Le lundi 10 novembre 2008 à 18:58 -0500, Jens Petersen a écrit : > > Actually we have 57 since Behdad requested FersiWeb fonts and no one > > I should not be that hard really to generate a script to generate a > skeleton spec file for any given font . > > I know some packaging people frown on automated packaging but this > might help lower the barrier to font packaging for which creating > rpm's is really quite easy compared to general packaging. Such spec > files would still need to go through being tweaked and polished during > review of course but it would make it easier for people to get started > I think. Since I do a lot of font reviews I'd like the polishing to be done before a spec hit bugzilla :) Anyway, with the experience of recent font reviews (un fonts in particular), I've written some macros and spec templates that push all the fc-cache scriptlet magic out of specfiles and should be a little easier for new maintainers to work on. They still require a human to 1. decide which font file goes in which (sub)package 2. decide which fontconfig generic family to associate with each font 3. write summaries and descriptions 4. do some legal auditing Quite frankly, appart from 1. that could possibly be automated by writing some script that uses fontconfig to tell people what font files declare the same font family, I don't see how we could go much farther (maybe generating the wiki page when it does not exist?) Please review the files at http://nim.fedorapeople.org/rpm-fonts/ and in particular the rpm-fonts package and how it is used by the other packages. I don't like the fontconfig file symlinking stuff much, if you can find a simpler way to express it I'd be happy to change it. > Perhaps it is something we (Fonts SIG) should consider working on? I don't really know what parts new font packagers find hardest, I'd love to see some feedback on the list. As I wrote before, I don't think we could win a lot by automating. But anyway, it's Fedora and everyone is free to work on what he likes, so if you think automating would help far from me to stop you :p I wouldn't mind being proven wrong. Regards, -- Nicolas Mailhot
Attachment:
signature.asc
Description: Ceci est une partie de message =?ISO-8859-1?Q?num=E9riquement?= =?ISO-8859-1?Q?_sign=E9e?=
-- fedora-devel-list mailing list fedora-devel-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/fedora-devel-list
