A place to discuss Joel on Software. Now closed.
This community works best when people use their real names. Please register for a free account.
Joel on Software
Business of Software
Design of Software (CLOSED)
.NET Questions (CLOSED)
Fog Creek Copilot
The Old Forum
Albert D. Kallal
Nice work if you can get it !
I'd happily design and build that project for 1% of $350m.
I'm guessing it's very complicated, but still a quater of a billion pounds sounds excessive. Because of this I'm also guessing the guy that owns the IT company that implemented this software went to school with the civil servant who had final say during bidding for the project.
you couldn't built it for 1% of the cost. It is not just creating "child" table and "child_info" table. There is alot more to it. Alot of data analysis is involved. Getting the data. Putting the data in the database. Application on top.
all the requirements from different agencies. Agencies changing their mind after you start and often after you finish.
doing work for a government is by its nature more expensive than a similiar application in the private sector.
Yeah, I was being facetious when I said 1%. I get that it is more than just a couple of tables.
However, I do wonder how you get to hundreds of millions for a system like this, or many, many billions for some government IT systems (NHS database, ID cards, etc).
Projects on this scale are nearly always late and over budget. Big business has learned that a big bang approach doesn't work, government hasn't.
I did some work for a government contractor a few years ago and although the code I created was correct and to specification I knew it was never going to go live. At that point this government department was on their fifth attempt to centralise disparate systems.
There were already three vendors supplying systems to different regions, the government would have been better off specifying interfaces for these products to exchange data then let the suppliers compete directly.
Monday, May 18, 2009
I agree with Contractor and Tony.
It's obvious but it bears repeating. 1% of $350 million is only $3.5 million and if you have a staff of any significant size, you will burn through that much money fairly quickly. In addition to the problems Contractor and Tony mentioned, you will also have to deal with employees, training, facilities, infrastructure, travel, presentations and general regulatory issues.
Now that I think about it, I'd hazard that $350 million was too cheap, and that may be why it was delayed twice and they're thinking about scrapping it.
Hopefully it won't work at all.
The idea is to track all children so schools, hospitals, police and social services are all on the same page.
It is of course totally secure even though an estimated 300,000 users will need to have access to the data. There is of course no chance that any bad people are among them, because they are all police, teachers, social workers, childrens group leaders etc. - and when have any of them ever done anything bad (it's not clear if priests have access!)
Although it is totally secure, about 60,000 'special' children will not be on it, any children in custody battles, any victims of abuse , and oddly politicians kids, will not be listed for 'security reasons'.
These government projects are a big joke. I'm just waiting for the public announcement of the new MPs Expenses System which will apparently cost $500m.
I stick with my assertion that I'd "design and build" it for $3.5m, but obviously all the other stuff mentioned above would increase the budget substantially (x100?).
I have actually worked on Government projects, for the military, and I remember thinking it was all a scam.
IMHO The objective of these projects is nothing other than to line the pockets of the directors/partners of the "consulting" companies employed by the dim-but-nice senior civil servants and the behest of our corrupt politicians.
I doubt if anyone senior involved actually cares whether it works or not - they win either way.
"I have actually worked on Government projects, for the military, and I remember thinking it was all a scam. "
Ironically, a lot of the extra cost is "quality", "transparency", "tracability", and other forms of documentation.
I had a friend who, in the 1990's, did some MILSPEC work. He said he worked at a "main" company that actually built the software. They were, like, 4 guys. They were a sub-contractor to a bigger company like Lockheed Martin or McDonnell-Douglas.
The general passed on the cost of the subs plus $20 per hour to the government, and also employed a:
- Project Manager
- Metrics Analyst
- "Quality" person (prevention/process/documentation)
- "QC" person ("verification")
- Documentation-of-the-product writer
- Contract writer/signer/verifier
- Gub'ment liason
Of course, each of these billed the gub'ment hourly. So a project that had, really, one or two developers for six to nine months would bill for 15 to 20 people, full time, for closer to a year.
Again, Ironically, about half the work of the guys at the general contractor was make-work, generally required by MILSPEC guidelines to prevent fraud.
This is not my story. I have no worked for the Department of Defense. It is my not my experience. I'm just passing it on FYI. I have, however, run into enough CMMI 'experts' to be ... skeptical of federal software development.
Wednesday, May 20, 2009
That sounds familiar. When I did my military project, I was the sub-sub-sub-contractor and you can bet every layer was adding their own overhead and more than doubling the price.
In fact, I was a student with no military clearance working on a complex system (hardware and software) that was going on a nuclear submarine for the British Navy.
When the representative from the "boat" came he was pretty shocked to see that it was just me doing the whole project. I had to pretend I was a highly qualified engineer who worked full time on military projects. In fact, I was an unqualified maths & physics student who just wanted some beer money.
As I mentioned earlier, this experience made me think that all government projects, especially military ones, are just gravy trains that rip off the tax payer.
That is what prompted me to post this thread. Hundreds of millions for what is essentially no more than a secure contacts database seem silly and I remain unconvinced that it represents good value for money.
This topic is archived. No further replies will be accepted.Other recent topics
Powered by FogBugz