reassign 314376 libgd2-xpm
thanks

On Thu, Jun 16, 2005 at 12:47:12AM +0200, MP wrote:
> Package: php4-gd
> Version: 4:4.3.10-15
> Severity: wishlist

> php4-gd from oldstable have dependencies:
> libgd1 (>= 1.8.4-11) or libgd1-noxpm (>= 1.8.4-11)

> but php4-gd from new stable have dependencies only libgd2-xpm and no 
> alternative for libgd2-noxpm
> This way it is forcing useless libraries to system 
> (apt-get install libgd2-xpm is telling these will remove libgd2-noxpm and 
> install 13.4Mb of useless stuff)

> Pretty bad for routers with small harrdisk (either some old piece of HDD  
> or CF-IDE reduction with some 256 mb compactflash or so)

Why the hell would you want php4-gd on a router?

The full set of packages pulled in by libgd2-xpm on a base system is as
follows:

  ucf file libmagic1
    -- all common dependencies of lots of other packages

  libfreetype6 libjpeg62 libpng12-0
    -- also installed by libgd2-noxpm

  libexpat1
    -- a dependency of all of the PHP4 binaries

  libfontconfig1 defoma fontconfig ttf-bitstream-vera
    -- these are not dependencies of libxpm4, and there's no obvious reason
       why this should be a difference between libgd2-xpm and libgd2-noxpm,
       as this has nothing at all to do with XPM images.

  libxpm4 xfree86-common
    -- these are the real dependencies needed for xpm support, and account
       for less than 1MB of space used.

  libx11-6 xlibs-data
    -- these are pulled in because of a spurious dependency of libgd2-xpm
       on libx11-6, and account for 9MB of the space used.

So, this last is a bug in libgd2-xpm that should be fixed.

> What point is with having alternatives, if there is no choice in 
> dependencies? I think the dependency should be changed to libgd2-xpm or 
> libgd2-noxpm

I don't believe there is any point in having these alternatives.  I would
love to see these extra dependencies fixed, and the libgd2-noxpm package
dropped for etch.

-- 
Steve Langasek
postmodern programmer

Attachment: signature.asc
Description: Digital signature

Reply via email to