[PD-dev] the future of [declare] and canvas_savedeclarationsto()
Hans-Christoph Steiner
hans at eds.org
Thu May 22 22:06:14 CEST 2008
To quote the whole text:
Avoid using this function. It is broken by design since (unless
using the non-
standard resolved_path == NULL feature) it is impossible to
determine a suitable
size for the output buffer, resolved_path. According to POSIX a
buffer of size
PATH_MAX suffices, but PATH_MAX need not be a defined constant, and
may have to
be obtained using pathconf(). And asking pathconf() does not
really help, since
on the one hand POSIX warns that the result of pathconf() may
be huge and
unsuitable for mallocing memory. And on the other hand pathconf()
may return -1
to signify that PATH_MAX is not bounded.
I think whoever wrote that man page is really splitting hairs there,
the FreeBSD/Mac OS X man page has no complaints about it. They are
complaining that PATH_MAX _might_ not be defined. But on GNU/Linux,
MinGW, and FreeBSD/Mac OS X, FILENAME_MAX _is_ defined. So just use
FILENAME_MAX to allocate the buffer that realpath() uses (which was
already happening anyway in s_main.c, but with MAXPDSTRING instead).
It's a widely implemented POSIX function, how much more standard do
you want? ;)
I think that Pd should give reliable information in the publically
defined API (ok, s_stuff.h, but still). For example, it makes life
easier in a lot of ways if you can count on sys_libdir being an
absolute path, rather than making every other bit of code that might
every use a path that is based on sys_libdir check to make sure that
it is indeed absolute. The vast majority of the time, sys_libdir is
an absolute path, and AFAIK, on Windows it is always an absolute path.
.hc
On May 22, 2008, at 9:44 PM, Miller Puckette wrote:
> I finally realized realpath() is a pre-existing function. (I imagined
> it was sitting in another source file that didn't make it to the
> patch)
>
> but.. "man realpath" says "Avoid using this function. It is broken
> by design..."
> so I'm sceptical of the idea. Anyway, why can't [import] just make
> the
> call if it needs it?
>
> cheers
> M
>
>
> On Thu, May 22, 2008 at 09:33:14PM +0200, Hans-Christoph Steiner
> wrote:
>>
>> The only real change to s_main.c in that path is the addition of
>> realpath(). It seems that the function there is just copying the
>> strings back and forth between sbuf and sbuf2, so I just added one
>> more iteration of that back and forth for realpath().
>>
>> .hc
>>
>> On May 22, 2008, at 9:25 PM, Miller Puckette wrote:
>>
>>> I might have missed it, but I didn't see that realpath() itself
>>> made it into
>>> the patch... ?
>>>
>>> M
>>>
>>> On Thu, May 22, 2008 at 09:22:12PM +0200, Hans-Christoph Steiner
>>> wrote:
>>>>
>>>> The realpath() is definitely related. If you start Pd using a
>>>> relative path, like I do when doing dev work (e.g. make && ../bin/
>>>> pd), then sys_libdir will be a relative path. That means it is
>>>> impossible to make absolute paths using sys_libdir, which is what I
>>>> need to do to make [import] work, or loading libdirs with
>>>> [declare -
>>>> lib] and the libdir loader.
>>>>
>>>> I can't see any problems that realpath() might cause.
>>>>
>>>> .hc
>>>>
>>>> On May 22, 2008, at 8:45 PM, Miller Puckette wrote:
>>>>
>>>>> OK, I took most of the patch (not the "realpath()" call which
>>>>> seems
>>>>> unrelated) and uploaded to SVN, unless I'm mistaken.
>>>>>
>>>>> cheers
>>>>> Miller
>>>>>
>>>>> On Wed, May 21, 2008 at 08:22:23PM +0200, Hans-Christoph Steiner
>>>>> wrote:
>>>>>>
>>>>>> I was looking into this a bit more, and from what I can tell, the
>>>>>> canvas-local path (ce_path) is always relative to the path of the
>>>>>> parent patch. Here's the commit for pd-extended that adds
>>>>>> support
>>>>>> for absolute paths:
>>>>>>
>>>>>> http://pure-data.svn.sourceforge.net/viewvc/pure-data?
>>>>>> view=rev&revision=9862
>>>>>>
>>>>>> Also, I submitted a patch:
>>>>>>
>>>>>> http://sourceforge.net/tracker/index.php?
>>>>>> func=detail&aid=1917574&group_id=55736&atid=478072
>>>>>>
>>>>>> .hc
>>>>>>
>>>>>>
>>>>>> On May 19, 2008, at 9:42 PM, Hans-Christoph Steiner wrote:
>>>>>>
>>>>>>>
>>>>>>> I think I fixed the full paths bug, here's the commit:
>>>>>>>
>>>>>>> http://pure-data.svn.sourceforge.net/viewvc/pure-data?
>>>>>>> view=rev&revision=9856
>>>>>>>
>>>>>>> .hc
>>>>>>>
>>>>>>> On May 19, 2008, at 8:03 PM, Hans-Christoph Steiner wrote:
>>>>>>>
>>>>>>>>
>>>>>>>> I think this issue is pretty clear, and the languages that I
>>>>>>>> know
>>>>>>>> would fall along the lines of "each patch/abstraction has
>>>>>>>> its own
>>>>>>>> namespace" or in other words "#include only affects the one .c
>>>>>>>> file", "import only affects the one .py file", etc. So I
>>>>>>>> agree
>>>>>>>> with Frank. Global settings are global, and canvas-local
>>>>>>>> settings
>>>>>>>> are local to the original file.
>>>>>>>>
>>>>>>>> The "semi-functional" part is the full paths that Marius
>>>>>>>> mentioned.
>>>>>>>>
>>>>>>>> Then the other question is how to use something like "#X
>>>>>>>> declare"/
>>>>>>>> canvas_savedeclarationsto() in externals. I'd like to
>>>>>>>> create an
>>>>>>>> [import] modelled after Python's import does, so I'd like to
>>>>>>>> use
>>>>>>>> "#X declare"/canvas_savedeclarationsto() with it.
>>>>>>>>
>>>>>>>> .hc
>>>>>>>>
>>>>>>>> On May 19, 2008, at 6:33 PM, Miller Puckette wrote:
>>>>>>>>
>>>>>>>>> Hi all,
>>>>>>>>>
>>>>>>>>> I use 'declare' all the time.. don't think it's
>>>>>>>>> semifunctional at
>>>>>>>>> all.
>>>>>>>>> I think the questions about how declares should act inside
>>>>>>>>> abstractions
>>>>>>>>> are hard to resolve; in my own usage (and in the way I suggest
>>>>>>>>> others might
>>>>>>>>> want to use declare) it's always in the main patch, as a
>>>>>>>>> way to
>>>>>>>>> show the
>>>>>>>>> patch what libraries, etc, it needs.
>>>>>>>>>
>>>>>>>>> cheers
>>>>>>>>> Miller
>>>>>>>>>
>>>>>>>>> On Mon, May 19, 2008 at 06:28:31PM +0200, Hans-Christoph
>>>>>>>>> Steiner
>>>>>>>>> wrote:
>>>>>>>>>>
>>>>>>>>>> Hey,
>>>>>>>>>>
>>>>>>>>>> So I am diving into the whole canvas-local namespace and
>>>>>>>>>> [declare]
>>>>>>>>>> issue these days. I like the new "#X declare"/
>>>>>>>>>> canvas_savedeclarationsto() functionality, I think it
>>>>>>>>>> could be
>>>>>>>>>> useful
>>>>>>>>>> for a lot of things. I was thinking of making an API to use
>>>>>>>>>> it in
>>>>>>>>>> externals, something like sys_register_loader(). I have two
>>>>>>>>>> questions, first, how entrenched is the current behavior of
>>>>>>>>>> [declare]? It currently is only semi-functional, and I
>>>>>>>>>> think few
>>>>>>>>>> people use it.
>>>>>>>>>>
>>>>>>>>>> The second is how to structure this for general use. I have
>>>>>>>>>> thought
>>>>>>>>>> of two ways:
>>>>>>>>>>
>>>>>>>>>> - make "declare" the key word and allow other
>>>>>>>>>> objectclasses to
>>>>>>>>>> have
>>>>>>>>>> their own custom "#X declare" data.
>>>>>>>>>>
>>>>>>>>>> - allow objectclasses to register their own declaration key
>>>>>>>>>> words,
>>>>>>>>>> like [import] could have "#X import".
>>>>>>>>>>
>>>>>>>>>> The first would mean changing the behavior of [declare], the
>>>>>>>>>> second
>>>>>>>>>> could lead to a big mess...
>>>>>>>>>>
>>>>>>>>>> .hc
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> -------------------------------------------------------------
>>>>>>>>>> --
>>>>>>>>>> --
>>>>>>>>>> --
>>>>>>>>>> -----
>>>>>>>>>> ----
>>>>>>>>>>
>>>>>>>>>> Man has survived hitherto because he was too ignorant to know
>>>>>>>>>> how to
>>>>>>>>>> realize his wishes. Now that he can realize them, he must
>>>>>>>>>> either
>>>>>>>>>> change them, or perish. -William Carlos Williams
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> _______________________________________________
>>>>>>>>>> PD-dev mailing list
>>>>>>>>>> PD-dev at iem.at
>>>>>>>>>> http://lists.puredata.info/listinfo/pd-dev
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> ---------------------------------------------------------------
>>>>>>>> --
>>>>>>>> --
>>>>>>>> --
>>>>>>>> -------
>>>>>>>>
>>>>>>>> News is what people want to keep hidden and everything else is
>>>>>>>> publicity. - Bill Moyers
>>>>>>>>
>>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> ----------------------------------------------------------------
>>>>>>> --
>>>>>>> --
>>>>>>> --
>>>>>>> ------
>>>>>>>
>>>>>>> You can't steal a gift. Bird gave the world his music, and if
>>>>>>> you
>>>>>>> can hear it, you can have it. - Dizzy Gillespie
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> -----------------------------------------------------------------
>>>>>> --
>>>>>> --
>>>>>> ---
>>>>>> ----
>>>>>>
>>>>>> News is what people want to keep hidden and everything else is
>>>>>> publicity. - Bill Moyers
>>>>>>
>>>>>>
>>>>>>
>>>>>> _______________________________________________
>>>>>> PD-dev mailing list
>>>>>> PD-dev at iem.at
>>>>>> http://lists.puredata.info/listinfo/pd-dev
>>>>
>>>>
>>>>
>>>> -------------------------------------------------------------------
>>>> --
>>>> ---
>>>> ----
>>>>
>>>> News is what people want to keep hidden and everything else is
>>>> publicity. - Bill Moyers
>>>>
>>
>>
>>
>>
>>
>>
>>
>> ---------------------------------------------------------------------
>> ---
>> ----
>>
>> "It is convenient to imagine a power beyond us because that means we
>> don't have to examine our own lives.", from "The Idols of
>> Environmentalism", by Curtis White
>>
>>
>>
------------------------------------------------------------------------
----
If you are not part of the solution, you are part of the problem.
More information about the Pd-dev
mailing list