gnugo-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [gnugo-devel] Membership request for group GNU Go


From: Stefan Pejovski
Subject: Re: [gnugo-devel] Membership request for group GNU Go
Date: Wed, 12 Jul 2023 13:32:09 +0300

Changelogs between 3.8 and 3.9.1

diff --git a/ChangeLog b/ChangeLog
index 61f1f17d..16532d91 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -2,6 +2,57 @@
 -- ChangeLog
 -------------------------------------------------------------------------

+3.9.1 released December 21, 2010
+
+- removed automake-1.4.patch from doc/Makefile.am
+- visual studio project files and config.vc removed in favor of cmake build
+- copyright notices updated for 2010
+- remove reading_cache_clear() calls between silent_examine_position() and
+  owl reading functions.
+- remove assertion from influence_erase_territory() to avoid crashing
+- view.pike updated from GTK 1 to GTK 2
+- cross-compilation support added to uncompress_fuseki.c
+- revised generation of hashvalues.
+- give small bonus to attacking vital moves in owl_determine_life()
+- allow tactically unstable owl lunches to be inessential
+- new static function test_unstable_lunch_essentiality()
+- break out new static function test_lunch_essentiality() from
+  owl_find_lunches()
+- capture big nakade to get out of atari in semeai even if tactical
+  reading disagrees
+- play common liberties in semeai with sufficient remaining liberties,
+  even if tactical reading believes it is unsafe
+- remove redundant function semeai_propose_eyespace_filling_move
+- do not play common liberties after having already passed
+- fall back to ordinary reading in attack_either and defend_both if the
+  two strings are in fact the same
+- never consider backfilling move for common liberty inside own eyespace
+  as risk free.
+- do_owl_analyze_semeai() can distinguish between outer liberties and
+  eyefilling moves, also chooses the latter correctly
+- recognize_eye() can optionally find out at which points eyefilling can
+  safely be done
+- new static function optical_graph_matcher() broken out of recognize_eye()
+- do not make domains for second owl in owl_make_domains()
+- do not consider also passing to get seki in semeai reading when stackp>0
+- new helper function bent_four_helper()
+- analyze_eyegraph improvements
+- TODO completely rewritten
+- revision of dragon_looks_inessential
+- fix overflow in computation of node limits
+- do not crash if gtp_aa_confirm_safety is fed an illegal move
+- score seki eyes correctly in aftermath scoring with Chinese rules
+- avoid overvaluation of eyespaces which are split by a single stone in atari
+- avoid playing inside own eyespace to capture lunches in semeais
+- do not crash on replacement moves conflicting with unconditional analysis
+- guess_eye_space() revised
+- major revision of eyes.db
+- improve_lunch_attack() in owl.c revised
+- filter out send-two-return-one moves in worm_reasons()
+- owl tuning
+- tuning
+- new and revised tests
+
 3.8 released February 17, 2009

 - added Wietze Brandsma to THANKS
diff --git a/Makefile.am b/Makefile.am
index bf1641e4..4a231afd 100644
--- a/Makefile.am
+++ b/Makefile.am
@@ -4,7 +4,7 @@ SUBDIRS = utils sgf engine patterns interface doc regression

 DISTCLEANFILES = *~

-EXTRA_DIST = gnugo.dsw config.vc makevcdist.pl WINDOWS OSX CMakeLists.txt config.h.cmake stamp-h.in
+EXTRA_DIST = WINDOWS OSX CMakeLists.txt config.h.cmake stamp-h.in

 maintainer-check:

diff --git a/Makefile.in b/Makefile.in
index 01561b61..84dbfdb8 100644
--- a/Makefile.in
+++ b/Makefile.in
@@ -1,8 +1,9 @@
-# Makefile.in generated by automake 1.9.6 from Makefile.am.
+# Makefile.in generated by automake 1.11.1 from Makefile.am.
 # @configure_input@

ke 12. heinäk. 2023 klo 6.04 Luis Felipe Strano Moraes (luis.strano@gmail.com) kirjoitti:
Hey folks, bumping this old thread here, would be really nice to get a release out there. If no one cares about being a maintainer, I can definitely go ahead and do the work there to get the release out and keep track of any other activities as time goes by, would be really good to not have this bitrot completely.

Best regards,


On Tue, Aug 30, 2022 at 8:24 PM Thien-Thi Nguyen <ttn@gnuvola.org> wrote:

NB: To and Reply-To set to gnugo-devel m.l; Cc trimmed.

() bump-math@sporadic.stanford.edu (Daniel Bump (bump@math.stanford.edu email))
() Fri, 26 Aug 2022 06:44:28 -0700

   > I hope that by including gnugo-devel in the discussion, we
   > can get more perspectives and that the GNU Go maintainers
   > can be convinced to change their position to that of "do
   > not keep generated files under version control", on the way
   > to eventually approving a merge of that branch and future
   > development along those lines.

   I think I was the one that was arguing for this position
   (mainly so that someone could build gnugo more easily after
   downloading from savannah versus a tarball), and since Gunnar
   was not in agreement, I would not insist on the point.

That's great news.  Thanks for reconsidering!  I would like to
propose the next steps for us to take, collectively:

- create top-level file HACKING, includes info on
  - branch discipline (e.g., push to ‘master’ vs dev br & merge)
  - how to bootstrap (run autogen.sh, etc)
  - how to make a release
  - copyright discipline (e.g., once per year vs only on change)
  - coding conventions (whitespace, indentation, etc)
  - other developer-only info/lore
  - etc

- merge branch ‘ttn-maint’ (and delete afterwards)

- gather / apply other fixes found in the wild

- make a test maintenance release (published on alpha.gnu.org)
  (info "(maintain) Test Releases") and solicit timely feedback

- tweak as necessary / work out the kinks

- make a "real" maintenance release

I'm not a maintainer, but i can take a crack at writing HACKING
(adapting the GNU RCS HACKING[1], basically, w/ input from
everyone), and help out w/ the other stuff.  It's been many
years since the last GNU Go release, so there's no rush (IMHO),
but i think it would be reasonable to aim for "real" release by
end of year.

What do the maintainers think?  Is this something you'd approve?
I don't want to step on anyone's toes!

[1] http://git.savannah.gnu.org/cgit/rcs.git/tree/HACKING?h=p

--
Thien-Thi Nguyen -----------------------------------------------
 (defun responsep (query)               ; (2022) Software Libero
   (pcase (context query)               ;       = Dissenso Etico
     (`(technical ,ml) (correctp ml))
     ...))                              748E A0E8 1CB8 A748 9BFA
--------------------------------------- 6CE4 6703 2224 4C80 7502

_______________________________________________
gnugo-devel mailing list
gnugo-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/gnugo-devel


--
Luís Felipe Strano Moraes
_______________________________________________
gnugo-devel mailing list
gnugo-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/gnugo-devel

reply via email to

[Prev in Thread] Current Thread [Next in Thread]