Fork me on GitHub
#clojure-europe
<
2021-06-09
>
dharrigan06:06:30

Good Morning!

pez06:06:27

Morning. Today is day 3 on my “hunt for a bug without the tiniest shard of clue about what is causing it”. I am expecting it to be as productive as day 1 and 2. It’s hard to be scientific without theory!

thomas07:06:56

I have a bug like that as well at the moment... want to share?

pez08:06:38

I’m dying to share! There is some info here https://clojurians.slack.com/archives/C0E1SN0NM/p1623144035106200 But at a higher level I’d summarize it as for reasons unknown to me some of the recent changes we’ve made to our mobile, React Native, app causes it to sometimes stop responding to taps on Android phones. I don’t know how to reproduce it. Sometimes it happens in development, but then the same thing that causes this also causes the REPL connection to die, making me blind. I can see in the system logs of the phone that it registers the taps, but nothing happens in the RN logs and neither my app logging. So, probably some infinite loop somewhere locking the JS thread. That loop is not in the changed code. It could be in some of our code, even if I doubt that. More likely it’s somewhere in the react native thins. We’re on a pretty old version. I am now seeing if I can make it happen on a branch where we have upgraded to latest stable react native. My hopes now are that it will happen on this branch as well, falsifying this hypotesis, because if it doesn’t happen I will not know if that’s just a coincidence. 😃 No idea how much sense this made. Your turn to share!

thomas09:06:25

being on an old version of RN doesn't help here I suspect.

thomas09:06:55

not sure about adb though.

thomas09:06:37

does it happen on the simulator as well?

pez09:06:08

Being on an old version of RN does not help. Thing is that the app is super slow on Android in the branch where we have upgraded to RN 64, so there is probably a large amount of work still to do there before we can upgrade…

thomas09:06:48

hmm that sucks.

thomas09:06:18

any ideas why the slow downs on newer version of RN?

pez09:06:30

It doesn’t happen on iOS. Haven’t tried on the Android emulator, if that’s what you mean. I use a phone instead and Vysor to make Android development bearable.

pez09:06:44

No idea yet. I’ve mostly been using that branch with iOS so this slowness is still new to me. Thinking it might be a neighbouring cause to the thing happening for Android only with the old RN.

pez09:06:35

It happens in release builds as well, btw.

pez09:06:08

Anyway, mainly sharing for the sympathy. 😃 What about your bug?

thomas09:06:38

I have an expo app and there is on crash on iOS only. and it is the OS killing the App.

thomas09:06:02

but only with a build through TestFlight.

thomas09:06:24

all other builds work fine (debug and optimized local builds)

pez09:06:33

So not with a relasebuild uploaded to the phone locally?

thomas09:06:49

haven't tried that...

pez09:06:14

Maybe that’s the same as “optimized local builds”?

thomas09:06:19

can you just download an IPA to the phone and install.

thomas09:06:36

in expo you can say optimise the build(minify etc.)

thomas09:06:18

I added bugsnag to my app... but it is not the app it self throwing an exception. it is the OS killing it. and I don't know why.

pez09:06:27

Does iOS monitor some metrics of the app maybe and killing it because of some threshold met, maybe? Is it easily reproducable?

pez09:06:29

Do you get a dump of your app-state together with the Bugsnag report? Something you can load into the dev version of the app and examine from there?

thomas09:06:41

not sure about iOS logging things... I don't know enough about iOS for that... I have been googling a bit about this this morning.

thomas09:06:11

I don't get a bugsnag report in this case... as the app gets killed by the OS I think.

pez09:06:10

Ah, yeah, maybe it is not given the chance to say goodbye properly.

pez09:06:11

If there is something that could tell you that it has happened when the app starts you can send the snag then?

thomas09:06:10

I think it is at the end of a network operation... I could add some bugsnag log statements to that..

thomas09:06:34

and I just found a page that explains where the logs are on iOS

pez09:06:54

Please share!

thomas09:06:37

hope that help

pez09:06:40

Ah, yes, I recall that from the days I hacked iOS apps using Objective C… 😃

thomas08:06:22

Hi @U0ETXRFEW I got a bit further yesterday... It is a EXC_CRASH (SIGBART) crash. but no idea how to debug that at the moment....

pez08:06:25

Progress! I googled it quickly. Seems like it happens because of a bad call to the system from the app. Is that what you have found out about it as well?

thomas08:06:01

yes... and I am trying to figure out how to symbolize the crash report... but as it is an Expo app that isn't straight forward it seems.

pez08:06:30

So that gets trickier because it’s an Expo app?

thomas08:06:54

I think so... it is build on the Expo build farm.

thomas08:06:12

and the xcode version I have is really old.

thomas08:06:37

ok, it looks like I need to eject from Expo and use a so called bare workflow.

slipset07:06:23

Morning. First shot done 🐄

💉 26
javahippie08:06:57

Congrats! Getting mine tomorrow

mdallastella09:06:07

Had mine yesterday! 🥳

gklijs09:06:21

Getting mine this evening.

genRaiy15:06:27

very tempted to get into that bug thread ... must resist

pez15:06:42

It’s yummy!

😬 2
gklijs16:06:05

Just got the vaccine, having to wait 15 minutes for it to kick in.

👏 15
2