, know beans about computers or, more
specifically and what is critically important here, how to write modern software.
Relying on the CIO for an analysis of the technology problems underlying
CCMS is roughly analogous to appointing the business office manager in a
hospital to the high-falutin sounding job of "Chief Medical Information Officer"
and then relying on her for surgical advice.
You can read the CIO report and I might suggest that you do so and get it up on
your screen in order to better understand what I am going to discuss below.
This is a report by a non-techie for non-techies. Every once in a while they make
an intelligent observation but that's about it. When they do note a serious
problem in CCMS, they gloss over it. They don't pinpoint what is wrong with the
technology because they don't understand what is wrong. They admit to being
puzzled about why certain things don't work because - let's face it - they are.
The report is written from the standpoint of a non-techie assessing the possible
business value of some gizmo they bought at an office supply store or some
service to which they have an ongoing subscription. The report never hones in
on the poor technology and how the task of creating electronic court records
and managing them could best be done.
Right up front, note the comments under Governance on page 5. Looking at the
situation from a $$$ standpoint, the CIO has at least caught on to the simple fact
that this nightmare software has to be spread over all the counties or pretty near
all of them or the dollar cost per county is going to be astronomical. Assuming
for the moment that the 51 counties who have not become involved in CCMS
look at what is occurring on a daily basis in Sacramento (the only true test
county so far) and intelligently decide to avoid CCMS, or at least until all
problems have been solved, then all the pitches made to the Legislature and
everyone else so far about this being a statewide system are undermined. And
note that, if the amount actually being spent keeps mounting (and it is, at around
$8 Million/month), this is going to be the most extravagant IT boondoggle in the
history of the state and probably the entire country.
The CIO argues the benefits to the state system as a whole should take priority
over the unique needs of individual courts. Hold it there! Is that some kind of
excuse warming up for trying to impose crippling software on individual
counties' courts? If it doesn't work for a particular county, what good is it?
In the first place, with minor exceptions, once good software is designed for one
county's courts, that same software can be used everywhere. The first county
might take a couple months to get fully going while the second county could be
done in days, perhaps a week. A good system could spread across the state in
weeks, from county to county. But it should be complete or at least nearly
complete and working like a Swiss watch in the first county before it is used
elsewhere.
CCMS, on the other hand, has proceeded with glacial slowness, creeping from
one county to another, with one part being tried here and another part or parts
being tried there, and inflicting pain everywhere it goes. The process has gone
slowly because the software is loaded with thousands upon thousands of bugs
(code execution errors) and the system architecture and specific software
architecture are just plain awful. The dumbos at Deloitte Consulting are
spending day after day, month after month, year after year, trying to find and fix
all their coding errors. There isn't even a single county that has electronic
CCMS records in all areas of law, from probate to criminal to family law to civil
to juvenile, etc. Different participating counties have different software. Out of 8
possible areas, the most any county has software for is 4. And note carefully
that they have not yet hooked up with any outside agencies, even though the
ability to do that was supposedly one of the big selling points for CCMS.
It is my understanding that each county court can decide whether or not to use
CCMS. And all the seven counties currently involved volunteered, although
some are obviously having second thoughts about that decision. That leaves 51
counties not affected yet by CCMS. However, none of the leaders of these 51
county courts, using ordinary common sense, will likely commit their courts to
using such a system unless everything is fixed in advance. And that is never
going to happen. It's impossible. Would you buy software or let your business
depend on it if you knew the software had thousands of bugs and myriad
problems?
Note carefully the lack of any rush by early testers of one or two parts of CCMS
to adopt the remaining parts that are being tested elsewhere. While publicly
some are trying to put on a smiley face and assert the pain isn't too bad, none
are rushing right out to find more needles to stick in themselves.
And even if you found and fixed all the coding errors, the CCMS
software/system would still be worthless because the overall architecture is like
something out of the dinosaur era of early computer technology. Parts are
incompatible with each and with current technology. Current technology
consisting of modern hardware and modern software can run rings around this
CCMS junk. Bear in mind that this project has been going for years. Everything
in it is way out of date.
Such projects only make sense if you can write the software in weeks at most
(which is possible), and get it up and running everywhere in months at most
(also possible). Then you need to remember that technology is going to
advance, so updating everything two or three years down the line should be in
your line of vision, assuming you want to keep taking advantage of advances in
computer technology. Software I wrote just a few years ago, while still running
smooth as silk (it has absolutely no bugs), is now outdated in comparison to
what can be done so I am encouraging customers to get their computers
updated so they can take advantage of new versions I am writing that take full
advantage of the huge advances in technology. And what I have written recently
will be dated in two or three years, just as surely as the sun rises and sets.
To see a report in another state on another failed Deloitte Consulting project,
also done by people who don't understand how to write modern software, take a
look at this Texas Comptroller of Public Accounts report on TIERS, Accenture,
Deloitte, etc. It talks about money and cost overruns, etc. but is written by
people who are clueless about the technology problems so, just like California's
CIO report on CCMS, it doesn't get to the heart of the problem.
Note that in Texas, the utterly failed TIERS project was handed over to
Accenture by Deloitte after the Texas government was conned into trying to get
yet another savior to fix the problems. I understand that at the time of the
handover which was in late 2005, TIERS had hundreds and hundreds of
problems. Deloitte can no more fix the problems in CCMS code than it could fix
the problems in the TIERS code.
Oversight of CCMS
There are supposedly a number of oversight groups or committees watching
with eagle eyes to make sure CCMS comes out right. One of these is the Court
Technology Advisory Committee, currently chaired by California Supreme Court
Associate Justice Ming Chin. Curious what the makeup and expertise of the
committee is? Look at this list. And look at this description. The makeup is set
by Rule 10.53 of the California Rules of Court. Now here's a simple question.
Where is the modern software-writing geek on this list? Here is the current
membership.
There are a few people on the committee who have tangentially "geek" job titles
but, as I have found out by talking to other people in such positions, they
typically have absolutely no clue about how to write modern software. Fix coffee
spills? Yes. Reset network passwords? Yes. Manage local networks? Yes.
Manage local Web pages? Yes. Handle general problems people have with
their computers or with specific applications? Yes. Write software like
SweeperScore or SuperSingles-n-Doubles or BowlerTracker or other similarly
complex modern software? They don't have a clue. But that's what the problems
with CCMS are all about. It's about software-writing. An attempt to create
electronic case records and an electronic system to handle them.
With all due respect to the people serving on this committee, they are in no
position whatsoever to evaluate the technology in CCMS unless they personally
understand how to actually write modern software. I realize the members of the
committee would probably rise up in righteous wrath if confronted with such an
assertion, but if you're a taxpayer being asked to foot the bill, I assure you it's
the key subject about which you should want to inquire. The failure to have
anyone who can actually write modern software (and I don't mean just some so-
called "IT person" who manages some local network) on the committee is a fatal
defect. All these people can do is generally realize that certain software doesn't
work worth a darn. They have absolutely no idea what to do about the situation
in terms of the technology fixes. Since the problems with CCMS are all
technological, this committee is absolutely worthless when it comes to the
matter of technologically solving the problem of obtaining software and
processes to make court records electronic and manage them. If they were to
use simple common sense about the situation (which I suspect they eventually
will), they could at least realize CCMS is going nowhere and they could cut off
the funding. And that would be a valuable action to take, even if they can't
specify the correct technology to be used to create a successful system.
In short, the problem with CTAC is roughly the same as in the hypothetical
situation I alluded to above of appointing a business office manager to the
position of Chief Medical Information Officer and then relying on her for surgical
advice. The CTAC members are people from the business for which the
software is to be written. And they understand the business. But the problem
isn't in the business, it's in the technology. It's in the tool they want to use in their
business. You don't need the business manager, you need the surgeon.
Unfortunately, neither the people on that committee nor anyone involved in
appointing them get it, or at least not yet.
What CTAC-type people can do intelligently is to evaluate from an end-user
standpoint whether or not a prototype actually works. It's like going to a store
that puts on demonstrations and trying before you buy. It's like taking a test
drive of a vehicle before you buy it. Ordinary people do this all the time with a
wide range of products.
The Future Of CCMS
This project is going to just plow through a bunch of money for the near future.
The people involved have no capacity to find and fix the coding errors causing
all the bugs, the software architecture isn't going to get fixed because a fix
would involve starting over, the system architecture isn't going to get fixed
because that again would involve starting over, the software conflicts are going
to remain because removing them would also involve starting over, the whole
thing is going to become more and more obsolete than it already is, and the
problems will just continue until the money runs out completely or people finally
come to their senses. Assuming this continues at around $8 Million/month, that's
going to consume an awful lot of scarce judicial branch resources.
Not sure that's going to happen, because, gosh, Deloitte Consulting has all
those "geniuses" working on this? If you want to get a good taste of how this will
play out, at least for the forseeable future, look at this summary of a settlement
reached in the Denver District Court in Colorado in December 2007, concerning
a now (in)famous IT project known as the Colorado Benefits Management
System (CBMS). And look at this page at the Web site for the Colorado Center
on Law & Policy. They sued in 2004 because the system was an absolute
disaster. The folks in Colorado are still waiting for the myriad problems to get
fixed while that could have been done in a matter of months a long time ago.
Note the current contractor carefully. OMG! It's the same folks running CCMS!
Originally it was EDS. The problems started immediately in September 2004 as
soon as the system went live, taking down the entire state welfare system. The
feds sued Colorado over erroneous payments. CBMS has been the subject of
hundreds of TV news pieces and newspaper commentaries.
Here is something for the PTB's in California to ponder, and they should
understand it because a lot of them are judges.
Settlement, there's the link to it.
Now don't get me wrong on this. These folks on the Court Technology Advisory
Committee are intelligent people. The problem is that they are way out of their
element. They just don't understand how to write modern software. They don't
understand the problems involved in trying to find and fix deeply embedded
coding errors. And they don't realize that the people to whom they have turned
for advice likewise do not have the expertise or experience required.
To put it bluntly, some of the people to whom they have turned for advice are
absolutely clueless.
As of this writing, those in charge of this project still resolutely refuse to accept
help. If you want to know why the Assembly Committee on Accountability &
Administrative Review held a hearing on October 28, 2009, and then did
absolutely nothing about the CCMS situation, ask them. I'm curious too but can't
get an answer. The chair is Hector De La Torre, and the vice-chair is Audra
Strickland. Their staff aide is Mark Martin. They have had a proposal for a good
solution in front of them for quite a while and have done nothing.
I will note here that some other states have obtained case management
systems for their courts and spent amounts like $30 Million instead of $1.75
Billion. But actually, they spent too much as well. The size of the state has little
to do with the legitimate cost so even though California has many more courts
and a higher population, $30 Million is about three times as much as this should
legitimately cost for California. Someone else who has knowledge in this area
estimated a maximum of $15 Million. At least we're in the same ballpark.
And if the initial cost isn't ridiculous enough, the AOC has been told it will cost
$79 Million annually just to maintain/service the system. And not knowing any
better, the AOC folks have bought into this! That is absurd beyond all belief. In
the first place, software to manage the court data should be flawless. That is
easily attainable. I have done that over and over and over with other complex
software I have written so I know beyond a shadow of a doubt it can be done.
What should annual maintenance cost? $1 Million absolute tops. The key to low
maintenance cost is flawless software. The second major factor is software that
is user-friendly. Between these two factors, you can cut the number of calls for
tech support by well over 99%. Furthermore, with flawless software, and easy to
use software, almost all calls for help can be handled locally by local IT
personnel. In my experience, most of the calls for help have nothing to do with
the software itself but relate to things like personal settings on users'
computers, lack of knowledge of the location of files on users' computers, minor
screwups with data entry, and the like. I have never had a user have a problem
due to a software flaw of any type with software supplied by RCP Software.
NEVER.
Oh, and about that annual maintenance cost? It should include taking
suggestions from end-users about features that could be added to the software
to help them with their daily tasks and actually adding those features into the
software from time to time.
So when will this end? When the California taxpayers, legislators, and others
finally run out of money and/or patience. Then, when someone who has the
power to do something comes to their senses, they'll have someone like me
come in, create a whole electronic records and management system in a couple
of months or so, and be done with it. And the whole solution will only cost
around $10 Million in vendor expense, and probably less than that, which will
include installing a modern system in every county in the state and educating all
the users. Then outraged citizens/taxpayers, suddenly indignant politicians, and
certain others will loudly and publicly ask how in the world could so many
supposedly smart people spend so much public money before figuring out they
were barking up the wrong tree.
As of now, late May 2010, the immediate outlook is not hopeful. People ardently
backing CCMS are still chasing rainbows and clinging to the vain hope that
CCMS will get fixed. So now it's probably time to go into hibernation. During
that waiting time, I'm guessing the pouring of money into the bottomless pit will
probably continue at around $8 Million/month. Politicians may be talked into
giving the judicial branch enough extra money to open courtrooms back up so
that the pouring of public money into CCMS can continue. But recent employee
layoffs at our superior courts will likely not be reversed. How long will we have
to wait for the realization to fully kick in that the CCMS software is not going to
get fixed? I'm guessing that a year from now we'll be about where we are now,
barring some developments in the political arena. Except of course for the
taxpayers being about $100 Million more out of pocket. I think at this point in
time it will take a skillful outside politician to bring the warring parties together
for a joint effort to help the California courts.
The Standoff And What Could Be Done
As a closing thought for the moment, I am aware as of mid-May 2010 of the
public standoff between two members of the judiciary over CCMS, one aligned
with the AOC and one being a presiding judge of a fairly sizeable superior
court. Neither party has sufficient knowledge of the technology to make an
informed decision about what technology to use or how to create modern
software. In spite of that, each side is making technology decisions and drawing
lines in the sand.
That is most unfortunate and I would hope that others who know these two fine
gentlemen can talk to them and convince them to set aside for the moment the
combative stances they have taken, for the good of the court system.
Interestingly, each side has a legitimate set of business demands or needs for a
court data management system and the heck of it is that, unless there is some
peculiar need of which I'm not aware, each side could actually get everything
they want in the way of business needs if they and others were to join together,
cooperate with me in the development of modern software, and set the CCMS
software/system aside as just an unfortunate lesson from which we can all learn.
How could it go if the warring factions were to get together and talk to me in
plain English (no technology people and no technology specifications) about
their business needs? They could then select a neutral ground to act as a
testbed. Preferably that would be some smaller Northern California county
within fairly quick driving range of my office that only has a few judges so there
would be fewer people to coordinate with during development and testing. With
good cooperation (much of which could be done during lunch time or before or
after regular court hours), prototypes of the various basic components could be
created and then tried out hands-on by actual court personnel on my laptop
computer. They could report on ease of use, reliability, speed, etc. Once people
see what can be done, I think a lot of the hard-line stances would melt away. I
realize that the human element here might induce some delay time while reality
sinks in at each stage.
Once the basics are done, one or more areas of law (e.g., probate, juvenile, or
criminal) could be fully fleshed out with electronic forms, fully enhanced data
management software, etc. and each area put into actual operation in the test
court. The court involved simply needs to have certain basic hardware and
standard software in place. The completion of the forms and expansion of the
software from prototype to actual working form wouldn't be free because it does
take some time to develop all the details but the cost would be a tiny fraction of
the cost currently projected for CCMS deployment into such a county. And, in
the interest of saving the court system from financial disaster, I would do this
first county at a heavily discounted price, essentially donating some of my time
for the good of the court system.
As reality gradually sinks in, I think the process could be speeded up to match
or nearly match what is actually possible. I can write this stuff many, many,
many times as fast as what people are used to seeing but I have noted people
are often slow to react to the production of prototypes because they assume that
what they have previously seen is a realistic model for how long it takes to
create specialty software. It isn't but they have a hard time accepting my
representations that the process could be done in the range of perhaps 10,000
to 100,000 times as fast as Deloitte Consulting has worked on CCMS. I recall
the somewhat startled reactions of certain people when prototypes of hugely
complex applications were presented to them for Alpha-1 testing within a day or
two of when they spoke to me about what they needed.
Once the first county is done, a second county could be selected and, by that
time, I think reality would have sunk to a great extent in so that county's entry
into the modern world of electronics could go much, much faster, perhaps in a
matter of just days. From there, the entire state could be done quickly, county
after county, producing a network of uniform case management.
It's Time To Stop Chasing Rainbows
When we were children, we read the enchanting fairy tales about the
leprechauns sitting at the end of the rainbow with the pot of gold. Of course, as
we grew up, we realized that no matter how far or fast you chase the rainbow, it
will always move away from you, although it may also just disappear. As the
end of the rainbow seems to be within reach, it moves again.
Those who are a bit cynical already about what has gone on have observed that
the completion time for CCMS has continued to advance further and further into
the future. A few years ago the end of the rainbow was 2009. That became the
Spring of 2010. But then that became September 2010. Now that has become
April 2011. And perhaps even more importantly, that has now become a date
for some type of "completion" that has become more and more vague as time
marches on because at the same time we are now hearing the date for "full
deployment" has been delayed over and over. I heard 2011, 2012, 2013, 2014,
and lately 2016. To put the matter simply, the 2011 date for completion of
something or other has no meaning at all and never did. And statewide
deployment of software after it has been written should take little time at all.
Weeks if you're fast; months if you're slow.
This has happened elsewhere so it's not as if we can't learn by example. In
Texas, they chased Deloitte's pot of gold at the end of a rainbow for years, with
Deloitte finally turning the problem over to others after hundreds of millions
down the drain. The bug-laden TIERS (acronym pronounced like "tears")
software was then adopted by others (Accenture) who in turn went down in
flames, turning the problem over to others (Maximus) who then, in turn, ducked
out by hiring a subcontractor (AutoGov - who incidentally only had a bare Web
site still under initial construction when I was writing this in May 2010 - but so
what, because they're politically connected) and referring all questions to them.
If you search the Internet you can find dozens of newspaper articles on the
subject, including one appropriately entitled, "A Trail Of TIERS".
And another crystal clear example of chasing rainbows is sitting right there for
all to see in a Denver court. Gee, maybe California judges could pick up the
phone, call the Denver court, and ask about the case of Davis v. Henneberry
which has been pending since 2004 with a stipulated settlement order in place
since December 2007. No one there in government seems to quite get it yet or,
if they do get it, they're too involved with the waste of money to want to do
anything about it. But maybe the plaintiffs are going to be back in court when
that 33 months provision hits the finish line tape in September. And Deloitte I'm
sure will be right there saying, "But we're soooo close!"
Time For Music & Politics
The claims of the backers of CCMS, TIERS, and CBMS should be set to music.
I might suggest a variation of the old "Over the Rainbow" sung so wonderfully
years ago by Judy Garland.
Now it's probably up to politicians with skills at getting people to retrench from
their hard line stands and work together with me to solve this situation.
See you in 2011!
And be sure to come back periodically to check the Demonstrations Page for
the latest demo and technical explanation.
Updates on CCMS, etc.
FWIW, I noted in late May or early June that the County of Marin finally woke up
and sued Delolitte Consulting over some software for managing certain
personnel data. They claim it makes piles of mistakes and has caused a lot of
grief. They finally bit the bullet and admitted they had made a huge mistake by
using Deloitte. I think it's going to take quite a while longer for the AOC to admit
they made a mistake. And it's possible they will never admit it.
Interestingly, on June 6th I watched a movie involving folks pulling off a huge
con job. And one of the observations made was that the larger the amount of
money being asked for, the fewer questions people ask. And there was also a
60 Minutes show on CBS about a lawyer who was arrested almost the same
time as Bernie Madoff. He had pulled off some huge multi-million dollar cons.
He got caught when someone in Canada was finally savvy enough to ask
questions.
An article appeared in a legal newspaper on June 4th revealing that confidential
data from Sacramento County Superior Court cases (they use CCMS for
probate and civil) was made public by the CCMS system. That would be a
problem emanating from the server located in Tempe, Arizona. It is not known
for sure how long the problem existed. In a well-designed modern system, the
court itself would have control over their own data. Such data would not be
accessible by the public because there would be no electronic pathway
between the public inquirer and the data. This type of problem is easily
prevented in a modern system.
June 9, 2010, KGO TV, ABC, Channel 7 San Francisco. Billion dollar
boondoggle report. One part of the video shows a computer screen pertaining
to some criminal case in Dept. 12, Sacramento, which is Judge Gilliard's court.
Although it is difficult to tell precisely without in-person inspection, it appears
that the technology being employed is in the range of 20 to 30 years old. The
screen presentation looks like what was common in the 1980's. The mention of
cutting and pasting rulings is caused by internal incompatibility of components
of the system combined with external incompatibility with software created this
century.
You see the personnel handling paper documents in the video and there are
comments about excessive expenditure of time filing documents. In a modern
system, paper would only be employed as a last resort and the vast, vast
majority of documents would be electronic. Filing of an electronic document
could take place in seconds. Many filings could take place across Internet and
Intranet connections. That would be routine for large volume filers such as the
district attorney's office, the public defender's office, county agencies, and large
law offices. Service of copies on other parties could take place within seconds
after the document is finalized. That is the power of modern electronics but,
sadly, the California court system is being bogged down now by fighting over
ancient technology, much of which was obsolete well over 10 years ago.
The mention of the Sacramento court's lack of control over their data is
unfortunate but in a modern system each county's court system would have total
control of their own data and would be able to do various types of statistical
analyses. That is easy to set up. I have already created such a system. Also,
each court would easily be able to designate which documents should be made
available to the public and which should not, thus avoiding disclosures of
confidential data such as took place in Sacramento in late May to early June
2010.
In mid-June, there were several follow-up reports by the news media, primarily
Courthouse News Service (www.courthousenews.com). These reports
mentioned numerous problems that the Sacramento County Superior Court was
having with CCMS, including disclosures of confidential data on public kiosks
located in the courthouse. The public could pull up litigants' medical records,
CPS records, bank statements, etc. just for the asking. Everyone blamed each
other. None of the people arguing had any substantial knowledge of the subject
matter they were arguing about.
Judge White also noted the problem of trying to pull up calendars for
courtrooms such as the law & motion courts, knowing there are dozens of cases
scheduled, and only getting a calendar with three items on it. That problem is
easily solved and I've already done it.
While the confidential data problem of course sprang from the totally inept work
done by Deloitte Consulting, the presiding judge in Sacramento missed the
mark in his analysis of the problem, blaming everything on the fact of the server
being in Tempe, Arizona. Any knowledgeable IT person who reflected on the
situation for only a moment would realize the server location affects primarily
speed and may also trigger some transmission errors causing incomplete data
transmission or errors in transmission, thus affecting speed and things like PCL
errors. But the failure to protect confidential data springs from poor design of
the system (failure to use isolation technology and techniques), failure to use
modern encryption tools, and just generally poor coding by the vendor - a
vendor which is famous for poor results.
The answer to the confidential data disclosure problem is sitting on the
computer I'm using to write this Web page. I have a complete demonstration
prepared, just waiting for someone in the court system to take me up on my
invitations to take a look at it. My suspicion at this time (late June 2010) is that
the CCMS project will eventually just wither away as the 51 currently non-
participating counties will simply decline to subject themselves to the
excruciating pain of being a CCMS county. Unless a new breed of politicians
calls a halt to this quickly after the next elections, CCMS will likely burn through
a few hundred million more taxpayer dollars before people come to their senses.
Also, as of late June 2010, I note that the State Auditor has their audit in
progress under investigation number 2010-102. (See CCMS audit.) It is
scheduled for a final report in October. Although it is highly doubtful they have
anyone on staff with expertise in writing modern software who can really give
them a good analysis on what has happened in the technological arena, I am
guessing they will probably figure out all of the following in the financial arena,
where they are the masters:
Spending any more money on CCMS is like spending a million dollars to get an
old typewriter to work so it can be used as a productivity tool in an office.
New Chief Justice
As of late July 2010, I note that the current Chief Justice has announced he will
retire on January 2, 2011, and a successor has already been nominated. As to
whether anything will change, that depends on whether the new chief justice is
willing to examine the faulty assumptions that have been made in the past and
is willing to try an alternative approach. I have no real idea how that will play out.
One very basic assumption that has been made, and which is totally erroneous,
is that any "IT person" is in a position to give advice about how to create a case
management system. Lawyers and the general public know better than to
believe that any doctor selected at random can do heart surgery. And they
understand about specialization in law. But the court administrative leaders in
California have blithely assumed that any "IT person" is qualified to give advice
on how to construct a case management system. They have not analyzed the
situation correctly. A correct analysis would almost certainly reveal that not a
single person at the Administrative Office of the Courts or at any of the county
court systems has the required expertise. Reliance on IT personnel who have
expertise in other areas besides software writing has gotten us to the present
situation of approximately half a billion dollars having been spent with nothing
productive to show for it.
Correction of this situation depends on the willingness of leaders to admit that
mistakes have been made and that it is time to start over and use modern
technology solutions created quickly and inexpensively by people who have the
necessary skills. Are they willing to do that? Only time will tell.
One safe bet to make, however, is that when September 2010 comes, the AOC
and Deloitte Consulting will once again announce that they are "so close"
(they've been there for years now) but they need more time to finish
writing/fixing the CCMS software. How much time will they say they need?
Beats me. Might make an interesting office pool subject. Six months? Eight
months? A year? And what will the excuses be this time? Everything but the
truth is of course a possibility. Again, here's a good excuse for another office
pool.
how our politicians are foolishly wasting hundreds of millions - even billions - of
taxpayers' hard-earned dollars on exorbitantly expensive IT projects that are
either already failed or are almost guaranteed to fail. How bad is the situation?
In California, the total amount wasted over the last 10 years on foolish IT
projects probably exceeds the currently projected (as of May 2010) state budget
shortfall of $20 Billion by quite a bit. The waste per year on exorbitantly
overpriced IT projects has increased to the point of several billion dollars. Most
of these projects were put into place with lobbying and are held in place -
against all rationality and dramatic budget shortages - by lobbying.
In many of the discussions, I am going to name names. If any of the politicians
involved don't like it, they're certainly welcome to publicly challenge me on what
I'm going to say here. Challenge me publicly that I can't write - at minimal cost
and in a short time - the software I'm describing and have it work well. Let's
have a competition. The voters (particularly in California where the state
government is dead flat broke and throwing away money at breathtaking speed)
need to hear what is going on with this incredible waste of their money so they
can make intelligent decisions in future elections.
And I may be in California, but this foolishness spans a number of states,
including Colorado, Indiana, Maryland, New York, and Texas for starters.
England probably had the most monumentally foolish IT project of all.
To get a handle on just how much of your hard-earned tax dollars are being
literally thrown away on doomed state IT projects, you might want to start by
reading two files linked to this page. Each file holds an entire series of articles
written for the Daily Recorder, Sacramento, California. I have also written a
number of freelance columns for the Daily Recorder on computer technology for
courts and lawyers. The two series, which were written about a failed project
called CalWIN and its spinoff, CBMS (Colorado Benefits Management System),
are:
The CalWIN project has been the subject of a number of TV news discussions
and the CBMS project, which is a spinoff of CalWIN, has been the subject of
hundreds of news media discussions in Colorado where it sparked lawsuits,
appointment of a special oversight person (a retired dentist turned politician
who didn't know beans about computers), reimbursement requests from the
federal government for erroneous payouts due to software miscalculations, and
a number of legislative discussions. The system is famous from coast to coast
as an example of a poor IT project and has even been the subject of business
graduate school classes at the University of Denver on everything not to do
when putting an IT project together. There has been a stipulated settlement
order in place since December 2007 in a lawsuit (Davis v. Henneberry) brought
against the state, but the current vendor, Deloitte Consulting, has been unable
to fix much of anything and the problems continue unabated. This idiotic IT
project has cost Colorado taxpayers around a quarter billion dollars as well as
causing enormous difficulties for their welfare system. The parent California
version? Pretty close to a billion.
Plenty of people in California had an opportunity to call a halt to the CalWIN
(parent) version. In Sacramento County, the County Board of Supervisors had
clear warning that something was drastically wrong. When the system went live
in their county, it had already gone live in neighboring Placer County, sparking
newspaper articles about the disastrous results. After the Sacramento County
Board of Supervisors heard from one worker after another, and were besieged
at their public meetings by welfare workers begging for the county to stop the
pain, I offered to help. But they weren't interested. Specifically, also, the head of
their Department of Human Assistance, Bruce Wagstaff, spoke with me. But he
wasn't interested in getting help, either. Note carefully that these folks poured
additional tens of millions of dollars into this foolish project, all the time refusing
offers of help. And now (2009 & 2010) their county is crying that it's broke. At
one point, one of my newspaper technology column readers intervened and
arranged a meeting with the newest elected supervisor, Jimmy Yee, because
he assured me Yee was honest and would be interested in fixing things. Yee
wasn't interested, either. TV news clips about CalWIN and its failures continue
from time to time, your tax money keeps getting poured into this bottomless pit,
and no politicians are willing to do anything about it.
The foolishness associated with such projects for creating exorbitantly
expensive poor quality software to do simple processing of unsophisticated
welfare data is not, however, limited to California, Texas, and Colorado. I
encourage anyone who reads these materials to pick up the phone and call the
people involved. Demand that they stop this incredible waste of your tax money.
One upcoming IT project in California is SOMS (Strategic Offender
Management System) for the California Department Of Corrections &
Rehabilitation. Look at http://www.cdcr.ca.
gov/News/2008_Press_Releases/June_19.html to see another IT project
disaster in the making. This one may well bleed the California taxpayers of at
least $100 Million. As a matter of fact, I heard in 2009 it was probably going to
cost about $250 Million. Here we go again! Prison inmate data, I assure you, is
not complex.
The California prison system, in combination with law enforcement agencies,
might actually be a logical starting place for a demonstration project for what
could be done to link up agencies with inexpensive, powerful, accurate, easy to
use software. I understand the computers on the medical side of the prison
system aren't even compatible with the system that tracks the inmates in
general and neither is compatible with computers at law enforcement agencies
such as sheriff's departments or police departments. A revamping of the entire
system is obviously in order and it wouldn't have to cost a mint. But then, of
course, there are the RFP's written by people who know absolutely nothing
about writing modern software, the lobbying, the kickbacks, etc.
And another huge money pit into which California's Administrative Office Of The
Courts is just aching to throw your tax money is a project called the California
Court Case Management System (CCMS) now being handled by a company
with a well-established track record of failures. (They're currently sitting on top
of Colorado's CBMS mess which is about to explode again into the news.) If it
weren't for the failed economy, and the current lack of state money, the
managers and backers of this crazy project would be pouring more than $1.75
BILLION into this junk at the rate of hundreds of millions per year. If you're
thinking that's a typo I assure you it's not. That's BILLION not million.
Temporarily they want to pour a mere $100 Million/year or so (what one person
who should know better by the use of simple common sense publicly termed
mere "life support") into this nonsensical project to "maintain" it while waiting for
the budget to recover in a few years so they can then follow that up with at least
another billion dollars of taxpayer funds. The administrators don't understand
the technology, have relied on the wrong people for advice, and are now stuck
with a monumental lemon. They are having difficulty coming to grips with their
situation, which is now being dictated more by the psychology of human
behavior than by technology.
Recently, with the announcements about impending closures of many courts in
Los Angeles, the Mayor of Los Angeles practically begged the court system not
to close courts. And there have been layoffs at other courts, resulting in long
lines of citizens waiting for service. Sacramento County has had long wait times
at the filing counter. One TV newspiece showed a lady who regularly files
pleadings as a legal runner sitting in a camping chair as she waits hours to get
to the counter at Sacramento County Superior Court. That has occurred also in
Alameda County. TV reports about San Diego have shown lines extending out
to the street. This is caused by the clerks being unable to work quickly with the
awful CCMS software. Absolutely none of this is necessary.
Want to see just how bad the results from such poorly done IT projects can get?
Contact Ginger Sylvester at the Sacramento Superior Court and ask for a copy
of the April 2010 CCMS briefing slides.
And if this waste of your taxpayer money just plain ticks you off, take note that I
offered to create case management software with stunning capabilities for the
California superior courts and do the first county for FREE as a demonstration
and still didn't have one taker. My wife told me to just withdraw that offer so it's
off the table now. I'll still do it, but not for free, although prototypes for people to
try out hands-on would still be free. The court leadership in one county told me
they couldn't accept help because their hands were tied. But the first county
court to take me up on the offer to create good case management software will
get it for peanuts compared to the $1 Million/judge price tag projected for
CCMS.
Perhaps it will take the utter collapse of some county court's old mainframe
(COBOL-based) system before anyone in charge of the funding will be willing to
admit that CCMS is a huge mistake and be willing to accept help. If that
happens, my number is (530) 306-6370. I can't guarantee the recovery of lost
data (although I would certainly try to recover it) but creating a system to work
for the future could be done on an emergency basis in probably a week to 10
days. Hooking up other counties into a true system with sharing of data could go
just as fast per county. And it's sometimes possible to recover current data even
if you can't get all the ancient software designed to run the files containing that
data to work.
Apparently we need the court system equivalent of a huge trainwreck with dead
bodies littered around before the people managing CCMS are willing to look at
better solutions. Please take note that I believe CCMS has been backed by
Ronald George, the Chief Justice of the California Supreme Court, only
because he has been misled by various people around him upon whom he has
relied for technology advice. Justice Ming Chin, head of the Court Technology
Advisory Committee (CTAC), is in the same position. He has been misled.
Some of these supposed "experts" - who apparently have been handing out
advice to the "suits" or "robes" - ought to know that they don't know how to write
modern software and should thus know they are not in a position to give advice
on the subject to anyone in charge, or to anyone holding the purse strings, or to
anyone whose backing they seek. As a prime example, Sheila Calabro, the
Director of the Southern Regional Office of the AOC, who was placed directly in
charge of the CCMS project, should know better than to believe she has any
knowledge of how modern software to run court case data management should
be written. I suspect the top leadership of the judicial branch has been given
bad advice by quite a number of people although I don't know exactly who all
those people are. Misleading the Chief Justice of a state supreme court to
obtain backing from such an official is a serious matter, at least in my opinion.
Those higher up the line than Ms. Calabro, including the Legislature and
ultimately the taxpayers, should force a termination of the CCMS (or at least a
time out) and just spend the approximately $10 Million or less in legitimate
vendor cost that it would take to create a system of electronic superior court
case records and a system to manage them. That is certainly a heckuva lot
better than continuing to spend over $8 Million/month on CCMS, with no end in
sight. This situation is a classic example of where we should follow the old
adage and not throw good money after bad.
The CCMS software is so bad that one of the pilot counties, Los Angeles,
essentially trashed it and now only uses it in some small court for minor matters.
And also please look at the justice system software pages where more
boondoggles specifically linked to the justice system will be discussed.
WHY??? For God's sake why???
So why are such government IT projects draining state coffers dry and forcing
hard-working taxpayers to pony up tax money so it can literally be thrown away?
Well, first there are several common misconceptions about the creation of such
government software and systems and these misconceptions foster such
monumentally foolish wastes of public monies.
But first, before we get into the technology issues, let's take a look at the
psychology at work in these boondoggles. These "multi-hundreds of millions"
projects are all essentially "con" jobs. As P. T. Barnum once put it, "There's a
sucker born every day." We are all subject to being conned, no matter how
smart we are. Sometimes it's much easier to con someone than in the average
situation because the "connor" takes advantage of a "connee" who is really
eager and is totally ignorant about the subject matter. If the potential pigeon has
already decided that "something must be done," then the setting is complete.
Pigeon drop schemes, Ponzi schemes, and the like all feed off of the desire to
get rich quick. We all have something we would really like to have
accomplished. The con artist takes advantage of the existing desire. If someone
wants some result really badly and is highly ignorant of the underlying subject
matter, then you have all the makings of a really good con. People want to
believe. Religious investment schemes are often such good cons that when
investors are presented with the facts, they go into heavy denial. "He was such
a nice man. He prayed with us every day." They can't believe they were taken.
And politicians have a big problem in this area because they have to figure out
a way to tell their constituents it wasn't their fault after they just spent huge
amounts of taxpayer money on something that was really simple and
inexpensive to do.
When confronted with the possible reality, human beings often go into denial.
Politicians and bureaucrats won't look at alternatives to what they have wasted
money on because that might result in the reality of the situation becoming
public. Then they would have to explain how they foolishly fell for a con job and
spent the taxpayers' (aka voters) money. So they look for another charlatan with
a new set of glowing promises instead of putting a stop to what they started.
And I have personally seen examples of this in the IT arena when simple
solutions were right in front of the politicians.
Politicians and government bureaucrats are often easily taken when it comes to
IT projects. Few know much of anything about computers. Fewer still know
anything about writing modern software. Thus it becomes fairly easy to talk
government bureaucrats into spending hundreds of millions of dollars and years
of time on something that could, in reality, be done in a few weeks or months for
just a few million dollars, with that price including full deployment, education of
thousands of users, and distribution of software, as well as the initial
development. Faced with the reality that something that has been in progress
for ten years could have been created in ten weeks, any politicians who were
the least bit involved will duck and hide rather than confront the situation. So it
often takes a new generation of leaders who can point back at those they
replaced and point out how utterly foolish they were.
Voters need to consider this in upcoming elections.
So what are the key ingredients to a really good con of government officials on
IT projects? Well, first you impress them with how you're going to fulfill lots of
supposed requirements. Get the people you're going to fleece in on creating the
method of picking their pockets. Have them set out detailed requirements. If
they can't seem to do it, give them some help. Have them set themselves up to
be taken. The best formula is to have someone in government who thinks they
know how to set out detailed specifications for an IT project ("I wrote some
mainframe software 25 years ago so I know how to do this"), then urge them to
get going on it. I saw that process in action on one project that has cost the
California taxpayers nearly a billion dollars and people involved in it are still in
denial. And it still hits the evening TV news from time to time as people note
how the software makes tons of mistakes handling simple calculations. And the
project expanded to a second state and the folks in the second state are also in
denial. And that state is on their fourth "savior" and still plowing through tons of
money. It is one of the most marvelous con jobs I have heard of. It's so good it
belongs in a college psychology textbook. So far it's only hit the college
classrooms (at DU) as a classic example of everything not to do when
designing an IT project.
When you're setting up "requirements," don't just describe what the software
should do in simple, straightforward terms like "software that will enable a
particular group of workers to do their jobs faster, more accurately, and easier."
That will never do for a good con job. Set down thousands of detailed "technical
requirements" to make it all sound really complicated. If you can create a list of
such "requirements" and then put it on paper, and the stack is at least two feet
high, you have the makings of a superb con. One that would make Bernie
Madoff envious. And, to take a line or borrow a word or two from linguistics
masters such as J. Clark Kelso, Deloitte Consulting, EDS, and others involved
in or being taken by these schemes in California, use key, important sounding
words like "strategic" and "system" and "statewide" and "integrated". Skillfully
blend words like those together and you have the makings of a con that can hit
nine figures easily.
For something that could have been designed in a couple of weeks and would
actually have worked.
Now, back to the technology involved in government IT projects and common
misconceptions that lead to these boondoggles.
If the project is going to serve many end-users, that necessarily makes the
software to handle it complex (and thus expensive), right? Wrong. Dead Wrong.
Same thing for projects that are going to handle lots of records. Same for
projects that are going to handle lots of money. Same thing for projects that are
going to handle "important" data instead of "unimportant" data. And the fact that
it's government data instead of data to track cookie sales for a national
organization of children? "But Richard, this is court data."
Doesn't mean a damn thing.
Well, if none of these commonly touted factors (vulture IT vendors love to
emphasize them to naive bureaucrats and politicians) have any direct
correlation to the complexity (and thus the legitimate design cost) of the
software required, what is important? It's the complexity of the functions
required to be performed by the software. Want to see complex? Look at
SuperSingles-n-Doubles.
Software that has to process complex functions for 500 people and that will be
handled by one person can be incredibly more difficult to design than software
to handle 1,000,000 records that will be handled by 60,000 workers. There is no
necessary correlation between the number of people or records involved and
the legitimate design cost of the software.
"But Mr. Politician, look here. This software has to be worth at least $300 for
each install, right? Surely you understand that value. We charge by the number
of butts in the seats. So, let's see now. $300 times 60,000 is $18,000,000 and
you can understand how $18 Million worth of software could take $100 Million to
design. Right? And you want to underwrite our design costs, of course, right?
And then, of course, we have to contribute something to your re-election
campaign, and your political party could use a couple bucks, and you wouldn't
really mind if we made a little profit, would you? So you can understand now the
$400 Million price tag, right?"
So if it's more complicated to design software to handle data for certain types of
complex bowling events (and it definitely is) than it is to design software to
handle court data or welfare data or unemployment insurance data, then how
can the prices you typically hear connected with government IT projects to
handle such simple data make any sense?
Well, they don't. And how can it make any sense for RCP Software to only
charge $79.95 or $99.95 for software that required more design work than
software to handle many government operations would require?
First of all in the software world, there is the factor of economies of scale. A
major software manufacturer that spends $100 Million on developing a
particular application but sells 20 Million copies to completely separate
purchasers, simply spreads the cost out and makes a profit on everything
above $5 + distribution and marketing costs per copy. Microsoft Office,
Microsoft Windows, etc. don't cost you a million bucks a copy. So I can still
make a buck off bowling software, even charging very low prices. And besides,
I'm not out to make a killing on this bowling software and I am definitely not
trying to gouge bowlers or league secretaries or others in the bowling world. I'm
trying to help them and I have noticed they definitely appreciate that. I have
received numerous compliments and thanks when it comes to the bowling
software. And besides, it's fun to write this stuff.
By way of contrast, there are usually fewer copies of government software that
actually need to be distributed than what the vulture vendors tell naive
government officials. Vulture vendors love to bamboozle ignorant government
purchasers with confusing license provisions, and put emphasis on the large
number of users, etc.
But the reality is that a large government office can often take distribution of a
few copies or even a single copy of specialized software and work from it. But
that's of course how an honest vendor would do it. So it's theoretically
necessary to charge more per copy but that still doesn't need to be very much.
Note carefully that many of the copies can be absolutely identical.
But that inability to spread cost among millions of independent users still
doesn't mean the total project has to cost hundreds of millions of dollars.
Overall costs should correlate much more strongly to design complexities rather
than to the number of users, copies, etc. Design complexity in turn correlates
strongly to the amount of programming time required. Let's look at that situation
a little more closely.
Many governmental IT projects priced at figures like $500 Million or even $1
Billion or more could be done for $10 Million or less. The end-users just need
current computers outfitted in a standard manner. The legitimate "design" cost
might be only $2 Million. Distribution, installation, and education costs, with
some left over for vendor profit would consume the other $8 Million. Ongoing
maintenance costs could be minimal, particularly if the software works
flawlessly.
One additional observation of the high pricing of many government projects is in
order here. The number of users doesn't logically relate to design costs at all
but, in the case of poor software that has lots of bugs, maintenance cost does
relate to the number of users. If the software is well written, there is no
maintenance cost problem. The difference between 2 calls in a month and 15
doesn't really mean much. But with many of the vulture vendors lobbying and
getting government contracts, they might have 6,000 calls for help in a month.
Outfits like Deloitte Consulting that provide extremely poor software have to
field tens of thousands of cries for help. Thus the irony is that the worse the
software is the more it costs.
Recently (end of April 2010) I had the privilege of sitting down with some nice
folks who work in a particular government agency in California. They are in a
certain amount of distress over a "system" that they have had to work with and
which doesn't work worth a damn. They have stated their distress publicly. I
heard their pain and talked to them. I think they got the message that I could
quickly create what they need in the way of data management software, and that
it would work like a Swiss watch compared to what they are suffering with now,
and that I could do it for a tiny fraction of what has already been spent. But they
are timid about doing anything about the situation. They feel their hands are
tied. I'm not going to name the local agency or people involved or where they
are located because they were all very nice and polite. But I hope they will
rethink what they told me about being unable to do anything really constructive
about their situation.
Interestingly, while I was sitting there, one of the attendees told the local IT head
honcho about problems she was having that day with the system. The IT head
honcho responded with an answer that I knew instantly was, to a 99% certainty
level or better, not the primary source of the problem.
So the people at this agency know they are immersed in a disaster but won't
speak up to anyone in government who could do something about it. This, dear
taxpayers, is what is wrong with California government. No one who could
influence anyone to get something constructive done, save money, and make
government work better speaks up. The state-level agency involved here is on a
dead certain path to a huge trainwreck. The state government is dead flat
broke. The taxpayers cannot afford to finance any more of such foolishness. The
project involved is grounded on colossal ignorance of modern technology. I
don't mean a little bit of ignorance, I mean absolutely breathtaking, stunning,
wall-to-wall ignorance. I would be willing to bet that not one single person
involved in the entire project (and that is hundreds and hundreds of people) has
any substantial knowledge of how to write modern software.
The local chapter of the agency involved has, probably in reliance on
assurances from local IT personnel, tentatively planned to do a workaround that
would incorporate the bug-laden software with which they are already working.
The risk involved in such an undertaking is HUGE. And I told these folks that.
The base software on which they will be relying apparently has thousands if not
tens of thousands of coding errors. And the local folks know it. I mean they have
to know it because they told the public all about their experiences, describing
problem after problem after problem. The agency involved deals with important
data and if it goes down, the impact on society can be devastating. And they
have already had several shutdowns. The tentative solution planned by the local
chapter of the agency is absolutely devoid of any common sense.
If this local agency has a catastrophe and they finally call in desperation and just
agree to do what I tell them to do, will I help them? Yes, I'll take the high road
and help them to the extent possible. The agency's task is too important to allow
them to fail. But if they need a panic bailout, it won't be free. And under no
circumstances will I spend any time whatsoever debating with them or any of
their IT people how to rescue them from disaster. I just hope they have the
common sense to make backup copies of all data they can save. And,
considering the road down which they are contemplating traveling, there is
absolutely no guarantee that anyone - myself included - can bail them out as to
their current data if they encounter certain types of smashups.
Among the ideas this local agency is apparently seriously considering to work
around their problems is the idea of buying nearly $300,000 worth of server
hardware to host the bug laden, ancient-type software that has already been
written. To describe this plan as spectacularly foolish barely does it justice. It is
the equivalent of buying a titanium-platinum container to hold yesterday's
garbage. Garbage in a high-priced container is still garbage. A server bought
over the phone from Dell or HP for less than $5,000, combined with modern
bug-free software, could do a far superior job and at far less total cost. For
$10,000 you could get a server from Dell with spectacular capabilities.
It is such a shame that people who ought (by use of simple common sense) to
know better enter into such foolish ventures. The evidence of the total
worthlessness of the software they have been dealing with is all around them.
They are immersed in that fact every day. It is a shame that apparently the only
way to really have that hit home is to have a total collapse of the system and
then have no help from the software manufacturer available.
Before venturing into the briar patch they are contemplating entering, I would
suggest to them that they have their CTO participate in a little test of her skills.
Since she is supremely confident that she can handle the situation in front of
her, she should be willing to demonstrate that to the folks in the suits, right? It is
a test to determine if the person understands and can detect dependent coding
errors. This problem is obviously present in the software they are presently
using, based on the plain English descriptions I heard by non-techie users of
the system of various phenomena they encountered.
Of course the problem with finding and fixing dependent coding errors is that
they cause random problems. The problems occur sometimes but not at other
times, depending on the combination of code being relied upon by the particular
action being taken by the end user. The mathematical possibilities for where the
parts of the problem could be located are staggering in some of these
situations. We are talking exponential numbers of possibilities. It can be so bad
that an army of Einsteins would have trouble finding the problem(s). And
although you may have some general idea of where the problem is located, you
don't even necessarily know how many such errors are present, what type they
are, how they interact with each other, or whether they are even visually
obvious. You can be looking right at the problem code and not realize
something is wrong. I have personally encountered such a situation and, if it
were not for the fact that I thoroughly tested every single sub-area before
moving on when writing the application involved, I doubt I would ever have
figured it out.
Fortunately, I knew better than to tempt fate because I don't have a 400 IQ, so I
routinely tested every single layer before moving on. Up popped a bug. I knew it
had to be something I had done very recently. So I looked at that code.
Carefully. Over and over. Then I did elimination tests. Everything seemed
correct. Then finally, bingo! I found the small rubber raft floating somewhere in
the Pacific Ocean. But even when I looked at the code string that I had narrowed
the problem down to, I couldn't see the problem. After I knew for sure where it
was located, I performed some alternative type invocations of code in the same
dependent situation. And then it hit me. Then I figured out how to fix it.
Want to get a taste of how tough the problem can be? Look at this sequence.
615, 616, 617, 618, 619, 620, 621, 622, 623, 624
There's a problem with that sequence. Don't see it? I know damn well there is,
at least in a certain situation, because I ran into it. One of the numbers in that
sequence is non-usable. Now imagine that you are dealing with the problem of
locating the problem because you were so confident of your abilities that you
wrote layer after layer of code without testing each small group and now all you
know is that the problem likely lies between 14 and 1,216. Maybe. Not really
sure. I mean, there's only an outside chance it's somewhere else. And gosh, if I
see the problem, I'll recognize it instantly. Right?
But hey, you lobbied the right politicians, and the government (read "taxpayers")
is paying you by the hour to find the problem, so what's the big deal, right? I
only found the problem in the situation described above because I knew the
problem had to lie in this short sequence. Elimination testing with alternative
invocations located the problem and indicated the fix.
The code that these folks would be dependent on likely has many, many errors
or problems, perhaps ten thousand or more. And many could be of the type I
personally encountered. Trying to fix such problems could take eons. And since
you don't know in advance if the problems you might encounter are going to be
crippling, you don't know whether you will find yourself up against a wall,
absolutely needing to find and fix the problem. And to top it all off, these folks
don't even have guaranteed access to the source code.
One can only hope these folks will come to their senses in time. If not, and if one
of those random, critical bugs hits them smack between the eyeballs, I
guarantee you will read about it in the newspaper, see it on TV, and be asking
how in the world could such supposedly intelligent people have been so foolish
as to tempt fate like this.
So have other attempts been made to get elected officials in California off of the
track of just throwing money away on what should be simple and inexpensive IT
projects? You bet. Quite a while ago I contacted my two supposed
representatives to the California Legislature and offered to help. I explained to
them about the potential savings reaching into the billions if foolish IT projects
were scrapped in favor of better approaches that would cost only a tiny fraction
of what was currently being spent. That would be Senator Dave Cox and
Assemblyman Ted Gaines.
Neither was the slightest bit interested. I learned later from an article in the
Sacramento Bee, and information the newspaper had collected and then posted
online, that both of my representatives had been lobbied by organizations
whose members are employees of IT vendors that have been ripping off the
state government right and left on these crazy IT projects. Free computers.
Boxes of chocolates. Is that the basis on which these two officials - both of
whom loudly proclaim they are out to cut the cost of government and shave
down taxes - are supposed to be making decisions? As a taxpayer who has to
foot the bill for all this, what do you think?
The May 2010 CIO Report On CCMS
In early May 2010, an e-mail came around from the State Bar. There was a link
to more information on CCMS and that included the CIO's report to the
Legislature on CCMS. Before taking the contents of that report as being a
serious technical analysis of what is wrong, the California Legislature needs to
take a deep breath and remind themselves of certain critically important facts.
First of these is that the position of CIO is a political appointment. Neither the
present CIO, nor her predecessors, know beans about computers or, more
specifically and what is critically important here, how to write modern software.
Relying on the CIO for an analysis of the technology problems underlying
CCMS is roughly analogous to appointing the business office manager in a
hospital to the high-falutin sounding job of "Chief Medical Information Officer"
and then relying on her for surgical advice.
You can read the CIO report and I might suggest that you do so and get it up on
your screen in order to better understand what I am going to discuss below.
This is a report by a non-techie for non-techies. Every once in a while they make
an intelligent observation but that's about it. When they do note a serious
problem in CCMS, they gloss over it. They don't pinpoint what is wrong with the
technology because they don't understand what is wrong. They admit to being
puzzled about why certain things don't work because - let's face it - they are.
The report is written from the standpoint of a non-techie assessing the possible
business value of some gizmo they bought at an office supply store or some
service to which they have an ongoing subscription. The report never hones in
on the poor technology and how the task of creating electronic court records
and managing them could best be done.
Right up front, note the comments under Governance on page 5. Looking at the
situation from a $$$ standpoint, the CIO has at least caught on to the simple fact
that this nightmare software has to be spread over all the counties or pretty near
all of them or the dollar cost per county is going to be astronomical. Assuming
for the moment that the 51 counties who have not become involved in CCMS
look at what is occurring on a daily basis in Sacramento (the only true test
county so far) and intelligently decide to avoid CCMS, or at least until all
problems have been solved, then all the pitches made to the Legislature and
everyone else so far about this being a statewide system are undermined. And
note that, if the amount actually being spent keeps mounting (and it is, at around
$8 Million/month), this is going to be the most extravagant IT boondoggle in the
history of the state and probably the entire country.
The CIO argues the benefits to the state system as a whole should take priority
over the unique needs of individual courts. Hold it there! Is that some kind of
excuse warming up for trying to impose crippling software on individual
counties' courts? If it doesn't work for a particular county, what good is it?
In the first place, with minor exceptions, once good software is designed for one
county's courts, that same software can be used everywhere. The first county
might take a couple months to get fully going while the second county could be
done in days, perhaps a week. A good system could spread across the state in
weeks, from county to county. But it should be complete or at least nearly
complete and working like a Swiss watch in the first county before it is used
elsewhere.
CCMS, on the other hand, has proceeded with glacial slowness, creeping from
one county to another, with one part being tried here and another part or parts
being tried there, and inflicting pain everywhere it goes. The process has gone
slowly because the software is loaded with thousands upon thousands of bugs
(code execution errors) and the system architecture and specific software
architecture are just plain awful. The dumbos at Deloitte Consulting are
spending day after day, month after month, year after year, trying to find and fix
all their coding errors. There isn't even a single county that has electronic
CCMS records in all areas of law, from probate to criminal to family law to civil
to juvenile, etc. Different participating counties have different software. Out of 8
possible areas, the most any county has software for is 4. And note carefully
that they have not yet hooked up with any outside agencies, even though the
ability to do that was supposedly one of the big selling points for CCMS.
It is my understanding that each county court can decide whether or not to use
CCMS. And all the seven counties currently involved volunteered, although
some are obviously having second thoughts about that decision. That leaves 51
counties not affected yet by CCMS. However, none of the leaders of these 51
county courts, using ordinary common sense, will likely commit their courts to
using such a system unless everything is fixed in advance. And that is never
going to happen. It's impossible. Would you buy software or let your business
depend on it if you knew the software had thousands of bugs and myriad
problems?
Note carefully the lack of any rush by early testers of one or two parts of CCMS
to adopt the remaining parts that are being tested elsewhere. While publicly
some are trying to put on a smiley face and assert the pain isn't too bad, none
are rushing right out to find more needles to stick in themselves.
And even if you found and fixed all the coding errors, the CCMS
software/system would still be worthless because the overall architecture is like
something out of the dinosaur era of early computer technology. Parts are
incompatible with each and with current technology. Current technology
consisting of modern hardware and modern software can run rings around this
CCMS junk. Bear in mind that this project has been going for years. Everything
in it is way out of date.
Such projects only make sense if you can write the software in weeks at most
(which is possible), and get it up and running everywhere in months at most
(also possible). Then you need to remember that technology is going to
advance, so updating everything two or three years down the line should be in
your line of vision, assuming you want to keep taking advantage of advances in
computer technology. Software I wrote just a few years ago, while still running
smooth as silk (it has absolutely no bugs), is now outdated in comparison to
what can be done so I am encouraging customers to get their computers
updated so they can take advantage of new versions I am writing that take full
advantage of the huge advances in technology. And what I have written recently
will be dated in two or three years, just as surely as the sun rises and sets.
To see a report in another state on another failed Deloitte Consulting project,
also done by people who don't understand how to write modern software, take a
look at this Texas Comptroller of Public Accounts report on TIERS, Accenture,
Deloitte, etc. It talks about money and cost overruns, etc. but is written by
people who are clueless about the technology problems so, just like California's
CIO report on CCMS, it doesn't get to the heart of the problem.
Note that in Texas, the utterly failed TIERS project was handed over to
Accenture by Deloitte after the Texas government was conned into trying to get
yet another savior to fix the problems. I understand that at the time of the
handover which was in late 2005, TIERS had hundreds and hundreds of
problems. Deloitte can no more fix the problems in CCMS code than it could fix
the problems in the TIERS code.
Oversight of CCMS
There are supposedly a number of oversight groups or committees watching
with eagle eyes to make sure CCMS comes out right. One of these is the Court
Technology Advisory Committee, currently chaired by California Supreme Court
Associate Justice Ming Chin. Curious what the makeup and expertise of the
committee is? Look at this list. And look at this description. The makeup is set
by Rule 10.53 of the California Rules of Court. Now here's a simple question.
Where is the modern software-writing geek on this list? Here is the current
membership.
There are a few people on the committee who have tangentially "geek" job titles
but, as I have found out by talking to other people in such positions, they
typically have absolutely no clue about how to write modern software. Fix coffee
spills? Yes. Reset network passwords? Yes. Manage local networks? Yes.
Manage local Web pages? Yes. Handle general problems people have with
their computers or with specific applications? Yes. Write software like
SweeperScore or SuperSingles-n-Doubles or BowlerTracker or other similarly
complex modern software? They don't have a clue. But that's what the problems
with CCMS are all about. It's about software-writing. An attempt to create
electronic case records and an electronic system to handle them.
With all due respect to the people serving on this committee, they are in no
position whatsoever to evaluate the technology in CCMS unless they personally
understand how to actually write modern software. I realize the members of the
committee would probably rise up in righteous wrath if confronted with such an
assertion, but if you're a taxpayer being asked to foot the bill, I assure you it's
the key subject about which you should want to inquire. The failure to have
anyone who can actually write modern software (and I don't mean just some so-
called "IT person" who manages some local network) on the committee is a fatal
defect. All these people can do is generally realize that certain software doesn't
work worth a darn. They have absolutely no idea what to do about the situation
in terms of the technology fixes. Since the problems with CCMS are all
technological, this committee is absolutely worthless when it comes to the
matter of technologically solving the problem of obtaining software and
processes to make court records electronic and manage them. If they were to
use simple common sense about the situation (which I suspect they eventually
will), they could at least realize CCMS is going nowhere and they could cut off
the funding. And that would be a valuable action to take, even if they can't
specify the correct technology to be used to create a successful system.
In short, the problem with CTAC is roughly the same as in the hypothetical
situation I alluded to above of appointing a business office manager to the
position of Chief Medical Information Officer and then relying on her for surgical
advice. The CTAC members are people from the business for which the
software is to be written. And they understand the business. But the problem
isn't in the business, it's in the technology. It's in the tool they want to use in their
business. You don't need the business manager, you need the surgeon.
Unfortunately, neither the people on that committee nor anyone involved in
appointing them get it, or at least not yet.
What CTAC-type people can do intelligently is to evaluate from an end-user
standpoint whether or not a prototype actually works. It's like going to a store
that puts on demonstrations and trying before you buy. It's like taking a test
drive of a vehicle before you buy it. Ordinary people do this all the time with a
wide range of products.
The Future Of CCMS
This project is going to just plow through a bunch of money for the near future.
The people involved have no capacity to find and fix the coding errors causing
all the bugs, the software architecture isn't going to get fixed because a fix
would involve starting over, the system architecture isn't going to get fixed
because that again would involve starting over, the software conflicts are going
to remain because removing them would also involve starting over, the whole
thing is going to become more and more obsolete than it already is, and the
problems will just continue until the money runs out completely or people finally
come to their senses. Assuming this continues at around $8 Million/month, that's
going to consume an awful lot of scarce judicial branch resources.
Not sure that's going to happen, because, gosh, Deloitte Consulting has all
those "geniuses" working on this? If you want to get a good taste of how this will
play out, at least for the forseeable future, look at this summary of a settlement
reached in the Denver District Court in Colorado in December 2007, concerning
a now (in)famous IT project known as the Colorado Benefits Management
System (CBMS). And look at this page at the Web site for the Colorado Center
on Law & Policy. They sued in 2004 because the system was an absolute
disaster. The folks in Colorado are still waiting for the myriad problems to get
fixed while that could have been done in a matter of months a long time ago.
Note the current contractor carefully. OMG! It's the same folks running CCMS!
Originally it was EDS. The problems started immediately in September 2004 as
soon as the system went live, taking down the entire state welfare system. The
feds sued Colorado over erroneous payments. CBMS has been the subject of
hundreds of TV news pieces and newspaper commentaries.
Here is something for the PTB's in California to ponder, and they should
understand it because a lot of them are judges.
- You can't order people who don't know how to do something to do it and
expect that it's going to get done.
Settlement, there's the link to it.
Now don't get me wrong on this. These folks on the Court Technology Advisory
Committee are intelligent people. The problem is that they are way out of their
element. They just don't understand how to write modern software. They don't
understand the problems involved in trying to find and fix deeply embedded
coding errors. And they don't realize that the people to whom they have turned
for advice likewise do not have the expertise or experience required.
To put it bluntly, some of the people to whom they have turned for advice are
absolutely clueless.
As of this writing, those in charge of this project still resolutely refuse to accept
help. If you want to know why the Assembly Committee on Accountability &
Administrative Review held a hearing on October 28, 2009, and then did
absolutely nothing about the CCMS situation, ask them. I'm curious too but can't
get an answer. The chair is Hector De La Torre, and the vice-chair is Audra
Strickland. Their staff aide is Mark Martin. They have had a proposal for a good
solution in front of them for quite a while and have done nothing.
I will note here that some other states have obtained case management
systems for their courts and spent amounts like $30 Million instead of $1.75
Billion. But actually, they spent too much as well. The size of the state has little
to do with the legitimate cost so even though California has many more courts
and a higher population, $30 Million is about three times as much as this should
legitimately cost for California. Someone else who has knowledge in this area
estimated a maximum of $15 Million. At least we're in the same ballpark.
And if the initial cost isn't ridiculous enough, the AOC has been told it will cost
$79 Million annually just to maintain/service the system. And not knowing any
better, the AOC folks have bought into this! That is absurd beyond all belief. In
the first place, software to manage the court data should be flawless. That is
easily attainable. I have done that over and over and over with other complex
software I have written so I know beyond a shadow of a doubt it can be done.
What should annual maintenance cost? $1 Million absolute tops. The key to low
maintenance cost is flawless software. The second major factor is software that
is user-friendly. Between these two factors, you can cut the number of calls for
tech support by well over 99%. Furthermore, with flawless software, and easy to
use software, almost all calls for help can be handled locally by local IT
personnel. In my experience, most of the calls for help have nothing to do with
the software itself but relate to things like personal settings on users'
computers, lack of knowledge of the location of files on users' computers, minor
screwups with data entry, and the like. I have never had a user have a problem
due to a software flaw of any type with software supplied by RCP Software.
NEVER.
Oh, and about that annual maintenance cost? It should include taking
suggestions from end-users about features that could be added to the software
to help them with their daily tasks and actually adding those features into the
software from time to time.
So when will this end? When the California taxpayers, legislators, and others
finally run out of money and/or patience. Then, when someone who has the
power to do something comes to their senses, they'll have someone like me
come in, create a whole electronic records and management system in a couple
of months or so, and be done with it. And the whole solution will only cost
around $10 Million in vendor expense, and probably less than that, which will
include installing a modern system in every county in the state and educating all
the users. Then outraged citizens/taxpayers, suddenly indignant politicians, and
certain others will loudly and publicly ask how in the world could so many
supposedly smart people spend so much public money before figuring out they
were barking up the wrong tree.
As of now, late May 2010, the immediate outlook is not hopeful. People ardently
backing CCMS are still chasing rainbows and clinging to the vain hope that
CCMS will get fixed. So now it's probably time to go into hibernation. During
that waiting time, I'm guessing the pouring of money into the bottomless pit will
probably continue at around $8 Million/month. Politicians may be talked into
giving the judicial branch enough extra money to open courtrooms back up so
that the pouring of public money into CCMS can continue. But recent employee
layoffs at our superior courts will likely not be reversed. How long will we have
to wait for the realization to fully kick in that the CCMS software is not going to
get fixed? I'm guessing that a year from now we'll be about where we are now,
barring some developments in the political arena. Except of course for the
taxpayers being about $100 Million more out of pocket. I think at this point in
time it will take a skillful outside politician to bring the warring parties together
for a joint effort to help the California courts.
The Standoff And What Could Be Done
As a closing thought for the moment, I am aware as of mid-May 2010 of the
public standoff between two members of the judiciary over CCMS, one aligned
with the AOC and one being a presiding judge of a fairly sizeable superior
court. Neither party has sufficient knowledge of the technology to make an
informed decision about what technology to use or how to create modern
software. In spite of that, each side is making technology decisions and drawing
lines in the sand.
That is most unfortunate and I would hope that others who know these two fine
gentlemen can talk to them and convince them to set aside for the moment the
combative stances they have taken, for the good of the court system.
Interestingly, each side has a legitimate set of business demands or needs for a
court data management system and the heck of it is that, unless there is some
peculiar need of which I'm not aware, each side could actually get everything
they want in the way of business needs if they and others were to join together,
cooperate with me in the development of modern software, and set the CCMS
software/system aside as just an unfortunate lesson from which we can all learn.
How could it go if the warring factions were to get together and talk to me in
plain English (no technology people and no technology specifications) about
their business needs? They could then select a neutral ground to act as a
testbed. Preferably that would be some smaller Northern California county
within fairly quick driving range of my office that only has a few judges so there
would be fewer people to coordinate with during development and testing. With
good cooperation (much of which could be done during lunch time or before or
after regular court hours), prototypes of the various basic components could be
created and then tried out hands-on by actual court personnel on my laptop
computer. They could report on ease of use, reliability, speed, etc. Once people
see what can be done, I think a lot of the hard-line stances would melt away. I
realize that the human element here might induce some delay time while reality
sinks in at each stage.
Once the basics are done, one or more areas of law (e.g., probate, juvenile, or
criminal) could be fully fleshed out with electronic forms, fully enhanced data
management software, etc. and each area put into actual operation in the test
court. The court involved simply needs to have certain basic hardware and
standard software in place. The completion of the forms and expansion of the
software from prototype to actual working form wouldn't be free because it does
take some time to develop all the details but the cost would be a tiny fraction of
the cost currently projected for CCMS deployment into such a county. And, in
the interest of saving the court system from financial disaster, I would do this
first county at a heavily discounted price, essentially donating some of my time
for the good of the court system.
As reality gradually sinks in, I think the process could be speeded up to match
or nearly match what is actually possible. I can write this stuff many, many,
many times as fast as what people are used to seeing but I have noted people
are often slow to react to the production of prototypes because they assume that
what they have previously seen is a realistic model for how long it takes to
create specialty software. It isn't but they have a hard time accepting my
representations that the process could be done in the range of perhaps 10,000
to 100,000 times as fast as Deloitte Consulting has worked on CCMS. I recall
the somewhat startled reactions of certain people when prototypes of hugely
complex applications were presented to them for Alpha-1 testing within a day or
two of when they spoke to me about what they needed.
Once the first county is done, a second county could be selected and, by that
time, I think reality would have sunk to a great extent in so that county's entry
into the modern world of electronics could go much, much faster, perhaps in a
matter of just days. From there, the entire state could be done quickly, county
after county, producing a network of uniform case management.
It's Time To Stop Chasing Rainbows
When we were children, we read the enchanting fairy tales about the
leprechauns sitting at the end of the rainbow with the pot of gold. Of course, as
we grew up, we realized that no matter how far or fast you chase the rainbow, it
will always move away from you, although it may also just disappear. As the
end of the rainbow seems to be within reach, it moves again.
Those who are a bit cynical already about what has gone on have observed that
the completion time for CCMS has continued to advance further and further into
the future. A few years ago the end of the rainbow was 2009. That became the
Spring of 2010. But then that became September 2010. Now that has become
April 2011. And perhaps even more importantly, that has now become a date
for some type of "completion" that has become more and more vague as time
marches on because at the same time we are now hearing the date for "full
deployment" has been delayed over and over. I heard 2011, 2012, 2013, 2014,
and lately 2016. To put the matter simply, the 2011 date for completion of
something or other has no meaning at all and never did. And statewide
deployment of software after it has been written should take little time at all.
Weeks if you're fast; months if you're slow.
This has happened elsewhere so it's not as if we can't learn by example. In
Texas, they chased Deloitte's pot of gold at the end of a rainbow for years, with
Deloitte finally turning the problem over to others after hundreds of millions
down the drain. The bug-laden TIERS (acronym pronounced like "tears")
software was then adopted by others (Accenture) who in turn went down in
flames, turning the problem over to others (Maximus) who then, in turn, ducked
out by hiring a subcontractor (AutoGov - who incidentally only had a bare Web
site still under initial construction when I was writing this in May 2010 - but so
what, because they're politically connected) and referring all questions to them.
If you search the Internet you can find dozens of newspaper articles on the
subject, including one appropriately entitled, "A Trail Of TIERS".
And another crystal clear example of chasing rainbows is sitting right there for
all to see in a Denver court. Gee, maybe California judges could pick up the
phone, call the Denver court, and ask about the case of Davis v. Henneberry
which has been pending since 2004 with a stipulated settlement order in place
since December 2007. No one there in government seems to quite get it yet or,
if they do get it, they're too involved with the waste of money to want to do
anything about it. But maybe the plaintiffs are going to be back in court when
that 33 months provision hits the finish line tape in September. And Deloitte I'm
sure will be right there saying, "But we're soooo close!"
Time For Music & Politics
The claims of the backers of CCMS, TIERS, and CBMS should be set to music.
I might suggest a variation of the old "Over the Rainbow" sung so wonderfully
years ago by Judy Garland.
- Somewhere, over the rainbow,
CCMS will fly,
Somewhere, over the rainbow,
Coders never lie.
Now it's probably up to politicians with skills at getting people to retrench from
their hard line stands and work together with me to solve this situation.
See you in 2011!
And be sure to come back periodically to check the Demonstrations Page for
the latest demo and technical explanation.
Updates on CCMS, etc.
FWIW, I noted in late May or early June that the County of Marin finally woke up
and sued Delolitte Consulting over some software for managing certain
personnel data. They claim it makes piles of mistakes and has caused a lot of
grief. They finally bit the bullet and admitted they had made a huge mistake by
using Deloitte. I think it's going to take quite a while longer for the AOC to admit
they made a mistake. And it's possible they will never admit it.
Interestingly, on June 6th I watched a movie involving folks pulling off a huge
con job. And one of the observations made was that the larger the amount of
money being asked for, the fewer questions people ask. And there was also a
60 Minutes show on CBS about a lawyer who was arrested almost the same
time as Bernie Madoff. He had pulled off some huge multi-million dollar cons.
He got caught when someone in Canada was finally savvy enough to ask
questions.
An article appeared in a legal newspaper on June 4th revealing that confidential
data from Sacramento County Superior Court cases (they use CCMS for
probate and civil) was made public by the CCMS system. That would be a
problem emanating from the server located in Tempe, Arizona. It is not known
for sure how long the problem existed. In a well-designed modern system, the
court itself would have control over their own data. Such data would not be
accessible by the public because there would be no electronic pathway
between the public inquirer and the data. This type of problem is easily
prevented in a modern system.
June 9, 2010, KGO TV, ABC, Channel 7 San Francisco. Billion dollar
boondoggle report. One part of the video shows a computer screen pertaining
to some criminal case in Dept. 12, Sacramento, which is Judge Gilliard's court.
Although it is difficult to tell precisely without in-person inspection, it appears
that the technology being employed is in the range of 20 to 30 years old. The
screen presentation looks like what was common in the 1980's. The mention of
cutting and pasting rulings is caused by internal incompatibility of components
of the system combined with external incompatibility with software created this
century.
You see the personnel handling paper documents in the video and there are
comments about excessive expenditure of time filing documents. In a modern
system, paper would only be employed as a last resort and the vast, vast
majority of documents would be electronic. Filing of an electronic document
could take place in seconds. Many filings could take place across Internet and
Intranet connections. That would be routine for large volume filers such as the
district attorney's office, the public defender's office, county agencies, and large
law offices. Service of copies on other parties could take place within seconds
after the document is finalized. That is the power of modern electronics but,
sadly, the California court system is being bogged down now by fighting over
ancient technology, much of which was obsolete well over 10 years ago.
The mention of the Sacramento court's lack of control over their data is
unfortunate but in a modern system each county's court system would have total
control of their own data and would be able to do various types of statistical
analyses. That is easy to set up. I have already created such a system. Also,
each court would easily be able to designate which documents should be made
available to the public and which should not, thus avoiding disclosures of
confidential data such as took place in Sacramento in late May to early June
2010.
In mid-June, there were several follow-up reports by the news media, primarily
Courthouse News Service (www.courthousenews.com). These reports
mentioned numerous problems that the Sacramento County Superior Court was
having with CCMS, including disclosures of confidential data on public kiosks
located in the courthouse. The public could pull up litigants' medical records,
CPS records, bank statements, etc. just for the asking. Everyone blamed each
other. None of the people arguing had any substantial knowledge of the subject
matter they were arguing about.
Judge White also noted the problem of trying to pull up calendars for
courtrooms such as the law & motion courts, knowing there are dozens of cases
scheduled, and only getting a calendar with three items on it. That problem is
easily solved and I've already done it.
While the confidential data problem of course sprang from the totally inept work
done by Deloitte Consulting, the presiding judge in Sacramento missed the
mark in his analysis of the problem, blaming everything on the fact of the server
being in Tempe, Arizona. Any knowledgeable IT person who reflected on the
situation for only a moment would realize the server location affects primarily
speed and may also trigger some transmission errors causing incomplete data
transmission or errors in transmission, thus affecting speed and things like PCL
errors. But the failure to protect confidential data springs from poor design of
the system (failure to use isolation technology and techniques), failure to use
modern encryption tools, and just generally poor coding by the vendor - a
vendor which is famous for poor results.
The answer to the confidential data disclosure problem is sitting on the
computer I'm using to write this Web page. I have a complete demonstration
prepared, just waiting for someone in the court system to take me up on my
invitations to take a look at it. My suspicion at this time (late June 2010) is that
the CCMS project will eventually just wither away as the 51 currently non-
participating counties will simply decline to subject themselves to the
excruciating pain of being a CCMS county. Unless a new breed of politicians
calls a halt to this quickly after the next elections, CCMS will likely burn through
a few hundred million more taxpayer dollars before people come to their senses.
Also, as of late June 2010, I note that the State Auditor has their audit in
progress under investigation number 2010-102. (See CCMS audit.) It is
scheduled for a final report in October. Although it is highly doubtful they have
anyone on staff with expertise in writing modern software who can really give
them a good analysis on what has happened in the technological arena, I am
guessing they will probably figure out all of the following in the financial arena,
where they are the masters:
- A lot of money has been spent. The total amount is not determinable for a
variety of reasons. - There has been little bang for the buck. To put it in more formal terms, there
has been little or no discernible net positive return on investment. There
have also been substantial negative effects on court operations in some
courts, particularly Sacramento. - The project has had continually receding completion dates and there is no
realistic one in sight. - The technology is obsolete now and will obviously be many years obsolete
by any completion date, if and when CCMS is ever done. - There is no realistic idea of how much the proponents of CCMS plan to
spend on CCMS. There is no spending cap. There is no contractual limit. - None of the courts currently involved in CCMS have implemented it past
what they had done as of quite some time ago and none will. There are 8
areas of law identified and no court has or ever will use CCMS for all 8. - One of the primary stated goals of CCMS - exchange of data between courts
and between the courts and other agencies - has not been achieved and it is
unlikely it ever will be achieved using CCMS technology. The data exchange
plan was based on the use of a remote server in Tempe, Arizona (earlier
Newark, CA). Only one court (Sacramento) currently uses that server and is
desperately trying to get away from it due to extreme performance and
reliability problems. Thus one of the primary justifications given for the
extraordinary price of CCMS does not exist, undermining the financial
prognostications for CCMS management of data. - The original plan of achieving uniformity statewide in court management of
data will not be achieved because of #6 above and because it is unlikely that
any of the 51 remaining counties will ever elect to come aboard. Each of
these 51 counties can and will elect to use other systems. Thus any
originally claimed savings in court operations due to uniformity and
exchangeability of data will never come to pass.
Spending any more money on CCMS is like spending a million dollars to get an
old typewriter to work so it can be used as a productivity tool in an office.
New Chief Justice
As of late July 2010, I note that the current Chief Justice has announced he will
retire on January 2, 2011, and a successor has already been nominated. As to
whether anything will change, that depends on whether the new chief justice is
willing to examine the faulty assumptions that have been made in the past and
is willing to try an alternative approach. I have no real idea how that will play out.
One very basic assumption that has been made, and which is totally erroneous,
is that any "IT person" is in a position to give advice about how to create a case
management system. Lawyers and the general public know better than to
believe that any doctor selected at random can do heart surgery. And they
understand about specialization in law. But the court administrative leaders in
California have blithely assumed that any "IT person" is qualified to give advice
on how to construct a case management system. They have not analyzed the
situation correctly. A correct analysis would almost certainly reveal that not a
single person at the Administrative Office of the Courts or at any of the county
court systems has the required expertise. Reliance on IT personnel who have
expertise in other areas besides software writing has gotten us to the present
situation of approximately half a billion dollars having been spent with nothing
productive to show for it.
Correction of this situation depends on the willingness of leaders to admit that
mistakes have been made and that it is time to start over and use modern
technology solutions created quickly and inexpensively by people who have the
necessary skills. Are they willing to do that? Only time will tell.
One safe bet to make, however, is that when September 2010 comes, the AOC
and Deloitte Consulting will once again announce that they are "so close"
(they've been there for years now) but they need more time to finish
writing/fixing the CCMS software. How much time will they say they need?
Beats me. Might make an interesting office pool subject. Six months? Eight
months? A year? And what will the excuses be this time? Everything but the
truth is of course a possibility. Again, here's a good excuse for another office
pool.
| Government Software, Boondoggle Government IT Projects, And What Could Be Done |