0.1 - 0.2:
* Internal versions.
0.3:
* Works somewhat. Can actually trace some routes, but most sites are
unknown. Also the heuristic for finding the sites' loation is imperfect.
0.4:
* Changed the spheremodel, and added support for NDG cache files,
improving the preformance somewhat. Also fixed about one bazillion bugs.
0.5:
* Now uses gtk instead of GLUT.
0.6:
* Rewrote the database support.
0.7:
* More database restructuring.
0.8:
* Changed the list/scrolled-window hack to the new great clist widget. That
solved quite a few of the things on my TODO list.
* Fixed one bazillion bugs.
* Ported to SGI, should have fixed most of the porting problems.
0.8.1:
* More bugfixes. Only real bad bug that I know of now is the error somewhere
in the string <-> numerical conversion routines for latitudes/longitudes.
* Ported all code to use gtk 1.0 and gtkGL 0.8.9.
0.8.2:
* Compiled and ran on Linux during the day. Found a few bad bugs that
crashed it there. 0.8.2 _should_ run correctly on Linux.
0.8.3:
* Got it to work on SGI again. I never quite got the Makefile to behave
like I wanted. Mostly due to SGI's _very_ picky make. This is sort of moot
since I'm going to use automake/autoconf soon anyway.
* First sort-of-public release.
0.8.4:
* Minor cleanups.
* Added new command-line option: --lod <level-of-detail>
0.8.5:
* Fixed (sort of; now it's as good as it gets.) the texturing bug near
the poles.
* Documentation and licensing updates.
0.8.6:
* Fixed the tar archive so that it won't untar into the current directory
anymore (sorry about that...)
* Done some linting and efencing. Looks good.
* Bugfixes.
0.8.7:
* Fixed a bug that made it eat 100% CPU on some platforms.
* You can now press the spinner to start a new trace, instead of having to
select "file->new" from the menu.
0.8.8:
* Started cleanup of glstuff.c. makeearth is much shorter now, and uses
some static functions.
* Fixed a bug where some sites sometimes didn't get plotted, if they had
ACC_NONE, and the trace timed out.
* The info-button now toggles wether the info window is visible or not.
0.8.9:
* Fixed bug where entries in the info-window was centered. Looked ugly.
* Rewrote, modularized, cleaned up the info-window code.
* Arranged so that selected sites don't go all white, now they just
brighten a bit.
* OK, makeearth looks reasonable now. Got the line-plotting into a separate
function, and that made it a lot shorter.
* Merged in two patches I got from Edouard Lafargu (thanks!), adding
support for a new kind of database, similar to the internal one but
runtime-loadable. Also he's improved the contry-suffix database no end.
0.8.10:
* Fixed bug in the database code. Edouard's new generic database wasn't
used.
* Added the type "site". What was previosly "boeps" is now "sites", an
array of "site"s
* Changed a lot of the code to use site* handles instead of indexes into
"sites". This should make the code for adding Edouard's locate feature
much cleaner.
* Started on some restructuring of the top-level DB code. Nothing big, but
the old system is getting a bit awkward.
* Changed the name resolv() to resolve(), added resolve_by_id(), which works
like the old resolv().
* Removed the word "boep" from the source tree alltogether. It was just
a placeholder in the beginning anyway. This meant renaming some functions.
Also, init_boeps() got changed to clear_sites(), since that what it does.
0.8.11:
* Doesn't print the true distance estimate if the accuracy is too low.
* Swapped the gtkGL widget out in favor of gtkglarea, which is more
actively developed.
* Fixed a few small bugs.
* Inserted some of Edouard's code. The notebook for example.
0.8.12:
* Uses GNU configure/automake. Hopefully it works.
* Has a bugfix for when the first site(s) in the trace doesn't reply.
* removed local_config.h, added tweaks.h instead, which is much smaller
and only contains the obscure performance-tweaks. (Nobody normally needs
to touch it, since all the important configuration happens thru GNU
configure.)
0.8.13:
* Made it work with gtk 1.1.x as well as 1.0.x.
* Fixed loads of small configure-related bugs.
* contries.c now knows the location of many more country-suffixes
(Thanks to Greg Louis)
* Doubleclick in clist to get the info window. (Martin Norb{ck (Norpan))
* Next- and previous-buttons in the info window (Norpan again)
* Changed texture to a correct one.
* Fixed bug where sites with no known location got plotted in the
atlantic, west of africa (at 0,0).
* Fixed typo in configure.in which broke --with-traceroute=PATH
* Added reading of traceroute data via stdin (for debug purposes)
0.8.14:
* Removed fluff from the code.
* Configure fix for SGI. (Look for glCore)
* Fix commandline parsing.
* Fixed annoying bug with Mesa (and some SGIs), where the sitemarkers
behind the globe was still visible. (Henrik Jonsson (Hej))
* Added manpage (Contributed by Stephane Bortzmeyer)
* Managed to merge Pablo Saratxaga's i18n patch, after some fighting
with Suns version of gettext(). This includes language support for
french and spanish and (apparently incomplete) japanese.
* countries.c got a bit more knowledgeable. Now almost complete. (Pablo
Saratxaga)
* Added xtraceroute.desktop file (something to do with Gnome) and a PNG
file for use as an icon (Pablo again)
* Added quaternion rotation stuff, suggested by many.
* Fixed a few small UI bugs/omissions.
* Merged old README and INSTALL files into README, included the standard
automake INSTALL file.
0.9.0:
* Merged new versions of .po files from Pablo.
* the .spec file should more or less autogenerate now.
* Fixed stupid configure bug. (Order of -l<lib> entries on link line.)
* Updated README slightly, added list of prerequisites, pointer to web
site.
* Changed "--stdio" to "--stdin" or just "-". (I don't know why it was
that way. I must have been very tired...)
* Fixed stupidity where "xtraceroute -h" didn't print the version number.
* Unportable function snprintf replaced with g_snprintf.
* Cleanups in db_*_gui.c, also fixed the alignment of some labels in
there.
* Added workaround for Irix compile problems (like -Xcpluscomm) to
configure.in.
* Threw out tifflib and am now using gdk_imlib. The texture is now a
PNG file. This file is 300k smaller than the old TIFF file.
* Fixed some trouble that came from this, with configure not finding it
etc.
* Fixed stupid bug re glxmakecurrent and select mode.
* Removed some ifdefs
* Fixed the elusive "can't select sites from the globe if I haven't
resized the window, sometimes" problem.
* Fixed problem with segfault if you killed the info_window via WM
instead of clicking "OK".
* Tricky bug with input handling fixed by Bert Driehuis.
* It now checks if the $HOME/.xt directry exists before trying to write
to it. (Reported by Timothee Besset)
* Integrated St廧hane Genaud's cool patch, giving me (in his own words):
>-------------------------------------------
>St廧hane Genaud (genaud@icps.u-strasbg.fr)
>-------------------------------------------
>* Moved the menu bar creation to the new build_menu() function
> that uses an ItemFactory. Changed Callbacks prototypes like addHost :
> they now return void instead of gint to prevent gcc warnings due to
> expected types in ItemFactory.
>
>* Added a "View" menu providing zooming in/out
> Zoom affects earth and sites but not the line respresenting the route
> (FIX ME)
>
>* Middle button lets you translate the world in the x-y plane
>
>* Transparency of sites. Seems useless and ugly at the moment but
> can be the base for some tests.
* Integrated Martin Schenk's patch, meaning you can zoom with the mouse.
* (pre-3 released here)
* Globe is now centered on user's location upon startup.
* (pre-4 released here)
* Replaced Imlib with gdk-pixbuf.
* Fixed error in the texture-coordinate-generation in sphere. The
texture would appear upside-down.
* Both 3 (RGB) and 4 (RGBA) component images work as textures now.
* RFC1876 baby! Via pipes and the "host" command. The unix way...
* The rfc1876 stuff is more general, I can basically hang in any data
source and have it work asynchronously.
* Satellites!
* Uses pipe/fork/exec instead of popen, even for traceroute.
* Major #include cleanup.
* Now gives traceroute "-m 40" for max hops. Let me know if it isn't
enough...