cluts daily reports 8/19

cluts daily reports 8/19

[Luka Marčetić]

[-- Attachment #1: Type: text/plain, Size: 1453 bytes --]

Hey.

I've contacted Rich via IRC to ask about pthread_eintr. Specifically, 
what pthread_* functions make syscalls. I asked because I was surprised 
to find out most of my tests were for noth. Turns out there are only 
about 4 for which I haven't written tests (about 2 written). But then 
Rich implied the task itself is low priority with his question about my 
will to finish the test. Rich, maybe you meant to imply that I should 
start doing tests before I complete the incomplete ones? I'd wish you'd 
tell me these things explicitly. Anyway, assuming that is what I should 
do, I moved on to the task of adopting autoconf/gnulib tests (btw. 
they're both GPL...).
Autoconf has about 4 trivial and/or obsolete tests for the whole of 
stdio. I think there's no merit in adopting those. Gnulib, on the other 
hand from what I've seen so far seems like it could be useful. Attached 
are potential files for adoption from gnulib/tests (see the grep). Other 
than that, there's also a certain 
gnulib/tests/unistdio/test-u8-printf1.h that seems to test wchar 
support. I'm not sure if it's utf-8 specific though. If you think this 
may be useful, I can whip up a few replacement functions and include 
that file too.
So basically, I'm waiting your ok to start adopting these things into 
cluts. In the meanwhile I may finish pthread_eintr and later start 
string.c again.
Thanks.
Luka

P.S. No real to-report today, hopefully there'll be one tomorrow.

[-- Attachment #2: printf.txt --]
[-- Type: text/plain, Size: 482 bytes --]

ls | grep printf | grep .c | grep posix
test-dprintf-posix2.c
test-dprintf-posix.c
test-fprintf-posix2.c
test-fprintf-posix3.c
test-fprintf-posix.c
test-printf-posix2.c
test-printf-posix.c
test-snprintf-posix.c
test-sprintf-posix.c
test-vasnprintf-posix2.c
test-vasnprintf-posix3.c
test-vasnprintf-posix.c
test-vasprintf-posix.c
test-vdprintf-posix.c
test-vfprintf-posix.c
test-vprintf-posix.c
test-vsnprintf-posix.c
test-vsprintf-posix.c
test-xfprintf-posix.c
test-xprintf-posix.c

Re: cluts daily reports 8/19

[Luka Marčetić]

sed 's/.*[ax]printf.*//g' #of course

Re: cluts daily reports 8/19

[Rich Felker]

On Fri, Aug 19, 2011 at 04:11:02AM +0200, Luka Marčetić wrote:
> Hey.
> 
> I've contacted Rich via IRC to ask about pthread_eintr.
> Specifically, what pthread_* functions make syscalls. I asked
> because I was surprised to find out most of my tests were for noth.

I'm not sure what you mean...

> Turns out there are only about 4 for which I haven't written tests
> (about 2 written). But then Rich implied the task itself is low
> priority with his question about my will to finish the test. Rich,
> maybe you meant to imply that I should start doing tests before I
> complete the incomplete ones? I'd wish you'd tell me these things
> explicitly.

On August 10 -- that's more than a week ago -- I gave you a list of 5
areas to prioritize, in order. Further work on the EINTR tests falls
under #4 on that list, but you've spent more than a week focused on
that and done very little on #1-3. Every time I've communicated with
you since then, I've been asking you to follow the priority list I
gave you, and furthermore not to spend all your time on one thing.
It's not that the EINTR tests are useless (although you've certainly
covered the most important ones already); it's that you're spending
disproportionate amounts of time on them when there's A LOT else to be
done.

Think of your work on GSoC as painting a portrait and yourself as the
artist. If a deadline were approaching and you kept adding details on
the nose without ever addressing the fact that the ears were lopsided
and the eyes and mouth hadn't even been started, how could anyone
evaluate your work?

Rich