03:05 <sb0> cr1901_modern, self.sync.nonexisting_clock_domain

03:11 <cr1901_modern> sb0: but if I do that, won't I get an error for assigning sequential statements to a clock_domain before it was defined?

03:11 <sb0> no

03:11 <sb0> the purpose of this is to allow the clock domain to be defined somewhere else

03:12 <cr1901_modern> Oh, I see. I ended up with multiple modules in a single source file, two clock domains, src and dest >>

03:13 <cr1901_modern> in each module, i did self.clock_domains.dest = ClockDomain()

03:13 <cr1901_modern> This ended up creating 4 unique clock domains

03:13 <cr1901_modern> when I meant all of them to share the dest clock domain

03:14 <cr1901_modern> Is that expected behavior? (Each clock domain was prepended with $MODULE_NAME_dest)

03:18 <cr1901_modern> Removing self.clock_domains.dest from each module except the one that I pass into verilog.convert fixed this, so as per usual, I was doing it wrong lol

09:26 <GitHub26> [artiq] sbourdeauducq pushed 1 new commit to master: https://git.io/vwA6R

09:26 <GitHub26> artiq/master bcfb01b Sebastien Bourdeauducq: gui: set application name in Gnome task switcher...

09:40 <bb-m-labs> build #401 of artiq-kc705-nist_clock is complete: Success [build successful] Build details are at http://buildbot.m-labs.hk/builders/artiq-kc705-nist_clock/builds/401

09:45 <bb-m-labs> build #159 of artiq-win64-test is complete: Success [build successful] Build details are at http://buildbot.m-labs.hk/builders/artiq-win64-test/builds/159

09:47 <bb-m-labs> build #664 of artiq is complete: Success [build successful] Build details are at http://buildbot.m-labs.hk/builders/artiq/builds/664

11:33 <rjo> sb0: browser/dashboard ui: are you saving the scheduling/submission options and arguments in the manager because the docks are ephemeral and you want to restore the state of a dock on close/reopen and on app close/reopen? isn't that a bit tricky if there are multiple docks for the same experiment?

11:33 <sb0> there are no multiple docks for the same experiment

11:33 <rjo> i was thinking of putting the arguments/logging etc into the docks and then a close/reopen would reset them which is fine imho. but it easily allows opening multiple docks from the same experiment.

11:34 <rjo> yes. not now. but i see people wanting them.

11:34 <sb0> that breaks keyboard shortcuts

11:56 <rjo> yes. or it makes them a bit more complicated.

11:56 <rjo> but the browser doesn't really need keyboard shortcuts for experiments.

11:57 <rjo> the fix could be to say that a keyboard shortcut will execute the experiment with the default arguments.

12:04 <sb0> stop trying to use the experiment dock and manager in the browser

12:04 <sb0> the behavior is different, the widgets are different (e.g. no pipeline/priority)

12:05 <rjo> i am not using them.

12:05 <rjo> but they are still experiment docks

12:06 <sb0> the common part is artiq.gui.entries and that's pretty much all afaict

12:06 <rjo> and there is state

12:06 <rjo> yes

12:06 <rjo> i am not using more

12:06 <rjo> i am just deciding on the behavior

12:06 <rjo> i am not changing the dashboard behavior at all.

12:06 <rjo> just the browser behavior

12:08 <rjo> and since i don't see a need to have shortcuts there nor the need to persist argument state across experiment dock closing, i can simplify it and don't need a manager at all.