Khaled Hosny <khaledho...@eglug.org> skribis: > On Tue, Jun 15, 2010 at 05:12:30PM +0200, Ben Laenen wrote: > > Khaled Hosny wrote: > > > On Tue, Jun 15, 2010 at 09:32:48AM +0200, Ben Laenen wrote: > > > > On Tue, Jun 15, 2010 at 3:53 AM, Peter Baker <b.ta...@gmail.com> wrote: > > > > > Couple of quick points. First, the FontForge format has always been > > > > > plain text. It works well with CVS, SVN, etc. > > > > > > > > Only if you cut out the unneeded bits like we do with DejaVu. If we > > > > would forget to run that script and commit a change to SVN we'd get > > > > something like a 2 MB patch. > > > > > > Not my experience, I do post-edit the file before committing, but it is > > > usually few cosmetics, nothing big, the last time I used the dejavu > > > scripts they broke my files, my be I did something wrong, but this was a > > > while ago and I didn't check again since then. > > > > FF remembers which windows were opened, several settings, which glyph > > points > > were selected etc. This can all be discarded. > > I know, I was actually commenting on the "2 MB patch" part, since I felt > it is a bit exaggerating.
Not for me, because I often use huge bitmap images, which sfd inconveniently stores in the one big file for no particular reason, and it's non-trivial to get stuff out of a Mercurial repository that you put there by mistake. I suppose just making an sfd breakable into distinctive parts would help solve a lot of issues. We've got to start making these fontforge changes ourselves, but I'm not happy working on a program that is written in glorified PDP-11 assembly language. Doing that is how I ended up disabled. :)
