> I'll just reregister my opinion that while open source is a great thing,
> it's not what we're selling here. We've mentioned that LyX (and latex) are
> open source. Linux et al. are doing just fine in the press and don't need
> our PR help. "LyX runs on Linux, FreeBSD, NetBSD, and most proprietary UNIX
> systems." Or even, "LyX runs on most UNIX platforms." (Is it really "most"?
> Do we know of a UNIX platform it *doesn't* run on?
We know of a few systems, where the provided compiler is not good enough. This
includes systems with gcc 2.6.x, but also some of the proprietary ones.
Jean-Marc is very helpful in compiling on DEC and other platforms with the
proprietary compiler, and thus we have relatively high portability on
compilers.
Besides the problems of compilers, our main portability bottleneck is XForms.
For instance, XForms does not exist for Amiga OS, and thus I doubt we'll see a
Amiga OS port anytime soon.
When we talk LyX 1.1 to become 1.2, things are different. We have decided to
grasp the future, and have begun using more modern C++ features, specifically
the revolution called STL. Sadly, most C++ compilers are obsolete and does not
support the STL well enough, so this might imply that we loose a few ports in
the process. However C++ has been standardized and won't change in the next
four-five years, so with luck, this loss is only a question of time. The
compiler writers have a stable goal to aim at, and this will hopefully imply
that we only loose portability to a very small group of users. (I think
Jean-Marc is trying to keep the gcc 2.7.x option open.)
Regarding the open source issue: it's not a question of Linux et al needing
our PR help. It's a question of us exploiting their existing PR. If we
associate with these high profile things, we might get a few extra points.
Greets,
Asger
