kowey at darcs.net
Sat Sep 19 05:59:36 UTC 2009
>> I think (if my memory serve me well) that it is one of David's
>> convention; that is: camel case for exported values and underscores
>> for internal ones. Which would make sense since you are exporting a
>> previously internal thing, right?
Your memory serves you well! But that's an old convention and even
David doesn't really follow it anymore. The problem is that it's
hard to be consistent about this all the more because there was still
lots of old code lying around that didn't respect this.
See also: http://wiki.darcs.net/CodingStyle
I think it's way simpler to just pick a simple rule and stick with it.
I'm sure there'll be an occasional useful underscore here or there,
but we should really just camelCase and be happy.
> I'm a mere user of darcs, waiting for the day when I can unreservedly
> recommend it for all SCM users, but...
I'm fervently hoping we get there.
> CamelCase is one of my pet loathings. It dramatically impacts
> readability in long names, consequently discouraging their use. That
> much, I think, is unexceptionable.
I too dislike it, but it's a de-facto standard in the Haskell community,
so we'd better just stick with it.
This is a part of a general trend of aligning Darcs with the wider
Haskell community: following common style, making more use of third
party libraries, taking up cool Haskell tools like hpc and hlint,
switching to Cabal, putting ourselves on Hackage.
PS. Before I drop my bikeshed paintbrush, I'll just throw in that
camelCasing probably means that acronyms should just squished
down for readability: httpPipelining and not _HTTPPipelining
or any other gymnastics.
Eric Kow <http://www.nltg.brighton.ac.uk/home/Eric.Kow>
PGP Key ID: 08AC04F9
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Size: 194 bytes
Desc: not available
More information about the darcs-users