[Prev][Next][Index][Thread]
Re: Evidence that Functional Objects is real?
"Bruce Hoult" <bruce@hoult.org> wrote in message
bruce-829D5C.13153105112000@news.nzl.ihugultra.co.nz">news:bruce-829D5C.13153105112000@news.nzl.ihugultra.co.nz...
> - the product won't suddenly stop working. It'll probably take at least
> two major Microsoft OS releases for that to happen.
Geez, not these days. I have three installs of Win2k now (laptop, here,
work) and all three are different. For instance on the laptop you can't
"pull right" the Control Panels, but at work you can. Joy.
> - you can start using Gwydion Dylan instead. No fancy IDE, but if it
> stops working -- or has a show-stopper bug -- then you can fix it
> yourself or say HEEEELLPP here and someone will pop up and fix it for
> you (possibly even in a timely manner if you offer to pay them).
But again, this is of little help to a boss. Consider what happens if all
three were to happen at once:
1) programmer in question leaves
2) existing compiler breaks
3) Gwydion isn't updated yet
This is a _very_ serious issue. It could mean months of downtime to fix a
trivial bug in an existing app. It's all well and good during development
time, but once you move onto maint these sorts of delays are like
fingernails on a blackboard. And "well if we ask real nice, someone might
fix it" isn't going to calm the willies in the project lead's stomach
either.
And I say this as one with an investment in a "marginal" language myself,
Obj-C. I face the very real possibility that my code will no longer work,
and that I'll have to port GNU-C in order to fix it. But I don't know how to
write compilers, or even port them. So that would basically be that.
Don't get me wrong, it shouldn't be like this. But it is.
Maury
Follow-Ups:
References: