This thread tangent has come along at a very convenient time for me,
because I've spent some time over my vacation planning out and designing
a GNUstep-based desktop. I'm going to write a bit about that here.
My computing needs are simple: I need a mail client, a calendar, an
address book, a web browser, and emacs. I need a few other things, but
I'll make those myself. The basics are listed here. GNUstep already has
excellent options for each of those categories, except a web browser.
But here's the funny thing: none of the other environments have a web
browser choice that's a compelling choice over Firefox anyhow.
There are a few necessities that are missing before I can claim
"complete desktop experience", and a lot of these things were mentioned
in the LiveCD comments thread. The first thing I noticed when I set up
my ThinkPad with WindowMaker + GNUstep is that the media keys don't
work. It's an easy fix, and I'll do that.
We need a few more preferences panes for configuring PulseAudio
input/output options, a small app for talking to NetworkManager for
configuring wifi, etc. After that, it's a matter of shipping a sane
"default configuration", much like Betrand describes. There are some
other things that David mentioned in a few mails, but it mostly boils
down to "make sure things implement the freedesktop specifications."
I have a parallel goal of getting some of my favourite mac apps
compiling for GNUstep. I'm starting with AntiRSI, for instance, which is
a small bit of code that reminds you to take breaks from typing. I would
really love to see something like Quicksilver working. Adium and
Colloquy would be amazing goals as well.
We're at an interesting crossroads right now: many mac folks are moving
back to Linux-based desktops, for a variety of reasons. Many of them are
bringing Cocoa + Objective-C knowledge and skills with them.
My vacation ends this week, so here's what I'm going to be spending the
next many months on:
1) a sensible default configuration of a GNUstep-based desktop. Focusing
on filling holes like "media keys that work" (xfce has great prior art
on this, and it's like 100 lines of code), and perhaps a GNUstep
implementation of xrandr (see ARandR for prior art). The things that
nobody thinks about, but *really* strong miss when they're gone.
2) a personal finance application for budgeting, and a GTD application,
both with the goal of also being available on Windows and Mac from the
same codebase. I'd like for these applications to be examples of how you
can do cross platform really well using GNUstep.
3) the integration bits that David talked about. We can use clever
theming to make sure that Qt and GTK+ apps fit in together, and that's
easy. But we really need to make sure that they "feel right" next to
each other.
4) obviously, fixing bugs as I find them in these two previous activities.
There are currently a few barriers to me making really good progress on
this:
1) HiDPI not being fully baked is causing me a lot of usability
problems, so I wind up working on my older systems. Which I don't always
have nearby. I would *really* like to help with this, but a lot of it is
way out of my depth. If someone who has the right knowledge would like
to discuss this with me and help me get a better understanding, I'd love
to take it on.
2) Bugs or "surprises" in Gorm and PB often have me scratching my head,
or shaving yaks. I'm starting to compile these in reproducible ways, and
I'll contribute where I can.
3) Documentation. There is a lot of missing API documentation. I am
basically just going to write this and submit patches. NSTableView is
the first thing I'm working on.
I've got a few other things on my list, and there's a lot of ocean to
boil. If some of you also feel like this is a good use of your time,
perhaps you'll want to join me, and we can set up a place for discussing
and sharing this. I'm doing this for myself, to scratch my own itches,
but I think it could have wider use.
The bottom line for me is that if I can have a development environment
that I *actually enjoy* using, as much as I did on NeXT, and as much as
I did for most of the time I did Mac and iOS development (Xcode bullshit
notwithstanding), then I think that's valuable to the entire community.
Anyway, those are my thoughts.
Happy hacking!
-Steven
On 31/07/17 07:12 PM, Ivan Vučica wrote:
On Mon, Jul 31, 2017 at 6:05 PM, Xavier Brochard
<[email protected] <mailto:[email protected]>> wrote:
Le 31 juillet, 18:46:22 Fred Kiefer a écrit :
> That just shows the main issue we have. People, even the one interested in
> GNUstep, are not aware of what is already there.
That's why a little how-to is necessary!
It would start with
For a good experience try with
- freeBSD
- Fedora
- SuSE
- forget Debian and derivative
Install this
Configure this and this
etc.
Side note: We seem to be having another potential centithread about
'what needs to be done', 'why is GNUstep not used more' and 'what is
the right thing for people to use'.
Before the whole thing escalates, posters, please consider to which
level is discussing this useful, and at which level we risk entering
the territory of 'this discussion is damaging'. Remember: we've had
these situations before, with key contributors bailing out.
Also, consider: depending on your point of view, this thread may or
may not have went slightly offtopic. Are we still discussing the Live
CD posting on OSNews? Perhaps, but it's not clear to me. :)
_______________________________________________
Discuss-gnustep mailing list
[email protected]
https://lists.gnu.org/mailman/listinfo/discuss-gnustep
_______________________________________________
Discuss-gnustep mailing list
[email protected]
https://lists.gnu.org/mailman/listinfo/discuss-gnustep
