|
Allegro 5.0 beta released! |
Peter Wang
Member #23
April 2000
|
http://sourceforge.net/projects/alleg/files/allegro-unstable/4.9.19/ Quote:
Changes from 4.9.18 to 4.9.19 (April 2010) The main developers this time were: Milan Graphics: - Implemented support for transformations for memory bitmaps. - Transformations now work properly when the target bitmap is a sub-bitmap in - Use the memory drawers when the source bitmap is the backbuffer with the - Make al_put_pixel clip even if the bitmap is locked, which was the reason - Added al_put_blended_pixel, the blended version of al_put_pixel. - Sub bitmaps of sub bitmaps must be clipped to the first parent. - Don't clear the transformation when setting the target bitmap in OpenGL. - Implemented ALLEGRO_NOFRAME and ALLEGRO_FULLSCREEN_WINDOW in WGL. - Set the ALLEGRO_DISPLAY->refresh_rate variable for fullscreen modes under D3D. - Make d3d_clear return immediately if the display is in a lost state. - Rewrote the function that reads the OpenGL version so it works for - Check for framebuffer extension on iPhone properly. - Fixed locking bugs on iPhone. allegro_ttf works now. Input: - Removed al_set_mouse_range. - Don't call al_get_mouse_state if the mouse driver isn't installed (Windows). - Send events even when the mouse cursor leaves the window, while any buttons - Allow mouse presses and accelerometer data simultaneously. (iPhone) File I/O: - Optimise al_fread{16,32}* by using only one call to al_fread each. - Optimise al_fgetc() for stdio backend. Path: - Fix an infinite loop in _find_executable_file when searching for the Primitives addon: - Made the software driver for the primitives addon check for blending - Made the D3D driver thread-safe. The whole addon should be thread-safe now. - The addon now officially supports 3D vertices, even though the software - Changed the way the primitives addon handles the OpenGL state (fixes a few Image addon: - Optimised BMP, PCX, TGA loaders. - Fix loading 16-bit BMP files. - Fix loading grayscale TGA images. - Nial Giacomelli fixed a bug where images could be corrupt using the native Audio addon: - Add al_is_audio_installed. - Fix al_attach_sample_instance_to_mixer for int16 mixers. - Implement attaching an INT16 mixer to another INT16 mixer. - Handle conversion when an INT16 mixer is attached to a UINT16 voice. Build system: - Add an option to disable Apple native image loader (iPhone and OS X). - Add ttf addon target to iPhone xcode project. Examples: - Special contribution from Shawn "the Progenitor" Hargreaves.
|
komons
Member #11,083
June 2009
|
Just please. Compile dynamic library for me ;( |
Dario ff
Member #10,065
August 2008
|
Is the title the only April Fool's joke and the 4.9.19 release is real? Or is something else going on? It compiled and works fine, so I suppose there's no problem(cool speedhack entry by Shawn BTW). TranslatorHack 2010, a human translation chain in a.cc. |
Todd Cope
Member #998
November 2000
|
It would be funny if Allegro 5 ended up being released on April 1. |
Milan Mimica
Member #3,877
September 2003
|
Quote: The main developers this time were: Milan Milan aha, that's the joke
-- |
Evert
Member #794
November 2000
|
I did have a "huh?" moment while looking through the list of contributors as well. |
KnightWhoSaysNi
Member #7,339
June 2006
|
In the 4.9.18 release the tga font examples looked blurry and messed up and the true type font example was fine. In this release its the opposite, the tga fonts are fine and the true type font font example is messed up. {"name":"600879","src":"\/\/djungxnpq2nug.cloudfront.net\/image\/cache\/e\/2\/e23048117578e1eaed914e2431742121.png","w":648,"h":509,"tn":"\/\/djungxnpq2nug.cloudfront.net\/image\/cache\/e\/2\/e23048117578e1eaed914e2431742121"} |
Evert
Member #794
November 2000
|
What happens if you change the example to use OpenGL instead of D3D? |
SiegeLord
Member #7,827
October 2006
|
I'll wager that this was fixed in the commit yesterday. TTF addon uses transformations internally (sets them to identity, to be precise) which resulted in the 0.5 pixel shift being applied twice, once for the D3D drawing code and once for the transformations. EDIT: ALLEGRO_TRANSFORM trans; al_built_transform(&trans, 0.5, 0.5, 1, 1, 0); al_use_transform(&trans);
"For in much wisdom is much grief: and he that increases knowledge increases sorrow."-Ecclesiastes 1:18 |
Trent Gamblin
Member #261
April 2000
|
4.9.19 binaries are now available on allegro5.org.
|
KnightWhoSaysNi
Member #7,339
June 2006
|
Sorry for the long delay in responding. Evert said: What happens if you change the example to use OpenGL instead of D3D? It looks even more messed up. {"name":"600922","src":"\/\/djungxnpq2nug.cloudfront.net\/image\/cache\/0\/e\/0ea433454136f34f6714ef31c3980307.png","w":652,"h":514,"tn":"\/\/djungxnpq2nug.cloudfront.net\/image\/cache\/0\/e\/0ea433454136f34f6714ef31c3980307"} However I just added: al_set_new_display_flags(ALLEGRO_OPENGL); al_set_new_display_option(ALLEGRO_DEPTH_SIZE, 16, ALLEGRO_SUGGEST); before i made the display; was there anything else I needed to do? SiegeLord said:
ALLEGRO_TRANSFORM trans; That fixed it and it looks normal now |
kenmasters1976
Member #8,794
July 2007
|
I've just built the plain 4.9.19 release (not the latest SVN) and I see an old issue with drawing rectangles: {"name":"600939","src":"\/\/djungxnpq2nug.cloudfront.net\/image\/cache\/b\/0\/b02331d8f6d1e273019d134423c22196.png","w":648,"h":514,"tn":"\/\/djungxnpq2nug.cloudfront.net\/image\/cache\/b\/0\/b02331d8f6d1e273019d134423c22196"} Maybe it's fixed in the latest SVN?, maybe it only affects old integrated Intel cards since rectangles in the other screenshots posted in this thread appear correctly?. Oh, and it only seems to happen with the Direct3D driver.
|
Michał Cichoń
Member #11,736
March 2010
|
This issue should be fixed right now. This is the proof. Direct3D is used to render this image. "God starts from scratch too" |
Elias
Member #358
May 2000
|
It's missing a pixel in the lower left corner. However I guess with thickness=1 it will look right. -- |
SiegeLord
Member #7,827
October 2006
|
kenmasters1976 said: Maybe it's fixed in the latest SVN? This is the same problem as the TTF fonts. Use the same workabout for it if it affects your program (only for 4.9.19). "For in much wisdom is much grief: and he that increases knowledge increases sorrow."-Ecclesiastes 1:18 |
Michał Cichoń
Member #11,736
March 2010
|
[post removed] Wrong thread. :} "God starts from scratch too" |
kenmasters1976
Member #8,794
July 2007
|
SiegeLord said: This is the same problem as the TTF fonts. Use the same workabout for it if it affects your program (only for 4.9.19). OK, fine. Thanks.
|
Kibiz0r
Member #6,203
September 2005
|
I'm getting an infinite loop here: The part between the first and second _al_fill_display_settings repeats several thousand times. ... #14516 0x00e6362c in _al_fill_display_settings (ref=0x16) at /Users/kibiyama/svn/allegro5/src/display_settings.c:161 #14517 0x00e7f3ae in tls_get () at /Users/kibiyama/svn/allegro5/src/tls.c:87 #14518 0x00e7f60b in _al_get_new_display_settings () at /Users/kibiyama/svn/allegro5/src/tls.c:258 #14519 0x00e62a98 in al_set_new_display_option (option=22, value=0, importance=1) at /Users/kibiyama/svn/allegro5/src/display_settings.c:33 #14520 0x00e6362c in _al_fill_display_settings (ref=0x16) at /Users/kibiyama/svn/allegro5/src/display_settings.c:161 #14521 0x00e63788 in _al_fill_display_settings (ref=0xb030d008) at /Users/kibiyama/svn/allegro5/src/display_settings.c:155 #14522 0x00e61cbc in al_create_display (w=800, h=600) at /Users/kibiyama/svn/allegro5/src/display_new.c:56 #14523 0x00c69d73 in ?? () #14524 0x00c69cce in ?? () #14525 0x00c624a3 in ?? () #14526 0x00c6234c in ?? () #14527 0x00cd508e in ?? () #14528 0x00e8fb04 in call_user_main [inlined] () at /Users/kibiyama/svn/allegro5/src/macosx/osx_app_delegate.m:213 #14529 0x00e8fb04 in +[AllegroAppDelegate app_main:] (self=0xebc3b4, _cmd=0xea6636, arg=0x0) at /Users/kibiyama/svn/allegro5/src/macosx/osx_app_delegate.m:225 #14530 0x909488d8 in -[NSThread main] () #14531 0x90948888 in __NSThread__main__ () #14532 0x97508fbd in _pthread_start () #14533 0x97508e42 in thread_start () Happens in SVN as well, with the only difference in the stack trace being display_new.c:57 rather than display_new.c:56. --- |
ImLeftFooted
Member #3,935
October 2003
|
What did Shawn do? |
Matthew Leverton
Supreme Loser
January 1999
|
His 1999 speedhack game was included as a demo. |
Milan Mimica
Member #3,877
September 2003
|
Kibiz0r can you retry the SVN now?
-- |
Todd Cope
Member #998
November 2000
|
I'm having some issues with this version:
|
kenmasters1976
Member #8,794
July 2007
|
I've also seen some assertion errors. This is what I get when trying to run ex_blend_test: {"name":"600983","src":"\/\/djungxnpq2nug.cloudfront.net\/image\/cache\/a\/b\/ab6077308df8823bef69289f9c63be5d.png","w":369,"h":250,"tn":"\/\/djungxnpq2nug.cloudfront.net\/image\/cache\/a\/b\/ab6077308df8823bef69289f9c63be5d"} I get a similar one whe trying to run ex_winfull. It seems kinda weird when I'm building Allegro and the examples with MinGW. Also, ex_config, ex_convert, ex_dir, ex_get_path, ex_memfile, ex_monitor_info, ex_path, ex_path_test, ex_user_events and ex_uft8 don't show any info on what they're supposed to do. ex_saw produces the sound but there's no info. ex_dualies crashes on my machine. I don't have two adapters, but it crashes instead of showing the error message and quit. ex_mouse_focus sometimes loses track of which window has focus (both windows are black even if the mouse is over one of them). This might be an error in the example code and not with Allegro. This is all for the plain 4.9.19 release under Windows XP. Sorry if this post is a bit annoying but I just went and checked all examples. I'll try to check the code for the examples to see if I can find a fix.
|
Peter Wang
Member #23
April 2000
|
Todd Cope said: Speed does not work properly. I can hear the music but the timing seems a little off. Nothing displays on the screen and I can't exit without force-quitting The display problem is very strange. There's nothing special going on there. Can you try other resolutions (i.e. on the command line, run "speed <w> <h>")? EDIT: wait, no, it's a windowed mode so that shouldn't matter With the music, do you mean the pitch is correct, but the notes are a little delayed? There is a background thread (sound_update_proc) that waits for timer events, which then fires new notes. Some latency processing the events could be the culprit, although there may be other reasons. Perhaps you could print out the differences in the event timestamp and al_current_time() and see if they are too large.
|
kenmasters1976
Member #8,794
July 2007
|
I have a crappy integrated Intel card and the Speed demo works fine on my machine.
|
|
|