Random Crashes While Performing

[fleahead]

During the past several weeks we have experienced an occasional SLM crash (iPad desktop appears and recording stops playing) while performing. Needless to say, this is not a good thing. Fortunately, it has only happened four times in the last 10 gigs, but it makes you a little apprehensive. There is no pattern or logic as to when it will happen. It's been on different songs. I run a memory cleaner before each job, and make sure no other apps are running. Argo, would it do any good to send a troubleshooting report? I thought there was a way to send an iPad crash log, but I can't remember where to find it. We are linked over wifi during the jobs, but it seems to be working OK.

[arlo]

Yes, please send a message from Help > Request Tech Support so I can see your device details and app settings.

[fleahead]

will do

[arlo]

I got your message. Have you turned on Privacy > Diagnostics & Usage > Automatically Send > Share With App Developers in the Settings app? That will allow me to log into Apple's system and look for your crash logs. I just looked through the recent logs, but didn't see anything that looked like it came from your device.

Otherwise, can you test this at home broadcasting remote control messages but not playing recordings, then playing recordings but not broadcasting remote control messages, to try reproducing the problem? The goal is to find out if the problem is related to one of those two functions.

[fleahead]

Sorry I had not responded to the last post. After double checking my partners iPad I discovered he was not current with the latest SLM patch and also not up to date with iOS version. After fixing that, we have not had any more issues of that type, but I don't understand how issues on his iPad would cause mine to crash???

I do, by the way, have my iPad optioned to automatically send diagnostic data to you.

[arlo]

I don't understand how issues on his iPad would cause mine to crash???

That's hard to answer without knowing what version he had, and going back through the code from that version to see what has changed since then.

If your real question is, "Is my problem really fixed now?" then I would recommend more testing with your setup to be sure. If you do see the problem again, then please see my previous comment about separating the device linking from the recording functions. The first step in understanding what's happening would be testing those two functions separately to see which one is causing the problem.

[fleahead]

I removed the duplicate post of this subject. I fully thought that this problem was resolved, as we hadn't had any more problems until the past couple of weeks.

It will be very difficult to to isolate the condition that may be at fault, since it occurs so infrequently. It has never crashed during the same tune, and I do know that, at least during one of the crashes, there was no automation activity (midi) occurring. The only condition that seems is be consistent is that we are linked over wifi, and I am the master iPad and the other guy is setup to see the lyrics when I start the recording. When we do have a crash, we lose our link, just to be safe, we do not try to re-establish it, so he manually starts his lyrics. I don't recall ever having a crash while we were not linked.

Like I said on the other post, I sent the troubleshooting report from the SLM app, but there was no data in the iPad crash history for SLM. I did send you some data from another file though that may connected to the crash since it was the same time and date of the crash.

Also, we have been running the same iOS version  and the same SLM patch. The only difference is that I use an iPad 3 and he has an iPad 2.

[arlo]

All right, the crash log you sent was incomplete, but I found a crash log from another user that looks like the same problem. It appears to be the same problem we discussed here:

http://forum.arlomedia.com/index.php/topic,961.0.html

I had implemented a potential fix last July, and it seems that that reduced the chance of the problem happening, but did not eliminate it. The problem is that iOS is releasing some objects from memory before it's supposed to. This is the same problem that required me to add the Help > Troubleshooting > Simplify List Loading setting as a workaround to avoid crashes when loading long lists.

Set List Maker and BandHelper still use the iOS memory management system that was current when I first built them. Since then, iOS has added a newer system, but I will have to do some significant rework to use it. I'm planning to do that as part of the next major update (Set List Maker 5.0 / BandHelper 3.0) in the coming months. Unfortunately, I will have to do all that rework before we can find out if that solves the problem.

Meanwhile, as a workaround, can you play your recordings from the slave device instead of the master device so the recording doesn't stop if the master device crashes?

Alternatively, I could send you a test version of the app that skips the part of the code where I think the crash is occurring. This code checks to see if the slave devices are still connected so it can update the little number in the top toolbar -- not essential.

[fleahead]

Loading all my music files  into the partner iPad, and editing all the songs and recording levels for the 70+ tunes we do, is simply out of the question. I would try using the modified app, though. Do you still have my information from when I was in the beta program? I'm getting ready to install the latest patch today. Will that have anything to do with the issue?

[arlo]

I don't think the latest releases will have any change relevant to this. I do have your device ID for an iPad 3. Please email me directly to request a test copy.

[fleahead]

Will do...

note: we did a 3 hr job last night with no crash.

[fleahead]

FYI... this is still happening. I just sent you another Help Desk Report, and I included the entire log. Apparently Apple has done away with Crash Logs. What I sent was a log called CoreTime-2016-03-16-201452.ips.synced. The time in the file was at the exact same time as the crash.

Same scenario for the crashes. Never the same song. Never the same time into the song. It is so infrequent (maybe one out of a hundred), that troubleshooting would be next to impossible, since it can not be readily duplicated. Hopefully, you can find something in the data.

[fleahead]

Set List Maker and BandHelper still use the iOS memory management system that was current when I first built them. Since then, iOS has added a newer system, but I will have to do some significant rework to use it. I'm planning to do that as part of the next major update (Set List Maker 5.0 / BandHelper 3.0) in the coming months. Unfortunately, I will have to do all that rework before we can find out if that solves the problem.

Frankly, I'm rather disappointed that resolution of a known issue that is effecting the professionalism of a paying customer, is not job priority #1. Approaching 6 months of on stage failures without a fix, can accurately be described as absurd.

Oh, and I am still getting the occasional crashes.

[arlo]

All right, I'll explain the situation again. I can't reproduce the problem, so I can't reliably fix it. I have released several updates that attempt to fix it, but since I can't reproduce it in the first place, I can't tell if a fix is effective without waiting for user feedback. The feedback I have received is that the problem is much improved but not totally gone.

My next tactic was to rework the memory management system of the app, which is a very large task and is likely to introduce new bugs, so could not be taken lightly. I did complete that in the last week. However, testing a similar problem after the rework indicates that it is not going to help with this problem after all.

During this rework I have seen more evidence that the problem occurs in a low-memory situation, so separating your audio playback functions to another device should help (your response to that suggestion earlier was that it was "out of the question"). Turning on Help > Utilities > Use Old Audio Engine should also help. Upgrading to a newer iPad should also help, especially if you are running iOS 9, which the iPad 3 sometimes struggles with.

Now that the memory rework is done, I could send you some test copies to test with. In order for that process to be effective, I would need you to test each version with your on-stage setup long enough to be reasonably sure if the problem is fixed or not, rather than just waiting until your next gig to see if the problem happens again.

[fleahead]

I could try the test version, but since the problem occurs so infrequently, (we had no crashes at all the first part of last month), pretesting prior to a gig will not prove much, other than making sure all the features work properly. I believe you told me you still have my beta testing information. Do you need anything from me?