Our pure JavaScript Scheduler component


Post by mats »

Very old post that one, not sure it applies today. We'll need to wait for guidance from Apple.


Post by nickcc »

I see Apple have posted a reply. They move at glacial pace and it is hard to be confident that they will be serious about assisting.

https://www.manchestereveningnews.co.uk/news/greater-manchester-news/girl-11-killed-indefensible-punch-29618832

I wass browsing this page when my iphone started getting rather hot!! Not a particularly pleasant story but if you zoom a few times it also crashes the brower which refreshes and then sometimes whitescreens similar to the issues we are facing... It is a newspaper's website so should be relatively simple compared to a calendar? Obviously the have done something not quite right which is causing their crashing in iphone safari. I am sure they also intend it to work on an iphone. As their html is visible I wonder if it is possible to work out and isolate what in it may be causing the crashes and then once identified perhaps there is something similar in the calendar code which could then be looked at and maybe adjusted?

I don't know. It is just a thought for a way to get to the bottom of it. I am clutching at straws here as really desperate to get this issue resolved!

Regards,

Nick


Post by nickcc »

Gents,

Another observation I had which you may be able to reproduce: I noticed that on the iphone the crashing issue is far worse when there are a number of other apps open in the background. Iphone users have apps open in the background all the time but it seems like the calendar is so resource intensive that if pretty much anything else is open it falls over more easily. Performance is considerably better when all other apps are closed first (although this should obviously not be necessary). This suggests the calendar is pretty resource heavy and runs the iphone out of resources and then it crashes on zoom and/or when other mathy functions are required. Yes Safari is crashing but is that the just effect or the root cause?? I am not a technical person but maybe there is a memory leak being caused or some sort or a loop that consumes resources? I don't know what to suggest. Could there be a systematic review of the code looking specifically for such issues in as much as you can see from your side? If you try it with other apps open in the background vs without you will see what I mean and that surely points to a resource use issue?

I know apple try to clamp all the diagnostics down (and then either don't help at all, or are very slow to help) but could an app like this perhaps be helpful to see what is going on under the bonnet? https://techet.net/
It is hard to accept there is nothing that we can do try and deal with this other than wait for assistance from Apple which we can't even be certain will even ever come. Hope is not a strategy!

I am sorry if I am beginning to sound rather desperate but it really impacts our plans given how heavily we have invested in the PWA concept and thrown our lot in with the calendar. The customer experience with constant crashes will regrettably be relatively poor.
So many users these days are increasingly on mobile devices with the iphone as the main device used. Surely it is in everyone combined interest to make best efforts to resolve these issues as quickly as possible?

Nick


Post by alex.l »

Hi Nick,

Thanks a lot for the researchment!

The thing is that when browser crashed, you loose all debug information and it's getting impossible to see last thing happened before crashing. Of course, the problem might be caused by memory leak, bot that seems like it's browser's memory leak. We don't have access to the browser code to debug why memory leak happened as well as fix that problem, this is possible to do from Apple side only.
I've spent a few days on debugging this problem and analysing Performance graphs. I didn't find any leaks on our side. I also was able to reproduce crash with empty Scheduler and without React! So it confused more than before and doesn't look consistent if the problem on our side. We do not really want to play ping-pong with Apple and interested in a fast fix as you. I've updated a topic with the question. Looking forward for reaction.

All the best,
Alex


Post Reply