The software is a "steaming pile of shit"? Of course, it was built by IBM Federal. Building software for U.S. gov't contractors is an awful, awful experience. The end result is the talented leave as soon as they can, because working in a red tape environment with ridiculous Information Assurance (that's gov't speak for their own ludicrous security theater that is their IT security process) policies makes you want to kill yourself.
My wife has a friend who barely made it out of high school, believes in psychics, believes that the U.S. gives an annual payment to England for our freedom (this was revealed on the 4th. I spit out my beer laughing, but then I realized she was serious). This person works as an Information Assurance person in charge of making sure software is "secure." She knows nothing about software, but she has passed a few tests, courtesy of prep courses which guarantee you can get the cert. Because she was in the Air Force for 6 months (she got pregnant and was honorably discharged a couple of months after her first duty), she is a veteran and is therefore fully qualified to classify software as secure. Or in reality, defense contractors have to meet quotas for hiring veterans, and they put her in the easiest butts in seats job they could find, IT security.
This, my friends, is the system you have to deal with in the federal fucking government.
I'm sure that IBM Federal is as bad as you suggest. But part of that is due to how government works, which is in turn due to how citizens react to things.
Think about it like a government employee. If you do your job perfectly well, nobody notices. Despite the bureaucracy, despite unclear success criteria, despite insane budgeting. Nobody notices. That's just what's expected.
When something goes wrong, though, you get hammered. God help you if something comes to the attention of the public or makes the news. Nobody will take the time to understand the context; everybody just looks for the most plausible person to blame. If that's you, then you've got a black mark for the rest of your career. Welcome to the basement!
It's the total opposite of a startup context. And in some ways it should be. But it does mean that government projects drown in red tape and politics and procedures up the wazoo. Which is absolutely a recipe for shitty software and overpriced contracts, whether you're in government or a megacorp.
I'm in startups for a reason, and I have a lot of sympathy and respect for the good people who keep plugging away in government despite the fucked-up incentives.
That paralleled my brief career in government. Managers are not promoted for efficient decision making—they are promoted for not having been associated with an unsuccessful initiative.
Given all the failed government projects, I feel like that people say, "As long as you can't prove it wasn't my fault" (as opposed to "my project") they won't get fired. After all nobody got fired from Big Government for following the rules... (Sarcasm intended)
Fired isn't the problem. If you're smart and ambitious and believe in public service, you don't want to spend the rest of your life in a basement cubicle re-verifying the presence of the new cover sheet on the TPS reports.
I think there's less reason to fear that. Some of the best parts of government are the science- and data-driven ones, and evidence-based medicine provides a lot of data. Plus, doctors are notoriously independent.
Personally, I'd feel much better about government-run healthcare than for-profit-insurer-run healthcare. At least with the government option, everybody works for me, either directly or indirectly. Plus, a bureaucrat's natural fear works in my favor.
I really like single payer in theory, but I'm with briandear on this one. There is already enough monkeying with what Medicare should and shouldn't pay for. My fear is the more money is involved, the less science/data/evidence based the decisions will be.
That's why I'm torn on the whole thing. I think for-profit should be able to do it better, but the current system clearly isn't structured in a way that encourages it. I have no idea whether the devil I know is better or not, and it's a big decision.
If (most of) the rest of the developed world can do "socialized" medicine reasonably well, then what's so special about the US that would prevent this? Just the size and scale for fuckups?
There are occasional health system scandals in Australia and the UK, for example, but even with this everyone gets coverage. So why the US needs to drag its heels in this is beyond me.
But if you do think the public service will make a dog's breakfast over the whole thing, I'm all ears.
My concern is the size of the entrenched interests would prevent things from being setup reasonably. If the government could truly start over, I would have more confidence. Instead I would think we'd get all sorts of pressure that, for example, treatments proven not to beat placebos (or do worse) should still be paid for because people want them, and unless it's so dangerous the FDA bans it.
Okay, to be clear though, exactly which segment(s) of the developed world are doing socialized medicine and not going bankrupt (or on the verge of going bankrupt) in spectacular fashion?
If the US adopted any of the systems of other nations, we'd be paying a lot less, so we'd at least be going bankrupt a lot slower. We pay way more than anybody else for outcomes that are basically the same.
wpietri: Yes, I read that series quite some time ago. I'm very familiar with the ridiculousness that is the US healthcare system. This, however, is totally irrelevant to my comment, which was regarding other developed countries that are going bankrupt over their entitlement programs.
The US is in the middle of the pack. Countries with more sane health care systems and better levels of debt include Ireland, the Netherlands, Spain, Germany, Finland, Denmark, Canada, Sweden, and Australia. I don't see any reason to believe that they're being driven bankrupt by their health care systems.
If your a government employee and something goes wrong you just claim that you didn't have enough money or resources to do the job. At least that's what everyone seems to do and they get away with it.
Agencies can claim that. Because we never fire or sideline agencies.
But individuals don't have the same out. Somebody who took a risk and failed may never get fired, but for somebody with ambition and vision, getting demoted to a pointless job with no power is worse than getting fired.
> defense contractors have to meet quotas for hiring veterans
Wow, that's gotta be one of the most boneheaded govt policies I've heard of. Is this because of some kind of patriotic "duty" to keep veterans off the streets? In my experience, a lot of veterans would have ended up on the streets even if they hadn't joined the military - in fact, many joined the military specifically so they wouldn't end up on the streets.
I don't believe there's an explicit quota (I could be wrong), but the Vietnam Era Veterans' Readjustment Assistance Act of 1974 added certain affirmative-action requirements relating to government contractors hiring Vietnam vets, which was later expanded to all vets.
From a brief skim of since-1974 stuff on the subject, it looks like much of it may result from bipartisan-coalition politics: you often see "disabled and veterans" joined together in laws and rulemaking as a group, probably as horsetrading where liberals get something pro-disabled, and conservatives get something pro-veteran.
They have HUGE incentives for hiring veterans, as well as any kind of "disadvantaged minority" (read: not white or Asian)
My best friend for 20 years, who is probably the smartest and most capable guy I've ever known, graduated with a software engineering degree (and a crappy GPA due to his pursuit of non-school software stuff at the expense of his school projects). He happens to be black, and he couldn't get hired anywhere due to his GPA. He goes to DC, immediately gets interviews. Now, here's the funny part: These assholes who called him for phone interviews were asking, multiple times!, if he was indeed "African American"? One of the females who called him from a very,very large contractor I shall not name said the words, "ok, I was just confirming because you don't sound like it." (meaning she didn't think he sounded black.)My buddy, ever the cleverest guy around, responded "I'sa do the numbahs too mastah." LOL
Working in a big corporate environment which I feel is similar to the way the government must work, you must understand that the "cost to develop the application" does not mean what you think it means. It's not the "cost of the development phase".
Coming up with the idea and getting it approved takes months. Wrapped into the project of creating the web site will be the "what should we create" phase and dozens of iterations on design and IA, which takes most of the work. There will be 20+ people involved in this process and one line of code hasn't even been written for the first 6 months of the project. The idea they had last week is different than the idea they are talking about this week, which will be different than the idea they will talk about next week. Imagine this goes on for 6 months.
Once the code is written, testing, security, infrastructure setup and all the sign-offs on that also takes months.
One person full time for a year will cost IBM $250,000 in real costs (salary, benefits, overhead) that they bill out to the Federal Government at $500,000 per person per year. Can I imagine 25 people working on this for one year, where only 3 of them are developers? And the rest of the money is hard costs and infrastructure? Yes I can. Easily.
Generally (in Australia anyway) government software also has to meet a series of strict accessibility guidelines before it is considered complete.
Things like high-contrast modes, adjustable font sizes, naming conventions (for screen readers, etc) all have to be taken into account on EVERY dialog.
I haven't researched the full contract award or pulled a copy of the original RFP, however, work on the Census FactFinder website was part of an $89.5mm contract awarded to IBM Federal in 2007 (for, potentially, 9 years) [1]. I wonder if the $33mm quoted in the FOIA request response is the total cost incurred to-date on that award (including services performed for the entire Statement of Work) rather than the actual cost of ONLY building/updating FactFinder.
I wouldn't be surprised if the $33mm price is correct ... but not 100% sure it is without spending more time looking around and, potentially, seeking further information through FOIA.
Awesome! I wrote the original FOIA that he references in the beginning of this article, I'm extremely pleased that somebody else took the same initiative!
Is there anything else about the government any of you guys want to know about? I'm getting pretty good at FOIA'ing now.
As much as I'd love to do that, I think casting a broad net in FOIA is very difficult and potentially counterproductive. FOIA offices have very limited resources already, so asking for literally everything at once or as lots of little requests performs a DoS attack on the one office of the government that I really like! It does make a good point philosophically, but practically, I think it might be counterproductive. The FOIA law itself also requires requests to be fairly specific.
I bet there are some _specific_ things that members of this community want to know (cost/sources of software, policy memos, etc.) that could be requested, received and processed in a more timely and useful fashion than if we just asked for "(asterisk).(asterisk)".
Or a better solution, everything that's eligible for a FOIA request should just be made available on the existing Open Government page [1] rather than making people request it and go through the process. What is the point of this security by obscurity nonsense? If it can be released under FOIA, there's no reason to not release it without a FOIA request.
there's no reason to not release it without a FOIA request.
How about the cost of digging up that information? Or the effort to ensure each piece of information released is indeed safe for release? Or the effort to put it all together in a coherent set?
The government has a TON of records. Releasing them is not as simple as picking the "public" checkbox in the settings pane.
There might also be a general concern that if someone had access to the entire database of records, they could mine that database and start to infer other information that is not supposed to be known. This sort of tactic has been used in wars past. Compiling supply chain records to infer troop movements, for example.
I accidently upvoted you. So please mentally -1 for this post.
I'd characterize your reply as concern trolling.
I've done my share of FOIA requests. I imagine I've heard most every excuse. Including "We lost the backup tapes."
Being charitable, the reason public records requests are hard, expensive to fulfill is because efficient records management is rare, making finding and retrieval difficult.
Someone may chime in suggesting CMS, workflow, sharepoint, whatever. Yea. If it was just that easy, everyone would be doing it.
Wow. I remember when you first posted that original FOIA, how me (and others) were saying, we should FOIA EVERY APP...maybe someone did. But I hadn't seen a reference to the crappy OSHA app until this post.
For folks interested in helping solve the root cause of these sorts of things ... the White House is looking for people to help build a streamlined procurement process:
In terms of streamlining a procurement process, you should definitely take a look at Cyber Fast Track which is a huge streamlining of the DARPA process for awarding grants (maybe procurement) done by Peiter Zakto (Mudge).
Some folks might consider it a boring timesink. But it also happens to be the important meta-problem that stands between us and increased government transparency and accessibility. Take almost any government data set, request / report interface, tax service, or help center you can imagine. It's very hard to believe that we'll ever have a quality online version without first fixing the procurement process.
I'm sure somebody thinks this is an interesting challenge to tackle. The Uruguayan equivalent (AGESIC) did make a pretty nifty website for government procurement:
It's a good initiative, but it has the side effect of making my blood pressure rise when I see the awarded projects (all relatively transparent, but still bad contracts)
I think it would be something awesome to work on. I'm in software to make a difference, and dropping a digit or two off the cost of government projects could be quite an impact.
If that's the democratically decided priority and it can be achieved without cutting other budgets, raising taxes or running a deficit, then that's a good thing.
Conversely, fighting a bloated military by wasting money on IT boondoggles is not a strategy with a whole lot of long term potential IMO.
Tell what to the people being bombed? That you're hard at work changing their fortune by refusing to reform the IT procurement process of the US government? They will be eternally grateful.
As a USian, sometimes it seems like the high cost of war is the only reason we haven't invaded yet more countries. The vast cost is certainly one of the few arguments that has resonance across the political spectrum.
Given the willingness of the US government to run huge deficits, I don't see how fighting waste in completely different areas of government would enable the country to go war.
Never mind that we're taking about saving millions, while the last two wars each cost trillions. Or 100,000 Census Bureau websites as discussed in the article.
There is a big difference between the act of allocating more money to a budget, and having the capability to do so because you're being more efficient. Which was my point.
I don't know how much it cost to redesign the Federal Register, but federalregister.gov is one of the most usable, data-intensive sites around. Especially compared to what it used to be.
GAO.gov is also a hallmark in usability and information-taxonomy.
One of the dudes who developed the Federal Register site (i think as a code for america fellow) is now at AirBnB too. This is a world we can and should get involved in.
several friends took jobs with the GAO out of school (data analysis type gigs) -- they have some of the coolest stuff to talk about; seems like a great branch of government to work for
For simplicity let's say that IBM charges $300/hr for a consultant (developer/PM/etc). That would be about 100,000 hours of work. Let's say a consultant bills 2,000 hours a year and it takes them 18-24 months to build this thing. That would be a team of 25-30 (developers, PMs, account managers, etc).
The time frame is realistic considering poor original design, scope creep and federal government counterparts not being particularly quick in gettings things done (getting access/fighting through bureaucracy, etc).
So why would the federal government hire IBM over you; who can build a totally bug free version of this in half the time for half the price? IBM is:
- Not likely go out of business anytime soon.
- Not be unavailable for support work because you decided to start/join a startup
- Not be able to deal with turnover since they have 100k+ employees, steady stream of college recruits and have been around for 100 years.
And you forgot the most important part. If the project goes to shit, people can say, "But we hired IBM!" If the government employee in charge hires a smaller company that does just as well as IBM would have, then everybody yells at the government employee.
This is disgusting. I've had to use it extensively for research, and while it 'gets the job done', its such a pain to use. I've asked myself who put this together, and I always assumed it was a team of interns working across a decade for each version.
I can't believe this is the best they could do for $33 million. I just can't. They better have some slick stuff under the hood.
Good UI comes from collaboration and iteration with users. Apple did something like 100 iterations on the first iPod. Typical consumer electronics companies do 3 to 5.
The government RFP and budgeting process is practically designed to prevent iteration and exploration. I can just hear Senator Blowhard of Cornhusk now: "So you want an unknown amount of money to keep building things until users are happy? That's an obvious boondoggle. Come back with a complete specification and put it out for bids from major contractors. We must be responsible with every penny of taxpayer money. Harrumph!"
So of course the software has terrible usability. Anything built like that will be terrible.
For the Apple figure, I'm pretty sure it was Inside Steve's Brain. This was some years back and I'm away from my books, though. I'm sure it was 100 iterations on some Apple device, but I won't swear it was the iPod. The 3-5 number I got from a consumer electronics designer I took a class with.
Yeah, and for $33M they couldn't build in some iterations and bring in focus groups or hire someone competent to get it at least partially right the first time?
These projects are handed out to reward the connected contractors, contributors and lobbyists that know how to work their target politicians and bureaucrats. There are ZERO incentives to do quality work. No one will lose his job over this software travesty. No one will get demoted. Rather than bemoan the waste of $32.9M on the job, politicians will simply push for higher taxes.
Putting iterations into a procurement process is very hard. How does one bid against that? How can one control costs? How can we know the cost of the project up front if we don't know exactly what will be done? And if there's a disagreement about whether the project is really done, how do you litigate without a clear spec?
It's not totally impossible to do something agile in a government context, but it requires much more skill and trust from all involved. It's risky. And nobody outside the government is very understanding when things blow up, which happens more when you pick risky choices.
NYTimes paywall cost $40m. Heard bank software refactor can easily exceed 100m and take many years. I've personally seen 25m project fail and the code thrown away. Its not uncommon especially when you have 3 to 4 "support" people per developer especially when dealing with a bunch of stake holders aka executives with fragile egos.
This isn't surprising. I've done work in the federal government. Recently we built some products that did more, looked better and cost far less ($200k) than the other big name consultants. Our risk for further funding is that it didn't cost enough although most people love what we've done. The federal government IT market works on a contracting model, which means the people working under the system do things at a minimum OR do things to ensure that they'll receive further funding down the line. It's incredibly time consuming and difficult to nail one of these contracts (usually requires a combination of relationship building, slow moving people and lots and lots of politics) so once you get into that zone, the incentive for the contractor is to layer as much stuff into the contract as possible.
Yet when anyone talks about cutting a dime out of the federal budget, the cries of "You're starving poor people!" and such dominate the political discussions.
There isn't even a little bit of common sense or honesty in the way the government spends our money and people wonder why some of us are so adamantly opposed to raising taxes.
This doesn't really seem like a public/private difference, but an enterprise/non-enterprise difference: Fortune500 firms are no more efficient than government contracting is, really. While it's a lot in other contexts, $33m for a custom enterprise software project, even one that seems like it ought to cost much less, is pretty middle-of-the-road as a price.
Having worked at a company that's top 20 in the Fortune500 and having a good friend who does government contract work... you have absolutely no clue about what you're talking about in terms of the relative amount of waste exercised by government vs corporate entities.
The fundamental reason why there's such a difference in efficiency between private efforts and government ones is this:
Private companies that are horrible at managing efficiency go out of business. People in private companies who are horrible at managing efficiency lose their jobs.
Government agencies that are horrible at managing efficiency get bigger budgets. Government employees who are horrible at managing efficiency rarely lose their jobs. The GSA scandal is the only one in recent history that has received any kind of real attention; and that's only because the idiots at the GSA made videos the went viral.
I worked for a successful Fortune 50 company whose products you use every day. I worked out of an office in NYC, and flew to the west coast every two weeks to attend a staff meeting that lasted about 2.5 hours. I literally earned enough frequent travel points that I didnt pay for a vacation from 2003-2009.
If you work for a company like Microsoft or IBM or Bank of America, you work in a bureaucracy at least as dysfunctional and Byzantine as an average US state. The Federal government is a whole other beast unto itself, but there are probably companies as screwy as they are too.
The fact that the private entity purges some folks doesn't make them better.
Well, you apparently haven't seen much of how enterprise contracting works. Companies can easily blow $100m+ on bespoke software development that produces crappy results and still be in business. Companies like Exxon, BP, etc. could, and do, waste billions and not go out of business. Sometimes they spend tens of millions and get nothing delivered at all, or scrap the result. And the incentives within the megacorp are fairly similar to what you describe: managers who are the worst at managing budgets get their group's budget increased, because they can argue to the relevant VP that they need more money for whatever boondoggle they're currently working on. Not spending your money is more of a problem, because it means you're less important.
I assume it's the same outside of oil as well, but I know the most about that sector. Someone elsewhere in this discussion mentions that the NYTimes paywall cost $40m to implement!
Really, you think this would have cost $33M if a Fortune 500 firm was paying? Enterprises overpay a little because managers are spending shareholders money. Governments overpay a lot because bureaucrats are spending tax payers money. There is a lot less accountability in the latter.
Yes, I really do think so. It's a nice theory re:accountability, but I don't think it holds up in the real world, where enterprise-contracting costs, for whatever reason, are extremely high. One possibility is that there isn't as much accountability for managers of large companies as you think (there's been a lot written on that problem); another possibility is that there are somehow inherent scale problems with doing bespoke software development for very large organizations.
Well, that's where a lot of the cuts are going. Nobody is saying, "Let's cut spending by shifting work away from major government contractors to smaller, scrappier companies." Nobody in Washington, anyhow.
And really, there's plenty of common sense and honesty in how governments spend money. Long ago I did a 6-month contract for a state government agency. I came away impressed. They were no more bureaucratic or wasteful than the large companies I've worked for. Like megacorps, many of the employees were clock-punchers who had all initiative beaten out of them. But many were sincere public servants who were getting good stuff done despite the bad incentives.
If you're interested in making actually good software for the government, an agency that looks like it's genuinely dedicated to that end is the Consumer Financial Protection Bureau.
I'm curious to see what caliber/capabilities HN thinks software that costs $33 million should offer. I have experience with cycle accurate emulators that cost $100k/floating license but never anything at the $33 million level!
Well, the company I work for has quotes from both CSC and Guidewire for their insurance packages (kind of an ERP for insurance companies), which would handle all of our processes, and they ran in the "tens of millions" category between licensing, implantation and other costs.
Guidewire at least has hundreds of people working on their insurance suite, some of then I've written to sound very capable and I really like their development process. They even created their own language based on the JVM (Gosu).
You don't think Apple or Microsoft or Adobe or Facebook have 300 developers working on their projects? Just saying, 300 man years isn't that much, considering every full time employee working on a project for a year is a man year by definition.
Yeah, but...any of those aforementioned companies produce products which each generate more usage in a day than the FactFinder likely generates in a month or even a year.
For $33 million, I would expect not to see "American FactFinder requires an Internet Browser with Javascript enabled." I would expect them to be able afford a developer who knows how to create links with an "a" tag rather than using Javascript to duplicate a core browser feature. Given that this is a federal website, $33 million should be more than enough money to comply with disability guidelines
that forbid such sloppy development practices on federal websites.
$33 million is 500 programmer man-years, minus a bit for overhead and profit. That should produce a lot of development.
I wonder if the $33 million cost includes the backend work done on the DADS II contract, or if that figure is just for the web frontend.
Where do you live where $66k/yr the total cost of employing a programmer? I don't know what IBM uses for its internal accounting, but based on what several other large tech companies use, $200k/yr is a decent guess (including salary+benefits+overhead). That'd mean 165 man-years, and only a fraction of those 165 would be programmers, since the contract also needs to pay for managers, sales, billing, legal, etc.
I don't know much about Kansas, but that is definitely not the going rate for enterprise programmers in St. Louis. But in any case, what proportion of software jobs are in the (non-Chicago) Midwest, anyway? The government is basically stuck with whoever submits a bid, which is most likely to be companies in the "IBM and competitors" category, and their cost structures are well into the $200k range (I wouldn't be surprised at $250k-$300k).
I have no idea why my parent comment was downvoted; I posted something based on my own (painful) experience and knowledge of what various other people I know made in programming jobs at the time.
There's actually a lot of programming jobs to be had in the midwest; courtesy of companies like Sprint, Oracle, credit card processing companies, large furniture service companies, and others.
Seems like a similar situation to the student info system at my school, which supposedly cost $8 million of our tuition money despite its terrible UI (scrollbars inside scollbars inside scollbars) and inability to store three-digit course numbers (making the entire university have to switch to four-digit course numbers).
That figure is clearly not for an Android app, that is a clear misrepresentation.
Who knows how it breaks down. This exposes a vast amount of queryable census data to the public (something which would normally be regarded as a good thing, particularly if you are a startup which uses data).
But, to be sure, we can just punish the government for ever releasing any data unless it had such a high fee attached that only huge corporations could afford to pay it.
You should FOIA some kind of itemization of how that huge number breaks down, rather than running to the press with a sensationalized version tailored more to partisan politics than to informing the public
Sure, I'd like to reduce government corruption and waste. However, I think most of the problems relate to scale and bureaucracy, not the public/private distinction. I've seen enough of how large multinational companies work to have my doubts about the private sector's efficiency and honesty, as well.
I suspect a lot of people opposed to "government" and "taxes" in some general sense have ulterior motives, mostly relating to libertarian ideology ("taxation is theft", etc.) or just plain not liking to pay taxes... as opposed to honestly assessing the efficiency of different organizations.
Well, since the government is the one that set up the whole game including how much they would pay for the work, how sub-contractors would be chosen, what methods would be use to ensure accountability and a quality product, etc. -- I think the target of your sarcasm is misplaced.
By the time IBM Federal got the project and could start coding, some bureaucracy of failure had already decided that $33M was a fair price for the specifications that they provided.
I know people who work on government projects like these. I know how the subcontractors are chosen. I can almost guarantee that the subcontractor who received this job had some special "in" with the decision maker on the government side through nepotism, trips to strip clubs, cash payouts, political connections, or something.
Sometimes the truth is partisan. The government really doesn't have any incentive to be efficient, and really does dump giant amounts of money into a bonfire on a regular basis:
Partisan nothing. It's only partisan if you think large private corps are somehow immune to wasting money. You think IBM charges fair, efficient prices for their private consulting? Big and bloated is big and bloated, whether it's Capital-G Government or the behaloed private sector.
I'd be willing to gamble real money on the following 2 facts about this project:
1) First thing IBM did was bill themselves for a couple hundred licenses of their own software.
2) The majority of the money was pissed away in meetings, and the majority of the dev time was spent on horrible multi-thousand line stored procedures for the ETL.
Probably not a very different price, and the end result might not be much better. Bespoke enterprise software is notoriously expensive, and the end results are often not good. Lots of reasons, many the same as with government: client has very particular needs, client has unclear specifications that are changed several times during the process, competition for the bid is weak, client doesn't have their internal shit together, client's existing systems that the new stuff must integrate with are poorly designed and/or bitrotting and/or broken, client's management/billing/procedures are so bureaucratic that you have to include some cost in the bid just to pay for that, etc.
I haven't done anything in that area myself, but from what I hear second-hand it's a gravy train once you've made a sale to a Fortune500 company. Sales is hard, but once you've made the sale, you can bill millions for years for all sorts of reasons, some of which are even legitimate (given what they're asking you to do).
I agree that makes it worse, but if indeed it'd cost similar at a large private company, it's not clear how we'd expect the government to do better. If the private sector can't manage to bring down costs on enterprise software, why would the government be able to get a better deal than they could?
I would be interested in how software for large organizations, public or private, could suck less. A handful of startups have attempted to disrupt enterprise markets, but the going seems tough.
That's what it should cost. But I have seen large companies blow similar amounts of money on even less useful systems. And heard enough stories from friends to know that such fuckups are common in the corporate world.
That may depend on whether that large private corporation is already running the US Census.
Hopefully that day will never come, it is enough that corporations are running the prisons.
I would also like to know how much said private corporation would charge. (But I don't have to wonder very much, because companies like Elsevier and Lexis/Nexis give a pretty good idea)
1) The people drawing up these contracts know nothing about software. They are often non-technical, elderly, upper-management types that have never written a line of code in their life. For all they know, adding a link to a webpage is a 2 week project.
2) Congress punishes federal agencies for not spending all of their budget each year by cutting their budgets. As a result, towards the end of the fiscal year every federal agency goes on a crap buying spree to make sure they've spent every last penny.
It's really hard to overstate how bad the government is at software development. I've worked in several contracts with various agencies, and I know people who still stay in the game, and the best ones I've seen are ten years behind everybody else and spending about 50x what they should to get things done. They're simply swamped in bureaucratic molasses.
For instance, even though the "business" of the government (things like tax collection or tracking shipping containers) uses much less data than say a Google does, not only do they have a CIO, they have an entire battalion of CIOs. There's a whole club of just government CIOs. Each of those, of course, has a full staff. When contracts are paid, it usually goes through some completely different set of people than the people who actually receive the value.
If you're a company trying to help the government, you'd laugh your ass off if it weren't so damned sad. Your programmers -- and you gotta love this -- are all considered little cardboard cutouts, like cookies. If you have a problem to solve, the correct answer isn't what technology you would use to solve it. Usually that's outside your control anyway. The correct answer is how many "standard" programmers it would take to get it done. The quality of a "standard" programmer varies, but it's usually somebody with a few weeks of training and a desire to be anywhere but here. Many times your government partners are incompetent and a danger to themselves and others when it comes to technology. They want the moon but they don't want to take any political risk at all. I had one guy just flat out tell me: you guys make all the tough choices and if it goes well I'll take all the credit. If it fails it's all your fault. At least he was honest.
I could go on, but I don't want to get into a rant. Ever wonder how the IRS spent billions on upgrades with nothing to show for it? Or how the FBI's new case management system was a total Charlie Foxtrot? Geesh. 33.3 million is getting off lucky.
Ex-employee of Accenture here. Accenture basically earns most of their money selling horrendously overpriced "solutions" to customers who don't care about cost.
The thing is, the bureaucracy that you mentioned, it's not just the government. Too many large companies, especially non-IT ones, assume that you can complete a project by throwing 100 mediocre programmers at it.
Your point is well taken. It is not just the government.
The thing is that most private systems operating like this go broke -- unless they have lots of federal protection. You can't spend billions in the private sector without people noticing. The same is not true in the government sector. While all big organizations are like this, the government wastes money on IT at another level entirely. (I know, I've worked with lots of both government agencies and large organizations)
Interesting use of the gif there. I wonder if there will be a migration of these gifs from meme-based humor sites to more "legitimate" blogs and news sites. They are, after all, the visual equivalent to using quotes by famous people.
But, seriously, someone is going to write a thesis on gifs one day. It is "literally" an innovation in communications. While we currently associate gifs with memes, etc, you can see quite a few very creative ones on ffffound.com, for example. And,as you say, TheAtlantic now uses them.
My wife has a friend who barely made it out of high school, believes in psychics, believes that the U.S. gives an annual payment to England for our freedom (this was revealed on the 4th. I spit out my beer laughing, but then I realized she was serious). This person works as an Information Assurance person in charge of making sure software is "secure." She knows nothing about software, but she has passed a few tests, courtesy of prep courses which guarantee you can get the cert. Because she was in the Air Force for 6 months (she got pregnant and was honorably discharged a couple of months after her first duty), she is a veteran and is therefore fully qualified to classify software as secure. Or in reality, defense contractors have to meet quotas for hiring veterans, and they put her in the easiest butts in seats job they could find, IT security.
This, my friends, is the system you have to deal with in the federal fucking government.