Every's fall behind

[apostolakisl]

I have noticed that repeat every x minutes/seconds/whatever fall behind. It would appear that it is counting off the time from one event to the next, which because of processor delays or whatever gradually falls behind.

 

It would be better if it timestamped the start and then made it a recurring event based on the clock.

 

In other words, if you ask for something to happen every minute, and the program first runs at 1:43:23, then it should subsequently trigger based on the clock hitting 1:44:23, 1:45:23, etc., not based on one minute passing from the last completed execution.

[Michel Kohanim]

Hi apostolakisl,

 

Thanks so very much for the feedback. We'll investigate.

 

With kind regards,

Michel

[apostolakisl]

For example

 

This program is now 30 seconds behind. In other words since 3 am local time (18 hours), it is 30 seconds behind.

 

If
  - No Conditions - (To add one, press 'Schedule' or 'Condition')

Then
       Wait  1 minute 
       Repeat Every  1 minute 
          $sMinute += 1

Else
  - No Actions - (To add one, press 'Action')

[Michel Kohanim]

Thanks apostolakisl,

 

Verified ... we'll need to figure out how to make changes without causing massive regression testing.

 

With kind regards,

Michel

[arw01]

Michel, curious why the regression testing? Why not put a note in the release notes that not guaranteed to work before version 3.3.10 or whatever the last official was on the 99i. Do a lot of ISY's update automatically as soon as a release is done? I thought we had to manually cause the update.

 

I expect updates to always break something, occasionally i cannot live without it and regress and stay, and certainly a working system does not need an upgrade unless there is a known issue, and that means they are touching it and can make a decision or modify the existing install to remedy if necessary.

[Michel Kohanim]

Hi Alan,

 

Any changes to any part of code require regression testing (i.e. to see whether or not we broke things that used to work before). This section of the code is pretty complex and thus any changes will require massive regression testing.

 

With kind regards,

Michel

[switches]

If it's any help I see the same thing - I have a program that starts at midnight and runs every hour for a few days. Most of the time it gets 10 seconds behind each hour, a couple days it got behind 15 seconds each hour and very few times it lags something other than 10 or 15 seconds.