* Daniel Kobras ([EMAIL PROTECTED]) [070610 23:46]: > On Sun, Jun 10, 2007 at 07:28:02AM +0200, Andreas Barth wrote: > > I uploaded an NMU of your package. > > > > Please see this as help to get the package into a releaseable condition. > > No need to emphasize which I'd never doubt anyway...
Well, this is more or less a standard text I use for any NMU (and one *has* to say something). > > diff -ur dx-4.4.0~/debian/changelog dx-4.4.0/debian/changelog > > --- dx-4.4.0~/debian/changelog 2007-06-09 22:37:00.000000000 +0000 > > +++ dx-4.4.0/debian/changelog 2007-06-09 22:38:16.000000000 +0000 > > @@ -1,3 +1,10 @@ > > +dx (1:4.4.0-2.1) unstable; urgency=low > > + > > + * Non-maintainer upload. > > + * Add linux-kernel-headers to fix FTBFS. Also Closes: #423280 > > + > > + -- Andreas Barth <[EMAIL PROTECTED]> Sat, 9 Jun 2007 22:37:25 +0000 > > but I wouldn't mind a bit more detail on your changes: Which FTBFS are > you referring to? From your "Also" I take it that it's an issue > different from #432280? Why should build-depending on > linux-kernel-headers be a proper fix for a standard application that > ought to have no business dealing with kernel interfaces? Why do you > claim that any change in your NMU version fixes #423280 when you didn't > touch any code and the bug log clearly states that it's actually a bug > in lesstif2-dev (or libxt-dev or whatever. There was a bit of confusion > about the true origin, and I'm unsure whether it's been settled in the > meantime, which is why the bug was still open)? Incidentially, I just > rebuilt -2 on i386 without a hitch, so the header problem looks fixed > indeed, and I've got no clue what this kernel-headers hack is supposed > to achieve. I also wonder what makes this leaf package so important that > you have to NMU a dubious-looking patch with four minutes of advance > notice. Next time, please spend an extra minute on the changelog and > save me an hour of hunting down what on earth you were trying to do. I'm sorry, that was indeed not documented best. Actually, the original bug has been fixed in libxt-dev. However, build on ia64 revealed that it FTBFS now because there was a Include on linux/sys.h. As you already said, the addition of linux-kernel-headers wasn't the right thing to do, but I should've better dropped the include - the package worked without that as well. (And of course document better what I did why.) If you want, I can upload a fix for that as well (as I broke it, I need to fix it as well) - but of course, that is totally up to you. Sorry for that. Cheers, Andi -- http://home.arcor.de/andreas-barth/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
