I'm working on some deallocators now. The idea is that when Eterm
exits and memory debugging is on, several routines get called to free
the in-use memory (menus, font cache, etc.) that we still know about.
Anything left after that would be either unavoidable leaks (like
environment variables...read the putenv() man page sometime...sigh)
or genuine memory leaks that need fixing. I'm down to about 4.5K of
leftover malloc'd memory now. Making progress....
SVN revision: 3295
Massive reorganization/rewrite to libmej. It should now be 100%
independent of Eterm. There still may be some gremlins in the memory
debugging code, so don't use too high a number with --debug....
SVN revision: 3282
Okay, first off I need to thank Marius Gedminas <mgedmin@takas.lt>.
Not only did he point out several issues, he was even willing to send
a patch. :-)
This commit includes Marius' patch which keeps MappingNotify events
from overriding users' modifier settings. It also includes a fix to
a seg fault that he pointed out when menus were loaded without the
menu imageclass having been defined.
He also pointed out that --pause tends to sit and spin, taking up 100%
of the CPU time. Funny how E-Cpu never seemed to get too bent out of
shape over it, but he was indeed correct.
Thus I have fixed it, and while doing so, I have changed the way it
works. It no longer waits for a keypress per se. Actually, it just
ignores the fact that its child went away and keeps right on taking
X events. There are a few exceptions though. Either ESC or Ctrl-C
will exit a paused Eterm. Any other input that doesn't have special
meaning to Eterm will be ignored. (Shift-PgUp and Shift-PgDn still
work however, as do any action bindings you may have.)
SVN revision: 2739
Added new shade/tint routines, including some done in MMX assembly, by
Willem Monsuwe <willem@stack.nl>. Thanks Willem! These should prove
to be faster than the old stuff, probably by quite a bit.
SVN revision: 2595
Fixed a bug in the handling of resize events thanks to a tip from
Teodor Zlatanov <tzz@iglou.com>. While I was at it, I also fixed a
couple remaining issues with the popup menu system. You shouldn't
get any more cases where items remain selected after you've moved your
mouse off of the menu, or where a submenu would stay up even though
another menuitem is selected.
SVN revision: 2594
After getting sick of hearing people whining about the obscure error
messages that resulted from trying to build without pixmap support, I
hacked it until it at least built without it. I also added warnings
to configure.in so that people would know if that happened.
It's still not very happy running that way, which is something I need
to fix. But hopefully it will stop the whining.
Yeah, right. =P
SVN revision: 2590
I fixed the problem reported by Joakim Bodin <bodin@dreamhosted.com>
with spawned processes. Also fixed Esetroot as reported by Wayne
Johnson <bigman1@alltel.net>.
Support for buttons higher than 5 was also added to the action
bindings code for those (like me) with IntelliMouse Explorers and
XFree86 4.0. :-)
SVN revision: 2494
Quick little option -q/--no-input. It keeps Eterm from accepting
keyboard input, and keeps the window manager from focusing it. Useful
for log tailers and such, maybe. This feature was requested by
Peter Ward <than@ilm.com>. I will be adding an escape sequence to
toggle this.
SVN revision: 2398
I added accelerated scrollwheel support. I also added a #define in
src/feature.h for specifying the number of lines of context to keep
when paging up or down.
Also, we're back to the old behavior where home on input won't be
triggered unless the key pressed has an associated string value. I
went back on this because Shift causing a jump to the bottom was
really annoying when trying to Shift-PgUp or Shift-PgDown. Sorry,
Darren, but you lose.
SVN revision: 2270
Some further fixes for inline functions, 2 new winop actions, brand
new and improved profiling macros, some miscellaneous fixes for SGI
from Kimball Thurston <kimball@sgrail.com>, and more robust checking
in the pasting code.
SVN revision: 2235
The buttonbar can now be toggled on and off both in the config file
and via an escape sequence. The themes in CVS use Ctrl-Shift-Button3.
You can also specify in the config file whether to dock the buttonbar
at the top or the bottom of the Eterm window. You can't move it on
the fly yet, but that will come.
I also fixed resizing so that the term window didn't redraw itself
unnecessarily. Hopefully I didn't break anything in the process. :-)
Plus, I fixed poor handling of X-generated ConfigureNotify events, and
the terminfo stuff is now done at install time instead of build time.
SVN revision: 2077
This is the first public availability of the work thus far on Eterm
0.9.1. There's quite a bit of new stuff here.
* Added scrollbar thumb support.
* Completely redid the terminfo/termcap stuff. The terminfo file is
now compiled (by tic) and installed by default (unless you specify
--without-terminfo). The config files still say xterm, though,
because some programs (like SLang and GNU mc) use the silly algorithm
of "Is $TERM set to xterm?" to detect mouse reporting support in a
terminal. =P But if you don't ever use xterm, you can use Eterm's
termcap and just name it "xterm" instead. Thanks to Marius Gedminas
<mgedmin@takas.lt> for his patch that started this whole revamp.
* Added the kEsetroot script for KDE users from Dax Games
<dgames@isoc.net>.
* You can now configure the Home and End emulation via --with-home=
and --with-end= options to configure. The --with-terminfo option is
also new, and --enable-xim is now the default.
* Added a new image state, disabled, for when Eterm loses focus. This
is supported by all widgets (well, all those that could possibly be
on screen when Eterm lost focus), even the background image. So you
could actually have all your images darken on focus out and restore
to normal on focus in.
* Widget colors formerly dealt with as colors (menu text color,
scrollbar color, etc.) are now handled by the imageclasses. Each
image state can have a foreground and background color defined. The
current exception is the background image; I hope to add that later.
The foreground is the text color and the background is the object
color (for solid color mode). So menu text color is set by the menu
imageclass. And again, for unfocused colors, use the disabled state
of the imageclass.
* Proportionally-spaced fonts are now handled much better. They are
still forced into evenly-spaced columns (it's a terminal for crying
out loud!) but at least you don't end up with Eterm's wider than your
screen. :-)
* Home on refresh is gone, as is home on echo. It's now much simpler.
There are two options: home on output, and home on input, the former
being a combination of echo and refresh. Also, keypresses that don't
necessarily have corresonding output can trigger a home on input,
like Ctrl-End or whatever...ones that don't have special meaning.
Credit to Darren Stuart Embry <dse@louisville.edu> for pointing out
this issue and the one with "m-" in font names.
* I finally got around to re-merging the new parser stuff from my
work on the Not Game. Closed up some old potential behavior quirks
with theme parsing.
* Added a new escape sequence to fork-and-exec a program. Also added
a scrollback search capability to highlight all occurances of a string
in your scrollback buffer. Use the new "Etsearch" utility to access
it. "Etsearch string" to search for a string, then "Etsearch" by
itself to reset the highlighting.
* And of course, the biggie. Eterm now supports a completely-
customizeable buttonbar. Not a menubar, a buttonbar. It can have an
arbitrary number of buttons, and each button can perform an action,
just like a menuitem. So a button could bring up a menu (like a
menubar) or launch a program (like a launchbar) or perform an
operation (like a toolbar). Each button can have an icon, text, or
both. And you can have buttons left- or right-justified in the
buttonbar. You will eventually be able to have an arbitrary number
of buttonbars, but I'm still working on that.
As with any change this big, things could very easily be broken. So
beware. :-) I have tested this myself, and everything seems to work,
but I can't test every possibility. Let me know if you find anything
that's broken, and enjoy!
SVN revision: 2048
I fixed a bunch of double-buffering stuff here. I'm going to use it
by default on my own Eterms for awhile, and if I don't hit any snags,
I'll make it the default in the themes.
SVN revision: 1708
I finally got around to implementing double-buffering, although it
seems to still have some issues with font changes. But if you don't
change fonts, it works great. :-)
I also fixed the multibyte font stuff with help from Sung-Hyun Nam
<namsh@lgic.co.kr>. There seem to be some new issues here, though,
with the background pixmap. But I'm to tired to look deeper tonight.
SVN revision: 1588
A good number of changes here. First off, since nobody reported any
bugs with the new font stuff, I switched the multibyte fonts over to
use it as well. They do use the same font index, however, in order to
keep the sizes matched up.
I also fixed up the modifier stuff so that Meta and Alt are matched
by KeySym rather than assuming Mod1. I also took care of the action
dispatcher so it would keep up with these changes.
To go along with this, I added 3 new options and config file
attributes which allow you to set the modifier that should represent
Meta, Alt, and NumLock. This overrides the automatically-detected
X server settings.
I also applied some fixes to the XIM code from Sung-Hyun Nam
<namsh@lgic.co.kr>.
SVN revision: 1482
Once again, I've rendered old themes obselete. :-)
I added a new config file attribute and command-line parameter. The
option is --default-font-index, but I wouldn't necessarily use it.
The config file attribute makes more sense. :-)
Anyway, your themes will now need to have a line like this:
font default <index>
in the attributes section. This tells Eterm which font it should use
on startup. (<index> is a number between 0 and the highest-numbered
font you define.) You can now have up to 256 fonts. Font 0 is no
longer necessarily the default font; it is the smallest font. And the
larger the font index, the larger the font should be. (Of course,
this assumes you want Ctrl-> and Ctrl-< to increase/decrease your font
size. In reality, you can have your fonts in any order, and those
keys will cycle through them in order.)
Before, font 0 was always the default, and you didn't have much
freedom in rearranging your fonts. Plus, you were limited to 5. Not
any more. :-) The new system is much more straight-forward, logical,
and powerful.
So please be sure to update your themes by hand, or remove your theme
directory before installing this new version. If your theme lacks
the "font default" line, your Eterms will start with the wrong font.
:-]
SVN revision: 1344
Okay, I think I have everything back in working order. PLEASE test
this thoroughly, as an awful lot has changed. Let me know what I
broke. :)
SVN revision: 1339
I've rewritten almost the entire scrollbar. It mostly works, but it
still has issues. I've put an #error directive in so that it won't
compile; take it out if you *really* wanna try it. It's much more
efficient than before.
SVN revision: 1333
Okay, I've fixed a lot here. First off, I fixed the bug Gnea
<gnea@rochester.rr.com> pointed out with toggling Max Size in E.
Turned out to be a symptom of a much larger problem, but it should be
fixed now. Then I corrected the XIM issues pointed out by Sung-Hyun
Nam <namsh@lgic.co.kr> and Jerome De Greef
<jerome_degreef@hotmail.com>, as well as a stupid typo on my part that
Jerome happened upon. Next I fixed the bug Adam Lucas
<ALucas@wcom.net> spotted with changing the scrollbar type. While
fixing that, I also fixed changing the scrollbar width. Both
operations are a lot smoother and cleaner now.
SVN revision: 1292
I fixed the popup scrollbar bug with transparency, but I fixed lots
more than that. I removed a lot of duplicate event handling, lots of
duplicate redraws, and lots of unnecessary transparency updates. In
doing so, I fixed the latency problem people had been reporting with
several shaded/tinted transparent Eterms while changing desktops. I
also threw solid color transparency support in the mix while I was at
it. All in all, you should notice significant speedups in transparent
Eterms.
SVN revision: 1031
Lots of cleanups here, including a couple fixes for bugs noticed by
Tom Gilbert <gilbertt@tomgilbert.freeserve.co.uk> and Martin Tyler
<martin@boo.org>. Also removed the requirement of glibc 2.1 for using
SVR4-style pty's (/dev/pts/*) under Linux.
SVN revision: 798
Shaved off around 100 KB of memory usage per Eterm by moving more
code into the shared library, removing lots of unneeded variables, and
fixing some small leaks here and there.
SVN revision: 710
Auto mode has been fixed to work with the newest E. If you don't have
the newest E, it won't work. So don't bug me about it; just upgrade.
Your version of E must be dated October 7th or later, or all bets are
off.
I also cleaned up some stuff to save lots of unnecessary redraws and
(if you're in auto mode) E IPC transactions.
A bug in the value set for $IFS was also fixed. Thanks to Kimball
Thurston <kimball@sgrail.com> for pointing out my oversight.
Furthermore, I added a parameter to the --enable-multi-charset option
for configure. You can now specify =utf-8 to get UTF-8 fonts for the
multi-byte mode stuff. If you pass =kanji (or no value at all), you
will still get the old Kanji fonts. If you use multi-byte Eterm and
have a set of fonts for other setups (Big 5, Cyrillic, etc.), please
let me know and I'll add a parameter for those too.
SVN revision: 706
Lots of changes here. First off, this should fix the background draw
bug with transparency that several people pointed out. While I was
at it, I also cleaned up a lot of other related stuff. Three-state
images should be a lot more robust now.
Then again, some stuff may be broken entirely from this, so let me
know. :-)
For one thing, the various image modes should work as expected now.
You can allow and disallow modes for the various widgets. The
fallback mode is "solid" now, rather than "image," so you can cause
a certain widget to refuse to use an image if you want to. If you
specify an image without specifying a "mode" line that allows the
"image" mode, your image will not appear. <-- READ THIS TWICE! I
had to go back and fix all the theme files because of this, so you
will need to remove your current theme directory and allow Eterm's
"make install" to put the new ones in place; otherwise, everything
will go back to being solid colors. =]
Anytime something changes this drastically, there are bound to be
problems. Let me know if you find any of them. :)
SVN revision: 348
Lots of changes here. First off, this should fix the background draw
bug with transparency that several people pointed out. While I was
at it, I also cleaned up a lot of other related stuff. Three-state
images should be a lot more robust now.
Then again, some stuff may be broken entirely from this, so let me
know. :-)
For one thing, the various image modes should work as expected now.
You can allow and disallow modes for the various widgets. The
fallback mode is "solid" now, rather than "image," so you can cause
a certain widget to refuse to use an image if you want to. If you
specify an image without specifying a "mode" line that allows the
"image" mode, your image will not appear. <-- READ THIS TWICE! I
had to go back and fix all the theme files because of this, so you
will need to remove your current theme directory and allow Eterm's
"make install" to put the new ones in place; otherwise, everything
will go back to being solid colors. =]
Anytime something changes this drastically, there are bound to be
problems. Let me know if you find any of them. :)
SVN revision: 345
Worked around a XF86 3.9.16 (perhaps Xinerama) problem. Also added
some XIM changes from Sung-Hyun Nam <namsh@lgic.co.kr> to handle
buffer overflows.
SVN revision: 274
Michael Jennings