| 1 |
On Thu, 2012-08-09 at 13:31 -0500, William Hubbs wrote: |
| 2 |
> On Thu, Aug 09, 2012 at 11:12:46AM -0700, Olivier Crête wrote: |
| 3 |
> > > Most ideas behind systemd are interesting, their current implementation |
| 4 |
> > > is sometimes completely wrong and given the experience with pulseaudio |
| 5 |
> > > we all know that they won't change even if you provide code for it. |
| 6 |
> > |
| 7 |
> > This is bullshit, if you have good reasoned arguments, Lennart is a very |
| 8 |
> > reasonable guy, but if you just say "your ideas are shit, you code is |
| 9 |
> > terrible", then yes, he'll just ignore you. |
| 10 |
> |
| 11 |
> Sorry to call you on this one, but that is not the experience I had. |
| 12 |
> |
| 13 |
> I proposed adding configure switches to their build system to accomodate |
| 14 |
> source base distros, such as gentoo, who at times want to use udev |
| 15 |
> without systemd. I even went out of the way to make sure that I didn't |
| 16 |
> change their default settings. |
| 17 |
> |
| 18 |
> Look at a thread on their ml called minimal builds along with their wiki |
| 19 |
> page on minimal builds for Lennart's answer. He even went so far as to |
| 20 |
> say that our package managers are broken, and there was absolutely no |
| 21 |
> negotiating this point. We are wrong as far as he is concerned. |
| 22 |
|
| 23 |
He has a perfectly reasonable argument that build time is really not |
| 24 |
something you should be optimising for. Build systems easily become |
| 25 |
overcomplicated if you try to make everyone happy, you do have to make |
| 26 |
choices. Anyway, I'm not sure how that's related to the quality or |
| 27 |
design of systemd. |
| 28 |
|
| 29 |
-- |
| 30 |
Olivier Crête |
| 31 |
tester@g.o |
| 32 |
Gentoo Developer |
Archives