* xfig-data leaked in gtk version as well
8 lines of code changed in 2 files:
* suppress several leaks caused by arbs motif usage
- previously these just were "ignored" in output by valgrind2grep (removed them there)
37 lines of code changed in 2 files:
* stuff leak
1 lines of code changed in 1 file:
* stuff leak in aw_create_shell
7 lines of code changed in 1 file:
* rename suppressions (cosmetics)
3 lines of code changed in 1 file:
* properly destroy !DbQuery before exit (done when DB is closed, because destruction removes DB callbacks)
106 lines of code changed in 3 files:
* reimplement GB_keep_string using template Keeper<char*>
46 lines of code changed in 4 files:
* suppress more motif-specific leaks (most caused by 'not freeing AW_window')
69 lines of code changed in 1 file:
* free AWT_configuration together with callback
4 lines of code changed in 1 file:
* use GB_keep_string for two dynamic awar names passed to callbacks
6 lines of code changed in 2 files:
* add GB_keep_string
- keeps an allocated c-string until program termination
- use this for allocated c-strings passed to callbacks
31 lines of code changed in 2 files:
* destroy all action callbacks
* destroy !CreateWindowCallbacks passed to window-poppers
19 lines of code changed in 3 files:
* each query leaked param.first_key
14 lines of code changed in 1 file:
* suppress many leaks (and two illegal memory accesses) caused by arbs motif usage (or by motif itself)
- previously most of these just were "ignored" in output by valgrind2grep (removed them there)
* valgrind always generates suppressions (copy&paste them from output if needed)
254 lines of code changed in 5 files:
* dump number of fonts tried/found
* insert "No suitable fonts detected" if NO font found (buttons completely disappeared in one scenario)
14 lines of code changed in 1 file:
- doc installing precompiled versions on debian (tested with wheezy)
11 lines of code changed in 2 files:
job "ARB-source-tarball" failed as expected with
{{{
[58:28.900883205] ------------------------------------------------ Make TEMPLATES
make[2]: Entering directory `/j/workspace/ARB-source-tarball/TGT/u1004/arbsrc_12326/TEMPLATES'
../SOURCE_TOOLS/build_info.pl
Missing file '/j/workspace/ARB-source-tarball/TGT/u1004/arbsrc_12326/TEMPLATES/svn_revision.h' at ../SOURCE_TOOLS/build_info.pl line 298.
make[2]: *** [TEMPLATES.a] Error 2
make[2]: Leaving directory `/j/workspace/ARB-source-tarball/TGT/u1004/arbsrc_12326/TEMPLATES'
make[1]: *** [TEMPLATES/TEMPLATES.dummy] Error 1
make[1]: Leaving directory `/j/workspace/ARB-source-tarball/TGT/u1004/arbsrc_12326'
make: *** [arb] Error 2
Build step 'Execute shell' marked build as failure
Sending e-mails to: ###
Finished: FAILURE
}}}
=> do not remove generated headers in target TEMPLATES/clean
1 lines of code changed in 1 file:
- update doc for compiling arb from source
33 lines of code changed in 2 files:
- default config now uses OPENGL=0
1 lines of code changed in 1 file:
- build from tarball fails after 'make clean' (test that!)
4 lines of code changed in 1 file:
* suppress leaks in TEST_tree_remove_add_all
- these leaks are accepted in PARSIMONY because they only appear when removing nodes and would require some sophisticated memory management. Problem exists since [2]
* now all unittest "pass" with valgrind
34 lines of code changed in 5 files:
* add valgrind suppression files
- debian-libc6-dbg.supp
- three arb-specific (common, motif, gtk)
31 lines of code changed in 4 files:
* add unittest to AW_selection_list (describing behavior of motif version)
- add dtor to AW_selection_list
- doesn't test delete operations (because they attempt to access the nonexisting awar)
124 lines of code changed in 3 files:
- dump build related environment variables
5 lines of code changed in 1 file:
- do not redefine macros provided by gmacros.h (glib)
6 lines of code changed in 2 files:
* include compiler-call into error if call fails
12 lines of code changed in 1 file:
* override CC / CXX in first invocation of main Makefile
* always export CC, CXX, A_CC and A_CXX early
Note: this fixes compilation with
{{{
cd $ARBHOME; make "CC=/opt/clang-3.3/bin/clang" "CXX=/opt/clang-3.3/bin/clang++" all
}}}
even if CC and CXX is defined via environment, e.g.
{{{
> echo $CC $CXX
/opt/gcc-4.8.1/bin/gcc /opt/gcc-4.8.1/bin/g++
}}}
7 lines of code changed in 1 file:
- fix !MacPorts compilation (in PERL2ARB)
- exports CC CXX from arb Makefile and from PERL2ARB/Makefile.main (avoiding fallback to sth present in environment)
11 lines of code changed in 2 files:
- rename variable CXX (was used to hold source files)
2 lines of code changed in 1 file:
* reduce differences between trunk/gtk in aw_select
* removes unused methods get_index_of_displayed, selectAll + deselectAll
22 lines of code changed in 2 files:
* help updates
- merged multiple help files referring to 'ARB_NT/Mark species/...' into one
- added help for 'new window'
- fix missing 'mark_colors.hlp' (using 'colorize.hlp')
109 lines of code changed in 8 files:
- fix 'mark complete': now marks all non-partial sequences even if 'ARB_partial' entry is missing
1 lines of code changed in 1 file:
- use explicit bool params
6 lines of code changed in 4 files:
* CC/CXX used by arb
- overwrite defaults after copying them into A_CC/A_CXX
- fix all errors (triggered by above change)
* remove fallbacks (probcons; MUSCLE; FASTTREE)
* set CC/CXX from A_CC/A_CXX (MrBAYES; PHYML20130708)
16 lines of code changed in 6 files:
* tweaked jump-feature (tree display)
- jump after tree- and treemode-change (previously only after species-change)
- allow to define behavior for both change-types separately
- added two jump-behaviors:
* always center vertically
* always center
- improved "keep-visible" behavior (=old auto-jump behavior):
* if selected species is "left of" viewport, center position depends on treemode
* IRS tree centers when selected species is inside bottom/top 10% of display
- feature activated in ARB_PARSIMONY
- when autojump was OFF, display did not refresh
172 lines of code changed in 7 files:
* JUMP button did not work in listmode (if no tree selected)
21 lines of code changed in 1 file:
- warn if no tree selected in several tree functions
18 lines of code changed in 1 file:
- do not compute 'no tree' (fixes #551)
4 lines of code changed in 1 file:
* pass down correct AD_MAP_VIEWER_TYPE to caller
* trigger refresh after selecting a new species (fixes #518)
9 lines of code changed in 2 files:
- check whether we have a tree AFTER checking whether a species has been clicked (fixes #518)
5 lines of code changed in 1 file:
* allow sequence decompression (by master seq) from inside (sequence) delete callbacks (fixes #546)
13 lines of code changed in 2 files:
- do not remove callback from inside delete-callback of deleted species (fixing #221)
6 lines of code changed in 1 file:
* list shared library dependencies (instead of listing glib); implements #547
* eval collect-CMD (instead of executing it in a subshell; allows to call local commands)
* shortened tags
35 lines of code changed in 1 file:
delete broken unused code. fixes #548
{{{
char num[6];
sprintf(num, "%i ", 10000); // crash
}}}
and num wasn't even used... sprintf=evil!
0 lines of code changed in 1 file:
* launcher also waits for keypress when arb terminates abnormmally (i.e. with error)
25 lines of code changed in 1 file:
* no longer limit number of cached search results (avoids problem described in #545)
1 lines of code changed in 1 file:
- attempt to fix compile error (centos5)
1 lines of code changed in 1 file:
* probe design results not generally differs between 32/64 bit
- u1004-32 behaves like 64bit
- u1204-32 produces different results (see also [11823])
16 lines of code changed in 1 file:
* added a bunch of arguments to CLI that allow to test various ways to terminate
* exit from signal handler with exitcode>=128
* arb_launcher knows about 3 types of exitcodes:
- 0 = success
- 1 = failure (no crash); no longer tell arb did crash in this case, fixing #538
- else = crash
22 lines of code changed in 5 files:
- replace magics
- fix two wrong exit codes
6 lines of code changed in 1 file:
- replace some suspect exit codes by GBK_terminate
21 lines of code changed in 3 files:
- no magic exit codes
15 lines of code changed in 9 files:
- no magic exit codes in WINDOW
8 lines of code changed in 4 files:
- no magic exit codes in RNACMA
- print proper errors
7 lines of code changed in 3 files:
* [http://bugs.arb-home.de/changeset/12157/trunk/EDIT4#file4 this change] introduced new problem:
- callbacks installed by cursor no longer were removed @ exit. was done by ED4_base_position::invalidate b4 above change.
- renamed ED4_cursor::invalidate_base_position into prepare_shutdown (to reflect intention)
- added prepare_shutdown to ED4_base_position
* hopefully fixes #529
7 lines of code changed in 3 files:
* wrap gtk/motif diff caused by [10010]
5 lines of code changed in 1 file:
* fixes #534
1 lines of code changed in 1 file:
* no fallback2default for selection list bound to AWAR_MOD_SAI_SCRIPT
- fixes comment:ticket:537:9
1 lines of code changed in 1 file:
* no fallback2default for selection list bound to AWAR_CONFIGURATION
- does NOT fix problem mentioned in comment:ticket:537:8:
* switching focus to selection list does various things (depending on what is selected before in the selection list). accepting behavior as is.
* using Shift-Tab to focus 'Abort' or 'Ok' works (wasn't broken b4 this patch)
- fixes two other differences to motif:
* renaming a config did "deselect config" (now selects renamed config)
* storing a config did not select that config if new
2 lines of code changed in 1 file:
* no fallback2default for selection list bound to AWAR_EDIT_CONFIGURATION
- fixes comment:ticket:537:6
2 lines of code changed in 1 file:
* elim unused function aw_input2awar
0 lines of code changed in 2 files:
* eliminate parameter check_fun from aw_string_selection and aw_string_selection2awar
- implement sole usage manually @ caller
* fallback to old awar value in aw_string_selection2awar when abort is pressed
81 lines of code changed in 5 files:
* remove_from_configs was broken (resulted in duplicated entries in existing configs)
* use abort feature introduced in [12244] in config manager
16 lines of code changed in 2 files:
* new_input_window()
- expect buttons to be passed
- '-Cancel' is handled as abort button 'Cancel'
- pass "Ok,-Abort" from callers which previously used the default (aw_input + aw_string_selection)
36 lines of code changed in 1 file:
* change value used for "no config selected" from '????' to {{{''}}}
- fixes: '????' matched all configs with 4 letter names (in GBT_find_configuration, i.e. in all but 'STORE' command)
* store config: selecting no config / entering empty name resulted in a config '<unnamed1>'. fixed.
75 lines of code changed in 3 files:
* define magics for
- no config selected
- no field selected
* DRYed AWAR_CONFIGURATION init
24 lines of code changed in 5 files:
* use typed CBs
65 lines of code changed in 5 files:
* pass 'fallback2default' down into awt_create_selection_list_on_configurations()
9 lines of code changed in 5 files:
* handle field-search-error in group_species
- triggered e.g. by some gtk-sellist-fallback
* assert 'most_recently_used_window' is set on use
9 lines of code changed in 2 files:
- fix next neighbour search of selected (gtk; #537)
1 lines of code changed in 1 file:
* disable selectionlist defaultfallback for filetype selection in tree-import-window
- fixes: it was impossible to load treefiles with extensions other than '.tree' and '.xml' in gtk-branch
1 lines of code changed in 1 file:
* fix #536
1 lines of code changed in 1 file:
* pass 'fallback2default' down into awt_create_selection_list_on_trees()
17 lines of code changed in 11 files:
* fix: impossible to create a new submission form (name always fell back to 'default')
- caused by selectionlist-defaultfallback via option-menu
2 lines of code changed in 1 file:
* use typed CBs
13 lines of code changed in 1 file:
* replace magic awar names by defines
27 lines of code changed in 1 file:
running unit tests of multiple WCs in parallel need an unreasonable amount of time to complete,
resulting in getting killed by 'deadlockguard'. Increase time for 'slow' tests from 30 to 45 min.
Reason for runtime is unknown; i guess it is somehow caused by tcp/pipe communication.
3 lines of code changed in 1 file:
* fix #535
1 lines of code changed in 1 file:
* [12215] was intended for trunk :(
1 lines of code changed in 1 file:
* no fallback for NDS field selection list (fixes #533)
2 lines of code changed in 1 file:
* pass down fallback2default to popup_select_species_field_window
21 lines of code changed in 4 files:
* create_selection_list_on_itemfields
- add and pass down new parameter 'fallback2default'
31 lines of code changed in 13 files:
* inlined default parameters of create_selection_list_on_itemfields
45 lines of code changed in 10 files:
* removed default for fallback2default
5 lines of code changed in 2 files:
* inlined default param for create_option_menu
81 lines of code changed in 36 files:
* add default param fallback2default to create_option_menu (see also [12185:12189])
10 lines of code changed in 2 files:
* reduce differences between gtk/trunk
- copy conditional sections (ARB_GTK)
- take-over some compatible changes
72 lines of code changed in 7 files:
- adapt to awartype changed in gtk-branch
4 lines of code changed in 2 files:
- make variable_name/type private + provide common interface for gtk+motif-version
10 lines of code changed in 2 files:
* eliminate all but 1st parameter from create_option_menu
9 lines of code changed in 3 files:
* do not use more than 1 parameter in calls to create_option_menu
- 2nd parameter 'label' is dominated by a previous call to label (use direct label calls instead)
- 3rd parameter 'mnemonic' is w/o function in motif and is superfluous in gtk anyway
* reset 'gone_tree_name'
48 lines of code changed in 11 files:
- belongs to [12192]
2 lines of code changed in 1 file:
* added 'fallback2default' to awt_create_selection_list_on_sai
* pass false (=disable fallback2default) for
- SAI/Create/Consensus
- SAI/Create/Max. Frequency
- SAI/Create/Positional variability + Column statistic
- ARB_PHYLO/File/Export Filter
- ARB_PHYLO/File/Export Frequencies
* fixes #520
17 lines of code changed in 9 files:
* AW_window::create_selection_list expects new parameter 'fallback2default'
- change calls outside WINDOW (pass 'true' everywhere, which previous default behavior in gtk)
36 lines of code changed in 18 files:
* AW_window::create_selection_list expects new parameter 'fallback2default'
- param is meaningless in trunk, but used in gtk-branch
- will not compile (calls outside WINDOW not changed)
- use fallback2default==false for
* help-window
* string-selection-dialog
* file-selection-dialog
9 lines of code changed in 5 files:
* remove default params from AW_window::create_selection_list
2 lines of code changed in 1 file:
* if an error occurs while building NDS expression, insert error into result instead of trapping
5 lines of code changed in 1 file:
* abort 'modify fields of listed species' when an error occurs while searching for target field (e.g. when key is invalid)
33 lines of code changed in 1 file:
- create dated backups of previous recordings
22 lines of code changed in 1 file:
- do not edit user-masks from callallcallbacks
3 lines of code changed in 1 file:
* fix needed for gtk:
- avoid to delete default-entry from multiprobe lists
7 lines of code changed in 2 files:
* fixes needed for gtk:
- fix error handling in create_consense_tree_cb()
- add default-entry in multiprobe ali-type selection list
- do not attempt to multifurcate missing tree
21 lines of code changed in 3 files:
* use typed cb for MP_delete_selected
12 lines of code changed in 3 files:
* gtk-file-selection boxes always had default-entry at top (caused by sorting)
- workaround: first sort, then insert default entry
5 lines of code changed in 2 files:
* AW_selection_list::move_content_to no longer deletes default entry (caused crashes in multiprobe sellists)
24 lines of code changed in 1 file:
* adapt gtk<>motif
- use delete_default
3 lines of code changed in 1 file:
* adapt gtk<>motif
- remove "always true" parameter 'clear_default' from AW_selection_list::clear (parameter vanished in gtk)
- add delete_default
- make AW_selection_list::get_entry_at const
23 lines of code changed in 2 files:
* reintegrates 'fix' into 'trunk'
- fixes #529
* adds:
- log:branches/fix@12147:12156
44 lines of code changed in 4 files:
* as we depend on glib now
- use G_BREAKPOINT instead of SEGV to trigger assertions
- continue using ARB_SIGSEGV
* from inside unittests
* when SIMPLE_ARB_ASSERT is defined
58 lines of code changed in 2 files:
- also skip dependencies in /usr/lib (as introduced by glib)
2 lines of code changed in 1 file:
pass down CXX_INCLUDES
3 lines of code changed in 3 files:
* hardcode minimum for callbacklist-elementsize and use that minimum for OSX
19 lines of code changed in 2 files:
* fix {{{ 17:29.56 Skipped 0 species (no data?) }}}
5 lines of code changed in 2 files:
* reduce gtk<>motif differences
- move check into is_wanted_callback()
38 lines of code changed in 1 file:
- be more verbose about marking callbacks as "called"
3 lines of code changed in 1 file:
* another fix for diff ignoring date-entries
- for each regexpr in difflineMode define whether it should match word boundaries
- replace month/day-names only if they are words (i.e. do not replace 'jun' in 'disjunct')
34 lines of code changed in 1 file: