[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: libnettle/libhogweed WIP

From: Ted Zlatanov
Subject: Re: libnettle/libhogweed WIP
Date: Tue, 27 Jun 2017 18:58:30 -0400
User-agent: Gnus/5.13 (Gnus v5.13) Emacs/26.0.50 (gnu/linux)

On Sat, 03 Jun 2017 10:23:39 +0300 Eli Zaretskii <address@hidden> wrote: 

EZ> There are a few TODOs in the documentation and in the code.  The one
EZ> in the docs should be either removed or moved to a comment so that it
EZ> doesn't appear in the manual.  Those in the code should be reviewed,
EZ> and in each case please decide whether the TODO will be handled any
EZ> time soon, or maybe should be simply deleted.

I hope we can resume work on this. I've set aside time for it. Sorry for
the repeated delays.

There is only one major TODO I can't resolve: I asked for help before,
and still can't make the autoconf code detect the GnuTLS functions
individually. So there are TODOs in the code and docs about it. I can't
merge without that--it would be a pretty bad situation for portability
and future compatibility to pin on just the GnuTLS version.

The other TODO below maybe can be answered here? Should I remove it or
leave it?

  // TODO: switch this to use a resize_string_data() function when
  // that's provided in the C core, to avoid the extra copy.

I think I've fixed the doc issues you noted. Thank you for the thorough
reading. I couldn't explain IV and other cryptographic terms in detail
in the Lisp reference manual so I pointed to the GnuTLS home page.

EZ> I think we still prefer the /* .. */ style of comments.

I changed all but the TODOs to that style.

EZ> You don't need a 'return' after calling 'error' (here and elsewhere),
EZ> as the latter doesn't return.

I got warnings for it in a few places so I added those returns
consistently. I also think it's cleaner to do it that way. Is this
important enough to change these instances?

EZ> Some of the lines in doc strings you wrote are too long, please make
EZ> sure they are no longer than 76 characters, and in any case fit on a
EZ> single 80-column line.

I tried. Several are really hard to condense further and I'm right
around 80.

EZ> No need for braces when there's only 1 line in the block.

I removed them, but it's not my favorite thing to omit the braces,
especially in the

if a
else if c

situation. Do you think I should reintroduce them in those places?

On Sat, 03 Jun 2017 11:00:16 +0200 Andreas Schwab <address@hidden> wrote: 

AS> Wouldn't it be even simpler to just pass each argument separately?

I want the format to be flexible to accomodate future expansion, and
passing arguments explicitly would defeat that purpose.


reply via email to

[Prev in Thread] Current Thread [Next in Thread]