apr-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "William A. Rowe Jr." <wr...@rowe-clan.net>
Subject Re: planning to T&R apr-util 1.3.11 later today
Date Thu, 14 Apr 2011 16:54:11 GMT
On 4/14/2011 10:32 AM, Jeff Trawick wrote:
> Speak up if you're working on something for apr-util.
> I had originally planned to T&R this at the same time as apr 1.4.3,
> but I may as well get started.  If folks want to wait for apr 1.4.3 to
> test apr-util, that's fine too.
> I still anticipate that apr will be ready for T&R of 1.4.3 within a
> couple of days ;)

Do you mean, shipping apr_crypto?  Ugh...

I have a laundry list of issues that were never addressed, misfactored
argument lists, unnecessary function args, pointers to pointers to pointers
where nested structures would have worked - and nobody else has really
even stepped up with their assessment of the overall design.  I don't
care if it is missing desired features, but do care if the features that
are implemented now are poorly thought out.

Essentially, it's a mess.  I understand it is derivative of some related
design flaws in apr_dbd etc, but that isn't a reason to propagate them.

Attached for your perusal is my very first draft edits of the function
prototypes, obviously this does very little good without all of the updates
to the common and crypto provider specific structures.  I had not reordered
the arg lists in the .h file, nor done a second pass to correct my own work,
since I'm paying more attention to apr, trunk issues, vc 10 and win64 issues
at this moment.

So I'm presenting this only as an illustration of how much work should happen
before I'd give this API a hearty +1.  I have some of this work done, but I'm
not going to be in a position of committing it in the couple of weeks.  If
nothing else, I can at least promise to dedicate time to complete and commit
my work during the May f2f in Wicklow.

If what you meant to type was "apr-util 1.3", I see no reason not to T&R that
immediately :)

View raw message