EOL2 - updated 23.12.2017
Moderator: Moporators
Re: Elma 2 - updated 20.9.2015
No, the desktop version is not yet updated and it's kinda hard to update the lgr file manually for the web version (it has packaged all required files in one file that is used to create the virtual file system for the browser). I can release 0.3.0 today for all platforms even though the replay things that I kinda promised are not done yet :)
Re: Elma 2 - updated 17.8.2016
Elma 2 0.3.0
Windows 32-bit
Windows 64-bit
Ubuntu 32-bit
Ubuntu 64-bit
OS X
Browser (online version)
Ubuntu 32-bit and OS X versions have not really been tested. I did try to make sure all required libraries are bundled correctly. If someone notices they don't work, post here.
What's new?
* Improved keyboard navigation
* Lgrzip format (see previous post for details)
* Web browser support
* Turn animation is now continuous
* Rotation speed is shown (requested by nick-o-matic :))
Bug fixes:
* Bike goes off screen with big zoom and non-centered camera
* If you're writing in a textbox and press a key that happens to be a shortcut for some button, the button gets clicked
* Non-centered camera does not get updated properly while playing replay
Known bugs/unfinished stuff?
* Some GUI controls are unreachable using keyboard (example: a checkbox next to a textbox)
* Flagtag mode doesn't work well (e.g. flag is never shown; touching killer won't restart)
* Speed, max speed and one wheel status are not shown correctly when playing a replay
* Old options.dat from the previous version file may cause a crash so you might have to delete it if you still have it
* Grass is sometimes too thin
* Window state (maximized/not maximized) is not remembered
* Non-ASCII characters in level filenames cause problems
* Some web-browser-specific issues
Windows 32-bit
Windows 64-bit
Ubuntu 32-bit
Ubuntu 64-bit
OS X
Browser (online version)
Ubuntu 32-bit and OS X versions have not really been tested. I did try to make sure all required libraries are bundled correctly. If someone notices they don't work, post here.
What's new?
* Improved keyboard navigation
* Lgrzip format (see previous post for details)
* Web browser support
* Turn animation is now continuous
* Rotation speed is shown (requested by nick-o-matic :))
Bug fixes:
* Bike goes off screen with big zoom and non-centered camera
* If you're writing in a textbox and press a key that happens to be a shortcut for some button, the button gets clicked
* Non-centered camera does not get updated properly while playing replay
Known bugs/unfinished stuff?
* Some GUI controls are unreachable using keyboard (example: a checkbox next to a textbox)
* Flagtag mode doesn't work well (e.g. flag is never shown; touching killer won't restart)
* Speed, max speed and one wheel status are not shown correctly when playing a replay
* Old options.dat from the previous version file may cause a crash so you might have to delete it if you still have it
* Grass is sometimes too thin
* Window state (maximized/not maximized) is not remembered
* Non-ASCII characters in level filenames cause problems
* Some web-browser-specific issues
Re: Elma 2 - updated 17.8.2016
It still happens on hotkeys (toggle map, zoom...)Smibu wrote:* If you're writing in a textbox and press a key that happens to be a shortcut for some button, the button gets clicked
Saving an internal rec in elma2 and trying to spy it in eol says the level file has changed. External recs seem to work normally.
Characters in key config textboxes show up as US layout equivalents (ä = ' etc)
Freelook keys could be separate from playing keys (but also pasibel to overlap) or just defaulted to arrows, since not everyone has brake/gas on the arrows.
then again i don't know anything
maybe easier not to think abouut alöl things thought than not things thought ... or something..=?
maybe easier not to think abouut alöl things thought than not things thought ... or something..=?
Re: Elma 2 - updated 17.8.2016
Not working on OSX 10.11.4 (15E65). Nothing happens when running elma2.app
Could be because the "libepoxy.0.dylib" under the Frameworks folder is only an alias and not the actual library file.
Edit:
Tried to install libepoxy with Homebrew
now I get this error
Code: Select all
dyld: Library not loaded: @rpath/libepoxy.0.dylib
Referenced from: /Users/sjurdur/Downloads/elma2.app/Contents/MacOS/elma2
Reason: image not found
Trace/BPT trap: 5
Edit:
Tried to install libepoxy with Homebrew
now I get this error
Code: Select all
[...]
Application Specific Information:
abort() called
terminating with uncaught exception of type utils::error_internal: CEGUI::RendererException in function 'void CEGUI::OpenGL3Shader::outputShaderLog(GLuint)' (/elma2/dependencies/cegui-0.8.7/cegui/src/RendererModules/OpenGL/Shader.cpp:244) : OpenGL3Shader compilation has failed.
ERROR: 0:1: '' : version '150' is not supported
ERROR: 0:1: '' : syntax error: #version
[...]
Thorze wrote:I just wanted to make a cool topic like Juish have cool topics..
Re: Elma 2 - updated 17.8.2016
Thx for testing, did not know about that int rec bug. First item is known but forgot to add it in the buglist. Same with third. My reasoning for the freelook keys being the same as playing keys is that you don't have to move your fingers around when toggling freelook mode, no matter what your playing keys are. Would there be any benefit for the separation?Lousku wrote:It still happens on hotkeys (toggle map, zoom...)Smibu wrote:* If you're writing in a textbox and press a key that happens to be a shortcut for some button, the button gets clicked
Saving an internal rec in elma2 and trying to spy it in eol says the level file has changed. External recs seem to work normally.
Characters in key config textboxes show up as US layout equivalents (ä = ' etc)
Freelook keys could be separate from playing keys (but also pasibel to overlap) or just defaulted to arrows, since not everyone has brake/gas on the arrows.
Strange :( because the real .dylib it references is included in the Frameworks folder. I will have to see how to fix it properly.Xiphias wrote:Could be because the "libepoxy.0.dylib" under the Frameworks folder is only an alias and not the actual library file.
This looks very similar to a Linux problem that I had a while ago, so I guess I know how to fix this. I'll send you a test version in PM when I think I've fixed the two problems.Xiphias wrote: now I get this error
Code: Select all
[...] Application Specific Information: abort() called terminating with uncaught exception of type utils::error_internal: CEGUI::RendererException in function 'void CEGUI::OpenGL3Shader::outputShaderLog(GLuint)' (/elma2/dependencies/cegui-0.8.7/cegui/src/RendererModules/OpenGL/Shader.cpp:244) : OpenGL3Shader compilation has failed. ERROR: 0:1: '' : version '150' is not supported ERROR: 0:1: '' : syntax error: #version [...]
Re: Elma 2 - updated 17.8.2016
Just intuitiveness, but yeah, it's not important, will get used to it eventually.Smibu wrote:My reasoning for the freelook keys being the same as playing keys is that you don't have to move your fingers around when toggling freelook mode, no matter what your playing keys are. Would there be any benefit for the separation?
then again i don't know anything
maybe easier not to think abouut alöl things thought than not things thought ... or something..=?
maybe easier not to think abouut alöl things thought than not things thought ... or something..=?
Re: Elma 2 - updated 17.8.2016
Xiph confirmed that the OS X version works now. Download link and version number are the same.
Re: Elma 2 - updated 17.8.2016
Tested on mac too and seems to work, now can finally play Elma on mac without hassle. ty
Once I didn't know what poo meant =D
Re: Elma 2 - updated 17.8.2016
Yeah works for me too. I still really don't like the UI, but the game itself runs very well.
<veezay> antti also gonna get stabbed later this month
<nick-o-matic> niec
<nick-o-matic> niec
-
- Beginner kuski
- Posts: 2
- Joined: 17 Sep 2016, 05:51
- Team: razer
Re: Elma 2 - updated 17.8.2016
Am I the only one getting no sound?
Re: Elma 2 - updated 18.9.2016
0.4.0
Download links in first post!
After the migration to modern OpenGL, it made sense to finish the graphics stuff, so:
* Pictures and textures are finally shown
* Pictures and textures can be animated!
Two bug fixes:
* Internal replays did not work in original Elma
* Elma 2 would crash if there was an object with invalid animation number in the lev
To make an animated picture/texture for LGR, make a multipage TIFF file. This was basically the only sensible option with FreeImage; it also supported GIFs but they are very bad (support only 256 colors).
I'm not sure what the best program for creating multipage TIFFs is, but after researching a bit, I found that http://www.graphicsmagick.org/download.html was the least sucky program when I wanted to combine multiple PNG images to a single multipage TIFF file. Another similar program (ImageMagick) did NOT work when I tried; it somehow lost the transparency during the conversion.
So the command is like:
I also updated the rest of textures in the default LGR (brick, stone1, stone2, stone3) so that they work in the browser version.
Edit: Kuski's body parts can't be animated just yet - I forgot to update one place.
Download links in first post!
After the migration to modern OpenGL, it made sense to finish the graphics stuff, so:
* Pictures and textures are finally shown
* Pictures and textures can be animated!
Two bug fixes:
* Internal replays did not work in original Elma
* Elma 2 would crash if there was an object with invalid animation number in the lev
To make an animated picture/texture for LGR, make a multipage TIFF file. This was basically the only sensible option with FreeImage; it also supported GIFs but they are very bad (support only 256 colors).
I'm not sure what the best program for creating multipage TIFFs is, but after researching a bit, I found that http://www.graphicsmagick.org/download.html was the least sucky program when I wanted to combine multiple PNG images to a single multipage TIFF file. Another similar program (ImageMagick) did NOT work when I tried; it somehow lost the transparency during the conversion.
So the command is like:
Code: Select all
gm convert frame*.png out.tif
Edit: Kuski's body parts can't be animated just yet - I forgot to update one place.
Re: Elma 2 - updated 18.9.2016
very sick! not long ago i asked for this possibility in irc channel and thought maybe see it in final version somehow but it already implemented :O so impsy nice work
Re: Elma 2 - updated 17.8.2016
This post was unapproved for a while in case mans missedrazerphynx wrote:Am I the only one getting no sound?
-
- Beginner kuski
- Posts: 2
- Joined: 17 Sep 2016, 05:51
- Team: razer
Re: Elma 2 - updated 18.9.2016
How come there's no sound?
Re: Elma 2 - updated 18.9.2016
this is a work in progress so sound is not implemented yet
Re: Elma 2 - updated 18.9.2016
I added a proper changelog to the first post. It also includes things for the not-yet-released version. I'll try to update that list regularly.
Hosp: Good idea, I wrote that down. Never even thought of that :)
razer: I guess sounds are not coming anytime soon. I added an issue about that anyway.
A roadmap for this project has always been missing (thanks 8-ball for reminding). My vision for "what to do next" is something like this:
Thanks to some new features in GitLab, I've made the project partially public: https://gitlab.com/Smibu/elma-2
This means:
- Anyone can see the list of issues
- Any logged-in user can create new issues and edit their own issues
- Any logged-in user can comment and up-/downvote any issues
In one tooltip it says "the project can be cloned without any authentication", but it's not true. They haven't yet updated the UI properly in all places...
Hosp: Good idea, I wrote that down. Never even thought of that :)
razer: I guess sounds are not coming anytime soon. I added an issue about that anyway.
A roadmap for this project has always been missing (thanks 8-ball for reminding). My vision for "what to do next" is something like this:
Code: Select all
- Make basic stuff work well:
- Flagtag mode (done for next version)
- Better support for replays
- Browse all replays
- List replays of selected level
- Autosave best time rides, maybe even all rides
- Ghost
- Show and save best times, import times to Elma 2 from state.dat and external levs with top 10 times
- Set cripple mode for playing (no-volt etc.)
- Start designing/implementing online mode
- First goals:
- Show players in current level
- Show list of online players
- Make chat work
- Investigate the possibility to use/integrate mila's/domi's client/server code
- Investigate if the current API is good enough or if something should be added/modified etc. considering new lev/rec formats
This means:
- Anyone can see the list of issues
- Any logged-in user can create new issues and edit their own issues
- Any logged-in user can comment and up-/downvote any issues
In one tooltip it says "the project can be cloned without any authentication", but it's not true. They haven't yet updated the UI properly in all places...
Re: Elma 2 - updated 18.9.2016
Features to request to potentially put on gitlab (?long-term)
1. More detailed replay files. Maybe 0.01 seconds per frame instead of 0.03 because I find that the current replay format loses a lot of detail for fine bike-driving and bounces and stuff. Or at least 60 fps
2. Scripting:
Some scripting thoughts and mongobraining. Would be nice to keep a ton of stuff accessible and modifiable for most flexible experience, and then other mans could copy/paste and also people could stop asking for feature request cause we can script it
modifiable variables:
triggers:
In theory if you take it very far you could override default object functions
onObjectTouch(object=flower/apple/killer/objecttype4/objecttype5/object[43]/):
any object of type flower/apple/killer/type4/type5 OR specific object 1-5000
onLevelStart
onCollisionwithPolygon(polygon=asd)
onHeadCollideWithPolygon(polygon=asd)
Can also override default "win" (e.g. onHeadCollideWithPolygon(): if(applecount>=numberofapples): Win())
[in editor/levelfile data]: custom objects (type 4-255 or whatever) (with possibly predefined objects like teleport and stuff)
Additional feature requests:
Additional polygon features (some may have been mentioned longtime in past but aren't added as features).
Dunno if this interferes with level rendering process though so maybe dynamically changing polygons have to be handled differently:
Polygon.Solid=True/False
Polygon.Visibility=Normal/FillWithSolidColor/FillWithTexture/Grass/None
Polygon.VisibilityColor="#FF00FF"/"brick"
Polygon.distance
Polygon/Object Tags
For scripting purposes, add free text variable or flags for each polygon/object?
polygon[x].tags="disappear,test"
onObjectTouch(apple): for all polygon x with tags containing "disappear", x.Solid=False, x.Visibility=None #disappear all polygons with tag "disappear" when any apple is touched
Level Info
level.info="sdjfdskjg"
Textbox containing details about level when selecting level (e.g. "hi guys you beat this level if you touch the flower ONLY when gravity is left")
1. More detailed replay files. Maybe 0.01 seconds per frame instead of 0.03 because I find that the current replay format loses a lot of detail for fine bike-driving and bounces and stuff. Or at least 60 fps
2. Scripting:
Some scripting thoughts and mongobraining. Would be nice to keep a ton of stuff accessible and modifiable for most flexible experience, and then other mans could copy/paste and also people could stop asking for feature request cause we can script it
modifiable variables:
Code: Select all
1. Active bike properties:
bike.x/y
wheel1/2 x,y
wheel1/2.rotationspeed
wheel1/2.rotation
bike.rotation
bike.rotationspeed
bike.speed ?dunno if vector or x/y
invicibilityflag***
gravity
applecount=4***
2. Bike function properties
relative_resting_position_of_wheel1/2/head_to_middle_of_bike
voltdelay
voltstrength(first&second part)
elasticity
gasstrength
maxgasspeed
3. Custom variables
customvariable[0-255]
4. Polygon/Object variables, e.g.
polygon[6].visible = Yes/Grass/FillWithTexture/FillWithColor (see below)
polygon[6].solid = Yes/No/Grass
object[5].hidden=Yes
In theory if you take it very far you could override default object functions
onObjectTouch(object=flower/apple/killer/objecttype4/objecttype5/object[43]/):
any object of type flower/apple/killer/type4/type5 OR specific object 1-5000
Code: Select all
onObjectTouch(object=apple): #overrides normal apple touching
bike.invicibility=True
defaultGetAppleFunction(object) (= level.applecount+=1; hide(object); object.touchable=False)
return
onCollisionwithPolygon(polygon=asd)
onHeadCollideWithPolygon(polygon=asd)
Can also override default "win" (e.g. onHeadCollideWithPolygon(): if(applecount>=numberofapples): Win())
[in editor/levelfile data]: custom objects (type 4-255 or whatever) (with possibly predefined objects like teleport and stuff)
Code: Select all
leveldata:
objecttype4.icon=file
objecttype4.roundobject_or_polygonshape=roundobject
objecttype4.radius=object.DEFAULTSIZE
objecttype4.minimapcolor=#FFFFFF
Code: Select all
onObjectTouch(object=objecttype4): #example custom object
bike.x/y=[x,y] #teleport bike to new location
bike.speed=0.5*bike.speed #slow down when portalling, because I feel like it
hide(object) #one-use portal
object.touchable=False #one-use portal
return
Additional feature requests:
Additional polygon features (some may have been mentioned longtime in past but aren't added as features).
Dunno if this interferes with level rendering process though so maybe dynamically changing polygons have to be handled differently:
Polygon.Solid=True/False
Polygon.Visibility=Normal/FillWithSolidColor/FillWithTexture/Grass/None
Polygon.VisibilityColor="#FF00FF"/"brick"
Polygon.distance
Polygon/Object Tags
For scripting purposes, add free text variable or flags for each polygon/object?
polygon[x].tags="disappear,test"
onObjectTouch(apple): for all polygon x with tags containing "disappear", x.Solid=False, x.Visibility=None #disappear all polygons with tag "disappear" when any apple is touched
Level Info
level.info="sdjfdskjg"
Textbox containing details about level when selecting level (e.g. "hi guys you beat this level if you touch the flower ONLY when gravity is left")
Re: Elma 2 - updated 18.9.2016
You know you can just register and post the issues yourself?
39:37,91
Re: Elma 2 - updated 13.3.2017
0.5.0
Download from first post!
Changes since 0.4.0:
- Replays of selected level(s) are listed.
- Any existing replays can be played, not just the latest ride.
- Unsaved replays can be saved later (but not after quitting the program).
- Replays of same play mode can be merged.
- Added support for ghost driver.
- Added option to toggle vsync preference.
- Added support for flag tag mode.
- Added Peek level -option.
- Browser version: Added support for downloading levels and replays via URL parameters.
- Browser version: Screen is now resized automatically to browser window size.
- XML format is now used for options file.
- Don't restart the whole application when changing full screen setting.
- Bugfix: Head was covered by suspension.
- Bugfix: Characters in key config textboxes showed up as US layout equivalents.
- Bugfix: Both in-game and menu actions were fired on keypress.
- Bugfix: Delete key did not work in textboxes.
- Bugfix: Alovolt and normal volt keys could conflict with each other.
- Bugfix: Arm animation sometimes went to incorrect direction in replay.
- Bugfix: Pictures in background -option was not being saved.
- Bugfix: LGRs without grass were not working.
- Bugfix: Driver body parts were not animatable.
- Bugfix: Browser version: Mouse wheel was inverted.
Peek level -option is basically a temporary zoom fill.
Some notes about keyboard shortcuts:
- The replay list is next to the level list. You can switch between the lists with left/right arrows. You can't yet visually see which of the lists is active.
- You can't yet select multiple levels/replays with keyboard - you need to use mouse and Shift/Ctrl + click.
- You can toggle the ghost status of replay(s) with Ctrl+G.
- You can save an unsaved replay with Ctrl+S (this changes focus to the textbox and you can start typing, and press enter to save).
URL parameter support is quite a nice feature. It allows to link to lev(s) and/or rec(s) and they will be opened automatically if possible. Automatic opening happens if:
- only one level is given, or
- several replays are given for the same lev (they will be merged and played) and the corresponding lev is given if it's not internal
In any case, the provided lev(s) and/or rec(s) will be loaded in the filesystem so they are visible in the tables.
Example URL to merge five best WCup703 replays: https://mkl.io/files/eol2/eol2.html?lev ... 3talli.rec
Cross-domain URLs don't work yet, but Kopaka and ville_j promised to configure their servers so that levs and recs can be downloaded from at least Recsource and EOLsite eventually.
Update: As a temporary workaround, I added a reverse proxy to my site, so URLs such as https://mkl.io/files/eol2/eol2.html?lev ... prec/37914 and https://mkl.io/files/eol2/eol2.html?rec ... euxyqgbjpm work now.
Download from first post!
Changes since 0.4.0:
- Replays of selected level(s) are listed.
- Any existing replays can be played, not just the latest ride.
- Unsaved replays can be saved later (but not after quitting the program).
- Replays of same play mode can be merged.
- Added support for ghost driver.
- Added option to toggle vsync preference.
- Added support for flag tag mode.
- Added Peek level -option.
- Browser version: Added support for downloading levels and replays via URL parameters.
- Browser version: Screen is now resized automatically to browser window size.
- XML format is now used for options file.
- Don't restart the whole application when changing full screen setting.
- Bugfix: Head was covered by suspension.
- Bugfix: Characters in key config textboxes showed up as US layout equivalents.
- Bugfix: Both in-game and menu actions were fired on keypress.
- Bugfix: Delete key did not work in textboxes.
- Bugfix: Alovolt and normal volt keys could conflict with each other.
- Bugfix: Arm animation sometimes went to incorrect direction in replay.
- Bugfix: Pictures in background -option was not being saved.
- Bugfix: LGRs without grass were not working.
- Bugfix: Driver body parts were not animatable.
- Bugfix: Browser version: Mouse wheel was inverted.
Peek level -option is basically a temporary zoom fill.
Some notes about keyboard shortcuts:
- The replay list is next to the level list. You can switch between the lists with left/right arrows. You can't yet visually see which of the lists is active.
- You can't yet select multiple levels/replays with keyboard - you need to use mouse and Shift/Ctrl + click.
- You can toggle the ghost status of replay(s) with Ctrl+G.
- You can save an unsaved replay with Ctrl+S (this changes focus to the textbox and you can start typing, and press enter to save).
URL parameter support is quite a nice feature. It allows to link to lev(s) and/or rec(s) and they will be opened automatically if possible. Automatic opening happens if:
- only one level is given, or
- several replays are given for the same lev (they will be merged and played) and the corresponding lev is given if it's not internal
In any case, the provided lev(s) and/or rec(s) will be loaded in the filesystem so they are visible in the tables.
Example URL to merge five best WCup703 replays: https://mkl.io/files/eol2/eol2.html?lev ... 3talli.rec
Cross-domain URLs don't work yet, but Kopaka and ville_j promised to configure their servers so that levs and recs can be downloaded from at least Recsource and EOLsite eventually.
Update: As a temporary workaround, I added a reverse proxy to my site, so URLs such as https://mkl.io/files/eol2/eol2.html?lev ... prec/37914 and https://mkl.io/files/eol2/eol2.html?rec ... euxyqgbjpm work now.
Last edited by Smibu on 22 Dec 2017, 23:27, edited 1 time in total.
Re: Elma 2 - updated 13.3.2017
levels/database should record most apples gotten in "best times" if level hasnt been finished. lots of big ff levels in EOL could use this feature and if level constraints are removed or increased by a lot people will probably make even sicker and longer levels where this apple record thing becomes useful and increases motivation to play even if unlikely to finish (i duno if apple results are shown in elma online site but missing obv from online table)
Re: Elma 2 - updated 18.9.2016
can you or someone who knows tell how the framerate for animation is determined? or the details, i assume all animations have same speed as apple animation but what is it exactly etc?Smibu wrote:* Pictures and textures can be animated!
Re: Elma 2 - updated 13.3.2017
Yeah so far everything has same anim speed, and that's 30 FPS. So technically the frame index is chosen like:Vermin Supreme wrote:can you or someone who knows tell how the framerate for animation is determined? or the details, i assume all animations have same speed as apple animation but what is it exactly etc?
Code: Select all
return *openGLTextures[static_cast<int>((time * 30)) % openGLTextures.size()].get();
Agreed. Related GitLab issue: https://gitlab.com/Smibu/elma-2/issues/101Vermin Supreme wrote:levels/database should record most apples gotten in "best times" if level hasnt been finished. lots of big ff levels in EOL could use this feature and if level constraints are removed or increased by a lot people will probably make even sicker and longer levels where this apple record thing becomes useful and increases motivation to play even if unlikely to finish (i duno if apple results are shown in elma online site but missing obv from online table)
Re: Elma 2 - updated 13.3.2017
Impsy to run Elma 2-0.5.0 on Linux Mint 64 bit
version 0.3.0 had worked correctly. Dunno how to delete whole elma 2 to reinstall it.
Code: Select all
dawid@dawid-linux ~/Pobrane $ cd /opt/elma2 && ./elma2
./elma2: error while loading shared libraries: libminizip.so.1: cannot open shared object file: No such file or directory
Re: Elma 2 - updated 13.3.2017
Before running, try:
It doesn't look from the lib dir for some reason...
Don't know about Mint, but on Ubuntu it's just:
Code: Select all
export LD_LIBRARY_PATH=/opt/elma2/lib
Don't know about Mint, but on Ubuntu it's just:
Code: Select all
sudo apt-get remove elma2
Re: Elma 2 - updated 13.3.2017
are those multipage tiff files similar to gif in that they update changes between frames, not load full image each frame? animating some big images 30fps would create sick large files (i didnt try multipage tiff yet since need that program so duno how compression or file sizes etc
Re: Elma 2 - updated 13.3.2017
Not sure, but I'd guess a multipage tiff doesn't do any kind of delta encoding between frames. At least couldn't find any hints from google. It's probably just "multiple files in one".
The zip format for lgr probably helps reduce the total size, but hard to say how much.
The zip format for lgr probably helps reduce the total size, but hard to say how much.
Re: Elma 2 - updated 13.3.2017
After reinstalling and
now I see this (i'm trying to install 64 bit version)Smibu wrote:export LD_LIBRARY_PATH=/opt/elma2/lib
Code: Select all
dawid@dawid-linux /opt/elma2 $ cd /opt/elma2 && ./elma2
./elma2: /usr/lib/x86_64-linux-gnu/libstdc++.so.6: version `CXXABI_1.3.8' not found (required by ./elma2)
./elma2: /usr/lib/x86_64-linux-gnu/libstdc++.so.6: version `GLIBCXX_3.4.20' not found (required by ./elma2)
./elma2: /usr/lib/x86_64-linux-gnu/libstdc++.so.6: version `GLIBCXX_3.4.21' not found (required by ./elma2)
./elma2: /usr/lib/x86_64-linux-gnu/libstdc++.so.6: version `CXXABI_1.3.9' not found (required by ./elma2)
./elma2: /usr/lib/x86_64-linux-gnu/libstdc++.so.6: version `GLIBCXX_3.4.21' not found (required by /opt/elma2/lib/libCEGUIOpenGLRenderer-0.so.2)
./elma2: /usr/lib/x86_64-linux-gnu/libstdc++.so.6: version `GLIBCXX_3.4.21' not found (required by /opt/elma2/lib/libCEGUIBase-0.so.2)
Re: Elma 2 - updated 13.3.2017
libstdc++ version is incompatible... there is some info e.g. here for how to fix it, but I'm not sure if those are applicable to Mint.
Re: Elma 2 - updated 13.3.2017
Can't install it anyway. In an earlier version I was possible to run Windows version with Wine but this doesn't work with the 0.5.0 version. So I stopped to run it cuz can't find a solution.
Anyway, is it possible to make browser version (when ran locally from disk) to save my control config? At this moment I have to change control setting each time I run Elma 2 Browser version.
And maybe for clarity you should place compilation version somewhere in an Elma 2 menu.
Anyway, is it possible to make browser version (when ran locally from disk) to save my control config? At this moment I have to change control setting each time I run Elma 2 Browser version.
And maybe for clarity you should place compilation version somewhere in an Elma 2 menu.
Re: Elma 2 - updated 13.3.2017
I've thought the same; should be possible. Issue here: https://gitlab.com/Smibu/elma-2/issues/45dawid wrote:Anyway, is it possible to make browser version (when ran locally from disk) to save my control config? At this moment I have to change control setting each time I run Elma 2 Browser version.
Hmm true; currently the version is not actually visible anywhere, not even in the log file (it has a build timestamp only).dawid wrote:And maybe for clarity you should place compilation version somewhere in an Elma 2 menu.
Did it give any error message?dawid wrote:In an earlier version I was possible to run Windows version with Wine but this doesn't work with the 0.5.0 version. So I stopped to run it cuz can't find a solution.
Re: EOL2 - updated 23.12.2017
I checked this topic quickly while doing research for my article. Some comments:
1) Good work! The cross OS support is amazing! I can't really play myself anymore because windowed Elma doesn't work in my Win10 and I just can't put that much time to playing that I would go full screen. Also not in work for same reason. The browser version works pretty well!
2) I'm not sure if replay autosave is good. Human must do the selection if the replay is good or not at some point so why not at this point? I don't really find it valuable that we have endless number of replays that no one is interested in. It makes it only harder to find the good ones.
3) Is cripple mode important? I wouldn't implement features because they are in EOL1 but because they are good or needed. At least I don't like those no-volt modes and to me it looks like they are not very popular, more like an interesting detail? Features like this can be added when everything more important is done. Just do the core playing as good as it can be done.
Update: added some more
1) Good work! The cross OS support is amazing! I can't really play myself anymore because windowed Elma doesn't work in my Win10 and I just can't put that much time to playing that I would go full screen. Also not in work for same reason. The browser version works pretty well!
2) I'm not sure if replay autosave is good. Human must do the selection if the replay is good or not at some point so why not at this point? I don't really find it valuable that we have endless number of replays that no one is interested in. It makes it only harder to find the good ones.
3) Is cripple mode important? I wouldn't implement features because they are in EOL1 but because they are good or needed. At least I don't like those no-volt modes and to me it looks like they are not very popular, more like an interesting detail? Features like this can be added when everything more important is done. Just do the core playing as good as it can be done.
Update: added some more
40:02,71 (152.) | WCup4: 8. | 3x WR | 3x GAA | 12x FEM | KOM | The History of Elasto Mania (1993-2018)
Re: EOL2 - updated 23.12.2017
I understand this project is going forward slowly and smibu doesn't have much time to work on it and he prolly gotten bored to listening our suggestions, but I agree with abula, at least for the first release the feature goals should be minimal, like try to replicate most of the current EOL features but for battle modes only implement norm battle + ff + apple battle without cripples.
Btw I'd like to contribute in form of testing but it feels very hard to do serious testing because all times done in EOL2 are obsolete.
Btw I'd like to contribute in form of testing but it feels very hard to do serious testing because all times done in EOL2 are obsolete.
Re: EOL2 - updated 23.12.2017
Thanks. Offtopic, but for best EOL Win10 compatibility, you should use DXGL; I just tested the latest version and it works fine on my Win10. It supports fullscreen and windowed mode. I've seen links to random ddraw.dll files but they don't necessarily work if their configuration is incompatible with your machine. Ropelli's DLL (which is more efficient than DXGL) doesn't seem to work anymore unfortunately.Abula wrote: ↑3 Apr 2018, 16:48 1) Good work! The cross OS support is amazing! I can't really play myself anymore because windowed Elma doesn't work in my Win10 and I just can't put that much time to playing that I would go full screen. Also not in work for same reason. The browser version works pretty well!
They're currently autosaved only in memory, which I think is fine. It's debatable whether they should be autosaved to disk. Would be good to have it configurable at least.Abula wrote: ↑3 Apr 2018, 16:48 2) I'm not sure if replay autosave is good. Human must do the selection if the replay is good or not at some point so why not at this point? I don't really find it valuable that we have endless number of replays that no one is interested in. It makes it only harder to find the good ones.
I feel it is kinda essential. Not very hard to implement either.Abula wrote: ↑3 Apr 2018, 16:48 3) Is cripple mode important? I wouldn't implement features because they are in EOL1 but because they are good or needed. At least I don't like those no-volt modes and to me it looks like they are not very popular, more like an interesting detail? Features like this can be added when everything more important is done. Just do the core playing as good as it can be done.
I have similar thoughts. Motivation is also a bit of an issue; I'd very much like to have modules in C++ already. I wrote about this in the first post and I'm tracking the status of some compiler bugs I've reported to MS.Zweq wrote: ↑3 Apr 2018, 18:30 I understand this project is going forward slowly and smibu doesn't have much time to work on it and he prolly gotten bored to listening our suggestions, but I agree with abula, at least for the first release the feature goals should be minimal, like try to replicate most of the current EOL features but for battle modes only implement norm battle + ff + apple battle without cripples.
Yeah, EOL2 testing is not very "incentivized" right now. Needs more work to get to that point.
Re: EOL2 - updated 23.12.2017
Wow, EOL2 is planned? or already done?
anyway, possible balle types:
1) apple battle with countdown
2) battle, where need to finish, with touching minimum amount of polys
3) battle, where wheels and head of diff players can interact. one player can kill other by wheel, smth like this pls, if poss)
4) ...
anyway, possible balle types:
1) apple battle with countdown
2) battle, where need to finish, with touching minimum amount of polys
3) battle, where wheels and head of diff players can interact. one player can kill other by wheel, smth like this pls, if poss)
4) ...
Re: EOL2 - updated 23.12.2017
Team TR
Multi WR in Labyrinth with GRob
Best Internal Total Times, Pipe stats & Pipe archive
World kuski map, World Cup stats
Re: EOL2 - updated 23.12.2017
Works! This is big! Thanks man!Smibu wrote: ↑3 Apr 2018, 20:38Thanks. Offtopic, but for best EOL Win10 compatibility, you should use DXGL; I just tested the latest version and it works fine on my Win10. It supports fullscreen and windowed mode. I've seen links to random ddraw.dll files but they don't necessarily work if their configuration is incompatible with your machine. Ropelli's DLL (which is more efficient than DXGL) doesn't seem to work anymore unfortunately.
40:02,71 (152.) | WCup4: 8. | 3x WR | 3x GAA | 12x FEM | KOM | The History of Elasto Mania (1993-2018)
Re: EOL2 - updated 23.12.2017
It's now been over six months since I did any EOL2 development (the latest thing I did in April was a CMake refactoring, so nothing user-visible).
Right now I feel I will not be able to finish EOL2 (let's say finish = reach feature parity with EOL) in any reasonable timeframe. My motivation is largely gone, so even though I might proceed every now and then, those moments are too rare.
Trying to think about the exact reasons for the lack of motivation, I would say they are:
* development needs to happen "in the shadows" because I cannot just publish the physics
* C++ is not a language I enjoy programming in (EOL2 is basically the only C++ project I've ever had, and I'm not planning to start new projects with it)
* huge amount of work still waiting to be done
* I don't really play Elma these days, so any Elma-related development doesn't feel very exciting anymore (this applies to Elmanager as well)
I kept wondering if there's anything I could do to fix motivation. So I started doing some small things:
As skint0r announced, I submitted some PRs to his elma-rust library. Utilizing his library, I rewrote the physics in Rust in July. That was the first step of solving #2.
Soon after that, at the end of July, I contacted Balazs. I basically told him I had figured out Elma (1.11a) physics and asked whether it would be okay for him if I published this sort of Rust library. On August 5, he gave the kind of answer I anticipated, which basically was "no comment", but he also said something more. Here's the full quote:
I'm not sure if there's much I can do about #4. I don't know why I was so hooked on Elma about 10 years ago. Likewise, I don't know why I'm no longer hooked. Brains change I guess.
So right now not much is happening. I might continue expanding the Rust code by e.g. porting the rendering code. But this also really depends on how things pan out with Balazs; I really wouldn't like to keep things in the shadows anymore.
To be clear, there is no definite plan for how things are going to move forward. Maybe I magically get motivation to continue with the C++ codebase. Maybe someone magically steps forward and offers to continue with the C++ codebase. Maybe things magically pan out nicely with Balazs. It's just that none of those options seem terribly likely at the moment, and I just wanted to give an update instead of being silent.
Right now I feel I will not be able to finish EOL2 (let's say finish = reach feature parity with EOL) in any reasonable timeframe. My motivation is largely gone, so even though I might proceed every now and then, those moments are too rare.
Trying to think about the exact reasons for the lack of motivation, I would say they are:
* development needs to happen "in the shadows" because I cannot just publish the physics
* C++ is not a language I enjoy programming in (EOL2 is basically the only C++ project I've ever had, and I'm not planning to start new projects with it)
* huge amount of work still waiting to be done
* I don't really play Elma these days, so any Elma-related development doesn't feel very exciting anymore (this applies to Elmanager as well)
I kept wondering if there's anything I could do to fix motivation. So I started doing some small things:
As skint0r announced, I submitted some PRs to his elma-rust library. Utilizing his library, I rewrote the physics in Rust in July. That was the first step of solving #2.
Soon after that, at the end of July, I contacted Balazs. I basically told him I had figured out Elma (1.11a) physics and asked whether it would be okay for him if I published this sort of Rust library. On August 5, he gave the kind of answer I anticipated, which basically was "no comment", but he also said something more. Here's the full quote:
So that was a tiny part of solving #1 and it's also related to #3.Balazs wrote:soon I plan to talk to some people of the elma community about the rights of the game. We had some talking going on some time ago but I stopped that talking because I had some open projects and wanted to wait for what happens to these projects. It seems my waiting will be over soon. Before these talks I do not like to comment and spend time on your rush [sic] project. Thanks for your email,
I'm not sure if there's much I can do about #4. I don't know why I was so hooked on Elma about 10 years ago. Likewise, I don't know why I'm no longer hooked. Brains change I guess.
So right now not much is happening. I might continue expanding the Rust code by e.g. porting the rendering code. But this also really depends on how things pan out with Balazs; I really wouldn't like to keep things in the shadows anymore.
To be clear, there is no definite plan for how things are going to move forward. Maybe I magically get motivation to continue with the C++ codebase. Maybe someone magically steps forward and offers to continue with the C++ codebase. Maybe things magically pan out nicely with Balazs. It's just that none of those options seem terribly likely at the moment, and I just wanted to give an update instead of being silent.
Re: Elma 2 - updated 28.4.2014
I still remember this reply, and I'd say you were pretty spot on with your assessment.milagros wrote: ↑25 Aug 2014, 01:32 the enthusiasm (and thus the update rate) typically follows this kinda of behavior
based on the current status of elma2, it looks like ~30% of the EOL has been reproduced (and who knows how much more was planned)
extrapolating current update rates, it will never be finished, not even close
Thorze wrote:I just wanted to make a cool topic like Juish have cool topics..
Re: EOL2 - updated 23.12.2017
Sad to hear, Smibu.
I hope Balazs will come forward and do something about this situation...
C++ would be the language which I like, but I don't have any interest or time to develop anything. Maybe I would have had interest 10 years ago but not anymore... And now when I have family, I only have time to play a battle now and then. So even if I wanted, and especially now if the development will be done in Rust, I won't be the saviour here...
I hope Balazs will come forward and do something about this situation...
C++ would be the language which I like, but I don't have any interest or time to develop anything. Maybe I would have had interest 10 years ago but not anymore... And now when I have family, I only have time to play a battle now and then. So even if I wanted, and especially now if the development will be done in Rust, I won't be the saviour here...
A winner of 4 GAA's (mc2 included), winner of mkup206, and a proud member of team TAP.
Play uni levels: http://koti.mbnet.fi/zebra/uni.html
Homepage: http://koti.mbnet.fi/zebra/elma.html
Play uni levels: http://koti.mbnet.fi/zebra/uni.html
Homepage: http://koti.mbnet.fi/zebra/elma.html
Re: EOL2 - updated 23.12.2017
What about publishing the EOL2 code without physics part? Maybe there is some person in elma community who know programming and will develop the project? IMO many mans here can develop it. When the EOL2 code will be done, then You will just add physics code and publish finnished EOL2 project <3
Re: EOL2 - updated 23.12.2017
I have considered that, but so far I've decided against it for a couple of reasons:
So I will try to find some time in the near future to make a compilable physics-less version that I can publish. I'm not yet sure if it will be just a zip source or a Git repo with full history (physics excluded). Making a Git repo is more work.
- Quite a few people (I count more than 10) already have EOL2 code (at least some version of it). All of them are people who know how to code, but they lack the time and/or motivation to continue EOL2. I base this assumption on the fact that after I published the status post, I have not once seen anyone express interest in continuing the development of EOL2. Not in public or privately. Ideally the new developer would be someone who spontaneously nominated themselves to be the new developer instead of me throwing the code in public and hoping that someone gets excited about it.
- Separating physics from the other code so that the project still compiles needs more work than just deleting a folder. The physics-less code should also have a separate repository and the process of merging the engine to the physics-less one should be automated, because of course it's not going to be done just once after "EOL2 code will be done". No code is ever 100% done.
So I will try to find some time in the near future to make a compilable physics-less version that I can publish. I'm not yet sure if it will be just a zip source or a Git repo with full history (physics excluded). Making a Git repo is more work.