Hi Allan,
Here are some tests and what I get in the log(I removed "date + TextMate[pid]" on every lines): OS X 10.4.8, TextMate1269 Tried in my account and in a fresh one. Same results.
If I press ^⎋: Nothing
If I press ⌘⎋: performKeyEquivalent: system gave us: ⎋ performKeyEquivalent: user actually pressed: ⌘⎋ performKeyEquivalent: match against: ⌃⎋ performKeyEquivalent: did match: NO
If I press alt+esc: Beep. Nothing.
If I press ⌃⌘⎋ performKeyEquivalent: system gave us: ⎋ performKeyEquivalent: user actually pressed: ⌃⌘⎋ performKeyEquivalent: match against: ⌃⎋ performKeyEquivalent: did match: NO
If I press ⌃⌥⎋ performKeyEquivalent: system gave us: ⎋ performKeyEquivalent: user actually pressed: ⌃⌥⎋ performKeyEquivalent: match against: ⌃⎋ performKeyEquivalent: did match: NO Any ^ + whatever or ^ + ⇧ + whatever: Nothing
Any ⌘ + whatever or ⌘ + ⇧ + whatever: Looks fine, e.g.: performKeyEquivalent: system gave us: ⇧⌘L performKeyEquivalent: user actually pressed: ⇧⌘L performKeyEquivalent: match against: ⌃⎋ performKeyEquivalent: did match: NO
Any ^ + ⌘ + whatever: Looks fine, e.g.: performKeyEquivalent: system gave us: ⌃⌘M performKeyEquivalent: user actually pressed: ⌃⌘M performKeyEquivalent: match against: ⌃⎋ performKeyEquivalent: did match: NO
Any ⌥ + ⌘ + whatever(not those used by the system.): Looks fine, e.g.: performKeyEquivalent: system gave us: ⌥⌘S performKeyEquivalent: user actually pressed: ⌥⌘S performKeyEquivalent: match against: ⌃⎋ performKeyEquivalent: did match: NO
Tell me if I can try something more...
-- FredB
On 5. Oct 2006, at 17:21, Fred B wrote:
Here are some tests and what I get in the log(I removed "date + TextMate[pid]" on every lines): OS X 10.4.8, TextMate1269 Tried in my account and in a fresh one. Same results.
If I press ^⎋: Nothing [...]
Thanks, so at least TM is not seeing the key in the performKeyEquivalent: method which *should* be called by the system.
Can you bind a bundle item to ⌃⎋?
If not, the problem is with 99% certainty outside TextMate.
On 10/6/06, Allan Odgaard throw-away-1@macromates.com wrote:
On 5. Oct 2006, at 17:21, Fred B wrote:
Here are some tests and what I get in the log(I removed "date + TextMate[pid]" on every lines): OS X 10.4.8, TextMate1269 Tried in my account and in a fresh one. Same results.
If I press ^⎋: Nothing [...]
Thanks, so at least TM is not seeing the key in the performKeyEquivalent: method which *should* be called by the system.
Can you bind a bundle item to ⌃⎋?
If not, the problem is with 99% certainty outside TextMate.
Thanks Allan,
I can't bind a bundle item to ⌃⎋. Maybe something eats ⌃⎋ before TM, although I can't see what. Specially in a new account without any input manager and no ⌃⎋ in the services shortcuts.
Anyway. I made some more tests and I finally got the menu to work with F3 and ⌘D (BTW, is ⌘D free?). The thing is I can't use ^ + whatever for the menu, even if I can use it for a bundle item and it doesn't appear in the log.
Key binding is a mysterious science. ;)
On 6. Oct 2006, at 02:52, Fred B wrote:
I can't bind a bundle item to ⌃⎋. Maybe something eats ⌃⎋ before TM, although I can't see what.
Another user who has the problem (on one of his machines) tried this program [1] and it did not report a key stroke on ⌃⎋.
[1] http://www.petermaurer.de/nasi.php?thema=keycodes&sprache=english
[...] Key binding is a mysterious science. ;)
Indeed -- many keys are “lost” when dealing with normal first responder stuff in Cocoa, for that reason bundle items have the sendEvent: method of the application object overloaded (so that “all” keys are available) -- this is AFAIK the most low-level you can get with Cocoa and key events, so if bundle items can’t bind to ⌃⎋, then there is really not much to do about the issue.
We should maybe make a poll for hardware/OS to see if we can find a pattern. So far I’m aware of 3 users (including you) which have the problem, one of them only on his PowerBook.
On 10/6/06, Allan Odgaard throw-away-1@macromates.com wrote:
Another user who has the problem (on one of his machines) tried this program [1] and it did not report a key stroke on ⌃⎋.
[1] http://www.petermaurer.de/nasi.php?thema=keycodes&sprache=english
Same here. Beside ⌃⎋, every keys seem to work except some used by the system (like ^F1 ^F7 ^F8 ⌘⌥⎋). So, there is a lower level than what Key Code can see...
[...] Key binding is a mysterious science. ;)
Indeed -- many keys are "lost" when dealing with normal first responder stuff in Cocoa, for that reason bundle items have the sendEvent: method of the application object overloaded (so that "all" keys are available) -- this is AFAIK the most low-level you can get with Cocoa and key events, so if bundle items can't bind to ⌃⎋, then there is really not much to do about the issue.
See previous remark. Either there is a bug somewhere in the system or something steel the keystroke at an absolutely ultra deep level (OMG, a Balrog! ;)
We should maybe make a poll for hardware/OS to see if we can find a pattern. So far I'm aware of 3 users (including you) which have the problem, one of them only on his PowerBook.
Same thing on my iMac G5 1.8Ghz (10.4.8) and my Powerbook G4 1.5Ghz (10.4.5! I don't use it much for now). The PowerBook had an older TM release version (pre 1200, I should have noted it before upgrading). Not sure it is hardware related as the PB is a (partial) clone of the iMac at the start and has the problem...
I'll install TM and test on my old G4 400 10.4.8 and my G4 Dual 1.42 (10.3.9) soon.
Thanks for your time.
-- FredB
Same on the G4 Dual 1.42 (10.3.9) :( This system is completely different. It's on Panther and is dedicated to audio only. Very few applications and nothing running in the background.
All systems are in english with belgian french azerty keyboards.(I tried with an US layout). Apple pro keyboards except for the PowerBook.
Looks like a dead end, at least I'm out of ideas.
--- FredB
-
Allan Odgaard -
Fred B