Gingerbread CM now for Archos – Proof of concept

Bubu sat down and made Cyanogenmod Gingerbread Android 2.3 working in his spare time. This is a proof of concept so don’t expect too much.

Currently working:

  • SGX/EGL 3D acceleration
  • WiFi
  • Touchscreen
  • adb
  • Android Market

Not working:

  • Audio
  • Bluetooth
  • Accelerometer
  • accelerated Video decoding
  • Suspend
  • TS calibration on A28/A32/A43

Known problems:

  • general instability / random crashes
  • Touchscreen has problems on the A101
  • hard to use on resistive TS models due to missing calibration

We’re looking for help to take this further. As this can’t be done by one person alone.

If you want to try it. It currently works on gen8 with latest openAOS boot menu:

You need to gunzip and copy it into the root directory of your device, then add it to your menu.lst.

Sources are here:
(Currently this is only vendor/archos/gen8 and device/archos/gen8, so you need to first “clone” the master branch of CyanogenMod)

Join us on IRC if you want to have a chat about it. This also has potential for the A5IT/A48IT but it needs volunteers that will help to make that happen.

Tags: , , , , , , , , ArchDruid, , , ,

11 Responses to “Gingerbread CM now for Archos – Proof of concept”

  1. Tom says:

    Wifi is not working on my device… a101 16gb
    and i am running the latest version

    each time i turn the wifi on, it just says ‘error’

  2. Bassam Bokhowa says:

    Thanks you so much. At last Arabic support is fixed. Where do I get a compatible Gingerbread Market apk for this gem?

    • tbr says:

      Nice that it is already a small improvement to someone!
      As this is based on CM7 their Market instructions should apply.

  3. [...] previous post and general gen8 instructions for [...]

  4. [...] previous post and general gen8 instructions for [...]

  5. Compyy says:

    Hi, Bubu and team, great work, we people from xda-developers have bounty about 105$ we want to donate you for 100% working version of CM7 on Archos.

    Its a great workkk well done.


  6. CalcProgrammer1 says:

    Boots on my A43 with the following boot menu config:


    I renamed the image to “gingerbread.img”. It does boot and shows the Android loading logo, but it messes up horribly when it gets to the main screen. It isn’t just a touchscreen issue, the entire GUI is misplaced and messed up. Besides that, the scrolling seems a bit smoother than Froyo in both the home screen and the menus. Awesome proof of concept, just the fact that it boots is quite awesome, and with the new boot menu working as well as it is it could easily be tri-booted with Froyo, Gingerbread, and Debian in the future.

  7. a5it-nogo says:

    Potential for A5IT/A48IT? I don’t think so. With this ancient (and broken) 2.6.27 kernel? Because nobody took care about it and tried to upstream the gen7 Kernel the A5IT became a pain in the ass and that’s exactly what will happen to gen8 if the kernel won’t get in upstream. All this “Proof of Concept” is meaningless. If the Kernel is too broken to make it into upstream then there is no way the userspace would be fine.

    • tbr says:

      Especially Android is not that picky about kernel versions. Else we wouldn’t be running Froyo, would we? Google might recommend a newer kernel version, but so far we haven’t really run into problems. If we really find something to be missing it might be much easier to just backport that.

      Also I warmly invite you to stop bitching and join the people who actually do work on newer kernel versions for Gen7 and Gen8.

      Thanks, Thomas

      • a5it-nogo says:

        No bitching, stating facts.

        Kernel is incomplete, uses depracted apis (even for 2.6.27) and is on parts simply broken (kexec?). Whatever you say, unless Kernel development doesn’t target mainline inclusion the developer are doing it wrong. Nicktime started with 2.6.35 didn’t he? Now it seems he is stuck in Development on 2.6.37 and mainline 3.0 is drawing near without him. It’s hillarious if you think about it, most code is already in mainline but the archos bits are missing of course. If someone actually CAN backport things to ancient kernel, why doesn’t he/she just port archos code for mainline? It doesn’t make sense, and I am getting depressed.

        • tbr says:

          You keep reiterating unsubstantiated and fuzzy claims. If you feel so strongly about it, then please come and help, send patches, work with us.

          Just to pick one of your claims: ‘kexec’ – this does not work for the majority of linux-omap kernel versions. I specifically talked to people in the know about this, so please don’t make it sound as if Archos deliberately broke it.

          PS: I spoke to some people and they estimate this effort in years if you work on it is being worked on in free time, to make it truly stable and fully supported. If we wait for this to happen we can then start working on CM9 or just go and buy an Archos gen11 device. Not everything that would be the ideal solution is also a viable solution.