[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Intent to hijack Bacula (Heads up, Get The Facts!) (long)



* Turbo Fredriksson (turbo@debian.org) wrote:
> Quoting Stephen Frost <sfrost@snowman.net>:
> > * Turbo Fredriksson (turbo@debian.org) wrote:
> >> But regarding the build system, I REALLY object to any major changes! Fixes yes,
> >> but not REPLACEMENT!!
> >
> > Uhh, or, not...  Sorry, but the build system was terrible and is
> > certainly something which should not be encouraged.
> 
> Rubish! There is nothing (major) wrong with it as it is now. Yes, hardcoding
> isn't good, but neither is 'an-hour-build'. Which you'd get if you do three
> full builds after each other...

There was quite a bit majorly wrong with it.

> > Honestly, though, I'm much more concerned about maintainability than speed of the
> > build.
> 
> It's not especially problematic to maintain as it is now, and I ask you
> to recognize the fact that we don't only ship amd64 or (fast) i386...
> Some of the arch's isn't that fast. Don't know how the m68k port is going
> or if it's still alive, but that would be a major point in getting speed
> increases in the build. I DO know that _I_ have been 'slapped around' for
> doing to extensive builds...

Sorry, but software is only going to continue to get larger and take
longer to compile.  Either the architectures need to find a way to
handle that (more buildds, distributed builds, whatever) or the
architecture it going to have to give up the ghost.  My understanding is
that the m68k folks have figured out some ways to improve things for
themselves such that they can handle larger builds, which makes me even
less inclined to hack up an unnecessairly complex build system.

No, "we have slow archs" is *not* an excuse for an overly complicated
and fragile build system.

	Thanks,

		Stephen

Attachment: signature.asc
Description: Digital signature


Reply to: