1 SimGrid (2.94-cvs) UNRELEASED
7 - Stop layering gras over msg?
8 - Pre-compilation of data_send/recv?
9 - Exception revolution?
10 All of them seem to be post-SG3 issues.
12 Rather finalize what's underway and finish the Big Cleanup:
13 - Check what can be cut from the xbt
14 - "pull" code diffusion in graspe
15 - document the gras' timing macro
22 * Update the main page and the FAQ. Adding references to gforge.
25 * Add a gras_os_getpid function.
28 * Add MSG_task_get_compute_duration() and MSG_task_get_data_size()
29 * Extend the logs so that they also print PID, hostname, date, ... if
31 * Convert the MSG example to the use of xbt_logs instead of PRINT_MESSAGE,
32 and kill the old version which were in testsuite/
33 * Rewrite tools/MSG_visualization/colorize.pl for using with logs instead
37 * Add xbt_os_time(). As the rest of xbt/portability, this is not public
38 for users. Instead, each programming environment (GRAS, MSG,...) use it
39 when needed to provide such a feature to users.
40 Don't shortcut the mecanism or you will also shortcut the virtualization
41 you need on the simulator.
45 * Cleanups in configury with regard to compile optimization/warning flags.
46 Also add -fno-loop-optimize to any powerpc since it's the optimization
47 killing gcc (< 3.4.0).
48 * Doxygen cleanups: move MSG examples, kill the second Doxygen phase
49 needed by MSG examples complications
50 * Borrow configury beautifications from PHP
53 * Bugfix: XBT_LOG_NEW_DEFAULT_CATEGORY now compiles without compiler
54 warning (thanks loris for stumbling into this one).
57 * Change SIMGRID_INSTALL_PATH to GRAS_ROOT in Makefiles generated for user.
58 * Rename gras_get_my_fqdn to gras_os_myname and implement it in the simulator
59 RL would imply a DNS resolver, which is *hard* to do in a portable way
60 (and therefore delayed).
64 SimGrid (2.93) unstable; urgency=low
66 Alpha 4 on the path to SimGrid 3 (aka the "neuf-trois" version)
69 - Use Paje properly where used. Still to be sanitized properly.
70 - Portability fix: Add an implementation of the contexts using pthread
74 - Add xbt_procname(): returns the name of the current process.
75 Use it to show the current process's name in all logging.
77 - fix detection of older flex version and the reaction, since we do
78 depend on modern ones (we use lex_destroy)
79 - Better separation of SG and RL in the libs: remove all simulation code
80 from libgras. As a result, this lib is now only 200k when stripped.
81 Some of the xbt modules may also be duplicated (two sets and such) and
82 should be cleaned/killed before SG3.
83 - Insist on using xlC on AIX because of weird problems involving gcc there.
84 - Cleanup the make remote stuff. This is now done by scripts
85 tools/graspe-{master,slave} (GRAS Platform Expender). This is still
86 mainly for our private use, but we're working on changing them to user
89 - Bugfix: flush the socket on close only if there is some *output*.
90 - Bugfix: flush idempotent when there's nothing to send (don't send size=0)
92 - Add MSG_task_get_name. The task names are mainly for debugging purpose,
95 -- SimGrid team <simgrid2-users@listes.ens-lyon.fr> Fri, 4 Mar 2005 14:32:37 -0800
97 SimGrid (2.92) unstable; urgency=low
99 Alpha 3 on the path to SimGrid 3
103 - New! First try of benchmarking macros.
104 - New! First try so that gras_stub_generator generate deployment and
105 remote compilation helpers.
107 - Bugfix: Initialization fix in msg_test.
111 - Bugfix: applied patch to lexer so that it doesn't need a huge heap.
113 - Bugfix: let dicts work with NULL content (_foreach didn't) and cleanups
115 - API Change: gras_os_sleep to take the amount of seconds as a double.
116 Accepting an int was error prone since it was the only location where
117 seconds were coded as such. It leaded to damn rounding errors.
118 - Bugfix: Hard to belive that timers ever worked before this.
120 -- SimGrid team <simgrid2-users@listes.ens-lyon.fr> Wed, 23 Feb 2005 22:09:21 +0100
122 SimGrid (2.91) unstable; urgency=low
124 Alpha 2 on the path to SimGrid 3
128 - Bug fix in the lmm_solver.
130 - New! Interface to Paje (see http://www-id.imag.fr/Logiciels/paje/)
131 through the function MSG_paje_output.
132 - New! Introducing two new functions MSG_process_kill() and MSG_process_killall().
133 - It is possible to bound the rate of a communication in MSG with
134 MSG_task_put_bounded() (was already in the previous version but I had forgotten
135 to write it in the changelog).
136 - Bug fix to let GRAS run on top of MSG until we move it directly on top
141 - Various cleanups to the autotools stuff
142 - Begin to move Gras examples to examples/gras/
143 - Let make distcheck work again (yeah!)
145 - documentation overhauled using doxygen.
146 gtk-doc-tools is dead in SimGrid now.
147 - Automatically extract all existing logging categories, and add the list
148 to the documentation (long standing one, to say the less)
150 - Cleanup the known architecture table. Reorder the entries to group what
151 should be, and use a more consistent naming scheme.
152 (some of the test dataset are still to be regenerated)
153 - New! Allow library to register globals on each process just as userdata
155 This is implemented using a xbt_dict and not a xbt_set, so we loose the
156 lookup time (for now).
157 Use it in msg and trp.
158 This cleans a lot the internals and helps enforcing privacy of the
159 headers between the gras components.
160 - New! Add a timer mechanism, not unlike cron(8) and at(1).
161 - Bugfix: gras_os_time was delirious in RL.
162 - Bugfix: gras_trp_select/RL don't run into the wall when asked to select
164 - Reenable GRAS now that it works.
166 -- Arnaud Legrand <Arnaud.Legrand@imag.fr> Mon, 14 Feb 2005 14:02:13 -0800
168 SimGrid (2.90) unstable; urgency=low
170 Alpha 1 on the path to SimGrid 3
172 * It is a long time since the last release of SimGrid. I'm sorry about
173 that but as I had told you, I was rewriting a lot of things. I apologize
174 to those who had been reporting bugs to me and that I had not answered.
175 If your bug is still in the new version, please tell me. Here is a
176 summary of the main changes.
178 * REVOLUTION 1: The SimGrid project has merged with the GRAS project
179 lead by Martin Quinson. As a consequence SimGrid gains a lot in
180 portability, speed, and a lot more but you'll figure it out later.
181 SimGrid now comprises 3 different projects : MSG, GRAS and SMPI.
182 I wanted to release the new MSG as soon as possible and I have
183 broken GRAS, which is the reason why, for now, only MSG is fully
184 functional. A laconic description of these projects is available
185 in the documentation.
187 * REVOLUTION 2: I have removed SG and I am now using a new simulation
188 kernel optimized for our needs (called SURF but only the developers
189 should use it). Hence, MSG is now roughly 30 times faster and I think
190 that by rewriting a little bit MSG, I could event speed it up a little
191 bit more. Beside the gain in speed, it is also much easier to encode a
192 new platform model with SURF than it was with SG. More to come...
194 * REVOLUTION 3: I have tried to change a little as possible the API of
195 MSG but a few things really had to disappear. The main differences
196 with the previous version are :
197 1) no more m_links_t and the corresponding functions. Platforms are
198 directly read from a XML description and cannot be hard-coded
199 anymore. The same format is used for application deployment
200 description. The new format is described in the documentation.
201 Have a look in tools/platform_generation. There is a tiny script
202 that converts from the old platform format to the new one. Concerning
203 the application deployment format, parsing the old one is tricky.
204 I think most of you should however be able to convert your files. If
205 it is really an issue, I can write a C code that does the conversion.
207 2) the toolbox tbx does not exist anymore. We now have a library
208 with much more data-structures but without the hash-tables (we have
209 dictionaries that are much faster).
211 -- Arnaud Legrand <Arnaud.Legrand@imag.fr> Mon, 31 Jan 2005 10:45:53 -0800
213 *****************************************************************************
214 * Follows the old GRAS changelog. It does not follow the same syntax, but I *
215 * don't feel like converting the oldies. (Mt) *
216 *****************************************************************************
219 Version 2.90: "the long awaited one"
220 - Finished rewriting and debugging MSG. Rewrote the documentation.
221 - disable GRAS for now since it needs to be ported to the newest SG
224 - Finish the port to windows (using mingw32 for cross-compile)
227 - Main loop and datastructures of SURF. A cpu resource object is
228 functional. Surf can thus be used to create cpu's with variable
229 performance on which you can execute some actions.
231 2004-11-15 Martin Quinson
232 - Port to ARM. Simply added the alignment and size descriptions. Should
233 work, but the ARM machines are so slow that I didn't had the opportunity
234 to 'make check' over there yet.
236 2004-11-15 Arnaud Legrand
237 - Trace manager now written. It uses a heap structure and is therefore
238 expected to be efficient. It may however be speeded up (particularly
239 when many events occur at the same date) by using red and black
240 trees. One day maybe...
241 - Max-min linear system solver written. It uses a sparse matrix
242 structure taking advantage of its expected use. Most operations are
243 O(1) and free/calloc are called as few as possible. The computation of
244 the minimum could however be improved by using a red and black tree
247 2004-11-03 Arnaud Legrand
248 - Rename every gras_* function that was in xbt/ to its xbt_
250 - Add a heap and a doubly-linked list to xbt
251 - Added a dichotomy to the dictionaries. make check works as well before
252 so I assume that the patch is correct. I do not know however if things
253 run effectively faster than before now. :)
255 Inclusion of the SimGrid tree in the GRAS one. The archive is renamed to
256 SimGrid, and the version number is bumped to 2.x
258 2004-10-29 Martin Quinson
259 - Introduction of the remote errors.
260 They are the result of a RMI/RPC on the remote machine.
261 ErrCodes being scalar values, you can't get the host on which those
262 errors did happen. Extending the error mechanism as in Gnome is possible.
263 No idea yet whether it is a good idea.
265 2004-10-28 Martin Quinson
266 - Interface revolution: the Starred Structure Eradication.
267 I used to do typedef struct {} toto_t; and then handle *toto_t.
268 Arnaud (and Oli) didn't like it, and I surrendered. Now, you have:
269 - ???_t is a valid type (builded with typedef)
270 - s_toto_t is a structure (access to fields with .)
271 - s_toto is a structure needing 'struct' keyword to be used
272 - e_toto_t is an enum
273 - toto_t is an 'object' (struct*)
275 typedef struct s_toto {} s_toto_t, *toto_t;
276 typedef enum {} e_toto_t;
277 Moreover, only toto_t (and e_toto_t) are public. The rest (mainly
278 s_toto_t) is private.
280 - While I was at it, all gras_<obj>_free() functions want a gras_<obj>_t*
281 so that it can set the variable to NULL. It was so for dicts and sets,
282 it changed for dynars.
284 - Fix a bunch of memleaks in dict_remove
285 - Fix a bug in sg/server_socket opening: it failed all the time.
287 2004-10-07 Martin Quinson
288 - Speed up dynar lookup operation a bit.
290 gras_dynar_get is dead.
292 Now, you can choose between gras_dynar_get_cpy (the old gras_dynar_get
293 but should be avoided for efficiency reasons) and gras_dynar_get_ptr
294 (which gives you the address of the stored data).
296 gras_dynar_get_as is an helpful macro which allows you to retrieve a
297 copy of the data using an affectation to do the job and not a memcpy.
299 int toto = gras_dynar_get_as(dyn,0,int); rewrites itself to
300 int toto = *(int*)gras_dynar_get_ptr(dyn,0);
302 It does not really speedup the dynar test because they are
303 setting elements all the time (and look them seldom). But the dict does
304 far more lookup than setting.
306 So, this brings the dict_crash test from ~33s to ~25s (200000 elms).
308 2004-10-05 Martin Quinson
309 - Allow to (en/dis)able the cycle detection at run time.
311 Whether we should check for cycle or not is now a property of each
312 datatype. When you think there may be some cycle, use datadesc_cycle_set.
313 datadesc_cycle_unset allow to remove this property when previously set.
315 Note that the cycle detection is off by default since it impacts the
316 performance. Watch the data you feed GRAS with ;)
318 This property is hereditary. Any element embedded in a structure having it
319 set have it set for the time of this data exchange.
321 You should set it both on sender and receiver side. If you don't set it on
322 sender side, it will enter an endless loop. If you forget on receiver
323 side, the cycles won't be recreated after communication.
325 - Header reorganization.
326 Kill gras_private.h, each submodule must load the headers it needs.
328 2004-10-04 Martin Quinson
329 - Interface revolution: do not try to survive to malloc failure.
331 Now, gras_malloc and friends call gras_abort() on failure.
332 As a conclusion, malloc_error is not a valid error anymore, and all
333 functions for which it was the only gras_error_t return value are
334 changed. They now return void, or there result directly.
335 This simplify the API a lot.
337 2004-09-29 Martin Quinson
338 - Re-enable raw sockets.
339 Created by gras_socket_{client,server}_ext;
340 Used with gras_raw_{send,recv}
343 It should allow to kill the last bits of gras first version soon.
345 This is not completely satisfactory yet (duplicate code with
346 chunk_{send,recv}; a bit out of the plugin mechanism), but it should
349 - Simplify transport plugin (internal) interface by not passing any
350 argument to _server and _client, but embedding them in the socket
353 2004-09-28 Martin Quinson
354 - Finish the port to AIX.
355 autoconf was my problem (segfault within the malloc replacement
356 function. No idea why)
358 2004-09-16 Martin Quinson
359 - Fix some size_t madness on 64bit architectures.
361 2004-09-08 Martin Quinson
362 - Reduce the number of system headers loaded, overload some more system
363 calls (such as malloc to cast the result of the system one, and work
365 - Fix and reintroduce the config support
367 2004-09-07 Martin Quinson
368 - Source code reorganization to allow Arnaud to surf all over there.
369 - Allow to document the logging categories.
370 - Remove all uppercase from logging categories and useless cleanup in names.
372 2004-08-18 Martin Quinson
373 Version 0.6.2 (protocol not changed; API changed)
374 - Interface cleanup: gras_msgtype_by_name returns the type (instead of a
375 gras_error_t), and NULL when not found. Functions expecting a msgtype
376 as argument (msg_wait; msg_send) deal with NULL argument by providing a
377 hopefully usefull message.
378 - Portability to prehistoric sparcs again
380 2004-08-17 Martin Quinson
381 Version 0.6.1 (protocol not changed; ABI not changed)
382 - prealloc some buffers to speed things up
384 2004-08-11 Martin Quinson
385 Version 0.6 (protocol not changed; ABI expended)
386 - The parsing macro can deal with the references, provided that you add
387 the relevant annotations (using GRAS_ANNOTE(size,field_name))
389 2004-08-09 Martin Quinson
390 Version 0.5 (protocol not changed; ABI changed)
391 - Allow to off turn the cycle detection code in data exchange at
392 compilation time. It should be at run time, but I'm short of time (and
393 the config stuff is still broken). That way, we keep dict out of the
394 critical path, which is good because the performance is poor:
395 - search not dichotomial yet
396 - dynar give no way to access their content and memcpy everytime
397 - In composed data description (struct, ref and so on), stop foolness of
398 keeping the subtype's ID, but store the type itself. This keeps sets out
399 of the critical path, which is good since they rely on dynar and
400 dictionnaries. The only loose of that is that we cannot detect the
401 redeclaration of a structure/union with another content (but I'm not sure
402 the code detected well this error before anyway). We still can detect
403 the redefinition discrepancy for the other types.
404 - Use a whole bunch of optimisation flags (plus -fno-strict-aliasing since
405 it breaks the code because of type-punning used all over the place).
406 This breaks on all non-gcc architectures (for now).
408 All those changes (plus the buffer of last time) allow me to gain 2 order
409 of magnitude on cruel tests consisting of 800000 array of integers on two
410 level of a hierarchical structure (200 secondes -> 4 secondes)
413 - the selector of reference must now return the type it points to, not
416 2004-08-06 Martin Quinson
417 Version 0.4 (protocol changed; ABI not changed)
418 - Allow to pass --gras-log argument to processes in simulation mode. Really.
419 - New debugging level: trace (under debug) to see effect of GRAS_IN/OUT
420 - Implement a buffer transport, and use it by default (it relies on tcp in
421 real life and on sg in simulation).
422 That's a bit hackish since I had a new field to the structure to store
423 its data without interfering with the subtype ones. Inheritance
424 is tricky in C. And that's a kind of reverse inheritance with one class
425 derivating two classes. Or maybe a game with java interfaces. Anyway,
426 that's damn hard in C (at least).
427 Moreover, I got tired while trying to ensure plugin separation and
428 genericity in SG mode. MSG wants me to do weird things, so let's go for
429 cruel hacks (temporarily of course ;).
430 See comment in transport_private.h:71
431 - do not include all the _interface headers in private but in the files
432 which really need them (to cut the compilation time when they are
435 2004-07-26 Martin Quinson
436 Version 0.3 (protocol not changed; ABI changed)
437 - Major overhault of the datadesc interface to simplify it:
438 - shorted the function names:
439 s/gras_datadesc_declare_struct/gras_datadesc_struct/ and so on
440 - add a trivial way to push/pop integers into the cbps without malloc.
441 This allows to make really generic sub_type description, which simply
442 pop their size of the stack.
443 - add a function gras_datadesc_ref_pop_arr() which does what users want
444 most of the time: Declare a dynamic array (which pops its size of the
445 stack) and declare a reference to it. Poor name, but anyway.
446 - kill the post-send callback, add a post-receive one
448 2004-07-23 Martin Quinson
449 Version 0.2 (protocol changed; ABI changed)
450 - add some testing for cpbs in the test cases, and fix some more bugs.
451 This invalidate again the little64 data file, since I cannot regenerate
453 - remove an awfull optimization in the logging stuff, allowing me to:
454 - understand it again
455 - learn gcc how to check that the argument match the provided format
456 - fix all errors revealed by gcc after that
457 - internal keys of dict are not \0 terminated. Deal with it properly in
458 loggings instead of segfaulting when the user want to see the logs :-/
460 2004-07-22 Martin Quinson
461 - Fix some stupid bug preventing cbps (callback postit) from working
463 2004-07-21 Martin Quinson
464 - Some documentation cleanups
465 - remove the useless last argument of msgtype_declare
466 - rename the Virtu functions to fit into the 'os' namespace
467 - move headers src/include -> src/include/gras/ and stop fooling with
468 gras -> . symbolic link
469 - make distcheck is now successful
471 2004-07-19 Martin Quinson
473 - Build shared library also
474 - Install html doc to the right location
475 - stop removing maintainer files in make clean
476 - build tests only on make check
478 2004-07-13 Martin Quinson
480 - No major issue in previous version => change versionning schema
481 - Re-enable little64 convertion test now that Abdou kindly regenerated the
482 corresponding dataset.
484 2004-07-11 Martin Quinson
486 - Get it working with any kind of structure (we can compute the padding
487 bytes remotely for all the architectures I have access to)
488 - Implement the structure parsing macro (still not quite robust/complete)
489 - Improvement to the remote testing toysuite
491 2004-07-10 Martin Quinson
493 - get ride of a bunch of deprecated macros
494 - actually run the test for two-compliment, not only compile it :-/
495 - test whether the structures get packed (and bail out if yes. Damn.
496 Alignment is a serious matter)
497 - test whether the structures get compacted (but respecting the alignment
498 constraints of each types)
499 - test whether the array fields of structures can straddle alignment boundaries
501 - Damnit, double are bigger than float (typo in creation of 'double' datadesc)
502 (took me 2 hours to find that bug, looking at the wrong place)
503 - Add gras_datadesc_declare_{union,struct}_close(). They must be used
504 before sending/receiving and are used to compute the offsets of fields
505 - Given that padding size depend even on compiler options, keep track of
506 alignment and aligned_size only for the current architecture. Not a big
507 deal since we send structure fields one after the other (seems
509 - Add the datastructure used for IEEE paper by the PBIO guys to the test
510 program, let it work on linux/gcc/little32. portability todo.
512 2004-07-08 Martin Quinson
513 - import and improve remote compilation support from FAST
514 - make sure make check works on half a dozen of machines out there
516 2004-07-07 Martin Quinson
517 Let's say it's version 0.0.3 ;)
518 - Implement conversions (yuhu!)
519 - Let it work on solaris (beside conversion, of course)
520 - Stupid me, using rand() to generate the conversion datatests in not wise.
522 2004-07-06 Martin Quinson
523 - Let make dist work, since I'm gonna need it to compile on remote hosts
524 - Let Tests/datadesc_usage write the architecture on which the file was
525 generated as first byte.
526 - Add PowerPC (being also IRIX64), SPARC (also power4) and ALPHA
527 architecture descriptions.
528 - Add datadesc_usage.{i386,ppc,sparc} files being the result of execution
529 on those architectures.
530 - Optimization: send/recv array of scalar in one shoot
532 2004-07-05 Martin Quinson
533 - YEAH! GRAS/SG and GRAS/RL are both able to run the ping example !
535 - Plug a whole bunch of memleaks
536 - each process now have to call gras_{init,exit}. One day, their log
537 settings will be separated
538 - Continue the code factorisation between SG, RL and common in Transport.
540 2004-07-04 Martin Quinson
542 - Redistribution between SG and RL.
543 We wanna have to accept in SG, so move accepted related parts of RL in
544 the common part. (more precisely, the dynar of all known sockets is no
545 more a static in transport.c, but part of the process_data)
547 [gras_stub_generator]
548 - Bug fix: Do call gras_process_init from gras_init (wasnt called in RL).
550 2004-07-03 Martin Quinson
551 - Create a new log channel tbx containing dict, set, log, dynar (to shut
552 them all up in one shot)
554 - Fix the ugly case of reference to dynamic array.
555 - New (semi-public) function gras_datadesc_size to allow the messaging
556 layer to malloc the needed space for the buffer.
558 - gras_socket_close now expect the socket to close (and not its address to
559 put NULL in it after it). This is because the socket passed to handlers
560 is one of their argument (=> not writable).
562 - propagate the interface cleanup from last week in datadesc, ie remove a
563 superfluous level of indirection. User pass adress of variable
564 containing data (both when sending and receiving), and not of a variable
565 being a pointer to the data. Let's say that I like it better ;)
566 The price for that is constructs like "int msg=*(int*)payload" in
567 handlers, but it's a fine price, IMHO.
569 - Let it work in RL (yuhu)
571 2004-06-21 Martin Quinson
573 - porting SG plugin and SG select to new standards (works almost).
574 - plug memleaks and fix bugs around.
577 - cleanup the prototype of data recv and force users to specify when they
578 want to handle references to objects. Test case working even for cycles.
579 - plug memleaks. Valgrind is perfectly ok with this.
581 2004-06-12 Martin Quinson
583 - cleanup the separation between plugin and main code in plugin creation
585 2004-06-11 Martin Quinson
587 - Reput hook for raw sockets, needed for BW experiments
588 - kill a few lines of dead code
589 [Data description] Interface cleanup
590 - gras_datadesc_by_name returns the searched type or NULL.
591 That way, no variable is needed to use a type desc once, which makes
593 - gras_datadesc_declare_[struct|union]_append_name is removed. The last
594 two parameters were strings (field name, type name), leading to
596 [Dicos] Interface cleanup
597 - gras_dico_retrieve -> gras_dico_get ; gras_dico_insert -> gras_dico_set
598 This is consistant with the dynar API.
600 2004-04-21 Martin Quinson
602 - Porting to new standards.
605 There is no bag anymore, no need to take extra provision to mask the
606 pointers behind "ID".
607 Better splitup of functions between files create/exchange/convert.
608 This is still a bit artificial since convert and receive are so
609 interleaved, but anyway.
611 - add a queued message list to procdata (the ones not matching criteria
613 - factorize some more code between SG and RL wrt procdata
615 - use gras_exit in example to track memleaks
616 - get rid of gs_example now that GS is properly integrated into gras
617 - update run_test to integrate the lastest tests (datadesc)
619 - rename WARNINGn macros to WARNn since it prooved error-prone
621 2004-04-19 Martin Quinson
623 - register init/exit functions within gras module mechanism
624 - send/receive function.
625 Convertion is not implemented, but short-cutted if not needed.
626 struct/array elements are sent one by one (instead of block-wise), but
627 nobody really cares (yet). Get a prototype before optimizing.
628 - tests (using a file socket) for DD send/receive on:
629 - base types: int, float
630 - array: fixed size, string (ie ref to dynamic string)
631 - structure: homogeneous, heterogeneous
632 - chained list, graph with cycle
633 Believe it or not, valgrind is not too unhappy with the results. The
634 cycle happily segfaults, but the others are ok. And I'm sick of pointers
638 - Bugfix when using a filename explicitely (instead of '-')
640 2004-04-09 Martin Quinson
642 - factorize more code between RL and SG in socket creation
643 - Complete the implementation and tests of:
645 o file (only in RL, and mainly for debugging)
647 I lost 3 days to design a portable address resolver, and then decided
648 that the prototype mainly have to run on my box.
649 Addressing portability too early may be like optimizing too early :-/
651 - use gras_init in the Tests instead of the crappy parse_log_opt
652 (the latter function is removed)
653 [Conditional execution]
654 - New functions: gras_if_RL/gras_if_SG (basic support for this)
655 [Code reorganisation]
656 - Get rid of libgrasutils.a since it makes more trouble than it solves.
657 Build examples against the RL library, since there is no way to disable
658 its creation for now.