On 11/04/10 22:16, Bram Moolenaar wrote:
> Christian Brabandt wrote:
>> Are you considering any patches from
>> for inclusion?
> Yes, but many of these patches are not mature. E.g., first one,
> "Improved regular expression engine", is still lacking the tests to
> verify that it doesn't break anything. That's a pity, because it can
> make syntax highlighting much faster.
> I want to avoid that I include something that triggers a long sequence
> of bug fixes. "Works fine for me" is not always a good indication.
> 7.3 is going to be a stable release, thus I don't want to take too much
> risc. Part of my work will be to estimate the risc, which involves
> carefully looking through the code changes.
It is true that they are in different stages of development. Here are my
top five; not in preference order.
#14 (Vince Negri's conceal/ownsyntax/cursorbind) already has a long
track record. I first heard about it when I first learned about Steve
Hall's Vim for Windows, that must have been in Vim 6.2 or 6.3 time, and
it was not new even then. Has documentation. Maybe too controversial
(not enough "mainline"-like) to be included by default? OTOH it has been
victim of bit-rotting in the past (i.e. conflict with "mainline"
patches) and of course bringing it in would eliminate that problem
forever. A compile-time option maybe (or two, or three)? You're the boss.
#13 (Access W32 clipboard from Cygwin "Unix" Vim) is interesting but
still in beta. IIUC ifdeffed by whatever FEAT_* corresponds to
has('win32unix'). Bring 'em in or let it bake some more?
#10 (Variable tabstops) sounds interesting. I haven't tested it.
Reportedly still in alpha. Probably wait some more (Vim 8.0 ?) but keep
an eye on it.
#9 (Relative line numbers) sounds interesting. I haven't tested it. Its
authors say "it works". I don't feel competent to evaluate it by
eyeballing the code.
#7 (Bill McCarthy's additional float functions). This one I've taken up
in my "Huge" Vim. Not a single problem AFAICT. Code examination shows
that it is done cleanly and simply, within #ifdef FEAT_FLOAT, and does
not interfere with other stuff outside the "call function -> return
value" codepath. IMHO this one is the most worthy of including into
mainline Vim (and perhaps the least risky). Maybe a one-time check in a
build with FEAT_EVAL on and FEAT_FLOAT off to make sure no #ifdef was
forgotten. (I already compile a Tiny build without +eval in addition to
my Huge build, from the same source, and no problems there either.)
Documentation exists and is well-written, as a separate helpfile to
avoid problems with rsync; probably merge that into eval.txt.
n = ((n >> 1) & 0x55555555) | ((n << 1) & 0xaaaaaaaa);
n = ((n >> 2) & 0x33333333) | ((n << 2) & 0xcccccccc);
n = ((n >> 4) & 0x0f0f0f0f) | ((n << 4) & 0xf0f0f0f0);
n = ((n >> 8) & 0x00ff00ff) | ((n << 8) & 0xff00ff00);
n = ((n >> 16) & 0x0000ffff) | ((n << 16) & 0xffff0000);
-- C code which reverses the bits in a word.
You received this message from the "vim_dev" maillist.
Do not top-post! Type your reply below the text you are replying to.
For more information, visit http://www.vim.org/maillist.php
To unsubscribe, reply using "remove me" as the subject.