Motif

[Thomas Orgis]

Hey, folks...

I have a question as a "packager" for a source-based GNU/Linux distribution.
Source-based means that we really want to build nedit from source and won't rely on the official binaries.

We used to have nedit dependent on openmotif, which is a bit troublesome as this is not really "Free", hence, not in the default package collection.
What also is troublesome, is the open question of compatibility of openmotif 2.3 with nedit.
Well, I guess we all know that the Motif question is a PITA, generally... now adding a new chapter:

The XPrint extension (libXp) has been removed from our distribution, following the development of X.org (I take this as a fact...).
Now OpenMotif rather seems to rely on that and I have yet to figure out how to disable libXp linking to get it working, prompting again the question of replacing it with lesstif (again).
Perhaps the age of the latest lesstif 0.95 (meanwhile patched in our distro to work without xprint) release is a sign of stability?

Well, my point is:
What is the recommendation of the nedit folks for keeping nedit and Motif in current GNU/Linux systems, buildable from source?

Should I fight for OpenMotif 2.3.1 to make it work?
Should I rather even get OpenMotif 2.2 instead and make that work (probably more compatible for nedit)?
Should I ditch the non-really-free OpenMotif and switch to lesstif 0.95?

That aside, people generally respond with suprise and incomprehension when I talk about needing that strange proprietary UNIX toolkit, even more more so for a dated strange editor that cannot even do Unicode (but well, the latter is also often just compressed to "not vim/emacs?";-).
The Motif bit being a rather basic one (probably best tackled by trying to work around lesstif bugs?) and Unicode being tricky, it would be nice if we could at least get a new release of nedit with various fixes/improvements that are in CVS?
Is anyone working on that?


Alrighty then,

Thomas.

[Joerg Fischer]

Post by Thomas Orgis
Should I fight for OpenMotif 2.3.1 to make it work?
Should I rather even get OpenMotif 2.2 instead and make that work
(probably more compatible for nedit)?

Unless the OpenGroup releases Motif as free software, I'd say no, don't
support them with contributions. (http://www.marutan.net/cde/)
Btw, OM versions 2.2.0 - 2.2.2 are listed as "known bad" in
http://nedit.cvs.sourceforge.net/viewvc/nedit/nedit/util/motif.c?revision=1.7&view=markup

Post by Thomas Orgis
Should I ditch the non-really-free OpenMotif and switch to lesstif 0.95?

Unfortunately, the latest lesstif versions contain more glitches than some
older ones wrt nedit.

Starting with 94.4 there is even more than just a glitch in there. The key
sequence Alt+H ESC lets nedit crash. But if someone can fix lesstif, IMO
it's more worthwhile than doing customization to get OM running.

Generally, all I ever critized was that some even big distributors
built nedit executables and released them w/o making sure they worked.
Often, these were builds using lesstif. However, I don't think I critized
using lesstif as such.

Post by Thomas Orgis
That aside, people generally respond with suprise and incomprehension when I talk about > needing that strange proprietary UNIX toolkit,

Just compare the start-up times of nedit, kedit/kate, and gedit, then tell me
what is strange.

Post by Thomas Orgis
it would be nice if we could at least get a new release of nedit with various fixes
improvements that are in CVS?

You claim to be a source distribution and you can't do a check-out of
nedit's sources? ;-)

[Btw, there is a tag "BETA-5-6".]

--Jörg

[Thomas Orgis]

Am Sat, 28 Feb 2009 21:08:52 +0100

Post by Joerg Fischer
Unless the OpenGroup releases Motif as free software, I'd say no

Agreed.

Post by Joerg Fischer

Post by Thomas Orgis
Should I ditch the non-really-free OpenMotif and switch to lesstif 0.95?

Unfortunately, the latest lesstif versions contain more glitches than some
older ones wrt nedit.
Starting with 94.4 there is even more than just a glitch in there. The key
sequence Alt+H ESC lets nedit crash. But if someone can fix lesstif, IMO
it's more worthwhile than doing customization to get OM running.

Interesting. Say, what would be the best lesstif version that's stable?
Hm. Also, I'd suspect nedit being a rather major application for lesstif (what else in the free software world uses Motif?) ... so it is rather disturbing to hear of new versions being even more troublesome, combined with the apparent lack of activity to fix the issues (or any activity at all).
Anyhow, fixing lesstif issues would make more sense than fixing openmotif... though, there is again this itch to somehow rewrite nedit to have my very own best editor in the world;-)

Post by Joerg Fischer

Post by Thomas Orgis
That aside, people generally respond with suprise and incomprehension when I talk about > needing that strange proprietary UNIX toolkit,

Just compare the start-up times of nedit, kedit/kate, and gedit, then tell me
what is strange.

...that I cannot scroll fluently in a (larger) text file with a CPU that can run Doom 2;-)
Honestly, I used kate before and I wasn't able to bear the lagging anymore.
That's the danger of the wish for nedit to get updated with needed features... you could end up with something that is just like the others:-(
I really hope it can be different.

Post by Joerg Fischer

Post by Thomas Orgis
it would be nice if we could at least get a new release of nedit with various fixes
improvements that are in CVS?

You claim to be a source distribution and you can't do a check-out of
nedit's sources? ;-)

shell$ nedit -version
5.6 [Under Development] HEAD
Jan 21, 2009

Built on: Linux, x86-64, GNU C
Built at: Jan 21 2009, 13:43:05
With Motif: (Untested) 2.3.1 [@(#)Motif Version 2.3.1]


Issue is that we at Source Mage have a rather strict policy to default to the stable upstream releases.
We offer development versions for some software packages as an option, only resorting to something like CVS snapshots if there really is no alternative.
I see the official release as a sign of life from nedit. So far, people think it's dead since 4 years.
Well, as current maintainer of mpg123, I can say from experience that it can take some time to convince people that a project is _not_ dead anymore, having been dormant for ... hm... hard to say, about 5 years.
Well, one should try to make a release from time to time, even if it's just a "Ping";-)

I think I'm gonna try what happens with lesstif 0.95 ... at least that predictable crash should be able to trace...?
But then, it doesn't feel ecomomic having to fixup a large toolkit for exactly one application on my system needing it.


Alrighty then,

Thomas.

[Thomas Orgis]

Am Sat, 28 Feb 2009 23:11:36 +0100

Post by Thomas Orgis
Anyhow, fixing lesstif issues would make more sense than fixing openmotif...
though, there is again this itch to somehow rewrite nedit to have my very
own best editor in the world;-)

I wonder how the people at Arch Linux apparently made the Clipboard
work with nedit 5.5 and lesstif 0.95 ... they still have the Alt+H+Esc
thing, though.

What I also want to add is that it's wrong that nedit is about the only
app needing Motif on my system.
There's xpdf, xdvi ... well, but I did a little research in our package
database and there are in total 23 "spells" (packages) that can make
use of some Motif (doesn't have to be a complete list with regard to
unnoticed optional dependencies).

But by far most of them have Motif as an option among other toolkits
like gtk or qt. So, in effect, the Motif GUI is about never used on
modern GNU/Linux.

Nedit's hard dependency on Motif separates it clearly from the crowd of
GNU/Linux applications. Furthermore, nedit being apparently the most
picky application for specific versions of lesstif / OpenMotif. But
then, the bug reports involving nedit and Motif (and not some other app
and Motif) may just due to the fact that nedit is among the very few motif
applications used in the free software world, outside corporate environments
(where people may have installs of CDE etc. running or some specialized
software they paid for).

There are xpdf and xdvik that I use, but regardless of their popularity
(compared to evince, kpdf, etc.), I don't recall having problems with
specific Motif versions or lesstif.

Then, I want to point out that our Solaris system at work also drops
CDE and Motif... they have it still there, but the default environment
is some GNOME2 thingy (the Sun Java Desktop); with other free software
(like KDE or several window managers) being available through
CSW/Blastwave.

Bottom line is that I see nedit being dragged down to death on newer
platforms (also commercial) unless it either separates itself from
Motif or learns to work with the current free Motif alternative;
heck, even the current
not-really-free OpenMotif (which I didn't manage to build yet on a
current modular Xorg system without libxp).
How tight is the dependency on Motif in nedit, btw? Would it be
possible to restructure nedit to work with another toolkit, without
killing it's leanness and speed?
I have to think of the dillo browser project... it was dead for some
time and has been resurrected recently, replacing the gtk1 toolkit with
fltk. Seemed to be manageable -- they even moved from C to C++ along
that (OK, the other way round would be more difficult;-).
I wonder how much the toolkit dictates for nedit's performance as an
editor (I guess one would _not_ use a default text widget...;-).

I see the petition to make CDE/Motif free... and I wonder how much it would
matter if it succeeds. Would Motif reclaim the world? Or would nedit
folks end up being the people finding fixing up Motif bugs (instead of nedit
bugs) on free software systems, being about the only app left relying on Motif
in that ecosystem. Isn't it a sign that lesstif seems to be rather dead
since 2-3 years, despite of a royal bug list?

I'll finish now, before the rant gets out of hand... but I am really worried
about my editor of choice staying a valid choice for me.
Dammit, others are sluggish, and I am so used to the shell interface the nedit
client offers me (I don't care about bugs in the file open dialog as long as
ncl works;-).


Alrighty then,

Thomas.

[Joerg Fischer]

Post by Thomas Orgis
How tight is the dependency on Motif in nedit, btw? Would it be
possible to restructure nedit to work with another toolkit, without
killing it's leanness and speed?

You aren't the first one thinking about switching the toolkit. Unfortunately,
for the pending 5.6 release there is nobody who kicks it out the door.
So, migrating, or rather completely rewriting, nedit to another toolkit
just needs some volunteers. The code is in CVS.

Post by Thomas Orgis
I have to think of the dillo browser project... it was dead for some
time and has been resurrected recently, replacing the gtk1 toolkit with
fltk.

I also thought about fltk since it uses already parts of nedit's
text editing widget. Moreover, fltk is small and fast. However, looking
at the short cuts handling or block selections for example shows that
an editor rewritten with fltk won't have too much in common with the
Motif/Xt nedit. fltk is simple, and this is at the same time its up- and
downside.

Post by Thomas Orgis
Seemed to be manageable

A browser isn't an editor even if it should be a bit more complex
than dillo.

This reminds me of vim and emacs. These editors are console editors.
There are versions of them with some X toolkit linked in, but this
doesn't change much IMO. (Ok, one can display pictures in the editing
buffer of emacs' X version, but the handling isn't different.)
Basically, for these editors the purpose for using X is to have some
more pixels.

Now sometimes folks come around here and wonder why nedit can't simply
switch the toolkit like vim or emacs can. The answer is probably
something like "because there is no console version of nedit".

Post by Thomas Orgis
I see the petition to make CDE/Motif free... and I wonder how much it would
matter if it succeeds. Would Motif reclaim the world?

My point is even me isn't too much interested in learning technologies
if they aren't free. (Where "even" means, I'm interested in nedit
development, and if one likes to do this right, some knowledge of
Motif is required.)

I don't care about "modern" (or to quote Otto Rehagel: modern is what wins).
The reason why gtk was developed wasn't because it is better technologically
speaking or more modern whatever this means (perhaps newer?). It was
because Motif wasn't and still isn't free. (Moreover, there are surely
different views of how a toolkit should be, as the collection of
available free toolkits shows.) Free software needn't be better than
proprietary stuff - that's more a matter of who created the programs.

It's just not inviting to live with the restrictions of proprietary
software. Obviously, even Sun seems to realize this when they drop
their own proprietary stuff from their proprietary Solaris in favour
of free software. The paradox is that they weren't helpful with
the petition to free CDE and Motif, although I believe they could be...

But I rather like to answer some real question: Personally I use
a lesstif checkout with the version number 0.93.95. However this
isn't the released lesstif version with this number, but was
grabbed from CVS a few days or so before the "official" release.

Now you may wonder how the official release could be different
if there were only a few days between "my" snapshot and the
release. It just was always the release policy of the lesstif
project until the end to follow a "The Cathedral & the Bazaar"
style of development, so that no release was actually stable,
which IMO was a mistake for a toolkit, where stability matters
more than to get the latest version number.

The reason why nedit sometimes worked surprisingly well with
lesstif was that a main developer of lesstif was also an nedit
developer. This doesn't hold for the later versions of lesstif,
though.

Post by Thomas Orgis
Issue is that we at Source Mage have a rather strict policy
to default to the stable upstream releases.

The idea is probably that these are really more stable in the
sense of the word. The example of lesstif shows how wrong one
can be with this. The example in the other direction is this
5.6 [Under Development] thing. It surely wasn't much further
developed for moons. It mainly contains fixes, and I wonder
how 5.5 still runs on Linux systems, where so much changed
in the mean time. Actually, the sources you got is the
stable nedit by now. Don't be fooled by the wording in
the version info. You may easily change this to not frighten
the naive users;-)

[Seriously, I run a patched (for extra features) version of
these sources stably since probably years, and I presume
I'm not the only one doing so...]

Now i've written even more than you for what it's worth.

--Jörg

[Thomas Orgis]

Am Tue, 3 Mar 2009 21:02:30 +0100

Post by Joerg Fischer

Post by Thomas Orgis
How tight is the dependency on Motif in nedit, btw? Would it be
possible to restructure nedit to work with another toolkit, without
killing it's leanness and speed?

You aren't the first one thinking about switching the toolkit.

I am somewhat aware of that. For sure these are not new questions.

Post by Joerg Fischer
Unfortunately,
for the pending 5.6 release there is nobody who kicks it out the door.

So somebody needs a kick? Would 20€ donation count as motivation?

Post by Joerg Fischer
So, migrating, or rather completely rewriting, nedit to another toolkit
just needs some volunteers. The code is in CVS.

Volunteers could do so much in this world... I don't know if I can
afford this, being busy with too many things already.
For sure I could not justify additional sleepyness at work due to
nights of coding yet-another-editor:-/
I'd trow in a donation, perhaps bet funds on a project on
http://cofundos.org ... And I'm quite sure that I could not resist
hacking in some bits myself once I feel it's going somewhere (at least
I have a patch of myself in use, facilitating different window names
for different nedit servers, giving my window manager better control
over my nedit window groups).

Post by Joerg Fischer
I also thought about fltk since it uses already parts of nedit's
text editing widget.

Yeah, and sadly, it seems to be unsure if FLTK 1 or FLTK 2 is there to
stay... unclear development situation.

Post by Joerg Fischer
Moreover, fltk is small and fast.

What I am trying to get insight on currently is the speed of nedit. For
some reason its text display is extremely fast compared to other
editors.
Medit (gtk) and kate (qt) are scrolling through a 2.5M kernel ChangeLog
a lot slower than nedit, for me at least.
It could be that my current Xorg setup and video driver play tricks
with 2D accerelation features, but the others feel so clearly less
responsive that I am wondering if people just have more patience than
me or if that slowness is inherent in the toolkits or display of
Unicode or using scalable fonts.
At least dillo2 seems to perform quite well with scrolling through the
huge text file... actually slightly faster than nedit.
That is using scalable fonts with anti-aliasing.
So it seems to be possible to get this done right, but strangely other
editor programmers don't seem to care (really, kate is an extreme
example: hold down the page down key for a while, release --- kate
keeps scrolling down, having buffered all those key events, rendering
X11 totally unresponsive until the scrolling is done).
Seeing how nedit and dillo2 get it right (perhaps dillo2 inherited
nedit's code somewhat through fltk?), there seems the others getting
something wrong about display and interaction. And rendering speed at
all.

Post by Joerg Fischer
Motif/Xt nedit.

Well, rewrite using plain Xt? Do what GIMP did once -- the nedit tool
kit! ;-)

Post by Joerg Fischer
something like "because there is no console version of nedit".

Though I insist on using it in conjuction with a console -- opening
files via the client;-)
But I know you didn't mean that and indeed you raised an interesting
point, in line with my justification for using a GUI text editor:
Exactly because it has controls beyond the text interface, easing
the manipulation of the text world from the outside.

Post by Joerg Fischer
But I rather like to answer some real question: Personally I use
a lesstif checkout with the version number 0.93.95. However this
isn't the released lesstif version with this number, but was
grabbed from CVS a few days or so before the "official" release.

If this is such a nice snapshot that has proven to be stable... could
you provide it to me (email is fine, ftp upload whatever)?
I guess one needs to adapt the current patch for libxp we have, but
maybe there's a chance to switch to that lesstif in our distro when I
can justify that it's more stable for the apps that actually use it.
Would solve the problem at hand...

Post by Joerg Fischer
Actually, the sources you got is the
stable nedit by now.

So one can just name it as such and put out a source tarball?
Binaries can follow... once I got a stable Motif I could provide some
for Linux/x86, x86-64 and Linux/Alpha ... heck, even Tru64;-) Oh, and
Solaris 9/10 (SPARC, x86-64), too...

Post by Joerg Fischer
[Seriously, I run a patched (for extra features) version of
these sources stably since probably years, and I presume
I'm not the only one doing so...]

That is a bit sad as in those extra features kept in private ... ?

[Joerg Fischer]

Post by Thomas Orgis

Post by Joerg Fischer
Motif/Xt nedit.

Well, rewrite using plain Xt? Do what GIMP did once -- the nedit tool
kit! ;-)

I heard this proposal before, too. But again, for lack of time even
a source release is pending, so who is talking about switching toolkits
or developing just another one? (20 months of spare time would count more
than some donation of money, btw ;-)

Post by Thomas Orgis
If this is such a nice snapshot that has proven to be stable...

Sorry, I didn't quite intend to suggest something like this. My fault,
I know. What I meant is, I have no idea how the newer releases behave
since I'm using something older for years, which I recall is somehow
different from the "official" release, but I forgot why I didn't switch.

Of course one can't distribute a program if you know it'll crash when
closing a drop down menu. The trouble with lesstif was that this
project released so much not really well-tested stuff, and if nedit
crashes the users blame it to nedit and not to some toolkit they
aren't even aware of. (I would do so too if I wouldn't know more.)

So, perhaps someone (you?) could trace the crash, and perhaps find
out that it's just a typo or a missing check for null pointers?

Post by Thomas Orgis

Post by Joerg Fischer
Actually, the sources you got is the
stable nedit by now.

So one can just name it as such and put out a source tarball?

You know the license. It's just a matter of wording whether to call
it beta, under development (and scare users away) or to find some
more neutral terminology which doesn't claim the thing would be
unstable so you couldn't use it for real work.

Post by Thomas Orgis
That is a bit sad as in those extra features kept in private ... ?

Come on, there is a patch tracker on SF, don't you know? It isn't
unusual, if you don't get things in a release to patch the sources
to your liking and use that instead. Come on, I'm sure you know
it's free software ;-)

--Jörg

[Thomas Orgis]

Am Wed, 4 Mar 2009 16:37:11 +0100

Post by Joerg Fischer
I heard this proposal before, too. But again, for lack of time even
a source release is pending, so who is talking about switching toolkits
or developing just another one?

Yeah, let's talk about Unicode support again;-)
Actually, does Motif pose a problem with that or is it only a matter of nedit pulling the right strings for text display (apart from the internal handling)?

Post by Joerg Fischer
So, perhaps someone (you?) could trace the crash, and perhaps find
out that it's just a typo or a missing check for null pointers?

I found the same bug report for xmgrace... but there a report pointed me to interpret the debugging output of lesstif with regard to the names of the menus being popped up/down.
Thing is, nedit crashes because it tells to pop something down that is not down.

XmRowColumn fileMenu: _XmMenuEscape()
XmRowColumn fileMenu: _XmMenuEscape()
XmRowColumn fileMenu: Menu fileMenu (XmMENU_PULLDOWN)
XmRowColumn fileMenu: Menu fileMenu (XmMENU_PULLDOWN)
XmRowColumn fileMenu: must be a popup
XmRowColumn fileMenu: activeChild: (nil)
Speicherzugriffsfehler

There is no active child of fileMenu to pop down, because in that case I have been in the edit menu instead.
Pressing ESC kills nedit for every menu _except_ the file menu... you get the idea already, I'm sure.
Let's look what happens when opening the file menu (hotkey or mouse, does not matter):

XmRowColumn fileMenu: MenuFocusIn
XmRowColumn editMenu: MenuFocusIn
XmPushButton paste: FocusIn()
XmRowColumn fileMenu: _XmVirtKeysHandler
XmRowColumn fileMenu: _XmVirtKeysHandler

... and later, on closing with ESC:

XmRowColumn fileMenu: _XmMenuEscape()
XmRowColumn fileMenu: _XmMenuEscape()
XmRowColumn fileMenu: Menu fileMenu (XmMENU_PULLDOWN)
XmRowColumn fileMenu: Menu fileMenu (XmMENU_PULLDOWN)
XmRowColumn fileMenu: must be a popup
XmRowColumn fileMenu: activeChild: (nil)

Well... that should editMenu, or shouldn't it? Could that be nedit's fault?
What would cause the fileMenu to get focus when popping up the edit menu?
(I am posting these questions rather here than on the lesstif bug tracker as here somebody might actually care. We can copy over conclusions later... for historic reference.)

I don't have time now to investigate more... perhaps someone here has an idea already.

The issue preventing any nedit usage (avoiding ESC in menus, even) for me now is that one:

NEdit warning:
XmClipboardCopy() failed: XmClipboardStartCopy not called or too many formats.
NEdit warning:
XmClipboardEndCopy() failed: XmClipboardStartCopy not called.

...every time I try to copy something with Ctrl+C . Middle-mouse copying works.
Those messages somewhat imply a failure on the application side (API usage)... is this correct?
Could it be that copying worked only by chance before?
I must add that someone on x86 with ArchLinux (nedit 5.5 and lesstif 0.95) did not have trouble with copying stuff.
Perhaps this bug is only surfacing on x86-64 Linux or with my specific set of modular Xorg versions.

Oh, and generally... try to run nedit with valgrind and you see _lots_ of conditionals that depend on uninitialized variables inside lesstif. Someone more familiar with the lesstif source would be far quicker than me in tracing these.
But, looking at the lesstif bug tracker... I wonder if there is anyone familiar with lesstif source and still caring.
Oh, and I fixed a memcpy() with overlapping memory in my local copy...

Post by Joerg Fischer

Post by Thomas Orgis

Post by Joerg Fischer
Actually, the sources you got is the
stable nedit by now.

So one can just name it as such and put out a source tarball?

You know the license. It's just a matter of wording whether to call
it beta, under development (and scare users away) or to find some
more neutral terminology which doesn't claim the thing would be
unstable so you couldn't use it for real work.

So, please, can someone with access to the nedit website just edit the version bit (5.6-beta ... or just 5.6!) and put up a tarball? 10 Minutes of precious free time for free software?
As I said, our distro wants to give people the stable upstream default if possible. The nedit-latest tarball is an official URL but it is volatile (we like to check for file integrity automatically), same goes for direct CVS.
I just verified again that any CVS snapshot cannot be worse than nedit 5.5 for my system ... it won't start up at all. Some X FillRectangle bad parameter thingy, being fixed in CVS.
If it is really impossible to fix up some kind of beta/rc tarball at least, I will have to grab a snapshot and upload it on my own to our mirrors, leaving upstream, which is plainly defunct for my current system.

Post by Joerg Fischer

Post by Thomas Orgis
That is a bit sad as in those extra features kept in private ... ?

Come on, there is a patch tracker on SF, don't you know? It isn't
unusual, if you don't get things in a release to patch the sources
to your liking and use that instead. Come on, I'm sure you know
it's free software ;-)

Well, of course I can hack stuff together, but as distro packager I care for users who want to have features in the release version eventually. As long as patches are either included or discarded after some time, it's fine, but I don't think it's good to keep patches around forever as a development model as mutt does (after a certain number of patches you have to encounter some conflicts at least, patches getting stale with newer releases ... ).
And also, as a user on different machines, managed by different admins, I'd like the version of nedit installed there to work with an acceptable feature set by default... while of course I have a patched nedit running atm. at work in my $HOME, too.
I see a danger of developers always using their hacked very own nedit (with their very own stable lesstif snapshot) and so forgeting about anyone trying to use the actual offifical version of nedit (and lesstif).
Soon there could be noone trying anymore... nedit is already obsolete enough because it cannot do Unicode, why make it worse by not releasing numerous fixes (and potentially well-tested added features) already in CVS?

Alrighty then,

Thomas.

[Erik de Castro Lopo]

Post by Thomas Orgis
Yeah, let's talk about Unicode support again;-)
Actually, does Motif pose a problem with that or is it only a
matter of nedit pulling the right strings for text display (apart
from the internal handling)?
From : http://en.wikipedia.org/wiki/Motif_(widget_toolkit)

"As of version 2.1 Motif supports Unicode, which has made it
widely used in several multilingual environments."

I don't know whether lesstif supports that or not.

Post by Thomas Orgis
Oh, and generally... try to run nedit with valgrind and you see
_lots_ of conditionals that depend on uninitialized variables inside
lesstif.

Yes, thats very disturbing.

Post by Thomas Orgis
Someone more familiar with the lesstif source

I am reasonably good at diving into large bundles of source code and
getting to the bottom of them. I once tried that with lesstif and
got exactly nowhere.

Post by Thomas Orgis
I wonder if there is anyone familiar with lesstif source and still caring.

I sort of doubt it.

Post by Thomas Orgis
Oh, and I fixed a memcpy() with overlapping memory in my local copy...

In lesstif or nedit? If it was nedit, patch please :-).

Cheers,
Erik
--
-----------------------------------------------------------------
Erik de Castro Lopo
-----------------------------------------------------------------
"Therapists typically base the nuttiness of a patient on the
strength of their convictions, on which basis this 43,000 word
opus alone stands as a kind of testament to Bill's (Gates)
madness." - The Register

[Thomas Orgis]

Am Thu, 5 Mar 2009 22:29:09 +1100

Post by Erik de Castro Lopo
"As of version 2.1 Motif supports Unicode, which has made it
widely used in several multilingual environments."
I don't know whether lesstif supports that or not.

Hm, I'd hope so,

Post by Erik de Castro Lopo

Post by Thomas Orgis
Oh, and I fixed a memcpy() with overlapping memory in my local copy...

In lesstif or nedit? If it was nedit, patch please :-).

Lesstif. It implemented left-shifting of an array using memcpy()...

Generally, the errors I see are on the lesstif side. Like with mpg123,
where I try to fix all possible memory issues, but lots of valgrind
errors show up for libalsa...


Alrighty then,


Thomas.

[Thomas Orgis]

Am Thu, 5 Mar 2009 22:29:09 +1100

Post by Erik de Castro Lopo

Post by Thomas Orgis
Yeah, let's talk about Unicode support again;-)
Actually, does Motif pose a problem with that or is it only a
matter of nedit pulling the right strings for text display (apart
from the internal handling)?
From : http://en.wikipedia.org/wiki/Motif_(widget_toolkit)

"As of version 2.1 Motif supports Unicode, which has made it
widely used in several multilingual environments."
I don't know whether lesstif supports that or not.

http://www.lesstif.org/FAQ.html#QU4.10

Does LessTif support Unicode?
Not yet.
There are more limitations than just a M*tif library which fails to do so - you may want to read the UTF-8 and Unicode FAQ for Unix/Linux.



... that settles the issue for my dream of upgraded nedit for unicode.
At least for the Free software world, that seems to imply a toolkit switch.
I don't see a lesstif unicode upgrade coming.


Alrighty then,

Thomas.

[Joerg Fischer]

Post by Thomas Orgis
I found the same bug report for xmgrace...
Thing is, nedit crashes because it tells to pop something down that is not down.

So does xmgrace.

Post by Thomas Orgis
Pressing ESC kills nedit for every menu _except_ the file menu...

Just like xmgrace.

Post by Thomas Orgis
Well... that should editMenu, or shouldn't it? Could that be nedit's fault?

No, we should face it, I didn't expect the last stable lesstif release to
be that bad. (On the other hand, somehow I tried to explain something
about so-called stable releases.)

Post by Thomas Orgis
XmClipboardCopy() failed: XmClipboardStartCopy not called or too many formats.
XmClipboardEndCopy() failed: XmClipboardStartCopy not called.
...every time I try to copy something with Ctrl+C . Middle-mouse copying works.
Those messages somewhat imply a failure on the application side (API usage)...

It's just another known issue with lesstif. NEdit calls XmClipboard and
relies on the toolkit working. The middle-mouse only transfers the primary
selection, which shouldn't be confused with the clipboard selection.

No, sorry, we have to find some working free tif for Source Mag.
Since openmotif was dropped by Fedora (for license reason, too),
I searched there. The point is, they fixed a ton of lesstif bugs,
so please try the lesstif sources from one of their mirrors, eg
http://fedora.patan.com.ar/linux/development/source/SRPMS/

(I hope you didn't do already ;-). On the Fedora project you're more
likely to find someone fixing bugs, btw. The copying issue at least
seems to be closed, too, see
https://bugzilla.redhat.com/buglist.cgi?component=lesstif&product=Fedora

Post by Thomas Orgis
I must add that someone on x86 with ArchLinux (nedit 5.5 and lesstif 0.95)
did not have trouble with copying stuff.

Could be by chance they grabbed the lesstif sources of Fedora?

Post by Thomas Orgis
So, please, can someone with access to the nedit website just edit ...

What I had in mind is that you change the [under development] part
to [CVS-05-03-09], or so.

Post by Thomas Orgis
As I said, our distro wants to give people the stable upstream default if possible.

No, your distro wants to give people something that works. In big
projects that are organized and deploy some sort of quality assurance,
one may rely on the releases marked as stable. For many low profile
projects the situation is different. (Take Fedora as example, they
don't rely on the "official" lesstif release.)

Post by Thomas Orgis
... but I don't think it's good to keep patches around forever as a
development model ...

I don't think it's a model, or that it was planned like this, it just
happened. So, it's real, not a model.

Post by Thomas Orgis
Soon there could be noone trying anymore... nedit is already obsolete...

It wasn't only me who predicted this roughly five years ago.

--Jörg

[Thomas Orgis]

Am Fri, 6 Mar 2009 19:53:28 +0100

Post by Joerg Fischer
No, sorry, we have to find some working free tif for Source Mag.

That's why I asked for your snapshot;-)

Post by Joerg Fischer
(I hope you didn't do already ;-). On the Fedora project you're more
likely to find someone fixing bugs,

Or debian and gentoo, for that matter. Of course I should have looked
there first for fixed lesstif versions.
Many distributions are notoriously patching stuff -- but sometimes it's
also necessary. I'll have a look at the fixes in the big distros and
port stuff over, if it helps.
For the greater good I may just end up hosting a source tarball with
the necessary patches applied myself.
That's one main grief I have with all the distros doing patching on
their own: Stuff tends to stay buried, not reaching upstream (which
doesn't always have to be dead).

Post by Joerg Fischer

Post by Thomas Orgis
So, please, can someone with access to the nedit website just edit ...

What I had in mind is that you change the [under development] part
to [CVS-05-03-09], or so.

OK, that about makes it clear that there is no "stable upstream" to
expect... I'll make a CVS snapshot one can verify, then.
Perhaps current CVS as an option (in case it will differ in future,
hopefully). This feels a bit like forking (my mpg123 maintainership
started with hosting a patch...).

Post by Joerg Fischer

Post by Thomas Orgis
As I said, our distro wants to give people the stable upstream default if possible.

No, your distro wants to give people something that works.

Yes and no. If it's really broken that it doesn't compile / work at
all, then yes, we want to make it work for our users.
But the main credo is the user's choice (of any optional features and
patches) -- with a minimum of default modifications.
There are enough "just works" distros out there;-)

Post by Joerg Fischer
In big
projects that are organized and deploy some sort of quality assurance,
one may rely on the releases marked as stable. For many low profile
projects the situation is different. (Take Fedora as example, they
don't rely on the "official" lesstif release.)

I rather draw the line between "alive" and "dead" projects, than "big"
and "low profile". mpg123 used to be in the "dead" state for some
years, mainly debian and gentoo took over with patching bugs and (few)
features for their own, possible others using that work or doing their
own share, too. Now it's "alive" again, and rather low profile... one
main developer (me) being active, plus one developler concerned with
win32/MSVC porting issues. I do stable releases, I do bugfix updates of
these... I do new stable releases after adding features and testing on
some configurations and for some time. I claim my stable releases to
earn their name - if not, then there's a bugfix release soon.

But, well, when you combine a big bit-rotting codebase with a small
developer team... stakes are high that people do changes that have side
effects they didn't think of and that they don't test for...

Post by Joerg Fischer

Post by Thomas Orgis
Soon there could be noone trying anymore... nedit is already obsolete...

It wasn't only me who predicted this roughly five years ago.

So what is the message here? Actually, I notice the absence of actual
developers in this discussion (people mentioned as developers in the
nedit version info).
Does Tony have an opinion about this (making a release of current
sources, at least)?

Well, I'll try to get some working lesstif to have the nedit as useful
as it gets in its current state, but for the longer term...
... I'll have to see how to push nedit development (get into it myself?
in what spare time?),
... or to choose some other (fast) editor and add the important nedit
features it misses,
... just write my own,
... learn vim,
... get a valium addiction and return to kate?


Alrighty then,

Thomas.

PS: Joerg: Sorry for stressing your patience with this repetitive
thread... the hope doesn't want to die.

[Carlo Graziani]

This is a fascinating thread, and somewhat hopeful to someone who would
grieve for Nedit, were the project to bitrot away into oblivion.

I have a perhaps naive observation. In part, these difficulties appear to
arise from the requirement that there should be _some_ working motif-like
libraries available where Nedit is to be installed. The
licensing/stability issues that attend *tif libraries make this a chancy
proposition.

Would it be out of the question to import the required parts of a
Nedit-friendly release of lesstif into the Nedit source, built as, say, a
smallish library ("neditif")? Presumably this would be accomplished by
extracting lesstif code that is called directly from Nedit, plus anything
downstream from those functions.

I am assuming Nedit does not use more than a fraction of Motif
functionality, although I don't actually know this. If this were true,
then this approach would have the advantage that the Nedit team would only
have to curate the aspects of Motif functionality that concern the
application, which is presumably easier than spelunking for bugs in the
full lesstif distribution.

It would be sort of like forcing a fork of lesstif, and nobody likes
forking. But sometimes forks are necessary. While I do not assume that
this would necessarily be an easy task to accomplish. I am suggesting that
it would have an attractive feature: once the one-time lifting of the
required functionality into neditif has been accomplished, Nedit could get
a divorce from Motif library development and licensing politics, and its
fate could be disentangled from Motif's.

Anyway, there's a suggestion, for what it's worth. Thanks for Nedit, I use
it every day.

Cheers,

Carlo
--
Carlo Graziani [phone] (Voice)
Department of Astronomy and Astrophysics [phone] (FAX)
University of Chicago
5640 South Ellis Avenue | Book Proposal:
Chicago, IL 60637 | "Combined-Arms Assault Crossings of
carlo-***@public.gmane.org | Major Rivers For Dummies"

[Thomas Orgis]

Am Sat, 07 Mar 2009 16:05:18 -0600

Post by Carlo Graziani
Would it be out of the question to import the required parts of a
Nedit-friendly release of lesstif into the Nedit source, built as, say, a
smallish library ("neditif")?

I guess unless you accomplish that yourself, noone wil try it...
Apart from the original nedit authors targetting commercial UNIX
platforms that have working Motif, such afork is a hell of a workload
and most probably that workload is better employed in just fixing the
lesstif that's there.
It is not broken beyond repair... it's some people missing that
actually fix bugs.
Actually, lesstif CVS has quite some fixes since the 0.95 release...
More on the other branch of this thread...


Alrighty then,

Thomas.

[Joerg Fischer]

Post by Thomas Orgis
That's why I asked for your snapshot;-)

If I would be certain to find the right thing after roughly 4 years, I
had no problem sending it.

Post by Thomas Orgis

Post by Joerg Fischer
It wasn't only me who predicted this roughly five years ago.

So what is the message here?

It's a bit late.

Post by Thomas Orgis
Actually, I notice the absence of actual developers in this
discussion (people mentioned as developers in the nedit version
info).

Some time ago I tried to gather some background information on
http://nedit.gmxhome.de/history.html

Understanding the history of nedit, you'll see this was never a Linux
project. It was developed on Unix machines at a time where Linux (not
to mention gtk or any other of the free toolkits) wasn't available.
The creator of NEdit wasn't related to free software, although he did
a lot to make it possible to release NEdit under GPL. (This was as
late as version 5.1.1.)

Most developers are no Linux guys, either. I think even today NEdit is
still more popular on Unix systems than on Linux. This is also related
to the toolkit. One argument of developers was why to switch the
toolkit if it is not readily available on Unix. Perhaps after Solaris
switched to gtk, the chances increase ...

Post by Thomas Orgis
PS: Joerg: Sorry for stressing your patience with this repetitive
thread... the hope doesn't want to die.

What rather stressed my patience was Vista, since the Linux kernel
caused a problem on my new laptop. Well, now after a BIOS update
things are working fine, and I don't feel stressed but just wondering
which desktop to use. Surely not KDE, rather Xfce or even just icewm,
let's see :-)

--Jörg

[Thomas Orgis]

Am Sun, 8 Mar 2009 11:24:06 +0100

Post by Joerg Fischer
Some time ago I tried to gather some background information on
http://nedit.gmxhome.de/history.html

Hm, right, you are the guy who did the great LaTeX mode. I used that
happily before (p.ex. the auto-brackets), but for some reason didn't
bother to install it on my current system.
I must say that I like nedit but think that the macro system feels
clumsy.
The chance to mess up everything with one nedit.rc and many different
code in it, the ugly \n\ at every line end (like gcc inline assembly...
it looks so redundant)... but that really is another topic, and one
where you most probably disagree with me;-)

Post by Joerg Fischer
Understanding the history of nedit, you'll see this was never a Linux
project.

ACK.

Well, I'll give you a short update on the situation I have on Linux now:
I grabbed lesstif CVS, which in fact includes quite some fixes (for
64bit systems, copy 'n paste, some security issues), and can build that
with our patch to remove libXp usage.
Well, copy and paste works, but I have the feeling that menus are even
more messed up.
The X focus has funny behaviour when a menu pops up... they menus don't
react to the keybard (since they don't get proper focus, I guess.
And of course, the Menu->ESC crash is still there.
This mess has to be tackled before one can use nedit on a Free Linux
system again... but at least copy 'n paste works again.

Also, I've seen lesstif commits up to a year ago. So it's not _that_
dead and rotting.
There may be someone who could do a commit to fix things... but
apparently noone bothering to make releases or update the website.
Hm, reminds me a bit of nedit:-/


Alrighty then,

Thomas.

[Thomas Orgis]

Current state: I updated the lesstif bug on sf.net:

http://sourceforge.net/tracker/index.php?func=detail&aid=1578451&group_id=8596&atid=108596

I also tried to send a message to the lesstif-discuss list... doubt if anyone listens, but perhaps...


Alrighty then,

Thomas.