Someone contacted me the other day trying to sell me an escrow contract for a piece of third party business software my IT team provides/supports for our users/customers (local government).
Fortunately, I’m not even a decision maker in such agreements - and when I conferred with senior management, I was told we don’t bother with any such things - which seemed fine to me, as software escrow looks at best like a waste of money and false hope, or at worst, a scam.
So I pay a company x hundred pounds a year, so that when the author of software product X falls under a bus, or their company goes to the wall, I find a heap of obscure, probably unmaintainable, possibly uncompilable source code dumped in my lap? Where does that get fun?
Are there even any such things as software escrow success stories? Seems like the last thing I want when a software supplier fails is the responsibility of taking over their job.
When my employer’s company upgraded a major piece of software that we use, the manager of our IT development team insisted the vendor put the source code – and compilers and maintenance iterations (v x.Nn) – into a software escrower’s repository. We, and other companies using the same version of that software, pay a portion of the escrow fee.
It’s basically insurance […which, ultimately, is a strangely legalized and widespread form of gambling]. If the vendor goes out of business, the companies that have contributed to the escrow fee get copies of everything in that repository. Our software developers, who customize the software for our employees to use for our business needs, are familiar enough with the base product in the repository to be able to rebuild and the application on our own host server, if that becomes necessary (i.e., the vendor goes ka-put). But it’s like Life and Accident insurance: You’re placing a bet that you will die or be horribly maimed within the next year. If you’re right, you win the bet (but, then again, you’re disfigured or dead). If you’re wrong, well, you’ve lost the bet but you’re heartily encouraged to place a new bet against the same outcome(s). In this case, we’re betting that the vendor will go belly-up within a year; if we win we’re still able to use and customize the application.
It’s also something of a hedge against *Abandonware *(q.v.), though the practice of distributing software way beyond its term of support by the originator seems more like an issue among video game enthusiasts than in the corporate world (corporations and other industries simply drag out the perseverence of IBM And COBOL systems despite the fact that most desktop PC’s eclipsed those things decades ago.)
As for success stories? I dunno…I just know I was told to pay our recent software escrow premium last month.:eek:
–G!
I win, I Win, I WIN! 
…but it’s not enough to cover the undertaker’s fees.:mad:
Sure, it’s insurance, but unlike many other forms of insurance, it doesn’t even look as though the payout would necessarily be all that useful. If it’s a bespoke bit of software that your company commissioned and had a hand in the design of, then I could see how it might be possible for you to hit the ground running and eventually get into a position where you can support and maintain the thing.
But in the case of commodity-type application/system software, it seems like the odds are heavily stacked against it ever being useful - even under the most favourable conditions (properly commented, well-structured source code), it’s very difficult to maintain someone else’s work - and if on top of that you also have to replicate their environment (i.e. the development platform and version, any libraries, third-party objects, tools, etc), and possibly skill up on an unfamiliar development platform (suppose they’re still using Delphi or some such), the escrow agreement is actually a negative benefit - you’d probably expend less effort recreating the software from scratch.
The only ‘success story’ listed on the website of the company who were trying to sell to me the other day is in fact a case study about one of their customers who has bought in, but never yet needed to trigger the escrow - so basically “company X feels all warm and fuzzy because they’re paying us a stack of cash, and we’re telling them everything will be OK”
I’m resurrecting this thread as I’m now in another role where I do in fact have a stake in the decision, and software escrow has come up again. Anyone got any success or failure stories now?
Mangetout, speaking from a US perspective the consensus seems to be not really. The following link is from some years ago, but it basically encapsulates (with legal cites) why it is probably just an expensive “feel good” method of insuring oneself. Additionally, it should only be for mission critical software, as far as I understand.