Re: Text collation

[Rob Browning]

Neil Jerram <address@hidden> writes:

>> Also, you definitely can't judge by the presence or lack of
>> documentation.  Guile's documentation has often taken a while to
>> catch up with the code.
>
> I'm not sure I agree with that.  I accept that documentation can
> sometimes take a while to catch up (although it shouldn't, and
> recently I think we've all been pretty good about writing doc at the
> same time as developing something), but in principle I think
> "presence in the manual" is a better way of indicating to users
> whether an API is officially supported than a coding convention.

Well, if that's how we want to proceed, then I think we need to be
very clear about it.  My understanding has been that we promse that as
a default, any C feature that's prefixed with scm_ instead of scm_i_
is fair game for the C programmer and that they can expect us to
maintain it going forward (plus or minus deprecation, *major*
(i.e. 2.0) releases, etc.).

In the past (at least), this was important because there were a lot of
things not mentioned in the documentation.  Further, as with the SRFI
code, the automatic documentation mechanism hasn't included the C
documentation, even when it exists, and even if our documentation
becomes (has become) more or less comprehensive, I still feel like the
scm_ vs scm_i_ convention is a good one.

-- 
Rob Browning
rlb @defaultvalue.org and @debian.org; previously @cs.utexas.edu
GPG starting 2002-11-03 = 14DD 432F AE39 534D B592  F9A0 25C8 D377 8C7E 73A4