Triggers are only saved 10s after being created

[pjrobertson]

Check it out:

1. Create a trigger
2. Quit QS (less than 10s)
3. Reopen QS and see the trigger isn't saved
4. Create a trigger
5. Quit QS (after waiting 10s)
6. Reopen QS and see the trigger is saved.

I found this when looking at Trigger stuff. Check out QSTriggerCenter.m:155

Anybody have any ideas WHY this might be, and if it was included post Alcor, or if Alcor implemented it?
To me, this is kinda stupid, no?!

I haven't testing it when delay is 0.0, because I'm working on some serious trigger stuff at the moment, so don't want to get away from that. Thought I'd document it here though

[HenningJ]

Etienne did that in commit 5185d45
(5185d45#diff-2).
The commit comment was:

"Fixed issues with some triggers being reactivated at relaunch.
Editing the hotkey no longer enables the trigger."

So there were some problem with triggers before. But I don't know how
this makes sense.

You can easily find that out with git blame. Just call up the source
code of a file on github and click blame on the top right of the file.
You get something like this:
https://github.com/quicksilver/Quicksilver/blame/master/Quicksilver/Code-QuickStepCore/QSTriggerCenter.m

On Sun, Apr 24, 2011 at 9:49 AM, pjrobertson
reply@reply.github.com
wrote:

Check it out:

1. Create a trigger
2. Quit QS (less than 10s)
3. Reopen QS and see the trigger isn't saved
4. Create a trigger
5. Quit QS (after waiting 10s)
6. Reopen QS and see the trigger is saved.

I found this when looking at Trigger stuff. Check out QSTriggerCenter.m:155

Anybody have any ideas WHY this might be, and if it was included post Alcor, or if Alcor implemented it?
To me, this is kinda stupid, no?!

I haven't testing it when delay is 0.0, because I'm working on some serious trigger stuff at the moment, so don't want to get away from that. Thought I'd document it here though

Reply to this email directly or view it on GitHub:
#240

[pjrobertson]

Interesting. I still don't understand though, unless it assumes you'd
relaunch after editing a hotkey, which doesn't make sense...

On 24 April 2011 18:14, HenningJ <
reply@reply.github.com>wrote:

Etienne did that in commit 5185d45
(
5185d45#diff-2
).
The commit comment was:

"Fixed issues with some triggers being reactivated at relaunch.
Editing the hotkey no longer enables the trigger."

So there were some problem with triggers before. But I don't know how
this makes sense.

You can easily find that out with git blame. Just call up the source
code of a file on github and click blame on the top right of the file.
You get something like this:

https://github.com/quicksilver/Quicksilver/blame/master/Quicksilver/Code-QuickStepCore/QSTriggerCenter.m

On Sun, Apr 24, 2011 at 9:49 AM, pjrobertson
reply@reply.github.com
wrote:

Check it out:

1. Create a trigger
2. Quit QS (less than 10s)
3. Reopen QS and see the trigger isn't saved
4. Create a trigger
5. Quit QS (after waiting 10s)
6. Reopen QS and see the trigger is saved.

I found this when looking at Trigger stuff. Check out
QSTriggerCenter.m:155

Anybody have any ideas WHY this might be, and if it was included post
Alcor, or if Alcor implemented it?
To me, this is kinda stupid, no?!

I haven't testing it when delay is 0.0, because I'm working on some
serious trigger stuff at the moment, so don't want to get away from that.
Thought I'd document it here though

Reply to this email directly or view it on GitHub:
#240

Reply to this email directly or view it on GitHub:
#240 (comment)

[pjrobertson]

This has been 'fixed' by reducing the time for triggers to be saved to 2s. writeTriggers and writeTriggersNow in QSTriggerCenter.m are what is important here.

[HenningJ]

I still don't understand why there has to be a delay at all.

[pjrobertson]

Because atm. triggers are saved whenever you click 'another' trigger in the list of your triggers, and when it saves triggers, it saves all of them to file (the manager can't just save one) so it's quite recourse intensive. If you're going about clicking loads of things rapidly QS is going to try to save, save, save, save...

[HenningJ]

Ah, ok. So the correct fix for this would be to make the trigger manager smarter, so he can save one trigger at a time and only the one that changed?

[pjrobertson]

Ah, ok. So the correct fix for this would be to make the trigger manager
smarter, so he can save one trigger at a time and only the one that changed?

Indeed :)
I need to improve my IB (I'm still gonna call it that, everyone knows what
I mean!) skills so I can get things saved when any button is pressed.
Plus the manager would need to do what you said.

That's why I stuck 'fixed' in inverted commas - I didn't actually 'fix' the
bug properly.

On 29 October 2011 09:07, Henning Jungkurth <
reply@reply.github.com>wrote:

Ah, ok. So the correct fix for this would be to make the trigger manager
smarter, so he can save one trigger at a time and only the one that changed?

Reply to this email directly or view it on GitHub:
#240 (comment)