[Pkg-e-commits] [SCM] Enlightenment DR17 graphical layout and animation library annotated tag, debian/0.5.0.042-1, updated. debian/0.5.0.042-1
Albin Tonnerre
albin.tonnerre at gmail.com
Thu Mar 20 13:20:56 UTC 2008
The annotated tag, debian/0.5.0.042-1 has been updated
to 891e45f30f4fc9186ff41ad6548ebf2819ac58b5 (tag)
from 125e3b3126f55bcce3194b25e2cf8e5f583b04b5 (which is now obsolete)
tagging cd5952ee2124318ba0c355233e20240771fe42eb (commit)
tagged by Jan Luebbe
on Thu Mar 20 14:22:37 2008 +0100
- Shortlog ------------------------------------------------------------
Debian release 0.5.0.042-1
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
iD8DBQBH4mUhioOL5NhIDy4RAj+8AKDnaxADON90nZG8KmUmkes95Z0N3QCbBhqL
LRQdLWIwtTfrRhoiNsWoEqU=
=7Yxu
-----END PGP SIGNATURE-----
Albin Tonnerre (4):
Freshen the changelog, generate the doc using the build/ target, update standard-version, use dh 6, freshen the manpages and fix the evas dependencies
Fix the clean target
debian/control: Vcs-*
libedje-dev depends on libeet-dev, libecore-dev and libembryo-dev when linking statically
Jan Luebbe (14):
Import upstream snapshot 0.5.0.040
Merge commit 'upstream-vcs/0.5.0.040' into upstream
Import upstream snapshot 0.5.0.041
Merge commit 'upstream-vcs/0.5.0.041' into upstream
Import upstream snapshot 0.5.0.042
Merge commit 'upstream-vcs/0.5.0.042' into upstream
Add Debian dir (thanks to Albin Tonnerre <albin.tonnerre at gmail.com>)
Update version to 0.5.0.042
Fix cleaning and doxgen
Add myself as uploader
Change doc-base section to Programming/C
Prepare for release
Apply patch from Niv Sardi <xaiki at debian.org> and change it to document the
Prepare for release (again)
Xavier Oswald (1):
* fix a typo in debian/control
andreas (1):
removed edje-config.in because of pkg-config
andyetitmoves (1):
Added cvsignore entries: mainly for *.spec files,
atmosphere (11):
Make a debian package of edje. I just ran sed on the ecore2 one and it
cmake instead of make, autogen instead of configure
apparently char is unsigned by default on my compiler
slight fixup
raster fixed this and it seems a tad more readable this way
get part geometry in embryo
as per raster's request, return values as ints
hook in get_geometry to the edje embryo calls
misinterpretted this, put it back in
doh, thanks dj2
more fixups for systems without ecore_evas_gl
azundris (5):
http://www.stefanmiller.de/nacktmull.html
* prelim edje.pc. emotion.pc wants this.
* more edje.pc
* don't depend on edb, we're not using it
* more fu
barbieri (17):
Use a more appropriate message for files not found than 'Unknown Format'
Messages are now passed to children groups (part.type=GROUP)
If object is swallow, apply color changes to swallowed object. Ack by rephorm.
Avoid doing math inside strings (quoted text), so I'm now enable to have part.description.text.text starting with "(".
Fix escaping by remembering state, allows "\\" to be used as text.
Support for selectable pointer_mode.
Fix segfault with part_recalc when no image_dir was specified.
Fix deletion of Edje with child "GROUP".
Fix edje_cc/edje_decc linkage in --as-needed mode
Fix edje calculation of max width for swallowed parts.
Set pass_events to 0 if "mouse_events: 1".
Fix calculation of maximum values for swallow parts.
Fix calculation of font size when text.fit_y is set.
Edje Emacs mode.
Add edje-mode.el to EXTRA_DIST
Edje mode improvements
Escape font name before writing it to fixed buffer.
boris (1):
Re-order params
codewarrior (1):
after hours of work, discusisons with rephorm, and raster, I found this!
davemds (1):
Update doxy style
devilhorns (6):
Fix typo
Fix typo spotted by odium :)
Patch from drexil to fix a leak in edje_data.c
Fix edje_color_class_list so that it's usable with enlightenment_remote again.
Properly fix edje_color_class_list.
Use memset on fdata for edje_color_class_list and edje_text_class_list based on
dj2 (3):
- don't segv if we can't open the file. this would die if you had a file with
- this needs to be initialized
- clean up the _edje_color_class_hash and _edje_text_class_hash on
doursse (14):
make edje binaries compile and allow the creation of the edje shared lib on windows. Remove some trailing spaces
revert that change
fix edje_cc on windows, remove some trailing spaces
_O_BINARY exists only on windows
Ecore_X and Ecore_Fb are not used
remove trailingspaces, include correctly alloca.h
remove trailingspaces
remove trailingspaces
implement a symlink function for Windows
configure improvements:
Do not use deprecated syntax anymore. Check versions of autoconf
* add mail in AC_INIT
update win32 port : use correct temp dir value and correct open function according to the compiler
fix typo in symlink (win32)
emte (2):
Added the change for BitBake the Standardized Openembedded Build System
Python is too incompetent to parse hyphens
englebass (18):
No need to alloc Edje_List_Foreach_Data
Typo.
No longer needed.
Remove debug comments.
Can't and shouldn't delete job on shutdown. The job handler might not get
Reset job incase the user reset's edje.
Remove unused file.
- Formatting.
Remove unused variable.
Remove padding when searching for character position.
Remove mysterious charachter.
Don't segfault on empty file.
Support spectr{a,um}
Don't clean up this job, ecore_job will clean it up itself.
add new keyword
Edje reference patch one (from dresb)
Edje reference patch two (from dresb)
formatting
glassy_ape (1):
Now colons are valid
horms (4):
build depenancy should be automake1.7, not automake-1.7
Various debian build cleanups
Need not depend on libc6-dev | libc-dev as it is in build-essential
The section of a library's -dev package is generally libdevel
kaethorn (7):
dependency and package content cleanups.
remove edje_ls from edje0-debug.
adjust dependencies. update to standards 3.6.2.2
fix include path.
fix include dir.
remove edje-config from install files.
remove duplicate line.
kwo (2):
Hush
Quiet.
lok (1):
Missing drag_step/page proto.
mej (14):
Sat May 22 00:13:33 2004 Michael Jennings (mej)
Sat May 22 00:19:51 2004 Michael Jennings (mej)
Thu Jul 8 13:45:01 2004 Michael Jennings (mej)
Sun Jul 25 17:45:53 2004 Michael Jennings (mej)
Wed Jan 5 00:03:30 2005 Michael Jennings (mej)
Sat Jan 15 20:55:10 2005 Michael Jennings (mej)
Sun Jan 23 22:30:28 2005 Michael Jennings (mej)
Thu Feb 3 18:33:51 2005 Michael Jennings (mej)
Mon Aug 29 12:27:11 2005 Michael Jennings (mej)
Thu Sep 8 17:12:14 2005 Michael Jennings (mej)
Sat Oct 15 00:09:36 2005 Michael Jennings (mej)
Mon Apr 10 17:05:40 2006 Michael Jennings (mej)
Fri May 19 14:52:44 2006 Michael Jennings (mej)
Tue May 30 12:32:21 2006 Michael Jennings (mej)
metrics (1):
Fix an issue where the button click is not correctly tracked.
moom (2):
* Add edje_object_part_object_get() to get the 'Evas_Object *'
* Correctly unswallow a swallowed-object when an Edje-object is deleted.
ningerso (2):
Use a hash for faster file global data lookup. If we hit a case where
awk portability fix in configure.in.
onefang (2):
.cvsignore++.
.cvsignore++
pfritz (4):
structure fits better the nature of descriptions, programms etc and looks better
add support for glow, far shadow and far soft shadow for text parts
sorry I forgot to commit this part of the patch
use a lookup table
quan74 (5):
Version number bumps
Makefile.am fixes for missing / renamed files in debian/
spec file autofoo'd
*** empty log message ***
make dist pkgs build again
raster (427):
in we go edje... no it doesnt do anything useful at the moment...
had wrong title & name/class!
Added more geometry calculation code. This function is beginning to get
more code & work on ejde. started on edje_cc.
edje_cc is beginning to do something. it parses the image section properly
parsing much nicer.. in fact starting to actually compile! :)
spread edje_cc code out a little... it's goign to get big now. put parse
EVEN better parsing! :) lots of NULL handlers need filling in. i need to unify
cleaner. centralise type parsing... and error checking
mem allocs guaranteed now... they are wrapped and errors handled centrally
a few more handlers. i think next i need to add a verbose mode for the parser
we parse.... MORE! :) order is now irrelevant of how you delcare images or
edje_cc can now fully parse my test edje input "test.edc" file and build a
some default "sane" values to save a lot of typing in edje .edc files + a
some of my other changes from this morning go into edje. more defaults. more
the edje test program displays an edje .eet file now! build it all then:
workin g on the code... looking cleaner. properly have shared data hashes.
Adding code to handle programs, signals, emissions, loops, animation, timers
1 copy & paste bug in edje_cc. fixed non animated programs
emissions now for the part collection soruce is now "". fixed a tinsy parser
to doubles...
a fixme list added, and 1 segv removed! :)
stop action supported. parser/compiler is accurate there in not needing more
signal emit action added
more fixme's added, more fixed up. external callbakcs are now in. calling
demo app now does its own internal edje resizing/moving.. redying for
edje test program now has no defaults. it requires you to provide all edje
fixed bug where instant changes didnt happen if no animation was happening.
freeze/thaw stuff in for more optimum recalculation
last commit for the weekend. some more fixmes.
I lied. 1 more commit for the weekend. don't segv if the edje part isn't
1 mega-file (edje_main.c) split up into saner-to-manage files.
more fixme's fixed. more fixme's added. edje is moving.
readme...
more useful README
ciabot...
ciabot.pl test?
Edje now has some semblance of basic documentation.
oops. edje! edje!
consistant signal names! :)
went through some more of the fixme list and fixed 'em
lets keep geometry of objects in the canvas in canvas units (double's) :)
nuke some FIXME's. smooth parameter now added for fills. program infinite
"random" ranged delays are now allowed before a program starts... in fact any
work has been done while i've been away. i just need to put this in cvs. many
api change... yes i know. sorry! but it's early days. this is much cleaner
build cross-compiled version. edje on ipaq's :)
more arm niceness
load printfs removed....
oops bug... :)
yaya! edje_cc total output statistics :)
tect & colorclasses - in theory, work. must test!
edje can now swallow other evas objects... including... other EDJES! :)
break out of while loops!
1. types declared as int, not uchar. thansk rbdpgn! :)
global freeze/thaw. much more optimial for COMPLEX edjed that take a fair bit
you can query the state of a part now
optimizing. Glassy's matrix.eet works MUCH faster now. basically edje was not
min.max size properties now per edje collection can be specified in the edje
handy conf script
build packages
0 time rpograms stop too now
errr stop proggies & emissions before starts
some notes for me about edje program expansions..
more features... working on fixme list.
fix 1 memleak...
memory corruption.. well just not cleaning up things in the cache hash... fix
stop pending proggies
more data
fix after programs to work p[roperly with delayed actions
respect minw/h on load of edje file...
you can add arbitary key/value pair strings to either 1 the edje file .eet as
cleanup printf's
optional output file spec
dang! crap. that was silly nwo wasn't it! FIX!
and allow sinstantaneous programs to have afters too...
oops. that was a fuckup! :)
fix segv on del! :)
make ending programs safe while iterating
separate x & y relative "to" params. the old .edc format is still supported
and an example in the test.edc
fix parser... :)
clip to swallower's clipper
the basics of dragables work. signals are generated properly. there's calls
IMMENSELY evil!
and remove printf's
no more MEH!
edje_cc now supports cpp, if you have it installed, so you can #include,
dragables are about complete now.. with new actions, page and step values....
safety thing... :)
setting extern min/max size overrides min/max sizes implied by objects
added geometry flags
oops - swallow went bad.
clean...
update docs...
match changes made to evas.
remove lots of warnings
load error fetching
oops. update.
atmos' patch! :)
fix dragables! eeek! :)
oops... fix!
bah! actually have entry in eet struct memebers to save/load for repeat :)
thanks tillsan! :) leak found. leak fixed. happy pants!
add : in as a delimiter so you can do:
examples change - test shwos 1 delimiter form, e_logo.edc, another
"*" matches ANYTHING including ""
cleanies
support for inlining ttf's in edje's so you can ship your fonts along too! :)
remove conf file, update README
aspect calc buggie. fix fix
and reset those vars when done
a possible leak plugged.
cross-compile cleanly!
fixie doccie
gendoc update
fix build deps
fixed a leak... patches from trill... need to go thru my mail....
yuri's math code for the parser!
init drag step on load
embryo support... started. :)
and.... edje_cc can take small code snippets in script{} areas and compile
remove that debug... dont need that.
some notes for myself...
notes for me on what apis to export...
busy adding some MEAT to edje's embryo script support. just added
string replacement with ID's in edje_cc for verbatim code snippets... see
cleaner...
faster lookups for programs & parts - saves a lot of list walking
alloca...
ok - examples of how to hook some embryo calls in - help appreciated with the
timers and animators are now supported from the embryo scripts - WOOO!
survive make distcheck...
oopsie - NOW its fixed! :)
embryo_cc line numebrs match up with edje .edc ones now - much better foir
abort edje_cc compile if embryo is not happy
not evas_coord safe! fix. :)
hmm - for now dont over-optimise...
make default clip smaller. we weem to be overflowing calculations to do with
well - we can go back to where we were before. type overflow issue
aaaaaaah #'s form cpp in embryo verbatim snippets - strip them.
hmm - dont delete swallowed child objects... too much implicit stuff..
clean up more...
clean up callbacks and smart stuff too
and put back some calls we lost in the patch/diff... :)
cross-compile fix
oops - double run - remove. my bad. mis-merge
just when you thoguht all was stable in edje land... i added a tag and field!
did i forget to remove the printf's again? DAMNIT!
hrrrrm - aaah this is better!
1 less printf too :)
dont need {}'s
protect against an infinite loop...
add stuff for evoak... and edje list vars... :)
edje arlo patches
embryo scritps update edje properly now?
autofool cleanups... argh! i hate autofools!
aqdd message_queue. i will move signal emits to go onto the message queue
1. fix a leak. i was naughty. spank spank
example of message handler in test.edc... a start.
werd up to ma homies yo... err - i mean - message queue now uses more
fully supported message queues now. that was a chunk of work!
update comments
oops - fix alignment just a little
BUGGERS.... squished. why didnt anyone spot this? :-(
get relative mouse pos relative to the edje... and add a -fill opt to edje
more container notes/skellies and allow aspect prefernce of BOTH (both axes
1. add SWALLOW object to be more explicit...
more comments
more container stubs...
edje done the same was as everything else...
aaaaaaaaaaaah message queue related fix.
more container fillage
forgot to emti program start/stop if the action is invalid/empty
1. more container work
oops got errors the wrong way
containers on hold for now...
minw/h didnt let given minw/h for item override text minw/h if the override
edje.inc was wrong (so was the comment)
edje decompiling works now! you can decompile an edje .eet back into source
isspace.. and oops infinite loopies
potential secuirty violations covered up.
rephorm missed another place fonts were being set, and forgot to include the
handy edje_recc t recompile an edje .eet from now on :)
this should make a lot of themers happy.
help fixed
if promoting compression we need to guess a new compression quality... :)
oops - if the edje ONLy has a swallow part.. bad things (tm) happen :)
use animators... :)
fix fonts in OTHEr states than default! :)
expose the ability to process the message queue any time :)
more message queue calls
no no. do it this way! :)
add some mroe comments for bored people who want to work on this :)
use Ecore_Evas info on engine availability :)
font size change
font size updates... again
1. note all the /* dup */ entries in edje_cc_handlers.c - we now are lenient
actually... error on any source file not being able to be found, decoded or
and dont use buf! :)
more compile options.. allow image sot be FORCIBLY rescaled so u can include
allow text source and text style sources to be specified.. this lets multiple
accessing wrong memmber. fixed.
limit embryo scritps form being able to go into finitie or waay too long loops.
and on file set - live in the RIGHT layer
dup tween list properly :) and dup lookups on text stuff
build cleaning
only do that once...
extar dist too for edje_recc
no ldflags
buildie cleanies
gawd that was bad!!!!
edje_cc supports -Ddefine=val style stuff - via cpp and like cpp
oops. edje text format cache didnt take into account alignment... fix.
more efficient :) (snprintf is a bit of a sledgehammer - if we can do it
no ldflags
out out ye demons of stupidity!
fixie fixie long time.
openembedded build files... this makes life so easy to build efl for
cleansies
oops.
cleanies
be nice!
fix initial dragable values
fix up dragables for panners panning element bigger than confine
fix page jumping!
hidden symbol support
able to force a calc if we want to...
8192 is as good as SIZE_MAX - its arbitary...
deferred message bug. fixeed.
actually detect x...
handle double/triple click again
testing the edje message queue.. revealed.... problems... and.. now....
use new extended event info
add edje_test to spec
dont SIGFPE
an edje thumbnailer executable... :) requires up to date ecore and evas
finx dir of input file, strip off file to find dir component and -I that when
oops arg order... fix.
fil! not filename! :)
and handle NULL text params
a workaround for myseterious job loss...
working on fixing job loss...
change versions. some went DOWN - because they arent releases and i'm trying
1. e17's init icons get put into an e_box for arrangement nad the init splash
help update
-fb too
suport fontset appends
asparagus time for some of the core efl bits...
asparagus!
support unfilled centers...
file and collection cache - this should remove a lot of file io... :)
make e17 pass make distcheck....
fix some leaks
asparagus!
spew out errors
full asparagus
commented out debug
fix segv
let programmers know they are NORTY!
fix potential bugs in msg queu processing
fix up fnmatch weirdness
asparagus...
aspara!
just in case....
dont need that anymore
it's been a while, so it was time for some ASPARAGUS on our plates
pavel's patch. :)
spec
ug callback that deletes the object that is bing walked for callbacks...
apps/e/enlightenment.spec CVS: apps/entice/configure.in
2nd asapargus for the weekend :)
dont segv on prog with no name (illegal anyway)
oops -- &&
max size respecting yo
temporary smart stacking workaround until we fix it properly with internal
and edje started supporting.... TEXTBLOCK2!!!!!
handle embedded fotns for textblocks
printf--
allow more tags for textblock obj's
fix min size calc
blame: TILMAN:
hmm - make min size calculation mroe accurate and handle corner cases...
comment--
handle textblock fixed with and variable height as a result formatting with
allow correct values for fill offset - too rtestrictive
god man that was a bad typo!
a bit of asparagus action for shits & giggles
fix typos in edje docs patch from martin :)
asparagus!
i dunno - be more paranoid and shit!
ok - be mroe paranoid!
be more paranoid... ?
after some quiet on the western front - asparagus.
code3 clean - move code to special files and maybe fix a bug?
asparagus. and make distcheck passes again.
embryo debug stuff from till
allow for xr engine for testing/display of edjes
asparagus - again. lots fo leak fixes and other fixes have been happening, so
add embryo api to get the state of a part (handy eh?) :)
the first wave of evas smart object modes. commit hell #1. :) other things
no more old textblock - gone. completely. tb2 renamed (as threatened) to
minor leak--
remove excess calcs.
dragables are back (broken them accidentally)
fix det leeek
add elipsis direct control
we dont need no steenkin move signals
edje no longer needs imlib2.
asparagus time! :)
no point disallowing sets.
handle NULL strings better - need to handle mroe NULLS though
more NULL string handling
clean up some null string stuff
dont dup empty strings :)
push/pop vm on the fly - saves a chunk of ram.
text badness fix
noneed for offset since its inside the object now
more fixes :)
asparagus!
no need to alloc path and part strings if its cached and thus shared
fix edje embryo scripts.
asparagus. there were issues. thsi syncs versions now too.
something is broken with text classes - disable for now.
use stringshare! :)
make more use of stringshare and eet alloc methods
oops - using free() instead of stringshare_del(). fixed.
null ptr checks
handle no name part.
dave's catch of null strings
different cpp invokation attempt
actually make symbol hiding work!
visibility -> off
big fat asparagus!
handle more returns with more info
handle more returns with more info
fix spec
pass distcheck...
asparagus!
asparagus!!!
cvs is back up.. time for some asparagus!
THATS what it should look like :)
fix up edje packaging deps
gasp! packages actually build!
edje_ls should be in the debug package - not bin
fix maintainer
another lib packages right.
fix edje packages to install
space --
dont set drag val if user already dragging
1. fix ecore to build again (xcomp - u need to cvs add and commit your 2 new
edje_decc works properly with default theme (as long as u rememeber to pass
1. autofoot patches.
fix blah-config includes
another good time for good asparagus :) snapshot.
remove openembedde pkg info - old and dead
zigs patch for edje adding more embryo script power to set state vals
cedrics font hash lookup speedup :)
asparagus time...
revert cedrics patch for font stuff - it breaks textblocks
asparagus time!
edje is now also prefix aware.
edje prefix fixes - oops - was broken
fix bin/lib prefix - pass as -D
missing a style sety when calculating text size... style padding was wrong.
aspara!!!
some prefix fixes for env vars
allow code to specify an aspect for swallowed objects
more external aspect support - all working nicely now
fix up aspect calcs for both axes controlling
release the asparagus... another snapshot! :)
add lib deps back in - seemingly i found a modern gcc/ld that doesnt seem to
fix some aspect cacls
fix escape handling in quotes! :)
fix segv
less noisy debug
try blake's suggestion.
same as previous commit.
premul fixes from jose
add max text size for text objects - object wont get bigger than its text
leaking spectrum colors man! :)
indenting
pass make distcheck.
asparagus! (finally after all the radical changes of late)
aspara!
asparagus - we have fixed a lot of things this week, so before the weekend of
debian packagnig mofo
remove data
removing...
remove...
remove even more
asparagus.
and actually allow "USER" images (ie don't encode them into the .edj - use an
lotys of changes - mayaswell asparagus - and the start of some of my fm
1. fixed some inefficient edje loading - moved checks to the cache code where
fix valgrind complaint - oops. unreffed before the file del!
solve the edje_decc file thing
tilman's edje font metrics patch
swallow parent recalc etc. in.
1. removed mosue_events 0 on swallows - as this now changes behavior. normalise
cedric's patches.
asparagus - pass distcheck.
less stack for the parser - scratchbox hates the big stack usage :)
ASPARAGUS!
save a few bytes by packing structures better... simple savings :)
1. fix a lot of things so they pass make distcheck - so many things have
constness patches from cedric
formattign and allocs...
adrunko's precise event patch
DONT set color - bad. see comments
data struct should match eet description - an enum may not be a char - it may
move a chunk of stuff to use evas_smart_class_new() - evas_smart_new() will
god damn.. chasing that bad mem access was tricky. found it after futzing
handle USER images properly
asparagus!
retain text styles fron theme if they are specificed
fix up edje to calc using the same font as the textclass.
fix edje to get the current default object size as initial size data.
cedric's ecore_job_init/shtudown patches.
bizarre - text sources were broken - but never appeared in e - somehow they
As requested -
terminate strings properly?
did some tests - this seems to be faster and seems to match correctly in some
remvoe fnmatch checks
textblock max size option in
edje data file patches in - not fill though.
asparagus! also pass distcheck and have common autofoo init that is
rbdpngn (17):
A simple program that lists collections in files, can take multiple files as
Slight re-organization of color classes, now can be set per-description rather
Spelling cleanup, though it feels kinda dirty "correcting spelling" on a word
Moved the text_class into the description's text section.
Global text and color class manipulation. Settings on individual edje's will
Avoid memory corruption by cleaning up parts at the correct time.
Portable version check for libtool.
The path for the libedje.la needs to indicate that it's in the same directory
Alloca check for the compiler.
Hackish, but it fixes the problem with stupid cpp's.
Rename edje.h in the src/bin so it doesn't conflict with Edje.h on
Make the enum typedef seperate to appease some annoying compilers.
glibtoolize
Set load errors in more cases.
Use the style features available in evas text objects to avoid creating more
Use the second updated patch for the cpp problem. Originally submitted by Romu
Stringshare free fix from David Stevenson.
rephorm (51):
"clicked" signals now emitted in both of the following cases:
Edje part geometry querying added.
stop after the first part with that name. (no need to go through the rest of list).
Make the edje clipper HUGE. Now edje's can go beyond their bounds. (i.e. relative, -.5, -.5)
edje_object_file_set() now returns 1 on success and 0 on failure.
The beginnings of Edje_Edit, a lib for editing edje .eet files.
gah! don't know how that one slipped by. sorry :)
fill in some embryo calls.
make it a bit more interesting...
lets not confuse people :)
ok. that definitely ain't the right way to do those. back it out...
Add a Font_Directory to Edje_File
Ok. Completing the commit from last night. Sorry about that (and thanks for cleaning up after me raster). Nothing functional changed between raster's last commit and this, just making things more consistent.
include edje_recc in EXTRA_DIST
actually delete ...edje_tmp at the start (add a -r since its a dir)
Check if descriptors exist (i.e. if edje_init() was run) before trying to use them.
ref and unref the edje when blocking/unblocking and creating/deleting messages to avoid accessing free'd memory if an edje is deleted from an edje signal callback.
Merge contents of Edje_Edit.h back into edje_private.h
use color classes when tweening.
color class blocks in edje.
color class blocks in edje.
Add edje_color_class_del(), edje_object_color_class_del() and edje_color_class_list().
Add edje_color_class_del(), edje_object_color_class_del() and edje_color_class_list().
mark edjes as dirty when their color class is deleted
damn wacked out indent style :)
Require that groups have a name.
don't free color class stuff in text_on_del, it gets done in color_class_on_del...
one more place that NULLS out the shared color class on delete.
Add initial support for edje gradients. See data/src/gradient.edc for an example.
oops. thanks jose :)
nothing to see here... *whistling*
add some validation to spectra
... more whistling
don't check spectra if we don't have any... (fix segv)
add linear gradient specific fill options
allow rel1.relative: 0 0; notation
convenience function for testing if groups are in an edje
add group aliases.
its possible for _pos_set() to be called after the final mouse move, but before the mouse up event when dragging a dragable. in this case, the calc immediately after would be incorrect. this fixes the bug noticable in winter sliders being jumpy. (the default theme did not exhibit the bug because the pulsing glow anim cause subsequent recalcs to immediately correct things)
fix deleting of signal callbacks with empty strings for signal or source (they're stored internally as NULL's)
if you set a file on an edje object that already has swallows, they will now be tracked and re-swallowed
expand a few docs
some more docs
docs
if requested scroll size is less than min, use min size.
only clear and re-apply gradients when the spectrum actually changes
Remove part list, using table everywhere instead
Add GROUP Parts to edje.
AUTHOR++
add cedric bail's fill.type patch.
yikes
sebastid (21):
viminfo
Silence
Remove .
Don't include $(includedir) in searchpath, find headers in this source
If the version from config has something after x.y.z, drop it.
Use AC_PATH_GENERIC, it's a nice macro!
We need the embryo includedir to build the binaries.
Export evas CFLAGS, because exported edje headers include Evas.h
Not needed?
Add missing headers, and define MIN
Add string.h for strcmp()
Fixes from Ryan Little
Unused
Don't use strlen to check if a string is empty.
Fix EAPI warnings.
Unused func.
filename is const
Fix warnings
If the list node is known, use evas_list_remove_list
Remove unused variable.
Remove unused variable.
shadoi (9):
Uncomment edje_edit stuff so iconbar will build again.
Fix dependencies
Merge the edje0-test package with edje-bin. Fixup dependencies a bit.
Buh-bye.
Add debian package files to EXTRA_DIST
Add a DEBIANSTAMP variable and generate the debian/changelog
Populates the version automatically
Merge the -test package with this one.
Found a better way to do this. Remove DEBIANSTAMP
shorne (3):
Add ability for text blocks to use text classes
Edje textblock cleanups fixes
Add negative font size handling in text_classes
sleuth (1):
.spec binary additions
stffrdhrn (2):
- If its not really edj, or if its corrupt dont crash
Edje bug fixes:
swielinga (7):
Big overhaul of a lot of the debian packages
Updated most of the debian packaging. Everything I have missed is out of date,
You forgot to add $(DESTDIR), buddy :-)
Now the edc vim syntax file for edje sources actually shows up in the deb.
Updated edje debian packaging.
Big debian packaging update.
Slight update for the debian packaging of imlib2, imlib2_loaders, ecore, edje;
technikolor (4):
Minor tweek to edje viewer. Adds font path for local directory "fonts/".
Fixed acouple syntax errors to avoid misleading new folks.
Atmos' excellent Edje EDC template, ready for all to use. All hail atmos.
Don\'t assume cpp and gcc are always in /usr/bin... absolute paths removed
tilman (12):
don't check for valid timers/animators in timer_cancel/anim_cancel. the gain:pain ratio didn't turn out as good as expected - it's a PITA to make sure you only cancel valid timers/animators in your embryo code.
edje depends on embryo
define datarootdir
switched edje over to pkg-config. almost. ecore is still detected by ecore-config, cause that one's gonna be pain.
use pkg-config to detect ecore
oops, don't reference ECORE_LIBS
edje requires ecore-job
back out the embryoincludedir stuff again. it broke distcheck, and the old way wasn't that bad ;D
move stuff to Libs.private
remove an unused variable
fixed some const char*/char* confusion.
hooked up the 'alternate font metrics stuff' in edje_cc.
tsauerbeck (151):
more memory management improvements. applications should call edje_shutdown() now.
more memory management cleanups. fixed some warnings.
edje is valgrind-clean now! :)
call edje_shutdown()
strcpy() doesn't allow copying overlapping strings
gcc, shut up!
call edje_shutdown()
call eet_init() and eet_shutdown()
added .cvsignore files
added valgrind suppression file for Edje
removed some unused variables
edje_init() shouldn't be called from edje_object_add()
fixed a memleak
don't segfault on a missing default description
updated documentation for --with-vim
fixed dist stuff
this was starting to annoy me :P
it's called edje, not eet
added text to cLabel
warning fixes
added SWALLOW to constants
i think we can get rid of the script debugging spam now
fixed a potential NULL pointer deref
add missing AM_CONDITIONAL call
increment EDJE_FILE_VERSION because of the recent font directory additions
fixed an endless loop caused by unexpected preprocessor commands. tweaked the code a bit so its more readable, hopefully.
when a font cannot be loaded (because -fd was omitted, for example), just print a warning instead of an error message. i.e., fonts are treated the same way images are.
added edje_decc to cvsignore
don't print constant data
smooth
part names need to be unique in a collection, too
warning fixes
String concatenation ("foo""bar" -> "foobar") now works when used in arguments to PART/PROGRAM/IMAGE lookups, e.g. PART:"foo""bar" will look up the part id of "foobar".
portability improvements
adapted to embryo_cc changes
boolean values can now be specified in the following ways: 0, false, off resp 1, true, on (case insensitive)
the 'low level' edje var list functions decrement id by EDJE_VAR_MAGIC_BASE, too, so make sure we increment it before to get a correct value in the end
warnings
int 'list' ops
make sure source file data is terminated properly
check that the input file exists. make sure that the input file doesn't equal the output file
tweaked font sizes
New feature: part descriptions can inherit attributes from other descriptions of the same part. All attributes are inherited, except description name and value, of course.
better vim modelines for raster code
deprecated description.border, use description.image.border instead, since border only applies to image parts.
image{} and fill{} attributes are only allowed for IMAGE parts. text{} attributes are only allowed for TEXT parts
catch missing targets in STATE_SET programs, for example
fixed a memleak
mingw portability, step 1
imlib2 is optional for edje_cc now. simplified autofoo
description.border -> description.image.border
open files properly
updated Doxyfiles
implemented string list calls. untested, beware.
cleaned up edje_cc's sanity checking for parts and programs. image parts need to have valid image attributes now.
include collection id in error messages
edje_cc now checks the number of arguments for each statement
fixed arg count checking for program.action
edje_cc is getting even more picky. arguments may not be separated by commas any more. simplified code a bit.
fixed some more edje_cc error messages
don't overwrite source file data with the terminating NUL
fixed edc source inclusion in edje_source edd.
only get the directory of the current file once
initialize the random number generator
programs may not specify 'target' before 'action'
implemented float list ops, replace_int and replace_str
doxygen fixes
there's no need to copy the string byte by byte, use snprintf instead
don't include mman.h
mingw fixes
'custom' is an invalid description name
fixed insert_float()
reduced code duplication
Oops
Oops2
tweak
fixed timer/animator mixup
uuh, i must have forgotten to commit this earlier
added basic support for custom state magic. only description.align can be changed at runtime so far.
don't invert part alignment when the part is smaller than the min size. this means that an alignment of 0.0 will *always* be top/left.
warning fix
these aren't needed
cleaned up autotools setup some more
minor cleanups
trying to make the code more readable
typo
minor cleanups
warning fixes
simplified _edje_part_description_apply(). one code block was executed twice for no reason apparently. removed some condition checks that aren't needed.
simplification. raster, have a look at that FIXME, please.
fixed a buffer overflow vulnerability
split _edje_text_recalc_apply()
unbreak edje_text_fit_x
more splitting
more cleanups
even more splitting
oops, orig_len is size_t now
snprintf goooooood :)
raster's snprintf() 'fix' applies here, too
added support for a few more properties for the custom state stuff
removed some unused variables
a nicer way of copying strings :)
formatting
force a recalc when set_state() has been called (Embryo)
some janitor work. no new bugs this time ;)
smarter _edje_dragable_pos_set()
grumble, i suck
edje_cc now knows about parens, which means that "(x + y + z)" is treated as *one* token.
switched to floating-point math. yell if you think this causes trouble.
removed support for the (so far deprecated) description.border keyword
only increment id_count if we could allocate memory for a new timer/animator
always append new animators to _edje_anim_list so we can have more than one animator in an edje/in multiple edjes
fix exit status in case of an error
add support for datadir to edje.pc.in resp edje-config.in, so 3rd party apps can find out where edje's data files (most important: edje.inc) are located
moved the include directory
updated debian files because of recent directory shuffle
output EET file can now be specified by -o, too. use that mode in build.sh
Decompiler improvements: Edjes now contain a 'compiler' entry, which is set to the executable that's used to compile an Edje. edje_decc does some sanity checking on this string, to prevent exploitation.
tell the user not to trust the build scripts that are generated generated by edje_decc
use proper english ;)
don't hardcode the root filename for decompilation
use proper long-option names
warning fixes
fixed text fitting on both the x and the y axis
no need to assign size at that point again, it's done in the loop
fixed a memory leak that was introduced with the recent decompiler changes
the default extension for Edje EET files is 'edj' now
updated cvsignore files
fixed a potential segfault
tweak. if edje_fontset_append_set() is passed NULL, free the global variable and set it to NULL instead of not doing anything
sanitize border scaling middle something logic. if the property needs to set to 'true' to be disabled, we better name it that way. this breaks eet compatibility, but since the previous commit is still young, i don't increment the version ids
optimization. to check whether a string is empty (len == 0), we don't need to run strlen, checking whether the first char is the terminating NUL is enough
quit on 'q'
removed unused function
don't die if embryo_cc brabbles about warnings in the embryo code
use getsarg()
fix a silly bug that caused a memleak (i think)
made edje_file_collection_list() check its argument before trying to do anything with it
the evas_hash_foreach callback is supposed to return a value ;)
evil hack to find include statements not at the beginning of the line, too
cleanup. formatting and stuff
added the 'fixed' keyword
textblock keywords
don't access freed memory
warnings suck
set walking_actions to 1 while we're walking the actions list. this fixes an embryo induced segfault, horray.
warn about cancelling non-existing timers, too
warning fixes
support empty includes
the custom state code was still using strdup, but it needs to use the stringshare api
custom state: i think we should inherit the parent's text style, too
urandom (1):
text_class_del/list functions
vacuum (1):
Respect config.h
vapier (10):
allow for setting of env var to prevent running ./configure like the older autogen scripts allowed
remove AC_CANONICAL_TARGET since it isnt used/needed
make sure we abort if autotools fail
fix underquoting bugs
fix multiple definition of EXTRA_DIST
have libtool copy files, not symlink them
EXTRA_SCRIPTS cant be substituted, so revert that change
add error checking to all autogen scripts
remove pointless [] around if tests
touchup help output
xcomputerman (9):
No more warnings! No more warnings I say!
- Fix warnings
Out, ye cursed critters!
Giving a vim syntax file its own module directory is just ridiculous.
Add configure option to manually specify vim directory
Fix gendoc script to include the right filename, and remove unnecessary find command
Add script statement to EDC syntax file
And remove C bitfield syntax so that new keyword notation (":" instead of ",") will work right
When outputting parse errors, give the correct line number - how did this whole offset thing happen anyway?
xenith (1):
If cvsignores were beers, I'd be dead by now.
-----------------------------------------------------------------------
--
Enlightenment DR17 graphical layout and animation library
More information about the Pkg-e-commits
mailing list