On Tue, Feb 10, 2009 at 8:49 AM, Robert Bradshaw
wrote:
>
> On Feb 9, 2009, at 11:34 PM, Simon King wrote:
>
>> Yes!
>> I could imagine:
>> 1. FOO.X searches for attributes that start with X (current
>> behaviour)
>> 2. FOO.X searches for attributes that *contain* X (new
>> feature)
+1! For ex
On Feb 9, 2009, at 11:34 PM, Simon King wrote:
> Hi!
>
> On Feb 9, 11:46 pm, john_perry_usm wrote:
>> What if there were a different trigger for the extended completions?
>> This way the user would have only one box to parse at a time.
>
> Yes!
> I could imagine:
> 1. FOO.X searches for attribu
Hi!
On Feb 9, 11:46 pm, john_perry_usm wrote:
> What if there were a different trigger for the extended completions?
> This way the user would have only one box to parse at a time.
Yes!
I could imagine:
1. FOO.X searches for attributes that start with X (current
behaviour)
2. FOO.X searches f
On Feb 9, 2009, at 2:21 PM, Nicolas M. Thiery wrote:
> On Fri, Feb 06, 2009 at 03:32:25PM -0800, William Stein wrote:
>> On Fri, Feb 6, 2009 at 1:01 PM, mabshoff
>> wrote:
>>> On Feb 6, 11:53 am, "Nicolas M. Thiery"
I just want to mention that I am precisely setting up a
standardi
On Feb 5, 12:33 am, mabshoff wrote:
> If you check outhttp://sagemath.org/mirrors.htmlyou see that many
> mirrors are out of date, mostly the ones that mirror off sage.math
> instead of sagemath.org. One reason is likely that the place on
> sage.math where to mirror from has changes, but the ol
What if there were a different trigger for the extended completions?
This way the user would have only one box to parse at a time.
john perry
On Feb 9, 9:44 am, Jason Grout wrote:
> Hi all,
>
> Currently, anytime there is a command that contains a noun and an
> adjective, there seems to be deba
> This is a bit arbitrary for my tastes, but I think it (or somthing
> similar) could work:
>
> If you type foo.X, where X is only one or two characters, then
> the extended completion list checks for *_X*. Otherwise (if X is
> three or more characters) then the extended completion list is *X*.
On Fri, Feb 06, 2009 at 03:32:25PM -0800, William Stein wrote:
> On Fri, Feb 6, 2009 at 1:01 PM, mabshoff wrote:
> > On Feb 6, 11:53 am, "Nicolas M. Thiery"
> >> I just want to mention that I am precisely setting up a standardized
> >> way in the category framework for setting up this kind of ge
On Feb 9, 2:08 pm, Jason Grout wrote:
> mabshoff wrote:
> > Well, svg on IE requires a plugin while the canvas element seems to be
> > support by all major rendering engines. So why make things
> > complicated? Anything requiring a plugin leads to usability problems,
> > i.e. I don't have to
On Feb 9, 10:39 am, mhampton wrote:
> I think this is a great idea. It might not solve the problem of
> clutter immediately, because if something like eigenvectors_right is
> removed it would break a lot of existing code. But perhaps we could
> remove such things in 4.0 after adding deprecati
mabshoff wrote:
>
>
> On Feb 9, 12:41 pm, Jason Grout wrote:
>> Rob Beezer wrote:
>>> When you follow the link to the demos, you get a graphic on the right,
>>> where if you move the mouse over the plot, text below continuously
>>> updates the coordinates. The vertical scale appears logarithmi
On Feb 9, 1:43 pm, Ondrej Certik wrote:
> On Mon, Feb 9, 2009 at 1:17 PM, mabshoff wrote:
> > * fix the "trivial" issues Tim reported a while back, i.e. #3686-
> > #3690
> > * fix the non-owner ought to be able to doctest issues, i.e. #5155 so
> > that users can run the regression suite o
On Feb 9, 12:41 pm, Jason Grout wrote:
> Rob Beezer wrote:
> > When you follow the link to the demos, you get a graphic on the right,
> > where if you move the mouse over the plot, text below continuously
> > updates the coordinates. The vertical scale appears logarithmic and
> > the coordinat
On Mon, Feb 9, 2009 at 1:17 PM, mabshoff wrote:
>
>
>
> On Feb 9, 12:40 pm, Ondrej Certik wrote:
>> On Sun, Feb 8, 2009 at 7:40 PM, Tim Abbott wrote:
>
> Hi Tim, Ondrej,
>
>> > I mentioned that I'd send mail here again once the segfault problems are
>> > fixed.
>>
>> I want to congratulate you
On Feb 9, 9:09 am, gedaliah wrote:
Hi,
> This looks like the relevant portion of the log --
>
> if /bin/sh /local/ftp/linux/sage/sage-3.2.3/spkg/build/gnutls-2.2.1.p1/
> src/build-aux/missing --run makeinfo -I ../doc -I . \
> -o gnutls.info gnutls.texi; \
> then \
> rc=0; \
> cd .; \
> el
On Feb 9, 12:40 pm, Ondrej Certik wrote:
> On Sun, Feb 8, 2009 at 7:40 PM, Tim Abbott wrote:
Hi Tim, Ondrej,
> > I mentioned that I'd send mail here again once the segfault problems are
> > fixed.
>
> I want to congratulate you once again for all the work you did. I
> noticed on your webpage
On Feb 9, 8:44 am, Jason Grout wrote:
> What do people think? I think this would finally answer the tension
> between the people that want useful tab completion and the people who
> want the function names to look "right".
FWIW I 'm pretty sure aliases are a disaster; whether you
can get most
On Mon, Feb 9, 2009 at 10:28 AM, Florent Hivert
wrote:
>
> Hi,
>
>> ValueError derives from ArithmeticError, so if anybody wrote
>>
>>try: something...
>>except ArithmeticError:
>> ..
>>
>> and you change the ArithmeticError to ValueError, then their code will
>> still work fin
On Mon, Feb 9, 2009 at 11:05 AM, Jason Grout
wrote:
>
> Carl Witty wrote:
>> And maybe the second list should be omitted altogether if it's too
>> big? For instance, if I type foo.e, I'm probably not interested
>> in the list of all method names that include an 'e' somewhere.
>
> Yes, but I'm no
Rob Beezer wrote:
> When you follow the link to the demos, you get a graphic on the right,
> where if you move the mouse over the plot, text below continuously
> updates the coordinates. The vertical scale appears logarithmic and
> the coordinates are reported correctly.
>
> No idea if this is s
On Sun, Feb 8, 2009 at 7:40 PM, Tim Abbott wrote:
>
> I mentioned that I'd send mail here again once the segfault problems are
> fixed.
I want to congratulate you once again for all the work you did. I
noticed on your webpage, that you were told that it would need 5 or 6
people working on it. It
Carl Witty wrote:
> On Mon, Feb 9, 2009 at 7:44 AM, Jason Grout
> wrote:
>> What if, at least for the notebook, and maybe for the command line too,
>> we listed two sets of completions: the first would be the list given
>> now, and the second would be the list given above?
>
> +1 for extending
I think this is a great idea. It might not solve the problem of
clutter immediately, because if something like eigenvectors_right is
removed it would break a lot of existing code. But perhaps we could
remove such things in 4.0 after adding deprecation warnings.
-Marshall
On Feb 9, 5:44 pm, Jas
On Mon, Feb 9, 2009 at 7:44 AM, Jason Grout wrote:
> What if, at least for the notebook, and maybe for the command line too,
> we listed two sets of completions: the first would be the list given
> now, and the second would be the list given above?
+1 for extending tab-completion to mention thes
Hi,
> ValueError derives from ArithmeticError, so if anybody wrote
>
>try: something...
>except ArithmeticError:
> ..
>
> and you change the ArithmeticError to ValueError, then their code will
> still work fine.
> So I think the impact of making this change isn't too bad.
G
On Mon, Feb 9, 2009 at 7:39 AM, Florent Hivert
wrote:
>
> Dear William,
>
> Sorry to bother you again with those stupid error messages... You'll
> probably think I'm a kind of lawyer after that :-) I still needs a
> confirmation... We agreed that:
>
>> 2. Any nx0 or 0xn (n != 0) matrix is
Dear William,
Sorry to bother you again with those stupid error messages... You'll
probably think I'm a kind of lawyer after that :-) I still needs a
confirmation... We agreed that:
> 2. Any nx0 or 0xn (n != 0) matrix is not invertible and should raise a
>**ValueError** "matrix mus
This looks like the relevant portion of the log --
if /bin/sh /local/ftp/linux/sage/sage-3.2.3/spkg/build/gnutls-2.2.1.p1/
src/build-aux/missing --run makeinfo -I ../doc -I . \
-o gnutls.info gnutls.texi; \
then \
rc=0; \
cd .; \
else \
rc=$?; \
cd . && \
$restore $backupdir/* `echo "
+1 to both commandline and notebook
On Mon, Feb 9, 2009 at 11:55 AM, wrote:
>
> +1!
>
>
> On Mon, 9 Feb 2009, Jason Grout wrote:
>
>>
>> Hi all,
>>
>> Currently, anytime there is a command that contains a noun and an
>> adjective, there seems to be debate about what order they should be
>> list
+1!
On Mon, 9 Feb 2009, Jason Grout wrote:
>
> Hi all,
>
> Currently, anytime there is a command that contains a noun and an
> adjective, there seems to be debate about what order they should be
> listed. For example, there have been debates over whether to name a
> function right_eigenvectors
+1 for the command line.
Currently to find all the plot commands is not as easy as it should be IMHO
On Mon, Feb 9, 2009 at 10:44 AM, Jason Grout
wrote:
>
> Hi all,
>
> Currently, anytime there is a command that contains a noun and an
> adjective, there seems to be debate about what order they
> In other words, if
> in the notebook, you typed m.eigenvectors, you would get a box
> containing two lists:
> ___
> | |
> | list of current completions|
> | |
> |-
Hi all,
Currently, anytime there is a command that contains a noun and an
adjective, there seems to be debate about what order they should be
listed. For example, there have been debates over whether to name a
function right_eigenvectors and eigenvectors_right. It seems the big
argument for
On Sun, Feb 8, 2009 at 8:38 AM, Georg S. Weber
wrote:
>
>
>
> On 8 Feb., 10:11, Robert Bradshaw
> wrote:
>> On Feb 8, 2009, at 1:04 AM, Minh Nguyen wrote:
>>
>>
>>
>> > Hi William,
>>
>> > On Sun, Feb 8, 2009 at 7:52 PM, William Stein
>> > wrote:
>> > [...]
>> >> Anyway, the bibtex entry was de
Hello folks,
3.3.rc0 will drop tonight and this is my status list:
Unresolved blocker:
#4942: find_root() is broken when interval borders cannot be evaluated
(Mike looked at it, it seems like a scipy issue)
#5100: worksheets: can't empty the trash (safari only?) (no lead)
#5210: gmp-mpir-0.9.rc
35 matches
Mail list logo
