Respect the 'run tasks in background' setting for triggers as well

[pjrobertson]

From #1477 I realised QS doesn't currently run triggers in the background, ever. That's not so good :/

I've fixed this now. It would have been nicer to tie it into the existing code that checks for this setting, but that's related to the interface, and this is related to triggers (which don't use the interface, so they'll have to be independent

[skurfer]

Weird problem I ran into. My trigger for Current Web Page (Safari) ⇥ Paste as Plain Text started pasting stale values. (The URL wasn't updating as I changed tabs.) Using the proxy manually shows the correct URL.

What's especially weird is that I can't reproduce it on this branch alone, but if I merge this with several other pending branches, it starts showing the incorrect behavior. Haven't had time to investigate further.

[pjrobertson]

Strange, I guess running on a background thread could theoretically cause problems resolving the proxy, but I don't see how… unless ScriptingBridge has to be run on the main thread?

I'll let you figure it out, since it might be a problem somewhere else

On 25 Mai 2013, at 02:12, Rob McBroom notifications@github.com wrote:

Weird problem I ran into. My trigger for Current Web Page (Safari) ⇥ Paste as Plain Text started pasting stale values. (The URL wasn't updating as I changed tabs.) Using the proxy manually shows the correct URL.

What's especially weird is that I can't reproduce it on this branch alone, but if I merge this with several other pending branches, it starts showing the incorrect behavior. Haven't had time to investigate further.

—
Reply to this email directly or view it on GitHub.

[skurfer]

I'll let you figure it out, since it might be a problem somewhere else

Yeah, like I said, this alone won't do it, but git bisect leads here. I might have to merge this with other branches one at a time to figure it out.

[skurfer]

It happens when you mix this with tempMeta (from #1218). I guess the “perform selector after delay” stuff doesn’t work when called in the background, because expireCache: never runs.

[pjrobertson]

OK, nice detective work.

That would make sense, since (I'm assuming) this line is called on the main thread, and as such it works, but by doing stuff in the background it messes it up.

From the docs about that method:

This method sets up a timer to perform the aSelector message on the current thread’s run loop

Have you tried running it on the main thread using performSelectorOnMainThread:withObject:waitUntilDone:?

Personally, I still think the tempMeta branch has a few problems... I've gone ahead and tried to get the ball rolling there, but I think this problem you mention should be addressed there rather than here, as the only way to fix the problem here is to not run triggers on a background thread - which defeats the purpose of the pull :P

[pjrobertson]

OK, nice detective work.

That would make sense, since (I'm assuming) this line is called on the main thread, and as such it works, but by doing stuff in the background it messes it up.

From the docs about that method:

This method sets up a timer to perform the aSelector message on the current thread’s run loop

Have you tried running it on the main thread using performSelectorOnMainThread:withObject:waitUntilDone:?

Personally, I still think the tempMeta branch has a few problems... I've gone ahead and tried to get the ball rolling there, but I think this problem you mention should be addressed there rather than here, as the only way to fix the problem here is to not run triggers on a background thread - which defeats the purpose of the pull :P

[pjrobertson]

OK, nice detective work.

That would make sense, since (I'm assuming) this line is called on the main thread, and as such it works, but by doing stuff in the background it messes it up.

From the docs about that method:

This method sets up a timer to perform the aSelector message on the current thread’s run loop

Have you tried running it on the main thread using performSelectorOnMainThread:withObject:waitUntilDone:?

Personally, I still think the tempMeta branch has a few problems... I've gone ahead and tried to get the ball rolling there, but I think this problem you mention should be addressed there rather than here, as the only way to fix the problem here is to not run triggers on a background thread - which defeats the purpose of the pull :P

[skurfer]

Yeah, I agree that what you've done here is correct and doesn't need to be changed to work-around something that isn't merged (or even agreed upon) anyway. Merging, even though it breaks my precious cache. 😭