* Vincent Lefevre ([EMAIL PROTECTED]) wrote:
> On 2005-07-17 12:09:46 +0200, Mike Hommey wrote:
> > If i got it correctly, this bug got introduced because of the changes to
> > the security model for XUL applications... so, for once, we can't really
> > blame the "it should have been a bug-fix release only", since it is what
> > it was, except that security fixes has side effects.
> 
> The Debian changelog just says:
> 
>   * New upstream release, fixes security issues. (Closes: #318061)
>   * debian/rules: Disable freetype in the build for the time being. This
>     *might* break printing in some cases.
>   * gfx/idl/nsIFreeType2.idl, gfx/src/freetype/nsFreeType.cpp,
>     gfx/src/freetype/nsFreeType.h, gfx/src/ps/nsFontMetricsPS.cpp,
>     gfx/src/ps/nsFontMetricsPS.h, gfx/src/x11shared/nsFontFreeType.cpp,
>     gfx/src/x11shared/nsFontFreeType.h,
>     layout/svg/renderer/src/libart/nsSVGLibartGlyphMetricsFT.cpp: Patch
>     from bz#234035 to try to get building with the new freetype. (Closes:
>     #314243)
> 
> I can see nothing there that this can break extensions.

Firefox is not a shared library. I don't remember any guarantee
things wouldn't break between releases. 

> And if due to the change of the security model for XUL applications,
> it is known to break some extensions, I really think that the conflict
> should be marked in the packages (in mozilla-firefox, since this is
> where the change happens). Hence the bug.

Umm, that makes no sense. I'm not going to track all the extensions
every release to see if they break, that's their maintainers job.

Anyway, this whole thing is moot, the breakage was not known until
after the release. 1.0.6 has just been released to fix the API.

> In a similar way, if a library changes its ABI, one doesn't expect the
> packages that use it to silently break just because of the upgrade.

Firefox is not a library, new versions can and will break things. You
can bet your ass Firefox 1.1 will break things.

> The user needs to know if an upgrade will break packages.

This is unstable, things break. If you can't handle it, don't use
unstable. 

-- 
Eric Dorland <[EMAIL PROTECTED]>
ICQ: #61138586, Jabber: [EMAIL PROTECTED]
1024D/16D970C6 097C 4861 9934 27A0 8E1C  2B0A 61E9 8ECF 16D9 70C6

-----BEGIN GEEK CODE BLOCK-----
Version: 3.12
GCS d- s++: a-- C+++ UL+++ P++ L++ E++ W++ N+ o K- w+ 
O? M++ V-- PS+ PE Y+ PGP++ t++ 5++ X+ R tv++ b+++ DI+ D+ 
G e h! r- y+ 
------END GEEK CODE BLOCK------

Attachment: signature.asc
Description: Digital signature

Reply via email to