CJones wrote:Afternoon.
I have a few questions on eValid, kind of a mix of questions, here goes.
The Lock/Unlock command is pretty neat ... is that a new concept?
What happens if the two commands get out of sync?
Thanks
Sure happy to answer some of your questions CJones.
No, the notion of a mutually exclusive operator -- called a "mutex" -- is not at all new with eValid.
Prevention of interactive lockup between two independent processes is a common engineering problem.
There are even mechanical equivalents of this kind of lockup prevention.
Think of the "safety catch" on various kinds of machines?
eValid implements the lock/unlock by keeping a token that is set ON by a Lock command, and other copies of eValid that ALSO use a Lock or Unlock command "know" about this and respect the value.
They check regularly (at the parameter value of the "retry interval" -- see Settings) and if the value is "ON" then they wait until the process that owns the lock turns if OFF.
And, yes, if the token is by mistake set to ON, and the process that sets it that way never turns it off...you can lock up everything.
Hence, caution is advised.
However, new instances of eValid always turn the lock token OFF.
Here is the manual page for Lock/Unlock:
http://e-valid.com/Products/Documentati ... nlock.htmlSo the only thing you have to worry about is a runaway process that sets the lock ON and never resets.
But each eValid instance in playback mode always has the upper time limit to prevent even THAT process from keeping control forever.
-- eValid Support
