[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-core-devel
Subject:    To Commit or Not to Commit?
From:       Ellis Whitehead <kde () ellisw ! net>
Date:       2001-10-25 16:06:35
[Download RAW message or body]

I'd appreciate some feedback on whether to commit now or later:

I'm ready to commit new KAccel and KGlobalAccel classes tomorrow.  I hate 
making large commits all at once, but the classes needed a clean re-write in 
order to begin to do what I eventually want them to (multiple key assignments 
to one action [Ctrl+C & Ctrl+Insert], implicit key assignments [Ctrl+Plus & 
Ctrl+Keypad_Plus], and allowing for a sequence of keys leading to a single 
action [as familiar from console text-editors]).

Problems:
- KControl's key configuration isn't in complete working order;
- It changes some of the rc-file key assignment names, meaning some custom 
key assignments will be lost;
- Any programs using KGlobalAccel outside of kdebase will require a 
modification: one being calling KGlobalAccel::updateConnections() after the 
keymap is populated in order to do the actual key grabbing.
- It'll certainly introduce bugs that I'm not aware of yet.

Advantage to committing later is that I'll have discovered more of the bugs 
on my own so that others don't have to contend with them.

The disadvantages to committing later include: that I won't be able to commit 
until Nov 16th, probably, since I'm going to the Usenix conference and will 
be gone the two fridays between now and the 16th.  I believe I'll still be 
able to debug in that time, though.  I'm not entirely happy with the idea of 
making such a new non-BC commit so late in the process.  It would also 
tighten the time-frame for working out any bugs.

Thanks,
Ellis

[prev in list] [next in list] [prev in thread] [next in thread]