Re: FontAnvil 0.1

On 6 March 2014 09:33, jon banquer <[email protected]> wrote:

> On 3/6/14, Dave Crossland <[email protected]> wrote:
> >
> > We have been adding many dependencies to make code simpler and easier
> > to maintain, since the old codebase had a lot of IFDEF stuff
> > splattered all over the shop. This also makes the GUI experience
> > better, as its more consistent.
>
> This is one of those things thats sounds like a good idea but in
> practice, it turns out to be awful. If the New Linux crowd had kept in
> spirit with the original developers, it would be great.
>
> But nowadays, New Loonix people break whatever they touch. What they
> don't break, they ruin. For example, gtk, to gtk2, to gtk3 ... a slide
> down the porcelain bowl into the murky waters of gcc and "It's open
> source ! If you don't like all the stuff we broke, then you can submit
> a patch !"
>

> Believe it or not, the original open source people went way out of
> their way to make code buildable on *everything*. The responsibility
> for good code is the *developer* and *maintainer*, not the user.
>
> Every time you turn around now some stupid dependency is broken or won't
> build.
>

The dependencies that have been made mandatory are all mature and portable.
Ensuring that dependencies work is definitely our (the developers')
responsibility, not the user's.

It's precisely because all these dependencies are unproblematic on all the
major user platforms that I've made them mandatory: there's no point
supporting the possibility of building with many different combinations of
dependencies.

Another thing that has changed is that installing lots of dependencies
these days is easy: you use a package manager. Hence, there's no need to
worry about the amount of extra stuff a user has to download. Further,
Fontforge itself is installable directly from many package managers, and we
build binary packages for some of the most popular operating systems that
do not have native package managers (Windows and Mac OS X).

The introduction of gnulib was exactly to make sure that Fontforge builds
on *everything*. But building on everything should not be the individual
responsibility of application maintainers: it should be shared across the
free software community. That is precisely what gnulib offers: when a
platform does not support standard POSIX APIs, or the support is buggy, it
fixes it. All programs benefit, and developers can simply write to standard
APIs, rather than individually have to reinvent the wheel.

Yesterday I was rereading some of Richard Stallman's original writings on
Free Software. One of the ways he suggested that free software developers
could support themselves was by contracting to user groups to, for example,
port programs to the systems they use. The FSF made much of its money in
the early days from donations to maintain, port and improve GNU programs.
Nowadays, the funding is almost entirely from big business. Only a few free
software projects attract significant user donations (GNU Lilypond is one
recent example).

As a result, most projects move forward like this one with a mixture of
commercial and volunteer support. Working for money, one's inclined to do
what one is asked. Working for non-monetary rewards, the motivations are
different. In particular, non-specific criticism ("Every time you turn
around now some stupid dependency is broken") and name-calling ("Loonix")
is hugely demoralising.

Yes, we developers like patches best, but if something doesn't work for
you, file a bug report! We fix bugs! Carping and name-calling we can't fix.
The responsibility for good code is indeed ultimately the user's: if there
were no users, code would never be written.

I'm an example of this: I'm a user of Fontforge, I want it to survive and
prosper long-term, so I jumped in to work on development for a bit. I bet
every one of the current crop of Fontforge developers would tell a similar
story.

-- 
http://rrt.sc3d.org

------------------------------------------------------------------------------
Subversion Kills Productivity. Get off Subversion & Make the Move to Perforce.
With Perforce, you get hassle-free workflows. Merge that actually works. 
Faster operations. Version large binaries.  Built-in WAN optimization and the
freedom to use Git, Perforce or both. Make the move to Perforce.
http://pubads.g.doubleclick.net/gampad/clk?id=122218951&iu=/4140/ostg.clktrk

_______________________________________________
Fontforge-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/fontforge-devel