SleepyHead "unstable" builds for advanced testers.

[Drowsy Dancer]

Dunno. I'll try it again tomorrow when I have some new data to import.

[jedimark]

The Statistics (under the pie chart) are off in 9.5.0 and 9.5.1 (9.5.1 is on the right, 9.2.0 on the left -- I didn't include 9.5.0 since it's the same as 9.5.1).

The Maximum record was accidentally switched in the code between the True Maximum and the 99.5% percentile in older versions.. this was a bug fix.
There is an option in General Preferences to change between them, but previous to 0.9.5, they were inverted in how the options were interpreted.

Thanks Mark, but now I'm confused. For example, the leak graph is empty as it always is (nasal pillow mask, mouth tape, no leaks) but 9.5.1 is showing a Max of 14.40 which doesn't seem right.

SleepyHead is reporting total leaks. A maximum leak of 14.40 is excellent..
The threshold is about 24.0 for a SwiftFX mask is before it's considered a large leak.

I find displaying the 99.5% as the Maximum option is useful, because it allows you to ignore "outlier" data caused by briefly pulling of the mask to sneeze/cough/scratch/whatever or adjusting the pillows for leaks.

You can turn it back on in preferences by changing the dropdown in Preferences -> General -> Preferred Calculation Methods -> Maximum Calcs if you prefer this behaviour.

It was just my intention in the code to show True maximum as default, but a typo caused it to do the opposite.

[kaiasgram]

Thank you, Mark.

[DKC_apnea]

After loading v0.9.5 I have been unable to make use of my previous preferences, and especially the daily "notes" stored from the previous beta release v0.9.4.

What am I missing to capture that information and use it in version 4.5??

[jedimark]

After loading 4.5 I have been unable to make use of my previous preferences, and especially the daily "notes" stored from the previous beta release 4.4.

What am I missing to capture that information and use it in version 4.5??

It's created another SleepyHeadData directory.. there will be two of them under your Documents with an added "-unstable" or "-testing" tag. (it could also be the without the tag, depending on the previous version you were using) I added the tags so it was possible to run two versions concurrently while it was being worked on it.

Each SleepyHeadData folder contains it's own group of profiles and system preferences.

When you first run you can choose which data folder to run from.

Hopefully this is what you meant. :-}

[DKC_apnea]

Okay, got it pointed to the previous Sleepyhead_data folder, and it has captured the notes, although it did not grab the previous preferences (e.g. graph settings).

thanks

[JohnO]

As a PR System 1 Series 60 user I've been using the PR Encore software as Sleepyhead didn't support my machine until recently. This afternoon I downloaded and tested 0.95-1-testing on my Mac. Very impressive! I did a quick comparison with the Encore numbers, and everything seemed to line up correctly. Being able to check my stats without launching my Windows VM is great.

I do wish the QT GUI toolkit was a bit more polished, but I know there isn't much you can do about that, Mark.

I hit the PayPal tip jar, and recommend others do the same!

John

[jedimark]

As a PR System 1 Series 60 user I've been using the PR Encore software as Sleepyhead didn't support my machine until recently. This afternoon I downloaded and tested 0.95-1-testing on my Mac. Very impressive! I did a quick comparison with the Encore numbers, and everything seemed to line up correctly. Being able to check my stats without launching my Windows VM is great.

I do wish the QT GUI toolkit was a bit more polished, but I know there isn't much you can do about that, Mark.

Thanks for the confirmation.. I'm glad the 60 series stuff is better behaving itself..

There is always StyleSheets. I haven't bothered to pretty things up because I've been focusing on underneath a bit.

It would be nice one day having a few loadable themes for the GUI stuff.

[Mykale]

I have been giving Build 0.9.5 a test run on my Windows 7 64 bit machine. I found that when I click on the calendar on the Daily Tab, to select a previous day and compare results, I am getting a frequent program crash (Sleepyhead.exe has stopped working).

Update on my previous post...

Prior to this morning I had only tried to use 0.9.5 with limited SD card data. I am talking about the data that you get when you just pop the SD card into the PR System One to download and then remove it. Well, I decided to check out what my waveforms looked like so I left the SD card in the machine last night. This morning when I retrieved the card and went in to check the results on the computer I was surprised. After uploading the SD card with "Full" data, I could click on the calendar as much as I wanted without any program crash. I re-launched the program several times and it appears very stable. The difference is that 0.9.5 had full data this time instead of partial data. Hopefully this will help Mark in his debugging....

[jedimark]

I have been giving Build 0.9.5 a test run on my Windows 7 64 bit machine. I found that when I click on the calendar on the Daily Tab, to select a previous day and compare results, I am getting a frequent program crash (Sleepyhead.exe has stopped working).

Update on my previous post...

Prior to this morning I had only tried to use 0.9.5 with limited SD card data. I am talking about the data that you get when you just pop the SD card into the PR System One to download and then remove it. Well, I decided to check out what my waveforms looked like so I left the SD card in the machine last night. This morning when I retrieved the card and went in to check the results on the computer I was surprised. After uploading the SD card with "Full" data, I could click on the calendar as much as I wanted without any program crash. I re-launched the program several times and it appears very stable. The difference is that 0.9.5 had full data this time instead of partial data. Hopefully this will help Mark in his debugging....

Ah ok.. something stray little bit is expecting waveform data to be there.. Strange though, because the Channel System only uses what's available. I'll try and replicate it by trashing all the waveform files in a fresh import.

[DKC_apnea]

Okay, having some issues with how beta v0.9-5 is now handling/interpreting some of the data from my ResMed S9 Elite. These may be known issues, or new I am not sure.

I've can e-mail attachments/screenshots if you wish.
Graphs showing Mask Pressure, Pressure, Insp Time, and Exp time have changed unexpectantly.

Somehow mask pressure and Pressure graphs are missing the maximum now, and both graphs have changed display for no apparent reason.
Insp time has dropped to near a zero baseline which seems very strange. Previously I was always in the 2 sec range.
Exp time looks questionable as well, as my baseline has been generally 2 secs, yet it seems to have dropped.

Am I doing something incorrect, or are these known issues with this new version. Did not have these problems with 0.9-4, but now if I try using that version I am getting same issues.

Getting to the point where I'm thinking of deleting everything, and re-installing new version again. Would re-install a previous release, but those seem to be gone now.

Any advice?

[jedimark]

Okay, having some issues with how beta v0.9-5 is now handling/interpreting some of the data from my ResMed S9 Elite. These may be known issues, or new I am not sure.

I've can e-mail attachments/screenshots if you wish.
Graphs showing Mask Pressure, Pressure, Insp Time, and Exp time have changed unexpectantly.

Somehow mask pressure and Pressure graphs are missing the maximum now, and both graphs have changed display for no apparent reason.
Insp time has dropped to near a zero baseline which seems very strange. Previously I was always in the 2 sec range.
Exp time looks questionable as well, as my baseline has been generally 2 secs, yet it seems to have dropped.

Am I doing something incorrect, or are these known issues with this new version. Did not have these problems with 0.9-4, but now if I try using that version I am getting same issues.

Getting to the point where I'm thinking of deleting everything, and re-installing new version again. Would re-install a previous release, but those seem to be gone now.

Any advice?

Does it do this in a fresh profile? It may need a purge and reimport from the Backup folder. I should have probably increased the version number so it forced a Purge.

I'm going to make Purge data automatically reimport from ResMed backups if they are available, so no one has to hunt for it anymore..

[DKC_apnea]

Decided to install v0.9.5 on a different PC with no previous SleepyHead versions or directories (i.e. like a fresh install). Created a profile, and then Imported my saved data. Now the graphs (mask pressure, pressure, Insp time, and Exp Time) for every day are messed up, whereas previously only the days since I upgraded were a problem.

To confirm it is this new version, I wiped that PC clean (of all Sleepyhead directories and versions) and extracted v0.9.4-1 which I managed to find. Created a new profile and the data is displayed correctly using v0.9.4-1. I know version 0.9.4-4 worked also though I've lost the zipped version, so can;t re-install it cleanly.

So my guess is, it is an issue somehow with the latest version, and nothing to do with data previously stored, backup folders or ResMed backups.

Side note (though doubt it is related).... with v0.9.4-4 (and all previous versions) I had never been able to display "snore" data. With version 9.5 it is showing up now. I thought my machine (the ResMed S9 Elite) did not record snore data. Maybe I was mis-informed, or maybe the software is importing this data and somehow adversely affecting how other data/graphs are interpreted?

[SleepyToo2]

Gidday, Mark,

Great to see you back on the board! Just thought I'd let you know I was having some problems with the previous version hanging every time I opened it - had to close it with Windows Task Manager, then reopen. Sometimes it took a couple of times. After that it seemed to be fine most days, with the occasional random crash. Anyway, this morning I updated to version 0.9.5 (testing), revision 3ec71cd (master branch). Bit of a problem in finding the data until I realized that I had stopped using an SD card because it would no longer work in my machine. Once I found the old card, it updated just fine, and I now have a complete data set since I started using this machine (11k+ hours).

Apart from a couple of anomalies, Like Sep 7, 2013, which has me with a respiration rate of 15,539.37 and a max leak rate of 97.00, it looks like it is working well for me. Good stuff - keep up the good work, mate.

[DKC_apnea]

Okay, I may have narrowed down what was causing my problems with the new version. The default for the preferences/profile is to save an SD Card backup (recommended for ResMed machines). In my case I have always downloaded the SD data to a backup file on my hard drive, and then have Sleepyhead import and read the data from the hard drive file. So when first running v0.9.5 it was looking to this new backup data folder that it had created but based on a different machine.

To resolve this, I created a new profile and made sure the "SD Card backup" option was turned off. Once the profile was created, I copied over the correct machine data from the SD card itself, and made sure the datafolder file had the correct machine data and all saved SD card data. That finally resolved my issues. I then ported over the journal files to my new profile journal directory and everything is back on track.

Note: I always kept the SD card write-protected when out of my ResMed S9 Elite, to avoid any possibility of overwriting data on the card. Each day I simply download any new data from the card and add it to my stored data on the hard drive. I have not reached the stage where I trust the machine or the software to maintain my data.

Obviously I'm not an "advanced" tester as yet, but I'm getting a better appreciation of how directories/profiles are created and data saved.

thanks