Arrests For Selling Poison-Ware In Spain

An anonymous reader writes "Spain's FBI equivalent has arrested the management of a software company (Google translation; Spanish original) for selling custom software to small and medium-sized businesses with 'controlled errors' that resulted in the software bombing on a predetermined date. They would then charge for fixing the problem and press the client into buying a maintenance contract. More than 1,000 clients were affected."

Comments

Re:Microsoft

[Anonymous Coward]

There's nothing the least bit controlled about Microsoft's errors, so I fail to see how this could apply to them.

Re:Shenanigans!

[Jade E. 2]

how about this one, does it contain bugs?

10 PRINT "Meshach is never wrong!";
20 GOTO 10;

Yes, BASIC doesn't use semicolons at the end of lines.

Re:Microsoft

[badboy_tw2002]

Mod parent up! Epic slam at the '$oft, brah.

Not entirely shenanigans!

[Ungrounded Lightning]

[Planned obsolescence] has been happening for generations, where have you been?

It's not always ENTIRELY shenanigans.

For instance: The "design lifetime" in the auto industry is not just about selling another car. It's also about not spending a lot of extra money making, say, the transmission good for 750,000 miles when several other major systems are going to go out at a small fraction of that time. (When you're making several million units a year, saving a nickel each adds up to enough to hire two more full-time engineers to figure out how to do it.)

Making mechanical parts that last can be tough and costly. (And half a century ago it was a lot tougher, without the major advances in materials science since then.) If you design all the parts to last for at least some design lifetime and not much longer you can accumulate a lot of savings. If some major system was going to unavoidably fail shortly after that design lifetime anyhow, having the rest not good for much longer doesn't appreciably affect the utility of the vehicle for the consumer. But the cost savings can be used to lower the price (and grab market share, for a net profit increase) - which DOES help him out significantly.

The ideal in the limit is the "Preacher's marvelous one-horse shay, which lasted a hundred years and all fell apart on the very same day."

Re:Does anyone know who it is?

[Anonymous Coward]

The company is CIPSA, is mentionend on the GDT news: https://www.gdt.guardiacivil.es/webgdt at the bottom of the page, under "Detenidos los responsables de comercializar software con "bombas lógicas"

Re:Yet another argument for Open Source. B-)

[Mike610544]

How often does anything that looks like an obfuscated C contest entry actually get committed to a repository ?

It happens all the time where I work. I maintain some old code written by an old hacker (he's got a credit in the K&R book!) Shit like this is not uncommon:
*(&z + z) |= ~tqq + m ? u9 >> 2: 741 | w & 0x8F ? ~(~t11) : foo

Re:Yet another argument for Open Source. B-)

[CAIMLAS]

Sure. And who, exactly, is going to contribute to an open source project written intentionally obfuscated? Nobody. Then the project gets the reputation of being shoddy, and nobody uses it.

Or, there's also the "we'll just rewrite this little obfuscation and fork it" scenario.

Open Source thrives on its quality and dies from crap like this. People don't contribute to dead projects: they fork them or reimplement them.

How dare you, my mother is a saint!

[SmallFurryCreature]

I wonder which programmer should be more worried, the one who can't read the above, or the one who can.

Re:How dare you, my mother is a saint!

[KiloByte]

In this case, the one who wrote that. And I don't mean just readability by novices.

*(&z + z) -- unless it's C++, this makes sense only for referring to the zth next variable after z. Like: int z, a, b, c; -- z=1 will select a, z=2 will select b, z=3 will select c. In an old compiler, this will always work. In an optimizing one, it's damn likely to break.

Mixing dec and hex numbers, and writing down constants for bit operations using decimal numbers in general is prone to mistakes.
So is using addition in an expression that consist mostly of bit operations, you want | there instead.

0x8F is a complex mask, it definitely should be a #define with a name. There's nothing wrong with masks like 0x7F or 0x1F, but for 0x8F, it's not obvious enough.

~(~t11) -- uhm, what's the point?

With these issues fixed, though, with a bit of comments such a code isn't that bad.

Re:Microsoft

[theshowmecanuck]

I call bullshit. For one, I was involved in Y2K related work for businesses that would have failed if their systems were not upgraded, namely the main enterprise business systems for daily newspapers (their circulation systems); for prepaid and wholesale (single copy) sales. And that included almost every newspaper in North America, and many overseas (not that our company worked on every newspaper's system, but considering there are not that many circ system software solutions, and they pretty much all had Y2K issues...). As well, the dot com bubble didn't burst until at least January 2000 which means that the Y2K issue was sorted out or companies were out of business before the dot com people were out of work. As well, the dot com people thrown out of work were mostly web developers, and the Y2K issue affected server side, and often COBOL related software... not exactly in the dot com programming skills bag. The company I worked for by the way provided a non-COBOL replacement system to fix our clients Y2K... complete new system instead of patching the existing system. I don't know where this 'myth of the Y2K' came about, but it seems to yet another conspiracy theory. I haven't seen or heard of one Y2K fix being worked on that wasn't solving a real critical issue. You're not a 'truther' or 'birther' by any chance too?

I'll play advocate of the devil

[mrjb]

Software bombing on a certain date, just so you can charge for "fixing" it is evil.
But that assumes that the software was paid for to start with.

I remember my father adding just this "feature" to the software
of a difficult client that only requested feature upon feature
but had a track record of being months late with their payments
(not very nice if you have a family to feed!)

When the payment was once again long overdue, the client was
faced with a friendly dialog stating that the software was
not paid for yet, and that it would only be re-activated after
payment in full. The payment cleared less than 24 hours later.

It probably would have held up in court, too.

Re:Nice

[mcoca]

The people were charged because it was a criminal case. Had it been a civil action, they would have gone after the company. Pretty sure it's the same in the US.

sometimes it is justified

[eennaarbrak]

A friend of mine works for a company that sells software to a government department a central African country (I want to keep the details vague to avoid incrimination). After completing the contract and delivering the software, reps arrived one day and simply stated "We're not going to pay full price for the software - we're not making as much money out of it as we thought we would." This country does not have much of a justice system to appeal to if you don't have a politician in your pocket, so my friend's company intentionally released code to make the system stop working if the payments are late. AFAIK that fixed the problem.

I'm just curios if these companies were perhaps faced with a similar situation...