Is The DOJ Using Obsolete Software To Subvert FOIA Requests? (theguardian.com)

"A new lawsuit alleges that the U.S. Department of Justice intentionally conducts inadequate searches of its records using a decades-old computer system when queried by citizens looking for records that should be available to the public," reports The Guardian. Slashdot reader Bruce66423 writes: An MIT PhD student has filed a suit in Federal court alleging that the use of a 21-year-old, IBM green screen controlled search software to search the Department of Justice databases...constitutes a deliberate failure to provide the data that should be being produced.
Ryan Shapiro's lawsuit alleges "failure by design," saying that the Justice Department records are inadequately indexed -- and that they fail to search the full text of their records when responding to requests "When few or no records are returned, Shapiro said, the FBI effectively responds 'sorry, we tried' without making use of the much more sophisticated search tools at the disposal of internal requestors." The FBI has a $425 million software system to handle FOIA requests, but refuses to use it, saying that would be "needlessly duplicative...and wasteful of Bureau resources."

Comments

Re:intent or consequence?

[beelsebob]

Political flamebait works both ways. The other side of the coin is that Democrats set up overly complicated systems that can't work without an ever-increasing price tag, then complain (loudly) that they just aren't getting the support they need.

The problem is that that assertion doesn't line up with reality. Go down to your DMV some time, and observe the kinds of systems that they're using. They're using databases built in the 80s and 90s on top of DOS, running on ancient computers with CRT monitors (at least around here).

What reasonable business do you know of that hasn't upgraded their systems since that time to allow for more efficiency savings, faster processing, reduced staff costs etc?

There's a lack of investment in this kind of system, plain and simple, being disguised as "government efficiency" by the republicans.

Re:intent or consequence?

[Sarten-X]

The problem is that that assertion doesn't line up with reality. Go down to your DMV some time, and observe the kinds of systems that they're using. They're using databases built in the 80s and 90s on top of DOS, running on ancient computers with CRT monitors (at least around here).

...And is that a problem? Does the thickness of the monitor really impact how legibly they can print your drivers' license?

What reasonable business do you know of that hasn't upgraded their systems since that time to allow for more efficiency savings, faster processing, reduced staff costs etc?

As one example, in the mid-2000s, I worked at a company whose main computer was built in 1988, with only minor upgrades (disk capacity, and a modem that was occasionally plugged in so it could be maintained remotely) since its construction. It had survived the obsolescence of its product line, the rise of DOS and Windows, and had only a minor stumble for Y2K. For a system whose primary purpose was tracking orders moving through departments, and tracking employees' time cards, it did the job perfectly well. That particular company was in the top 10% of the industry by order volume and profits, so it seems to have done just fine by most standards of "reasonable".

There's a lack of investment in this kind of system, plain and simple, being disguised as "government efficiency" by the republicans.

Again, to show the other perspective, there is grossly excessive spending in other kinds of systems, being disguised as "upgrades" by the Democrats.

I'm not promoting any particular political party here. Rather, my point is to illustrate that every partisan criticism in this thread has an equally-valid counterpoint that is too-often glossed over. When the Republicans shout about "spending", the Democrats shout "obsolescence". Nobody ever seems to want "get what's useful and nothing more", or "review the cost/benefit analysis for every component in the system".

I've worked for the federal government before, notably on one particular system whose lifespan was about 20 years. The system was designed and built to be state-of-the-art, using top-of-the-line COTS hardware available at the time (as a cost-saving measure, naturally). Ten years into the system life, those original components were obsolete, and being replaced with new top-of-the-line hardware, with the promises you mentioned: efficiency savings, faster processing, reduced costs, et cetera.

However, the basic workflow hadn't changed at all, and the software hadn't been rewritten (as that'd be prohibitively expensive), but only ported up to newer technologies. Even though each part of the process was indeed faster, the system as a whole hadn't changed significantly. It could run perfectly fine on modern (for the day) mid-grade or even low-end hardware, but because "upgrades" were seen as desirable, the system continued to be built with top-of-the-line parts, for about triple the cost.

Towards the end of the project lifespan, there was an effort to re-engineer it using minimal hardware, but by that point the idea had grown into something of a legend. The managers (and bureaucrats) who had seen the system's early versions and knew its original cost couldn't believe the system could actually run on such a low hardware budget. Every actual test was successful, but the mantra that "you get what you pay for" had become such an integral part of common sense that actually getting approval for a cost-efficient system was impossible. Eventually, my team ended up inflating our quoted costs to get approval, then delivering a working system under budget and getting extra praise.

That tale doesn't meet my idea of "reasonable", but it was definitely the reality that I saw.