SleepyHead 0.8.8 "Almost Beta" Release

Mark, I finally figured it out! I deleted the sleepapp folder and deleted the offending file from the test at the DME, which was the the first file that was conflicting with the last file from the previous CPAP machine. One thing that I did notice though, when I import a new day of info, I have to do it all over again. I thought that the import of new data looked at just new data and not data that was previously there. Do I, should I, delete the info off of the SD Card? That's whats causing the import to burp. I know that I could delete the last day off of the old machine, but that is real info.

cpaptex wrote:Mark, I finally figured it out! I deleted the sleepapp folder and deleted the offending file from the test at the DME, which was the the first file that was conflicting with the last file from the previous CPAP machine. One thing that I did notice though, when I import a new day of info, I have to do it all over again. I thought that the import of new data looked at just new data and not data that was previously there. Do I, should I, delete the info off of the SD Card? That's whats causing the import to burp. I know that I could delete the last day off of the old machine, but that is real info.

Now I can get back on topic again. I am sure people here will agree with me when I say NEVER delete the data from your card. In fact do the opposite. Back it up and back it up often!

ThomasMcKean wrote:

cpaptex wrote:Mark, I finally figured it out! I deleted the sleepapp folder and deleted the offending file from the test at the DME, which was the the first file that was conflicting with the last file from the previous CPAP machine. One thing that I did notice though, when I import a new day of info, I have to do it all over again. I thought that the import of new data looked at just new data and not data that was previously there. Do I, should I, delete the info off of the SD Card? That's whats causing the import to burp. I know that I could delete the last day off of the old machine, but that is real info.

Now I can get back on topic again. I am sure people here will agree with me when I say NEVER delete the data from your card. In fact do the opposite. Back it up and back it up often!

Thomas, I KNOW that I shouldn't delete the file, but the data on the new machine is all 0000000, it represents only that they turned on the machine to show me that it was working, nothing more. It happened at 10AM and causes a problem with the previous day because of the way Sleepyhead works. Maybe a routine can be developed to reject the conflict rather than to crash. But I thought that Sleepyhead ignored old data when it imported new data based on the date. Perhaps not.

Thomas, I KNOW that I shouldn't delete the file, but the data on the new machine is all 0000000, it represents only that they turned on the machine to show me that it was working, nothing more. It happened at 10AM and causes a problem with the previous day because of the way Sleepyhead works. Maybe a routine can be developed to reject the conflict rather than to crash. But I thought that Sleepyhead ignored old data when it imported new data based on the date. Perhaps not.

I see! That is one for our developer, Jedimark. I sure can't answer it...

My Bad! I was copying over the data from the SD Card and then importing. Duh, I should have een importing from the card. As always, operator error.

Mark,

Just tried the newest version on my W7 64bit machine and it is really slow. First off, it is only importing 31 days from the directory (last 31 days) even though I have over 90 days. And half the time when I go to import and tell it to use a new location, I have to kill the program as it goes unresponsive.

Any ideas on what is going on? On XP it works just great.

John

Using latest version Windows 7...64 bit and seems to be working fine. I had one episode where I had to point to a new location when I had already thought it should have been there.
This is with the VPAP machine data. I also imported the prior PR S1 BiPap data without issue so all my data is there.

Perrybucsdad wrote:Mark,

Just tried the newest version on my W7 64bit machine and it is really slow. First off, it is only importing 31 days from the directory (last 31 days) even though I have over 90 days. And half the time when I go to import and tell it to use a new location, I have to kill the program as it goes unresponsive.

Any ideas on what is going on? On XP it works just great.

John

The only things that really slows it down is a crummy video card and lack of system memory.. I doubt your 64bit W7 box suffers from either of those conditions.

ctrl-shift-escape should bring up the task monitor in windows.. not sure how to check per app memory usage..

Most of my development work is on an old P4 Dell laptop with crummy intel integrated graphics.. It runs very well on this machine.

So something else will be causing these issues.. As for what.. I'm not sure. How big is your SD card?

As for the import issue, I think this is a Qt custom file dialog bug coming back from the grave.. (Qt 4.8 was supposed to have fixed it )
Killing the non-native file dialog means losing multi-folder select, but this now has two strikes against it.. different issues on both on windows & mac.

I've just written an AutoUpdate module for SleepyHead.. I need to give it a thorough thrashing before it goes live.. but shouldn't be too long before it's ready.
It will make future upgrading fully automatic. (It asks first of course..)

cpaptex wrote:My Bad! I was copying over the data from the SD Card and then importing. Duh, I should have een importing from the card. As always, operator error.

So you have it all sorted out now?

Sorry for not responding, I seem to have stopped receiving emails for this thread again..

I still need to put more thought into handling the machine crossover day..

In my adventures with SleepyHead and Flex, I found that when I do a clean import, all sessions list the Flex setting used by the most recent session, not the setting that corresponds to the session in question.

tetragon wrote:In my adventures with SleepyHead and Flex, I found that when I do a clean import, all sessions list the Flex setting used by the most recent session, not the setting that corresponds to the session in question.

At the moment, for days with multiple sessions, I'm just picking the highest value flex setting.. An admittedly lazy approach. (I do the same with the humidifier settings)

It shouldn't make any difference on fresh import or not... ? (unless the day cache object is being corrupted somehow, this gets rebuilt when you move an import slider in preferences)

I could list each individual sessions Flex/humidifier settings separately in the Details tab, but that might look a little messy data wise.. it's already quite full, especially with the Pie chart switched on.

I could also just show the most commonly used (for the longest session) setting, or a (rounded?) weighted average?

I can't remember which method I used in the Overview Chart.. think it's just max again..
EDIT: Overview shows the weighted average of all sessions for a day.

jedimark wrote:

tetragon wrote:In my adventures with SleepyHead and Flex, I found that when I do a clean import, all sessions list the Flex setting used by the most recent session, not the setting that corresponds to the session in question.

At the moment, for days with multiple sessions, I'm just picking the highest value flex setting.. An admittedly lazy approach. (I do the same with the humidifier settings)

It shouldn't make any difference on fresh import or not... ? (unless the day cache object is being corrupted somehow, this gets rebuilt when you move an import slider in preferences)

I could list each individual sessions Flex/humidifier settings separately in the Details tab, but that might look a little messy data wise.. it's already quite full, especially with the Pie chart switched on.

I could also just show the most commonly used (for the longest session) setting, or a (rounded?) weighted average?

I can't remember which method I used in the Overview Chart.. think it's just max again..
EDIT: Overview shows the weighted average of all sessions for a day.

For the past three days, my Flex setting has been C-Flex+ 1, with the possibility of a couple short sessions at 2. Prior to that, all sessions were C-Flex+ 3. Currently, all days display as C-Flex+ 1, as I ditched the data and re-imported it all. Prior to the re-import, each day had the current setting for the import.

tetragon wrote:

jedimark wrote:

tetragon wrote:In my adventures with SleepyHead and Flex, I found that when I do a clean import, all sessions list the Flex setting used by the most recent session, not the setting that corresponds to the session in question.

At the moment, for days with multiple sessions, I'm just picking the highest value flex setting.. An admittedly lazy approach. (I do the same with the humidifier settings)

It shouldn't make any difference on fresh import or not... ? (unless the day cache object is being corrupted somehow, this gets rebuilt when you move an import slider in preferences)

I could list each individual sessions Flex/humidifier settings separately in the Details tab, but that might look a little messy data wise.. it's already quite full, especially with the Pie chart switched on.

I could also just show the most commonly used (for the longest session) setting, or a (rounded?) weighted average?

I can't remember which method I used in the Overview Chart.. think it's just max again..
EDIT: Overview shows the weighted average of all sessions for a day.

For the past three days, my Flex setting has been C-Flex+ 1, with the possibility of a couple short sessions at 2. Prior to that, all sessions were C-Flex+ 3. Currently, all days display as C-Flex+ 1, as I ditched the data and re-imported it all. Prior to the re-import, each day had the current setting for the import.

I must be going blind.. My flex line is suddenly all 1 too in overview.. thanks for the report.. :}

Ugh.. the PRS1 FlexSet code was accidentally mixed up with the Humidifer Status code.. (which is always usually set to 1 (true) )

All fixed now in GIT.. Though you will have to purge and reimport..

jedimark wrote:Ugh.. the PRS1 FlexSet code was accidentally mixed up with the Humidifer Status code.. (which is always usually set to 1 (true) )

All fixed now in GIT.. Though you will have to purge and reimport..

Which explains why one of my days had Flex set to 0.98 . It's working for me now. I just had to redo the change I made to my local copy of calcs.cpp. I was curious about how your calculated graphs would look with greater resolution than purely truncated integers (I think C++ truncates when casting from double to int), so I multiplied br by 1000.0 on line 294. Now my respiratory rate and minute vent graphs are off by a factor of 1000, but they're more interesting to look at.