in reply to "As software grows up" .. what do you do?
The “first adopter” scenario is in some ways the most interesting to me because it brings forth the question of what exactly should you do with an application ... now in-service and now revenue producing ... when the technology of the entire world (front-end) is moving headlong into tablet computers, but hasn’t completely done so yet and possibly never will.
As an aside, we used a very interesting new tool on a couple of other projects lately ... check it out ... http://www.haxe.org. It is a true cross-platform development language. We haven’t suggested it in this case, but it’s the sort of situation I think we’re all faced with.
Generally, the problem here is ... how to teach an old dog new tricks ... how to replace the engine while you’re driving down the road ... how to drive on two roads at once (possibly forever) ... and, how to persuade a client with rapidly declining revenues that money needs to be spent and spent in entirely new directions. This concern, at least for us, is growing quite rapidly in proportion.
As a (21-year old...) business, we mostly deal with applications that we did not write, very often after the original developer(s) have “gone on to greener pastures.” Therefore, reverse-engineering and continuing to support legacy applications is very much a part of our business strategy and always has been. The question here, then, really is strategy. When you are wedged against, on the one hand, shrinking revenue because you are not in the lonesome position that you once were in, and on the other hand, uncertain future costs because only a (maybe, quite small) part of what you're running now seems applicable to the future ... and yet, you cannot stop running it, not even for a moment ... well, that’s certainly a conundrum for a business owner.
One not-so-small element of the whole thing is that ... hardware is back, namely in the form of tablets. And there is no clear winner. (Hence the vital importance to us of haXe.) A “web site” is no longer enough. A “classic” Web 1.0 interface is a serious disadvantage, and it is rather tightly-wedged together in the name of the “mod_perl efficiency” that of course really was a major factor when dealing with smaller machines. Therefore, I am particularly interested in hearing (here...) the real-world experiences of people who have done and/or who are now doing projects of this nature.
Discussions are welcome; down-votes are merely annoying. Please refrain. I need to start a serious discussion-thread here. (If this sort of situation does not appear to be too-important to you now, trust me, it soon will be. Our target is moving (again) and I think we all need to collectively be discussing how to serve our various customers’ needs.p