MMS users: call for testing

Emphrath

@lduboeuf haven't tried multiple mms yet

lduboeuf

lduboeuf

@emphrath are you trying to send a MMS to yourself ?
Any screenshot available ?
[Edit] : no need for screenshot. So there are two mms notifications instead of only one. Maybe @jezek have an idea

Emphrath

@lduboeuf no, i was sending from another phone.

jezek

@emphrath said in MMS users: call for testing:

Tested this morning on FP2 halium-devel. It works, although not perfectly. First, it seems like I had to toggle WiFi along with cellular to receive the images. Might be related to data being limited to H+ on this sim.

I'm pretty sure WiFi has nothing to do with MMS. Unless some new technology I missed, or the operator is very brave to put user messages on public internet. Or am I wrong?

For each message sent, the receiving FP2 often (not always) displayed two different error messages: the usual one with a download button, and a new, shorter one, with the size of the attachment, and a download button. If i tap on the usual one after toggling data back on, it downloads, and deletes the other message. If i tap on the new one, it also downloads, but the other message is still displayed, and tapping the download button is useless.

I've tried halium-devel on my FP2 yesterday and I can't reproduce the double message error behavior.

How many times does it happen?
Have you rebooted your phone after applying the ubports-qa commands?
Does any message got lost (you didn't receive the message in the end) aside of the double error message?
~~After restarting the phone does the useless error message disapear?~~ Edit: Ignore this question. There is no logic for such a behavior. The answer should be always "No".
Could you provide logs (nuntium.log, dbus.log, application-legacy-messaging-app-.log) from the case when the double error message happened?
Note: It would be helpful, when you delete the logs before the test to not pollute the test logs with noise (logs with events before test). I prefer to:
1. Delete (relevant) logs.
2. Reboot phone.
3. Do tests until the error occures.
4. Copy/paste logs to some service (pastebin, gist) and report the link.

Thank you for your cooperation.

Emphrath

@jezek unfortunately i won't be able to test again. I used a friend's sim in the fp2, which is not my main device, and my present sim is stuck in my oneplus tray. I can try to get the logs from the fp2, but they won't be "clean". The fp2 was having some issues of its own, too, so maybe that's that. Sorry I can't help further

jezek

@emphrath said in MMS users: call for testing:

@jezek unfortunately i won't be able to test again. I used a friend's sim in the fp2, which is not my main device, and my present sim is stuck in my oneplus tray. I can try to get the logs from the fp2, but they won't be "clean". The fp2 was having some issues of its own, too, so maybe that's that. Sorry I can't help further

It's OK. Just provide the logs, you have. But don't wait too long, the logs can get more polluted (and it will be more pain to go through them). Also keep in mind, that logs get rotated, so the relevant parts may be already zipped in another file (remembering the time of test will help). And try to answer the other questions, please. Thanks.

lduboeuf

@jezek just as a reminder, the issue that i've found last time:
https://github.com/ubports/nuntium/pull/8#issuecomment-861480575

Emphrath

@jezek I don't use that phone, and it's just sitting there with no wifi or cellular, so I guess what you'll see in the logs will be rather plain, mostly background daemon routines. I'll try to get them tonight. Now for the questions:
-I think I sent four mms in total and all of them displayed this behavior.
-I did reboot, since ubports-qa explicitly asked me to (and you did as well) but I saw some errors while passing the commands. I can't remember properly, but sth about sth not being sandboxed, and also about the system being ro... I didn't have to make it
rw beforehand, right ? Anyway i'll get you these errors asap
-no message got lost

lduboeuf

@emphrath ubports-qa script often output errors even if ok

@jezek it seems nuntium received two wap push don't you think ? That could explain the two mms notif.
In case of bad signal if the wap push couldn't be acknowledged, the provider might send it twice t

Emphrath

@lduboeuf Quite frankly I wouldn't look any further. My fairphone has so many hardware issues it's hard to get a proper idea about the culprit - noticeably network issues, so...

lduboeuf

For the use case "cellular data off", it is working nicely except when i'm with bad signal ( see issue above ) .

This is already a huge step for MMS users, in its worst case scenario we will have the same state as it is today on Devel/RC/Stable ( just ask user to resend)

jezek

@lduboeuf said in MMS users: call for testing:

For the use case "cellular data off", it is working nicely except when i'm with bad signal ( see issue above ) .
This is already a huge step for MMS users, in its worst case scenario we will have the same state as it is today on Devel/RC/Stable ( just ask user to resend)

I think it is not a showstopper for deploying to next OTA. It's working better than before and there are no regressions. Also it will motivate me more to fix the bad signal issues, when the changes land to stable. I will definitely tackle the issues, but I had to change my HW and installed manjaro on it and now I have a little fight with crossbuilder and the lack of apt and dpkg[-architecture,-parsechangelog] commands on the system.

Thank you all for support.

jezek

I've discovered, that the patches are already in RC, so if you are on RC channel (and dev channel too), there is no need to apply the above mentioned ubports-qa commands. This also means that the patches will probably be in the next OTA. Just saying. I'm happy. Thank you all.

I've been wrong. Read furher, why.

AppLee

@jezek said in MMS users: call for testing:

Thank you all.

Thank YOU :thumbs_up:

lduboeuf

@jezek humm i probably miss something, i don't see the PR already merged

jezek

@lduboeuf Or am I wrong? Is there a possibility that the patch is not in RC (and dev) and the applied ubports-qa changes persisted 2 RC upgrades and I drew a bad conclusion? Can anyone answer this conundrum, or do I have to do full reinstall to check?

lduboeuf

@jezek

I'm not 100% sure, but debian packages could be not totally overwritten. I already noticed that when working on Lomiri, where RC or devel updates didn't overwrites my modified files. It would overwrite it though if the same file was modified.

jezek

So I red the OTA-19 call for testing and after browsing though the github QA page I have come to the conclusion, that I've been definitely wrong about the patch landing in RC and maybe in OTA-19 update. I'm sorry for putting your (and my) hopes up. Maybe next time.

Does anyone know how to be in the next OTA update?

lduboeuf

@jezek

Does anyone know how to be in the next OTA update?

idk, maybe @UniSuperBox or @Flohack can say ?