Closing In...

5 views
Skip to first unread message

Jim Shargo

unread,
Aug 28, 2010, 5:46:04 PM8/28/10
to edwin-...@googlegroups.com
It's been a whirlwind few weeks, and I'm sure you can all tell by my
number of commits. In short, I'm sorry--I should have given you all more
of a heads up than I did, kept you more up to date, and explained the
situation more clearly over the group. I am ashamed and won't let it
happen again. And the worst part is...

I couldn't have chosen a worse time to do it! Work was picking up steam,
we were inching ever closer to our milestone, and the Emacsen Tower of
Babel might be getting a new foundation.

We only need a few things to get into the next, more exciting stage of
production:

*TTY -> kbd parser*
This might end up being tricky depending on how Duncan's
termcap/terminfo works now. I personally haven't been able to get
Duncan's meta-key test to work with nuscsh, and this is fairly critical
to getting blindwin.

*(scaffold) environment*
The more advanced we make blindwin, the more we're going to want
critical features like... multiple buffers. Getting the environment of
current-datastructure is something that shouldn't take long to establish
and will provide us with an entry point to the program. Think editor.scm

*reasonable file-system*
For sanity's sake, a more reasonable means of holding the files together
is required. How's this?

+ Docs -- Documents
+ Core -- Core Edwin code that makes up the heart of the editor
+ Scheme -- Scheme libraries, srfis, and the basic bits that get edwin
working
+ Modes -- There are so many modes... they need to go here and named
conventionally
+ OS -- Whatever OS specific files should hide here
+ Misc. -- Things that we don't need to handle, things we can defer
until way later.

Should this happen now? Or after we make all these sweeping changes?

*integrate Roderic's pantene system*
This is in the same vein as the environment. We need to be able to
handle all sorts of fault states and C-g enterings. I'm by no means a
master on this, so someone else could be awesome and hack it in in small
bits.

*implement basic*
We're going to work on this incrementally, bringing features in
piecemeal and bloating up our editor. Beyond self-insert-key, this is
low priority.

And there we go! Thoughts? Until merges happen,
g...@github.com:mustermark/edwin48.git will be our hub.

-- Jim Shargo

Roderic Morris

unread,
Aug 29, 2010, 9:36:28 PM8/29/10
to edwin-...@googlegroups.com

Good stuff! If you could point me at some code you think pantene should be used in, I'll work on integrating it. This'll be my last week out of Boston, so I'm gonna try to ramp up my scsch / Edwin hacking from now.

-Roderic

Reply all
Reply to author
Forward
0 new messages