Day Light Savings Time

sc0ttt wrote:

Krelvin wrote:

chunkyfrog wrote: Or, move to Arizona where we don't play with or clocks... MST all the time.

I used to wonder why that was... and then I realized that life in Arizona in the summer pretty much consists of waiting for the sun to go down so you can go outside.

I'll take 115F and 8% humidity over 89F and 80% humidity anywhere anytime.

So I changed my machine's clock this morning, and my data moved in a way that's backward!

On Date Start End Duration
ResMed Session #1414903470
11/1/14 23:44:31 01:18:32 01h 34m 01s
ResMed Session #1414909131
11/2/14 01:18:52 01:56:53 01h 38m 01s
ResMed Session #1414918625
11/2/14 02:57:06 04:28:07 01h 31m 01s
ResMed Session #1414924117
11/2/14 04:28:38 05:01:39 00h 33m 01s
ResMed Session #1414926259
11/2/14 05:04:20 06:35:21 01h 31m 01s
ResMed Session #1414936450
11/2/14 07:54:11 07:59:12 00h 05m 01s

The actual times were:
23:44:31 CDT -- 01:18:32 CDT
01:18:52 CDT -- 01:56:53 CST (which is a one hour thirty-eight minute time span)
01:57:06 CST -- 03:28:07 CST
03:28:38 CST -- 04:01:39 CST
04:04:20 CST -- 05:35:21 CST
--- changed clock on A10
07:54:11 CST -- 07:59:12 CST

The times shown on the x-axes of my Sleepyhead graphs:
22:44:31 -- 00:18:32
00:18:52 -- 01:56:53 (not real one-hour gap between 01:56:53 and 02:57:06)
02:57:06 -- 04:28:07
04:28:38 -- 05:01:39
05:04:20 -- 06:35:21
--- changed clock on A10
07:54:11 CST -- 07:59:12 CST

Sleepyhead is doing this weird thing where the x-axis on the graph goes up linearly starting at 22:44, but when I mouse-over an x-axis on a graph and look at the time on the bottom of the window, it starts at 23:44 and is an hour off until I get to the point where the graph labels say I am at 00:59 while the time on the bottom says 01:59, and then when I move one more minute over the graph is at 01:00 and the time on the bottom has "fallen back" to say 01:00 as well.

On the unix system clock in central time zone, localtime goes from 02:59:59 to 02:00:00 in the fall, which is wrong -- it should have changed an hour earlier. But Sleepyhead seems to be doing it from 01:59:59 to 01:00:00 instead, and the time stamps are pretty weird...

I got mis-quoted again--twice.
Never been to Aridzona, but I imagine it's nice.

cathyf wrote:On the unix system clock in central time zone, localtime goes from 02:59:59 to 02:00:00 in the fall, which is wrong -- it should have changed an hour earlier. But Sleepyhead seems to be doing it from 01:59:59 to 01:00:00 instead, and the time stamps are pretty weird...

I'm not sure what you're *trying* to say there, it looks like maybe some words got missed.

proper time change is 1:59:58, 1:59:59, 1:00:00, 1:00:01, etc. (in the US)

palerider wrote:

cathyf wrote:On the unix system clock in central time zone, localtime goes from 02:59:59 to 02:00:00 in the fall, which is wrong -- it should have changed an hour earlier. But Sleepyhead seems to be doing it from 01:59:59 to 01:00:00 instead, and the time stamps are pretty weird...

I'm not sure what you're *trying* to say there, it looks like maybe some words got missed.

proper time change is 1:59:58, 1:59:59, 1:00:00, 1:00:01, etc. (in the US)

Finally something we can agree on

palerider wrote:

cathyf wrote:On the unix system clock in central time zone, localtime goes from 02:59:59 to 02:00:00 in the fall, which is wrong -- it should have changed an hour earlier. But Sleepyhead seems to be doing it from 01:59:59 to 01:00:00 instead, and the time stamps are pretty weird...

I'm not sure what you're *trying* to say there, it looks like maybe some words got missed.

proper time change is 1:59:58, 1:59:59, 1:00:00, 1:00:01, etc. (in the US)

If I run the mouse across a graph, a green vertical line appears and on the bottom of the screen the time that goes along with the location on the x-axis is displayed.

You see here in the picture, the time stamp on the bottom of the window says 01:55:43, but the x-axis underneath the green line is 00:55:43. When I start at the left side and run right, the time stamp on the bottom says 23:44:31, but the label on the x-axis is 22:44. If I run across the part of the x-axis that is labelled 00:59:58, 00:59:59, 01:00:00, 01:00:01, the numbers in the bottom time stamp count up/around 01:59:58, 01:59:59, 01:00:00, 01:00:01.

The other thing is there is a 1-hour gap between the pressure graphs at 2am and 3am, and shows me sleeping until 6:30. In fact I was asleep from 2:00-5:30.

Mine is weird too, but not the graphs or anything, it just has me going to sleep an hour later than I actually did. No gaps or weirdness other than that but something's definitely not right with the time shift calculations. PRS1 460P

Amenite wrote:Mine is weird too, but not the graphs or anything, it just has me going to sleep an hour later than I actually did. No gaps or weirdness other than that but something's definitely not right with the time shift calculations. PRS1 460P

if anybody can think of a good way to graph the one hour jump back, write it up and suggest it to mark. he may put it on the 'something nice to have someday' list. since it's only a once a year issue. (or he may assign it higher priority, or he may just start reporting all the times in GMT as he's threatned at me a few times *lol*

palerider wrote:

Amenite wrote:Mine is weird too, but not the graphs or anything, it just has me going to sleep an hour later than I actually did. No gaps or weirdness other than that but something's definitely not right with the time shift calculations. PRS1 460P

if anybody can think of a good way to graph the one hour jump back, write it up and suggest it to mark. he may put it on the 'something nice to have someday' list. since it's only a once a year issue. (or he may assign it higher priority, or he may just start reporting all the times in GMT as he's threatned at me a few times *lol*

The reason I know how the unix system times behave is that we have learned through the school of hard knocks that we can't schedule any overnight batch jobs to run between 1:59am and 3:01am. Otherwise they run twice on time-change night in the fall and don't run on time-change night in the spring. So I feel your pain! I think that if I had reset the clock last night everything would have behaved a lot better. Which WAS the original poster's instructions!

palerider wrote:

Amenite wrote:Mine is weird too, but not the graphs or anything, it just has me going to sleep an hour later than I actually did. No gaps or weirdness other than that but something's definitely not right with the time shift calculations. PRS1 460P

if anybody can think of a good way to graph the one hour jump back, write it up and suggest it to mark. he may put it on the 'something nice to have someday' list. since it's only a once a year issue. (or he may assign it higher priority, or he may just start reporting all the times in GMT as he's threatned at me a few times *lol*

It's a twice a year issue, but hey, who's counting? To graph it the x axis would simply show the extra hour in November and be missing an hour in the spring. Not sure what the data looks like from these machines, but presumably if their clocks were properly set in the beginning the machine's firmware would handle the time/date correctly; it's not a new problem - but I wouldn't be surprised if it's a mess. I don't have the manual to check my machine's date/time, or if I do it's buried somewhere and I'm not going digging just now

Daylight Saving Time!

Amenite wrote:It's a twice a year issue, but hey, who's counting?

it's a once a year issue, as mentioned, in the spring, you just have a missing hour gap, since there's no two o'clock times, going straight from 1:59:59->3:00:00.

two different display issues, each happening once a year.

RogerSC wrote:Daylight Saving Time!

pfft, nobody's saving anything, it's really Daylight MOVING time!

In the earlier years of the Respironics machines and the Encore software, there was always the 1-hour gap in the Spring and a missing hour in the Fall. And, many users on the forum were going "Hey, what happened?" In the old "MyEncore" software add-on (which I still use), it knows the exact number of hours and reports them correctly. Like last night, my Encore Pro shows 9 hours of use and my MyEncore shows 10 hours of use. In subsequent years and software releases, I think the Encore Viewer and maybe others were modified to compensate for those hour jumps. I had Encore Viewer 1.0 and it seemed to compensate for them, but it didn't store the data in a database and that was a "deal-breaker" for me, so I only used it occasionally.
If it's an issue in Sleepyhead, then maybe that needs to be addressed in some way.

It's just something you learn to live with.


Den

.

I am a severe night owl, so I typically go to be between 4 and 6am. I get up for work at 12:40pm on work days, and may sleep until 2 or 3pm on days off if I don't have to be somewhere early. Since I sleep past noon on a regular basis, I set the time on my machine so that noon is at 3/4pm depending on the time of year. Since it is not correct any day of the year, I see no reason to change it twice a year. I rarely sleep past 3pm, so either way, I get a graph and data that is not split, and that is what is important to me.

Also, since my sleep time fluctuates on a daily basis, I am not usually bothered by the time change. I ay go to bed at 4am one night, 6am the next night, then 4am, 5am, 3am, 7am, etc. Last night, I went to bed at the old 7:52 which is unusually late for me. I got up at the new noon. So, less than 5 hours of sleep. No idea why. I usually do not go to bed extremely late and get up on such short sleep without an alarm clock. I was really looking forward to sleeping in today.