Port Requests


How about Discord, via Electron?: https://github.com/XNBlank/discord-linux I use this on my laptop and it works perfectly! Is Electron able to run on the Pandora? I might try porting this. But I may need some help.
[doublepost=1461817098,1461814538][/doublepost]Yeah I can't figure Electron out. It doesn't seem to have any normal build process and I can't find build instructions.
 
@Silent-Hunter : strange github repo. There is just a buch of pre-compiled x64 stuff along with some nacl binaries, and no sources at all...
There are sources for Electron though, that's a separate repo. https://github.com/electron/electron will have the source for those x86 bits.
[doublepost=1461855636,1461855461][/doublepost]https://github.com/XNBlank/discord-linux/issues/30 for more info. I'm not sure how to actually build Electron, I'll look into it more though.
[doublepost=1461865508][/doublepost]It does require clang, I don't know if that will be a problem or not.
[doublepost=1461870487][/doublepost]Hmm... yeah I'm gonna need some help. I'm not sure what's going on here:

Code:
Cloning into 'vendor/requests'...
remote: Counting objects: 17315, done.
remote: Compressing objects: 100% (26/26), done.
remote: Total 17315 (delta 2), reused 0 (delta 0), pack-reused 17289
Receiving objects: 100% (17315/17315), 6.87 MiB | 491.00 KiB/s, done.
Resolving deltas: 100% (10216/10216), done.
Checking connectivity... done.
Submodule path 'vendor/requests': checked out 'e4d59bedfd3c7f4f254f4f5d036587bcd8152458'
Failed to recurse into submodule path 'vendor/breakpad'
None
Traceback (most recent call last):
  File "./script/bootstrap.py", line 212, in <module>
    sys.exit(main())
  File "./script/bootstrap.py", line 39, in main
    update_submodules()
  File "./script/bootstrap.py", line 95, in update_submodules
    execute_stdout(['git', 'submodule', 'update', '--init', '--recursive'])
  File "/home/silent/electron/script/lib/util.py", line 176, in execute_stdout
    raise e
subprocess.CalledProcessError: Command '['git', 'submodule', 'update', '--init', '--recursive']' returned non-zero exit status
 
Here Codecombat source code, it's a game for learning how to code:

https://github.com/codecombat/codecombat

Thanks for the link. Always interested in looking at learning tools for programmers (it's part of my job).

But after playing through the first 50 levels or so I think the actual content is not very good at teaching programming (will still force some of my apprentices to work with it to test my opinion). The underlying model is flawed (subtle racing conditions - that don't affect the system much but will teach bad coding practices - , badly done randomization so that often the suggested solution won't work reliably; IMO a less imperative approach - especially keeping the main loop out of the students code, at least at first - would work much better) and the IDE isn't smart enough to detect when you are not solving the level as intended (e.g. you can present a solution without a loop when the task explicitly given is to construct a loop; furthermore all methods are available right from the beginning even without the corresponding item, help texts contain sample code that uses features not yet introduced, etc). The most glaring problem IMO is that even in this minimal sandbox there is much to much syntactical magic involved (e.g. it is never explained when you need full stops, colons, or parentheses; to make it even more confusing all the methods are available as both self.xxx and hero.xxx and the documentation mixes both forms without any explanation, especially the need for self. or hero. isn't explained anywhere). In my experience these syntactical hurdles are the biggest problems for beginners and one of the main reason people just copy solutions from somewhere else instead of thinking about their own. This is solved much better in e.g. htdp.org with racket-lang.org (even though they use the square bracket notation for cond, in my experience it is even easier for the students to just use normal parens instead, because then there is only one primitive grouping syntax and constant notation). Also Racket has the better stepper and syntax analyzer (and isn't as dog slow as this web framework; it was very difficult to complete the flag placing level).
 
This just came to my mind, but how about a port of mpv?
Some old stuff was removed as a result of the fork, not sure if any of that was critical to get decent mplayer performance on the Pandora.
Would be interesting to see how it holds up here. It's my preferred player on PC (although in combination with SMPlayer there).
 
I just stumbled upon this http://www.openra.net/news/release-20160508/
I noticed a port request for it was made quite some time ago but I guess it didn't work out... Maybe things changed and it's possible to port now?
The cross platform multiplayer thing appeals to me :) They seem to support a lot of OS's already..

I will try to compile it, but the zoom feature would need a new shorcut, ;)
 
OpenRA need shaders, and use mono. It either needs a glshim with a GLES2 backend, or a conversion to of the mono code to GLES2. I plan to work on the 1st option, but you can give it a try too.
 
This just came to my mind, but how about a port of mpv?
Some old stuff was removed as a result of the fork, not sure if any of that was critical to get decent mplayer performance on the Pandora.
Would be interesting to see how it holds up here. It's my preferred player on PC (although in combination with SMPlayer there).

I have a rescent running MPV Git Version on my CodeBlocks SD Card and tried it to bundle with the SMPlayerGui from the SMplayerRepoPND without success.
I can try it again..maybe i get it working and can send you a PND with it ;)
 
Is there anyone who actually tried Starbound on the Pandora? If there is, how slow is it? (with Exagear of course)

Also, sorry if I asked this before.
 
I have a rescent running MPV Git Version on my CodeBlocks SD Card and tried it to bundle with the SMPlayerGui from the SMplayerRepoPND without success.
I can try it again..maybe i get it working and can send you a PND with it
mpv support in SMPlayer is somewhat recent (since version 15.9.0, which was released last September), so I don't think any SMPlayer version we current have as a PND would be able to use it. So a fresh compile of that would be useful as well.
mpv can work nicely as standalone as well, though, so I wouldn't mind that either. :)
 
ok i will try to compile the latest Git from today.
Maybe i can make a Version which simply open a Terminal where you can use MPV from with the normal Commands ;)

Edit: Sorry my failure...i have it but not complete compiled and not working..sorry. :(
Tried it 7 times to rebuild yet:(
 
Last edited:
I have built mpv, with GLES2 renderer. But I don't find it very fast. It skip frame (for movie encoded as 6xx*3xx using h264, but not sure, I have to check back). This one also need Streaming Texture to be resonably faster, but I haven't checked how difficult it can be to add that.
(it needed quite a lot of hack, and to updated a couple of libs, no an easy plg'n play build in fact).
 
Last edited:
Afaik mpv still supports using SDL as output renderer. Wouldn't that be a more reasonable choice on the Pandora?
 
Would anyone be willing to take a look at Puyo Puyo VS ?

It is not open source, but there are Windows, Mac, and Linux ports. The developers say they are open to further porting too. It seems like they have a Raspberry Pi 3, as they said they would personally help support an ARM Linux port to that platform.
 
Looks to have been abandoned in 2008.. due to internet dramas on their forums.

Can't say that's worthy of a port but you could probably try it with GLShim
 
Back
Top