Re: APBuild

"Jan Niklas Hasse" <[email protected]> Wed, 12 Nov 2008 19:04:29 +0100
Newsgroups gmane.comp.autopackage.devel
Message-ID <[email protected]>
On Wed, Nov 12, 2008 at 6:26 PM, Justin Karneges <[email protected]> wrote:
> That was it!  The installer now works.  So I guess it was all fine and I
> misunderstood how it worked.  I might suggest that you make this more
> idiot-proof.  Specifically:
>  - have apgcc abort if being invoked on x86_64
>  - have makepackage refuse to package an x86_64 executable if it doesn't
> target AP 1.4

I added the latter and will commit it later this evening, thx.

> Btw, I got this output at the end of makepackage:
>
> /home/jkarneges/.local/share/autopackage/apkg-funclib: line
> 2149: /tmp/autopackage.865614527/reporting: No such file or directory
> rm: cannot remove directory
> `/tmp/autopackage.865614527/payload/@barracuda.com/imclient:3.3.2/share/barracuda/gfx/.svn/prop-base':
> No such file or directory
>
> Maybe a race in some cleanup code?  Permissions look fine.  Anyway, this
> didn't seem to affect creation of the installer.

Thanks for reporting this. I will try my best to fix it. Could be
related to http://trac.autopackage.org/ticket/104

> I suppose this should really be: have apgcc abort if the target arch is
> x86_64.  The host arch should not affect it, so that cross-compiling stays
> possible.

Why shouldn't apgcc be able to compile for x86_64 machines?

---------------------------------------------------------------------
To unsubscribe, e-mail: autopackage-dev-unsubscribe-OfajU3CKLf1/[email protected]
For additional commands, e-mail: autopackage-dev-help-OfajU3CKLf1/[email protected]