investigate XTMODKEYS in the absence of kitty keyboard protocol #2135
Comments
dankamongmen
added
documentation
|
Your biggest problem will be handling shifted keys. XTMODKEYS is completely unspecified in its behavior and this will mean that pressing say alt+shift+x may or may not give you the upper cased or lowercased variants. |
|
so it looks like we want something like
issued at the beginning. hey @kovidgoyal , the doc says |
|
On Tue, Oct 05, 2021 at 12:24:20AM -0700, nick black wrote:
so it looks like we want something like
`echo -e '\e[>4;2m\e[>1;2m\e[>0;2m\e[>2;2m`
issued at the beginning.
hey @kovidgoyal , the doc says `In the interests of interoperation, the XTerm specific sequences CSI > 4; 1 m and CSI > 4; 0 m are treated as CSI > 1 u and CSI < 1 u.` is that supposed to be `CSI=1u` and `CSI=0u`? it seems like the line is absent in the current git.
Yes, it was removed. See #4075 for the reason.
|
|
got it, thanks for the heads up. my plan right now is to emit several XTMODKEYS codes followed by a kitty-specific code. so long as other terminals are ignoring the kitty specific code, this strategy ought be fine for kitty, right? |
|
On Tue, Oct 05, 2021 at 12:46:34AM -0700, nick black wrote:
got it, thanks for the heads up. my plan right now is to emit several XTMODKEYS codes followed by a kitty-specific code. so long as other terminals are ignoring the kitty specific code, this strategy ought be fine for kitty, right?
Yes, it is fine. The only downside I can see is that the XTMODKEYS ones
will add an entry to the kitty keyboard states stack (on older kitty
which still respects XTMODKEYS), however given this is a very minor isse
and transient anyway, I wouldnt worry.
|
Add prep_xtmodkeys() to handle modified function keys. Unify several paths into load_ncinput(), eliminating several paths that missed statistics and drain checks. Process mouse clicks outside of critical section. Handle XTMODKEY sequences with parameter 5. Interpret 0x8 as NCKEY_BACKSPACE on all paths. Send XTMODKEYS of 2;1 and 4;1 #2135.
|
Alright, I'm now sending XTMODKEYS of 2;1 and 4;1, and processing most of the new keycodes. We now correctly and precisely identify a good number of modified sequences on XTerm that we didn't before, huzzah. This is in the |
Add prep_xtmodkeys() to handle modified function keys. Unify several paths into load_ncinput(), eliminating several paths that missed statistics and drain checks. Process mouse clicks outside of critical section. Handle XTMODKEY sequences with parameter 5. Interpret 0x8 as NCKEY_BACKSPACE on all paths. Send XTMODKEYS of 2;1 and 4;1 #2135.
|
actually, @kovidgoyal , the expression i outlined above is no good; i need to do the XTMODKEYS after pushing the current keyboard state, because apparently XTMODKEYS gets pushed/popped with your
|
Add prep_xtmodkeys() to handle modified function keys. Unify several paths into load_ncinput(), eliminating several paths that missed statistics and drain checks. Process mouse clicks outside of critical section. Handle XTMODKEY sequences with parameter 5. Interpret 0x8 as NCKEY_BACKSPACE on all paths. Send XTMODKEYS of 2;1 and 4;1 #2135.
See #2131. Most terminals don't support the Kitty protocol, but many might support
XTMODKEYSfor basic improvements. XTerm certainly does. Look into using it.The text was updated successfully, but these errors were encountered: