[Replicant] Replicant 6.0 0004 RC4 images are ready for testing.

doak doak+list at posteo.net
Fri Feb 5 21:20:22 UTC 2021


Since I have access to some spare i9300 now as well, I could test.
Let me know if it is by any interest, if I should test someting special or another version.

Regards,
Korbinian



On 18.01.21 02:35, doak wrote:
> Hello Denis,
> 
> the same happens even if I skip the fancy stuff.
> 
> What I did:
>        * Wiped data and cache
>        * Flashed 0003
>        * Installed and configured some apps:
>            * Added contact in local addressbook
>        * Wiped Cache
>        * Flashed 0004 RC4
>        * Booted
> 
> Doesn't work:
>        * Several "Unfortunately, the process android.process.acore has stopped." messages
>        * Contacts does not show contact (it fails at startup) and triggers above error as well
> 
> 
> Best regards,
> doak
> 
> 
> On 18.01.21 02:03, doak wrote:
>> Hello Denis,
>>
>>> - From Replicant 6.0 0003 to Replicant 6.0 0004 RC4
>>
>> I have tested this on an i9305.
>>
>> What I did:
>>       * Wiped data and cache
>>       * Flashed 0003
>>       * Enabled ADB
>>       * Copied firmware blobs for wifi and bluetooth
>>       * Installed and configured some apps:
>>           * Added contact in local addressbook
>>           * Configured wifi access point
>>           * Installed Nextcloud and configured account
>>           * Installed Corona Contact Tracing Germnay, granted all permissions and let it update keys.
>>           * Started browser and visited some site
>>       * Wiped Cache
>>       * Flashed 0004 RC4
>>       * Booted and installed firmware blobs for wifi and bluetooth again
>>
>> Works:
>>       * Configured account for Nextcloud works
>>       * Previous data for Corona Contact Tracing Germany is still available
>>       * Browser has history from last time.
>>
>> Doesn't work:
>>       * Several "Unfortunately, the process android.process.acore has stopped." messages
>>       * Contacts does not show contact (it fails at startup) and triggers above error as well
>>
>>
>> Regards,
>> doak
>>
>>
>>
>> On 15.01.21 17:53, Denis 'GNUtoo' Carikli wrote:
>>> Hi,
>>>
>>> I've just pushed the Replicant 6.0 0004 RC4 images.
>>>
>>> I didn't have the time to test them at all since I was (and am
>>> still) working to integrate the telephony stack in Replicant 11 to test
>>> it (it's not ready yet either but I'd like to test it to see if the
>>> work in progress I have so far really works somehow).
>>>
>>> We need help for testing:
>>> =========================
>>> Beside booting and telephony, what really needs to be tested is the
>>> upgrade between different releases: it needs to work without wiping the
>>> data partition, else most people won't be able to upgrade.
>>>
>>> If the test goes wrong you might have to wipe the data partition and/or
>>> try to recover the data from it (we have scripts for that too but it
>>> also needs to be more widely tested).
>>>
>>> To do the tests you can simply install a Replicant 6.0 release and if
>>> possible install some free software APKs from f-droid (that can be done
>>> with adb install path/to/file.apk), and then install a more recent
>>> Replicant 6.0 release without wiping the data.
>>>
>>> If the second installation fails and that you need to wipe the data,
>>> then we have an issue that needs to be fixed.
>>>
>>> What needs to be tested the most are the following combinations:
>>> - From Replicant 6.0 0003 to Replicant 6.0 0004 RC4
>>> - From Replicant 6.0 RC1 to Replicant 6.0 0004 RC4
>>> - From Replicant 6.0 RC2 to Replicant 6.0 0004 RC4
>>>
>>> Why we need to test the upgrades:
>>> =================================
>>> During the build of an image, signing keys are generated and these keys
>>> are also used to sign the system applications that are built and shipped
>>> in the image.
>>>
>>> The data that an application generates is stored in a location that is
>>> only accessible by an application of the same technical name (like
>>> fil.libre.repwifiapp) and signed with the same key, or by the Android
>>> system. So when signing key changes we have to migrate the data to be
>>> used with the new key.
>>>
>>> The Replicant 6.0 up to the 0003 images were signed by the same keys
>>> (generated by Wolfgang Wiedmeyer).
>>>
>>> For the 0004 RC1 I generated new keys. As I didn't connect the dots, I
>>> didn't keep these keys and I generated new ones for the 0004 RC2.
>>> Hopefully at that point I understood the issue and I kept the keys, so
>>> the RC3 and RC4 are signed with the same keys. I wrote code to generate
>>> a migration script that is now part of the RC3 and RC4 images.
>>>
>>> Since this hasn't been extensively tested yet and that people's data
>>> are at stake, it's a really good idea to do test.
>>>
>>> Most visible changes between the RC:
>>> ====================================
>>>
>>> Changes with RC4:
>>> ----------------
>>> - More nonfree software was found and removed.
>>> - The OMAP4 devices are now fast enough again (but without complete
>>>    OpenGL compatibility).
>>>
>>> Changes with RC3:
>>> ----------------
>>> - Telephony was fixed
>>> - A script took care of the data partition migration.
>>> - The OMAP4 devices were still too slow.
>>>
>>> Changes with RC2:
>>> ----------------
>>> - The OMAP4 devices booted fine but the graphics were way too slow.
>>> - Broken telephony / cellular network.
>>> - The data partition needed to be reformatted.
>>>
>>> Issues with RC1:
>>> ----------------
>>> - The OMAP4 devices (Galaxy Tab 2 and Galaxy Nexus) had a black screen
>>>    at boot.
>>> - The data partition needed to be reformatted.
>>>
>>> Other less visible changes were also done between the RC.
>>>
>>> How to report test results:
>>> ===========================
>>> You can reply to this mail. Alternatively you can also add the
>>> information in the https://redmine.replicant.us/issues/2207 bug report.
>>>
>>> Giving enough information to enable other people to do the exact same
>>> test also helps a lot. For instance reporting the device used for the
>>> test, the images, which APKs are installed (if possible or applicable)
>>> could help reproduce tests if there is any issue.
>>>
>>> If some information is missing we often can still manage to reproduce a
>>> similar result but it often takes a lot of time as we need to test many
>>> combinations until we reach the same result.
>>>
>>> Denis.
>>>
>>>
>>> _______________________________________________
>>> Replicant mailing list
>>> Replicant at osuosl.org
>>> https://lists.osuosl.org/mailman/listinfo/replicant
>>>
>>
>>
>> _______________________________________________
>> Replicant mailing list
>> Replicant at osuosl.org
>> https://lists.osuosl.org/mailman/listinfo/replicant
>>
> 
> 
> _______________________________________________
> Replicant mailing list
> Replicant at osuosl.org
> https://lists.osuosl.org/mailman/listinfo/replicant
> 

-------------- next part --------------
A non-text attachment was scrubbed...
Name: OpenPGP_signature
Type: application/pgp-signature
Size: 840 bytes
Desc: OpenPGP digital signature
URL: <http://lists.osuosl.org/pipermail/replicant/attachments/20210205/fd16b686/attachment.asc>


More information about the Replicant mailing list