I have decided to join Python Applications Packaging Team to help
maintaining Cython for Debian. While looking at 0.13 I have detected
that 3 dependent projects fail to build from source with cython 0.13 (or
current HG tip (3755:646494) although were fine with 0.12.
Since there is no changelog and, unfortunately, I am not yet very verse
in Cython, it is difficult for me to judge either following failure is a
feature or a bug (well, Doc/special_methods.html still lists this API as
valid):
Error converting Pyrex file to C:
------------------------------------------------------------
...
if ptrptr != NULL:
buf = <void*>ecore_evas_buffer_pixels_get(self.obj);
ptrptr[0] = buf
return self._get_buf_size()
def __getsegcount__(self, int *lenp):
^
------------------------------------------------------------
Error converting Pyrex file to C:
------------------------------------------------------------
...
f = flags
else:
f = NULL
evas_object_image_save(self.obj, filename, k, f)
def __getsegcount__(self, int *p_len):
could you please advise?
Thanks in advance
--
.-.
=------------------------------ /v\ ----------------------------=
Keep in touch // \\ (yoh@|www.)onerussian.com
Yaroslav Halchenko /( )\ ICQ#: 60653192
Linux User ^^-^^ [175555]
Now Cython correctly uses Py_ssize_t for these "len" arguments. Code
generated by Cython 0.12 was broken on 64 bits arch. Unfortunately,
the docs are not updated to reflect this change. At list
http://docs.cython.org/src/userguide/special_methods.html#buffer-interface-no-python-equivalents-see-note-1
, you have to replace int -> Py_ssize_t
PS: I have a patch updating the docs, but for some reason I'm not able to push.
--
Lisandro Dalcin
---------------
CIMEC (INTEC/CONICET-UNL)
Predio CONICET-Santa Fe
Colectora RN 168 Km 472, Paraje El Pozo
Tel: +54-342-4511594 (ext 1011)
Tel/Fax: +54-342-4511169
Great, thanks!
> While looking at 0.13 I have detected
> that 3 dependent projects fail to build from source with cython 0.13 (or
> current HG tip (3755:646494) although were fine with 0.12.
>
> Since there is no changelog
We try and list every backwards incompatible change in the release
notes http://wiki.cython.org/ReleaseNotes-0.13 , but this one seems to
have slipped through the cracks. (In general, we try to stay backwards
compatible, but this is a pre 1.0 project). A slightly lower level
view (though still woefully incomplete) can be found at
http://trac.cython.org/cython_trac/query?group=status&milestone=0.13 .
This should be the changelog you're wanting. Even lower than that,
there's the actual revision control, but that's probably too detailed.
On that note, hopefully this will provide a good testbed for release
candidates.
> and, unfortunately, I am not yet very verse
> in Cython, it is difficult for me to judge either following failure is a
> feature or a bug (well, Doc/special_methods.html still lists this API as
> valid):
Looks like Lisandro tracked this one down.
- Robert
William's been doing a bunch of work on the machines our hg repos are
hosted on -- I suspect that's why you couldn't push. I think the
machines are temporarily back up, but will be down again at some point
this week ...
-cc
Actually, for some reason Lisandro just weren't in the committers
list, which I've fixed. (Put you in there too :). It's all on a VM,
the same one that's hosting the main sagemath website, so less
affected by all the other work going on with those machines.
- Robert
Until we have something more formal, pulling a dump of those wiki
pages would be a good idea. (Actually, our announcement is pretty much
the same content, formatted as plain text--the main reason to put it
on a wiki is for collaborative editing.) Good point about making them
easier to find, I'll try to think of the best place to put that.
- Robert