darwin2k-devel Mailing List for Darwin2K (Page 2)
Status: Beta
Brought to you by:
xrayjones
Menu
▾
▴
darwin2k-devel — A list for developers of Darwin2k
You can subscribe to this list here.
| 2000 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
(2) |
Oct
(2) |
Nov
|
Dec
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2001 |
Jan
(7) |
Feb
(2) |
Mar
(4) |
Apr
|
May
(1) |
Jun
(4) |
Jul
(1) |
Aug
|
Sep
|
Oct
|
Nov
(1) |
Dec
(2) |
| 2002 |
Jan
|
Feb
(1) |
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
(1) |
Sep
|
Oct
(1) |
Nov
(1) |
Dec
|
| 2003 |
Jan
|
Feb
|
Mar
(3) |
Apr
(1) |
May
(3) |
Jun
(1) |
Jul
|
Aug
(1) |
Sep
(6) |
Oct
|
Nov
|
Dec
|
| 2004 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
(1) |
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2007 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
(1) |
Dec
|
|
From: Christopher L. <ch...@MI...> - 2001-05-22 15:49:18
|
Hi -- In getting my new XML stuff to work within the Darwin2K tree, I was forced to use automake conditionals (for the first time) to get it all to work. I found that the AM_CONDITIONAL() approach was much better than making all those EXTRA_* variables in the Makefile.am's and defining lots of AC_SUBST() variables in configure.in to customize the build. So after trying to fix the bug in configure.in that Priyan found, I went through and updated the rest of Darwin2K's build system to use automake conditionals. The new configure.in and Makefile.am's are in CVS. I think this makes the configuration/build system much more understandable and hopefully more maintainable. I'd appreciate having this tested-out on some different systems to see if anything is broken which I can't simulate here. -Chris -- Christopher Lee <ch...@mi...> (617) 253-5095 <http://web.mit.edu/chlee/www> Postdoctoral Associate of Mechanical Engineering Field and Space Robotics Laboratory Massachusetts Institute of Technology 77 Massachusetts Ave, Room 3-472m Cambridge, MA 02139 |
|
From: Chris L. <cl...@ea...> - 2001-03-20 06:09:50
|
Hi, In case any of you have tried to get in touch with me and are wondering why your mail has bounced, our email server (actually, our entire subnetwork) at JPL has been down for the last week. I'm not sure when it will be back up, but in any case I'm reachable at cl...@ea.... Chris |
|
From: Adam R. <ada...@ya...> - 2001-03-12 19:53:30
|
--- Chris Leger <cl...@ea...> wrote: > Here are two files that should fix the problem you > had. Hi Chris, The updated file src/d2k/db/comp_grammar.yxx solved my compile problem. I replaced the other file as well, though that was not necessary. I could compile now, and looking forward to the end of the workday, so I can try it out. Thanks for the fix! cheers, Adam __________________________________________________ Do You Yahoo!? Yahoo! Auctions - Buy the things you want at great prices. http://auctions.yahoo.com/ |
|
From: Chris L. <cl...@ea...> - 2001-03-09 17:23:47
|
Hi, Here are two files that should fix the problem you had. One is src/d2k/db/comp_grammar.yxx, and the other is src/d2k/pm/df_grammar.yxx. Let me know if they work for you. Chris |
|
From: Adam R. <ada...@ya...> - 2001-03-09 05:46:52
|
--- Chris Leger <cl...@te... > Wow, I didn't know I even had a tracker option for Darwin2K! Yes, I know the feeling! That's why I sent an email as well. > could you send me the comp_grammar.cxx that > is created in your build directory, in src/d2k/db/ ? Sure. I tried with the newer version, 0.85, but it's the same. The error message is below, and I attach the comp_grammar.cxx and comp_grammar.yxx files. thanks, Adam make[4]: Entering directory `/home/adam/darwin2k-0.85/src/d2k/db' byacc -d -pcompDB comp_grammar.yxx && mv y.tab.c comp_grammar.cxx if test -f y.tab.h; then \ if cmp -s y.tab.h comp_grammar.h; then rm -f y.tab.h; else mv y.tab.h comp_grammar.h; fi; \ else :; fi c++ -DHAVE_CONFIG_H -I. -I. -I../../.. -I../../../src/ddcore/misc/ -I../../../src/ddcore/matrix/ -I../../../src/d2k/misc/ -I../../../src/dyno/mechanism -I../../../src/d2k/modules -I../../../src/d2k/metrics -I../../../src/d2k/cfg/ -g -O2 -c comp_grammar.cxx comp_grammar.yxx: In function `class componentDB * parseComponentFile(const char *, int = 0)': comp_grammar.yxx:221: implicit declaration of function `int compDBparse(...)' make[4]: *** [comp_grammar.o] Error 1 __________________________________________________ Do You Yahoo!? Get email at your own domain with Yahoo! Mail. http://personal.mail.yahoo.com/ |
|
From: Christopher L. <ch...@MI...> - 2001-02-01 19:16:30
|
On 01 Feb 2001 09:31:56 -0800, Chris Leger wrote: > Those changes sound good. The command I've been using for doc++ is: > > doc++ -f -d darwin2k-docs --html `find . -name '*.h'` > > So, I guess yours would be > > doc++ -f -d $(top_builddir)/html --html `cat doc++.file` OK -- This is basically what I do now, but my results don't look as nice as yours so far. You can try a 'make doc++' to compare the results, anyway. > > World domination by > > killer robots will inevitably follow, I'm sure. > > I'll make sure you're among the spared by my army. Similar immunity > from yours would be appreciated on this end. :) Of course -- with the obvious caveat that they may decide not to listen to me and kill everyone with impunity. We can always hope, though.... -- Christopher Lee <ch...@mi...> (617) 253-5095 <http://web.mit.edu/chlee/www> Postdoctoral Associate of Mechanical Engineering Field and Space Robotics Laboratory Massachusetts Institute of Technology 77 Massachusetts Ave, Room 3-472m Cambridge, MA 02139 |
|
From: Christopher L. <ch...@MI...> - 2001-02-01 02:51:36
|
Chris --
I've made a first draft at a small patch for automating doc++ file
generation through the makefiles. Note that it doesn't use any special
trick of automake -- it just adds an extra make rule or two. I can add
configuration smarts later. My documentation looks less nice than
yours, so I am assuming you have a better way of invoking doc++ than I
do (I just downloaded it and tried it for the first time myself). I'm
including a diff, and will commit the changes if you give the OK. I
have no problem re-doing it if you'd prefer things done differently.
How it works:
* Only *.h files are checked for doc++ information so far -- I don't
know if documentation might be found in *.cxx files too, but this is
easily changed.
* Put a comment starting with '/*MAKEDOC++' in a *.h file to have it
included in the set of files which are processed by doc++.
* Running 'make doc++' in the top-level directory will
* scan all *.h files for the trigger comment and make a list of
these files in "$(top_builddir)/doc++.files"
* run doc++ on the files in $(top_builddir)/doc++.files" placing the
generated html files in directory "$(top_builddir)/html" (I assume
the doc++ command is what needs to be made more like what you use
yourself)
* I haven't yet added the hooks which will install the generated
documentation in a global documentation directory such as
${prefix}/share/doc/darwin2k/html/, and pre-build the documentation to
automatically include it in the distribution, but this is easy when
everything else is to your liking.
I think our work on the NASDA project will really get serious now that
our visiting scientist has arrived from Japan, and the turborover
project is getting to the point where serious simulation is needed, so I
expect to be teaching the lab all about D2K very soon and providing the
world with a few D2K users. Your schedule for getting the new dynamics
engine in working order is amost perfect for us. World domination by
killer robots will inevitably follow, I'm sure.
-Chris
--
Christopher Lee <ch...@mi...>
(617) 253-5095 <http://web.mit.edu/chlee/www>
Postdoctoral Associate of Mechanical Engineering
Field and Space Robotics Laboratory
Massachusetts Institute of Technology
77 Massachusetts Ave, Room 3-472m
Cambridge, MA 02139
|
|
From: Chris L. <cl...@te...> - 2001-01-31 17:50:17
|
Hi, Takayuki asked me the other day if there was a tool for analyzing source code to give the relationship between different C++ classes. I didn't find anything in that vein, but I did find a tool called Doc++ which can generate documentation from commented source code. I've started reformatting the source code for doc++, and I have put the resulting documentation at: http://darwin2k.sourceforge.net/darwin2k-docs/index.html It's definitely not complete, but it's a good start and I will be updating it on a daily basis. One note of caution, though: this is for the next release of Darwin2k, which has all of the dynamic simulation stuff separated into a separate package. The high-level stuff is still the same (configurations, modules, and so on), but everything in the src/d2k/mechanism directory has been moved to src/dyno/mechanism, and all of the dynamic simulation code in the src/d2k/cfg directory has been moved there as well. I hope to have the new version released this week, as some people here at JPL are waiting on it. Chris -- [ Chris Leger cl...@ro... ] [ Phone: (818) 393-4462 Fax: (818) 393-3254 ] |
|
From: Christopher L. <ch...@MI...> - 2001-01-29 18:57:12
|
On 29 Jan 2001 10:35:24 -0800, Chris Leger wrote: > I was pleasantly surprised to find out about doc++, a tool for > automatically generating documentation from C, C++ & Java source code. > It's very easy to use: you just format your comments a certain way, and > then run doc++ to produce html or LaTex documentation. Here is a sample > directory with some of the generated documentation: > > http://darwin2k.sourceforge.net/documentation/html/ This looks like it will be very useful! I find this kind of automatically-generated online documentation very helpful when I code using Gtk and Glib libraries, and I wish some other projects did their development documentation similarly (e.g., OpenGL libraries). > I installed it from an RPM off of rpmfind.net. I'll 'apt-get install ' it myself :). > Chris, can you modify the automake scripts to include rules for > generating documentation? Don't worry about modifying all the > Makefile.am's; I can do that, but I don't have a clue how to go about > specifying rules for automake. Sure -- I'll look into it. It should be simple. > Also FYI, I'm working on getting the dyno code back into Darwin2K; I > probably won't commit anything until I can at least get the entire > system to compile. It looked like this might happen based on your recent cvs commit messages. I'm trying to get to my stuff, but am currently writing proposals in work-time and house-chores in nonwork-time. I'll get some coding done before too long because proposal-writing by itself will make me insane. -Chris -- Christopher Lee <ch...@mi...> (617) 253-5095 <http://web.mit.edu/chlee/www> Postdoctoral Associate of Mechanical Engineering Field and Space Robotics Laboratory Massachusetts Institute of Technology 77 Massachusetts Ave, Room 3-472m Cambridge, MA 02139 |
|
From: Chris L. <cl...@te...> - 2001-01-29 18:38:32
|
I was pleasantly surprised to find out about doc++, a tool for automatically generating documentation from C, C++ & Java source code. It's very easy to use: you just format your comments a certain way, and then run doc++ to produce html or LaTex documentation. Here is a sample directory with some of the generated documentation: http://darwin2k.sourceforge.net/documentation/html/ I haven't finished converting the comments in that file, and there are a few bugs, but it gives you an idea of what doc++ can easily do. I'll be gradually doc++-ifying the code, which should make things less painful for everyone. The URL for doc++ is http://www.zib.de/Visual/software/doc++/index.html I installed it from an RPM off of rpmfind.net. Chris, can you modify the automake scripts to include rules for generating documentation? Don't worry about modifying all the Makefile.am's; I can do that, but I don't have a clue how to go about specifying rules for automake. Also FYI, I'm working on getting the dyno code back into Darwin2K; I probably won't commit anything until I can at least get the entire system to compile. Chris -- [ Chris Leger cl...@ro... ] [ Phone: (818) 393-4462 Fax: (818) 393-3254 ] |
|
From: Christopher L. <ch...@MI...> - 2001-01-26 22:59:13
|
I think it is fine to keep the dynoDisp stuff separate from the d2k/disp for now, eventually it would be nice if the d2k stuff used the same multi-purpose modules as everything else, just so that if code was improved for one purpose, all d2k/dyno-derived projects would benefit. I haven't gotten too much done yet, but if I speed up then I'll be thinking of how to beef up the graphics in dynoDisp and make it useful for multiple purposes. I'm glad to hear that other groups at JPL are interested in the code -- the more people using it there the more good excuses you have to hack. -- Christopher Lee <ch...@mi...> (617) 253-5095 <http://web.mit.edu/chlee/www> Postdoctoral Associate of Mechanical Engineering Field and Space Robotics Laboratory Massachusetts Institute of Technology 77 Massachusetts Ave, Room 3-472m Cambridge, MA 02139 |
|
From: Chris L. <cl...@te...> - 2001-01-26 19:16:26
|
I did'nt have any problems working on things yesterday or today, as far as your changes go. I haven't quite decided what to do about the display code for Darwin2K after I complete the separation of Dyno--my original plan was to keep the d2k/disp directory as the main display code, since there are other things besides links and joints to display, though it would be possible to just subclass the dyno display code, which I originally intended for use by people who are only using dyno, not the whole darwin2k package. It looks like there will be two other groups here at JPL using Darwin2k and/or Dyno; there've also been a few people from the net who've gotten it up and running after some recent changes. After I finish the dyno split-off, I think I may have to break down and work on documentation again... Chris -- [ Chris Leger cl...@ro... ] [ Phone: (818) 393-4462 Fax: (818) 393-3254 ] |
|
From: Christopher L. <ch...@MI...> - 2001-01-26 18:46:57
|
Hi --
I finally committed my changes to split the XForms code out of
dynoGLDisp.{cxx,h} and subclassing it in dynoFormsGLDisp.{cxx,h}. It
looks like mtestGL runs OK after the change, so I hopefully didn't mess
anything up in the transition. Now that this is done and my computer is
more-or-less fixed, I should be able to write dynGtkGL.{cxx,h}. I
realize that I updated the Makefile.am in dynoDisp before I added the
changes to dynoGLDisp.{cxx,h} -- I hope that didn't mess-up compiles too
much yesterday. I also made a slight change to mtestGL to try to make
it toolkit-agnostic.
-Chris
--
Christopher Lee <ch...@mi...>
(617) 253-5095 <http://web.mit.edu/chlee/www>
Postdoctoral Associate of Mechanical Engineering
Field and Space Robotics Laboratory
Massachusetts Institute of Technology
77 Massachusetts Ave, Room 3-472m
Cambridge, MA 02139
|
|
From: Christopher L. <ch...@MI...> - 2001-01-11 16:35:29
|
Hi -- I just saw this at Sourceforge and thought I would mail it to the developers list so that the development email archives will show activity for January '01 (and so that the other developers will see it too): Posted By: xrayjones Date: 2001-Jan-10 10:37 Summary:0.83 released Version 0.83.0 includes automake/autoconf support, the ability to save animations, and dlopen() support for dynamic libraries. Get the latest version at http://download.sourceforge.net/darwin2k BTW: I'm another Chris L. from CMU (a.k.a. und...@us... but that mail address isn't working yet), and am interested in using D2K, especially the new dynamics core libraries XRayJones is developing, for simulating rovers and free-flying robots. I just showed Chris' d2k demos to my lab group and a space robotics researcher at NASDA (Japanese space agency) and they are all impressed and interested. I wrote autoconf/automake scripts for D2K, and may write Gtk/libglade interfaces to some of the current programs in Darwin2K to make it a little more portable (and consistant with the GPL). I know XRayJones is working on a new dynamics core for Darwin2K --- does anyone else have specific development plans underway? -Chris -- Christopher Lee <ch...@mi...> (617) 253-5095 <http://web.mit.edu/chlee/www> Postdoctoral Associate of Mechanical Engineering Field and Space Robotics Laboratory Massachusetts Institute of Technology 77 Massachusetts Ave, Room 3-472m Cambridge, MA 02139 |
|
From: karl r. <kar...@do...> - 2000-10-27 00:36:07
|
sorry, ive gotten horribly busy recently and havent had a chance to get much of anything done.... karl ____________________________________________________________ woof! woof! I'm @dog.com Get your free @dog e-mail at http://www.dog.com |
|
From: Chris L. <cl...@he...> - 2000-10-26 23:05:32
|
Hi, I've made lots more changes lately to Darwin2k; you can update or grab the latest tarball if you're interested. I've been contacted by Peter Bentley, author of "Evolutionary Design by Computers". He said he'd like to include Darwin2K on the CDROM for his next book, and he needs the data by mid-November. It'd be good if the automake/autoconf stuff was ready by then; any chance you'll have time for that, Francesco? Mark, the robot design GUI would be cool to include too, but I'm not sure how realistic that is schedule-wise--if you think you'll have time, let's talk and hash out some plans for the tool. Chris |
|
From: Chris L. <cl...@he...> - 2000-09-27 01:04:38
|
Ok, I finished the installation instructions (see "INSTALL" in the top-level directory) for the current system, and got the demos working reasonably w/ OpenGL. (Instructions for running the demos are also in the INSTALL file.) Francesco, I think the INSTALL and README files should contain most of the general information you'll need as far as what libraries are needed for the system. Also, the Imakefiles are set up to build the standalone targets first, so that if you don't have RTC (the communications package) you can still run standalone. I talked to the author of RTC, and he gave me the OK to distribute RTC headers, libraries, and binaries with Darwin2K. He may also release it via sourceforge, which would be good--it's an extremely robust, efficient, and easy-to-use package. Anyway, that's it for now! Chris |
|
From: Chris L. <cl...@he...> - 2000-09-25 22:41:37
|
Hi, I've put a first-cut INSTALL file in the d2k/ directory; if you do a 'cvs update darwin2k', you should get all the updated files. As Francesco suggested, I've also made a branch for the version of the code that uses Imakefiles, as there may be extensive changes once automake/autoconf is fully in place. Chris |
4 messages has been excluded from this view by a project administrator.
