Future of NEdit?
Joachim Lous
joachim at lous.org
Mon Nov 26 13:12:12 CET 2007
On Nov 25, 2007 1:04 AM, Erik de Castro Lopo <nedit-erikd at mega-nerd.com> wrote:
> nedit at irotas.net wrote:
> The idea of porting Nedit to a new GUI toolkit is also probably a
> non-starter as the GUI code is too tightly interwoven with the
> editing code.
I'm not sure that is the biggest showstopper. The main edit widget is
mostly its own code.
The reist of the interface is indeed a large job to port, but not
insurmountable, especially
considering we want to rework a lot of it anyway.
The biggie with toolkits is that no toolkit satisfies a big enough part of the
user community, and for very real reasons; not just personal or aesthetic
preference: A signifcant part of nedit's usage is there because we're the only
graphical editor of similar sophistication that is easy to get to work
on a lot of
platforms do have good-quality Motif implementations but nothing supported
by no other toolkits. Then there are the open-source-only platforms, where
acceptable-quality Motif is not available for most.
I can only really see two alternatives, and none are very nice:
a) split into two forks, one using Motif and focused on leagcy platforms, the
other moving to a modern toolkit, and probably ending ujp as a near-total
rewrite. Of course this would probably kill both branches, since we're hardly
moving as it is, and the linux-oriented branch might face tough compatition
from other established editors there, especially since it would take a couple
of releases until the port catch es up wiht the original in important details.
OR
b) Move down from Motif to work directly against Xt. This probably means
coding our own lightweight toolkit. In order to replicate the GUI we have in
this way would be a large implementation job, and the design work would
be big too, so the risk of project death would be huge. The only way I can
see it happen and succeed is if someone makes the toolkit privately and
secretely, and shows up with it good to go; someone who isn't then diverting
effort from core NEdit.
Like I said: not very inviting. But then again, neither is the third
option: keep
on floundering along until the project stops completely.
More information about the Discuss
mailing list