Dingoo A320 forum

Heya
I've had my Dingoo a couple weeks now and I'm about ready to start doing something with it... thing is, seeing as Linux is making it's steady way over to the beastie, should I start with what's currently available, or hold out a bit?

I'm guessing this is a "fear" that a few devs may have just now.. should we start with what we have, which when Linux comes along, may be forgotten about... or should we hold off a bit and ride the first wave of Dingoo Linux Apps...

What I'm planning on doing is porting over my prototype game engine - SGZEngine. It basically acts as a virtual machine to Lua and runs on a number of platforms already, though as I've not gotten off my lazy arse to fix some of the nastier bugs to release it yet ( which I was meant to do on the GP2X just before Christmas *cough* ), this would be perfect to whip me into shape to do something about it

Just wondering what the general concensus is, and if Dingoo Linux is closer than I think.

stuckie wrote:Heya
I've had my Dingoo a couple weeks now and I'm about ready to start doing something with it... thing is, seeing as Linux is making it's steady way over to the beastie, should I start with what's currently available, or hold out a bit?

I'm guessing this is a "fear" that a few devs may have just now.. should we start with what we have, which when Linux comes along, may be forgotten about... or should we hold off a bit and ride the first wave of Dingoo Linux Apps...

Yes I have a similar question - some insight from developers who have worked on other similar consoles like the GP2X would be welcome here.

The core Dingoo firmware doesn't seem to be updated much, and there's irritations like the button issue, which could be fixed in minutes with an open source replacement (providing it's not a hw issue). So on that note, I'd rather code for a well supported community OS - rather than a closed firmware where updates are irregular and not guaranteed.

Having said that, I'm keen to hear how fast the Linux libraries are, and what overheard Linux is likely to cost (memory footprint/performance of libraries etc.). Still waiting for a crappy MiniSD card so I can actually try the Linux port on my device you see..

Well I've been slowly battering my engine with the graphics libraries flatmush developed for the astro lander and minesweeper games... it'll "just run" under Dingoo Linux as it has an SDL backend anyway, so that's not much of a challenge

Though with booboo's advancements with getting dual booting working, I can see support for the default OS trailing off rather quickly now.

I say do what you can to drum up interest in Dingoo Linux as a firmware replacement both to existing an potential users. Every App we get ported to, or develop for Dingoo Linux is another win that draws more attention to the platform and inspires more developers to do the same.

I would suggest that, at this early stage, it's all the more important to see developers doing incredible things with Dingoo Linux, even if it does mean more work for them in the long run. People with little to no knowledge of programming/hacking are stuck out there with their stock firmware looking to Linux as the Holy Grail of the Dingoo community, we need to keep it alive and show developer support lest the whole project die off and never quite come to fruition.

I'm not saying those out there smart and capable enough to work on Dingoo linux need more motivation than the challenge itself, but it can't hurt.

Furthermore, attempting to target Dingoo Linux now should help to work out kinks in the toolchain and the system itself, if there are any, particularly if you build upon the rootfs itself and fill in some of the common executable gaps and configuration shortcomings that it has.

Development at this stage will also help to arrive at a set of Dingoo Linux conventions for storing games, game data and game saves in such a way as they can be enumerated by a GUI.

It wouldn't hurt to have some "for dummies" package management, either.

My thinking is that I'll develop for the installed OS, but keep it agnostic enough to build a linux version at a later date.

I'm not that familiar with linux, particularly in this kind of device, and I don't entirely see what it gets me, as a developer. It will make porting existing linux/sdl apps easier, I guess, but I'm more interested in writing my own stuff, and from that point of view I don't see a lot of difference.

On the other hand, of the dozen or so developable handhelds I own, to date I have only managed to write anything for one, so who am I to give advice?

Well if there is going to be some sort of package management, it may be an idea to get it sorted before too many things get released, otherwise there could be a fair amount of hassle to package things up right. Could be as crude as a zip file with a txt file detailing what it is, or perhaps something as fancy as some desktop distros have. Perhaps we can even ask the Pandora peeps nicely for their PXML format

Seeing as Linux is becoming the focus for the Dingoo now, I think I'll scrap my idea of getting my engine running on native firmware, and just patch it up properly for Dingoo Linux ( as well as fix the mountain of oddities among the code, and actually release it!) as well as seeing where I can help out.

( Off Topic, anyone know how to hide that Save button when posting? I keep clicking it instead of Submit! )

You said your app was virtual machine to Lua, you mean Lua the language or Lua as in something else? Just wondering if you would need Lua actually installed in the toolchain/rootfs since I do think I remember seeing it as an option I could add in

-------

Well lua was a quick one, installed easily into the rootfs. Ive never actually used it so I have no idea if its working or if you need any additional lua libs for it, but its possible to have it in there if you need it.

That works, thanks

Generally when I was messing with the GP2X and PSP for my Honours Project ( this engine is my final year project from University ), I just compiled Lua within the engine as it saved a lot of hassle. But having the library there to do it "properly" is a lot easier
Lua should've been very easy to compile, as it's ANSI C and very small, so just about anything should be able to run it.. and I don't use any extra Lua libraries - I just bash Lua directly so that's all I'd need, thank you.

Anyway, I'll patch up some more things and try get a release done this week - though bear in mind I've been trying to patch this up and release it on the GP2X since Christmas! That said I've almost sorted out the codebase so once that's done, I'll release as is and then start working on the extra features it should've had.
It does work in it's current state, but it's not exactly documented on how to use it at the moment.. making it a bit pointless and hostile for other people to use

Alright, well it may be a while until i release this rootfs with lua in it, i guess I could always post the lua executable seperately if you need it

I've got my engine compiled and running under Dingoo Linux ( using dingoo-uclibc-toolchain-05-28-09 ) just need to document it.

Though SDL seems to be jumpy? I'm going to assume this is more SDL_mixer with it's OGG support ( as that's what I was testing at the time ) as when the music stops, the jumpyness stops too. Probably vorbis spamming the CPU a bit too much while decoding.

It's still going to take me a while to document and clean the code up a bit.. and a working example game on it's use wouldn't go amiss either
I'll be aiming to get this done for the end of the month.. next weekend is a mini-LD and hopefully I will have the engine working enough for then to take part - which'll be where the example game will come from!