Dnia 09-09-2007, N o godzinie 21:27 -0400, Jim Jewett napisał(a):
> If python handled small ints itself, and only farmed out the "large"
> ones,
If GMP is used, it's definitely worth to have a non-GMP representation
for small integers, because GMP itself does not do it. A GMP integer
is represent
I have a patch to add __format__ to datetime.time, .date, and .datetime.
For non-empty format_spec's, I just pass on to .strftime. For empty
format_spec's, it returns str(self).
I think this is the only reasonable interpretation of format_spec's for
datetime. Does anyone think otherwise?
E
> I have a patch to add __format__ to datetime.time, .date, and .datetime.
> For non-empty format_spec's, I just pass on to .strftime. For empty
> format_spec's, it returns str(self).
>
> I think this is the only reasonable interpretation of format_spec's for
> datetime. Does anyone think o
Martin v. Löwis wrote:
>> I have a patch to add __format__ to datetime.time, .date, and .datetime.
>> For non-empty format_spec's, I just pass on to .strftime. For empty
>> format_spec's, it returns str(self).
>>
>> I think this is the only reasonable interpretation of format_spec's for
>> da
On 10/09/2007, Eric Smith <[EMAIL PROTECTED]> wrote:
> Martin v. Löwis wrote:
> >> I have a patch to add __format__ to datetime.time, .date, and .datetime.
> >> For non-empty format_spec's, I just pass on to .strftime. For empty
> >> format_spec's, it returns str(self).
> >>
> >> I think this is
Eric Smith wrote:
> Martin v. Löwis wrote:
>>> I have a patch to add __format__ to datetime.time, .date, and .datetime.
>>> For non-empty format_spec's, I just pass on to .strftime. For empty
>>> format_spec's, it returns str(self).
>>>
>>> I think this is the only reasonable interpretation of
On 9/10/07, gregory.p.smith <[EMAIL PROTECTED]> wrote:
>
> Author: gregory.p.smith
> Date: Mon Sep 10 01:55:55 2007
> New Revision: 58068
>
> Modified:
>python/branches/py3k/Doc/library/exceptions.rst
>python/branches/py3k/Doc/library/socket.rst
>python/branches/py3k/Doc/whatsnew/2.6.rs
> svnmerge should come with svn, nowadays, or you can download it separately
> (as svnmerge.py, probably; it's just a Python script.)
It comes with version 3 of svn.
Or
http://svn.collab.net/repos/svn/trunk/contrib/client-side/svnmerge/svnmerge.py.
Bill
_
> It comes with version 3 of svn.
Sorry, that should be 1.3. But I see I've got version 1.4.4 installed,
and no svnmerge. Of course, this is Apple's XCode version of svn.
Bill
___
Python-3000 mailing list
Python-3000@python.org
http://mail.python.org/
On 9/9/07, Nicholas Bastin <[EMAIL PROTECTED]> wrote:
> On 9/9/07, Guido van Rossum <[EMAIL PROTECTED]> wrote:
> > On 9/9/07, Nicholas Bastin <[EMAIL PROTECTED]> wrote:
> > > I'm not suggesting that Python handle small ints itself and then farm
> > > out large integer computations, I'm suggesting
On 9/10/07, Thomas Wouters <[EMAIL PROTECTED]> wrote:
>
>
> On 9/10/07, gregory.p.smith <[EMAIL PROTECTED] > wrote:
> >
> > Author: gregory.p.smith
> > Date: Mon Sep 10 01:55:55 2007
> > New Revision: 58068
> >
> > Modified:
> >python/branches/py3k/Doc/library/exceptions.rst
> >python/branc
On 9/10/07, Eric Smith <[EMAIL PROTECTED]> wrote:
> I have a patch to add __format__ to datetime.time, .date, and .datetime.
> For non-empty format_spec's, I just pass on to .strftime. For empty
> format_spec's, it returns str(self).
>
> I think this is the only reasonable interpretation of form
On 10/09/2007, Thomas Wouters <[EMAIL PROTECTED]> wrote:
> svnmerge should come with svn, nowadays, or you can download it separately
> (as svnmerge.py, probably; it's just a Python script.)
It's not part of the Win32 binary distribution for Subversion - but I
found it at http://www.orcaware.com/s
On 10-Sep-07, at 8:33 AM, Thomas Wouters wrote:
> Alternatively, if you know what you're doing, you can edit the
> svnmerge-integrated property on the branch directly -- but don't
> mess it up :)
>
svnmerge also has a handy -M flag that marks a (set of) revisions as
merged, but doesn't act
On 9/10/07, Guido van Rossum <[EMAIL PROTECTED]> wrote:
> On 9/9/07, Nicholas Bastin <[EMAIL PROTECTED]> wrote:
> > On 9/9/07, Guido van Rossum <[EMAIL PROTECTED]> wrote:
> > > On 9/9/07, Nicholas Bastin <[EMAIL PROTECTED]> wrote:
> > > > I'm not suggesting that Python handle small ints itself and
On 9/10/07, Nicholas Bastin <[EMAIL PROTECTED]> wrote:
> > > I did redo my benchmark using 200 as the increment number instead of
> > > 1, to duck any impact from the interning of small value ints in 2.6,
> > > and it made no discernible difference in the results.
> >
> > I'm sorry, I've lost conte
On 9/10/07, Guido van Rossum <[EMAIL PROTECTED]> wrote:
> On 9/10/07, Nicholas Bastin <[EMAIL PROTECTED]> wrote:
> > > > I did redo my benchmark using 200 as the increment number instead of
> > > > 1, to duck any impact from the interning of small value ints in 2.6,
> > > > and it made no discernib
Martin v. Löwis wrote:
> The first right of the user is to get the source code
> easily, without having to beg for it. Only then it is also
> the user's right to modify it, and use the modified version
> in the application.
Where does begging come into it? As long as the user
is provided with inf
Greg Ewing wrote:
If there's a link on the same web page that works
when the user clicks on it, I don't think they're
even going to notice the difference.
They'll notice the difference when they want to redistribute Python,
when they note the new licensing-based restrictions ("GMP must be in a
Marcin 'Qrczak' Kowalczyk wrote:
> The major technical problem with GMP is that an out of memory condition
> during computation is a fatal error, GMP does not provide a way to
> recover from it.
If using GMP itself is not feasible, then perhaps
some algorithms could be extracted from it in
areas w
On 9/10/07, Greg Ewing <[EMAIL PROTECTED]> wrote:
> Martin v. Löwis wrote:
>
> > The first right of the user is to get the source code
> > easily, without having to beg for it. Only then it is also
> > the user's right to modify it, and use the modified version
> > in the application.
>
> Where doe
On 9/10/07, Larry Hastings <[EMAIL PROTECTED]> wrote:
>
> Greg Ewing wrote:
> If there's a link on the same web page that works
> when the user clicks on it, I don't think they're
> even going to notice the difference.
>
> They'll notice the difference when they want to redistribute Python, when
Paul Moore wrote:
> I'd like to see the default format specified (somewhere). I note that
> the default format for datetime values seems to differ for me (on
> 3.0a1 on Windows)
>
> Python 3.0a1 (py3k:57844, Aug 31 2007, 16:54:27) [MSC v.1310 32 bit
> (Intel)] on win32
> Type "help", "copyright",
Paul> The date and time defaults (which appear to be %Y-%m-%d and
Paul> %H:%M:%s) seem perfectly acceptable, on the other hand.
I would like to see an analog to %S which preserves fractions of a second as
the default formatting for time and datetime objects does:
>>> print(now)
2
Right. It's odd that there's nothing explicit that exactly produces
the default. (Though floats have this issue too -- I wish it could be
fixed there too.)
On 9/10/07, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
>
> Paul> The date and time defaults (which appear to be %Y-%m-%d and
> Paul>
I'd like to see Travis's response to this. It's setting a precedent
regarding locking objects in read-only mode; I haven't found other
examples of objects using LOCKDATA (the only mentions of it seem to be
rejecting it :). I keep getting confused by the two separate lock
counts (and I think in this
"Nicholas Bastin" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]
| Yeah, my point was mostly an aside to anyone that might have
| questioned my earlier results of a 2.3x slowdown on integer-sized
| values because I used 1. A quick switch to 200 netted the exact same
| results,
Curr
On 9/10/07, Terry Reedy <[EMAIL PROTECTED]> wrote:
>
> "Nicholas Bastin" <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]
>
> | Yeah, my point was mostly an aside to anyone that might have
> | questioned my earlier results of a 2.3x slowdown on integer-sized
> | values because I used 1
Guido van Rossum wrote:
> I'd like to see Travis's response to this. It's setting a precedent
> regarding locking objects in read-only mode; I haven't found other
> examples of objects using LOCKDATA (the only mentions of it seem to be
> rejecting it :). I keep getting confused by the two separate
>> The first right of the user is to get the source code
>> easily, without having to beg for it. Only then it is also
>> the user's right to modify it, and use the modified version
>> in the application.
>
> Where does begging come into it? As long as the user
> is provided with information which
> If python.org agreed to host the GMP source, that would suffice for
> all people distributing python binaries (they could then just refer to
> the GMP source download as a link).
It would not if they don't distribute the binary through download.
If they put it on some media, or preinstalled on a
> Interesting, I didn't look at the code (obviously), but my
> understanding was that it was only positive integers below 100.
See NSMALLPOSINTS and NSMALLNEGINTS. It's 257 positive ints since
r42552, contributed through bugs.python.org/1436243.
Regards,
Martin
___
Nicholas Bastin wrote:
As for the user-replaceable shared library part, that's up for
considerable debate. It's unlikely that static linkage legally
creates a derivative work (that would be pretty unreasonable in
computer science terms), but it's never been tested in court, so
static linking wou
33 matches
Mail list logo
