[neomutt-users] Vim-Lover Wanted
z+mutt+neomutt at m0g.net
Sun Feb 26 23:50:38 CET 2017
On Sun, Feb 26, 2017 at 09:05:39PM +0100, Pietro Cerutti wrote:
> On 26 Feb 2017, at 13:18, Guyzmo <z+mutt+neomutt at m0g.net> wrote:
> > For example in [line 344] there's a redefinition of the options_t types,
> > as an enum and then as a struct? So all the new types we've added are
> > actually non-documented.
> Please double check: the default is just fine in most cases.
I'll do, I was just warning as I noticed a copy of the types there.
> > For efficiency, I could try rewriting makedoc.c in python, using the
> > pyparsing library (and some C source code introspection) to extract
> > needed information out of the source. If someone doesn't think it's a
> > good idea (@ghar?) I'll abstain, otherwise I can put that in my TODO
> > list.
> Yup, so in six months' time to build mutt we'll need lua & python, great ;)
yup there's nothing wrong with that ☺
first: python is available per default on all systems installed in the
last 10 years.
second: we already depend on M4 for the build stuff. Is that really
better? (that's rhetorical)
> No, seriously, what is so wrong with poor makedoc.c that requires
> total annihilation and recreation in a different programming language?
There are a few things that are wrong with it, IMHO:
* It's not DRY. you have repetition of data from the original source code.
* It's ignoring functions.h and globals.h that might be as well automated documentation the same way.
* It's using a half baked nroff language unseen elsewhere.
* It's a stack of stack machines that's hard to maintain in the long term
> Let's not diverge too much yet again and let's see if anybody volunteers for [topic].
yup, I'd say we should make it a ticket (rather low priority) and
replace at some point in the future.
More information about the neomutt-users