Poll: Making sage-mode a standard package

[Ivan Andrus]

Dear all sage and emacs (or not) users,

Since I've been working on sage-mode a bit recently I thought I would look at trac and try to fix any sage-mode related bugs there. I found #2666 which wants to make sage-mode a standard package.

If people want this then I'd like to get it done. But if not then I'd like to close the ticket. Personally, I'm not sure it should be standard since people who wish to use it still have to add something to their .emacs, and hence they still need to know about it and take some action. However, I'm open to being persuaded. The spkg is 272K.

[ ] Make it standard at some point in the "near" future.
[ ] Don't make it standard and close the ticket.
[ ] Don't make it standard now, but don't close the ticket.

-Ivan

[Benjamin Jones]

I've been using the sage-mode spkg you made and I like it, but I'd say
this should be an optional package. So I vote:

[X ] Don't make it standard and close the ticket. ** Include it as an
optional package.**

--
Benjamin Jones

[David Roe]

[X ] Don't make it standard and close the ticket. ** Include it as an
optional package.**

+1
David

[William Stein]

On Fri, May 25, 2012 at 2:24 PM, Ivan Andrus <darth...@gmail.com> wrote:

[x] I thought it was already standard, and am surprised to hear that it isn't.

>
> --
> To post to this group, send an email to sage-...@googlegroups.com
> To unsubscribe from this group, send an email to sage-devel+...@googlegroups.com
> For more options, visit this group at http://groups.google.com/group/sage-devel
> URL: http://www.sagemath.org



--
William Stein
Professor of Mathematics
University of Washington
http://wstein.org

[Justin C. Walker]

On Fri, May 25, 2012 at 2:24 PM, Ivan Andrus <darth...@gmail.com> wrote:

> [X] Don't make it standard and close the ticket. Include as optional package.

Justin

--
Justin C. Walker, Curmudgeon at Large
Institute for the Absorption of Federal Funds
-----------
If it weren't for carbon-14, I wouldn't date at all.
-----------

[William Stein]

On Fri, May 25, 2012 at 3:41 PM, Justin C. Walker <jus...@mac.com> wrote:
> On Fri, May 25, 2012 at 2:24 PM, Ivan Andrus <darth...@gmail.com> wrote:
>
>> [X] Don't make it standard and close the ticket.  Include as optional package.

I'm curious -- why are people against sage-mode being a standard package?

-- William

[Keshav Kini]

Ivan Andrus <darth...@gmail.com> writes:
> Dear all sage and emacs (or not) users,
>
> Since I've been working on sage-mode a bit recently I thought I would look at trac and try to fix any sage-mode related bugs there. I found #2666 which wants to make sage-mode a standard package.
>
> If people want this then I'd like to get it done. But if not then I'd like to close the ticket. Personally, I'm not sure it should be standard since people who wish to use it still have to add something to their .emacs, and hence they still need to know about it and take some action. However, I'm open to being persuaded. The spkg is 272K.
>
> [ ] Make it standard at some point in the "near" future.

This. It is very standard for programs to include relevant emacs modes
and vim syntax/indent/whatever files in their source distribution.
Cython does this, for example, as does GAP.

> [ ] Don't make it standard and close the ticket.
> [ ] Don't make it standard now, but don't close the ticket.

-Keshav

----
Join us in #sagemath on irc.freenode.net !

[Volker Braun]

[X] Make it standard at some point in the "near" future. 

Its tiny, too...

[Justin C. Walker]

On May 25, 2012, at 16:27 , Volker Braun wrote:

> [X] Make it standard at some point in the "near" future.
>
> Its tiny, too...

"Tiny" is not a good reason to increase the number of moving parts.

--
Justin C. Walker, Curmudgeon at Large
Institute for the Absorption of Federal Funds
-----------

I want to die, peacefully in my sleep, like my grandfather;
not screaming in terror, like his passengers.

[Justin C. Walker]

For me, it's not about "sage-mode": it's got more to do with the number of moving parts. Fewer is better, all things equal, and we ought to put things into Sage if there are really good reasons.

If it's standard, tests are needed.

Also, emacs, as delivered on a given platform may not change much, but a number of people use GUI versions of the app, which are separate from the regular distributions (Mac OS X, for example). Version change for those apps is unpredictable, and based on the user.

Is emacs even a standard part of Linux distributions?

Finally, I think most emacs users would have little problem installing an optional package, while most non emacs users aren't interested :-} That's a rash generalization, but it resonates...

One guy's opinion...

Justin

--
Justin C. Walker, Curmudgeon-At-Large
Director
Institute for the Enhancement of the Director's Income
--------
"Weaseling out of things is what separates us from the animals.
Well, except the weasel."
- Homer J Simpson
--------

[Benjamin Jones]

On Fri, May 25, 2012 at 4:57 PM, Justin C. Walker <jus...@mac.com> wrote:
>
> On May 25, 2012, at 16:07 , William Stein wrote:
>
>> On Fri, May 25, 2012 at 3:41 PM, Justin C. Walker <jus...@mac.com> wrote:
>>> On Fri, May 25, 2012 at 2:24 PM, Ivan Andrus <darth...@gmail.com> wrote:
>>>
>>>> [X] Don't make it standard and close the ticket.  Include as optional package.
>>
>> I'm curious -- why are people against sage-mode being a standard package?
>
> For me, it's not about "sage-mode": it's got more to do with the number of moving parts.  Fewer is better, all things equal, and we ought to put things into Sage if there are really good reasons.
>
> If it's standard, tests are needed.
>
> Also, emacs, as delivered on a given platform may not change much, but a number of people use GUI versions of the app, which are separate from the regular distributions (Mac OS X, for example). Version change for those apps is unpredictable, and based on the user.
>
> Is emacs even a standard part of Linux distributions?
>
> Finally, I think most emacs users would have little problem installing an optional package, while most non emacs users aren't interested :-}  That's a rash generalization, but it resonates...
>
> One guy's opinion...
>
> Justin
>

I agree with Justin; well put.

--
Benjamin Jones

[Julien Puydt]

Le vendredi 25 mai, Keshav Kini a écrit:

> This. It is very standard for programs to include relevant emacs modes
> and vim syntax/indent/whatever files in their source distribution.
> Cython does this, for example, as does GAP.

This is a very important point which bears repeating : sage should try
to conform as much as possible to usual open source practices and
standards.

Snark on #sagemath

[Dima Pasechnik]

On 2012-05-25, Ivan Andrus <darth...@gmail.com> wrote:
> Dear all sage and emacs (or not) users,
>

> [ ] Make it standard at some point in the "near" future.
> [ ] Don't make it standard and close the ticket.
> [ ] Don't make it standard now, but don't close the ticket.

[X] Don't make it standard and close the ticket. Include as optional
package.

Also due to the painfully obvious next move: bundle in an emacs distro
compatible with the said sage-mode...
No, really, we'd hate to see a stream of complaints like "oh, it does
not work on my ZSuperDuperEmacs version 666.666.6..."

Someone mentioned that GAP has such a mode in a standard distribution,
but in fact it's horribly broken and obsolete,
$ ls -l /usr/local/src/gap/gap4r4/etc/emacs/
total 248
-rw-r--r-- 1 dima wheel 58933 Aug 15 2001 comint.el
-rw-r--r-- 1 dima wheel 4542 Aug 15 2001 gap-mode.doc
-rw-r--r-- 1 dima wheel 35909 Aug 15 2001 gap-mode.el
-rw-r--r-- 1 dima wheel 17587 May 8 2005 gap-process.el

Dima

[P Purkayastha]

For that matter even the vim plugins are broken. See

http://archives.gentoo.org/gentoo-science/msg_a74bfb4eb0abd88584885bab90d7e891.xml 

[P Purkayastha]

What I meant to say is that, in the long term there is a possibility that the package might become unmaintained. No offense to Ivan, but this is a fact of the way OSS works. :)

[Ivan Andrus]

On May 26, 2012, at 1:11 AM, Keshav Kini wrote:
> Ivan Andrus <darth...@gmail.com> writes:
>> Dear all sage and emacs (or not) users,
>>
>> Since I've been working on sage-mode a bit recently I thought I would look at trac and try to fix any sage-mode related bugs there. I found #2666 which wants to make sage-mode a standard package.
>>
>> If people want this then I'd like to get it done. But if not then I'd like to close the ticket. Personally, I'm not sure it should be standard since people who wish to use it still have to add something to their .emacs, and hence they still need to know about it and take some action. However, I'm open to being persuaded. The spkg is 272K.
>>
>> [ ] Make it standard at some point in the "near" future.
>
> This. It is very standard for programs to include relevant emacs modes
> and vim syntax/indent/whatever files in their source distribution.
> Cython does this, for example, as does GAP.
>

> -Keshav

That's true. Often they have bash-completion and man pages as well. Which now that I think about it is an interesting idea. Maybe someone should improve http://wiki.sagemath.org/bash-completion and create an spkg. Actually, when we switch to argparse we should just use optcomplete or genzshcomp or something similar. Of course we might still have to write some custom stuff to properly complete certain subcommands (like sage --gap) which do there own parsing.

OTOH, most programs don't have an easy way to install optional pieces like Sage does with spkg's.

-Ivan

[Ivan Andrus]

On May 26, 2012, at 11:50 AM, Dima Pasechnik wrote:
> On 2012-05-25, Ivan Andrus <darth...@gmail.com> wrote:
>> Dear all sage and emacs (or not) users,
>>
>> [ ] Make it standard at some point in the "near" future.
>> [ ] Don't make it standard and close the ticket.
>> [ ] Don't make it standard now, but don't close the ticket.
>
> [X] Don't make it standard and close the ticket. Include as optional
> package.
>
> Also due to the painfully obvious next move: bundle in an emacs distro
> compatible with the said sage-mode...
> No, really, we'd hate to see a stream of complaints like "oh, it does
> not work on my ZSuperDuperEmacs version 666.666.6..."

I'll just have to use the time travel capabilities of emacs to ensure compatibility [1] :-)

> Someone mentioned that GAP has such a mode in a standard distribution,
> but in fact it's horribly broken and obsolete,
> $ ls -l /usr/local/src/gap/gap4r4/etc/emacs/
> total 248
> -rw-r--r-- 1 dima wheel 58933 Aug 15 2001 comint.el
> -rw-r--r-- 1 dima wheel 4542 Aug 15 2001 gap-mode.doc
> -rw-r--r-- 1 dima wheel 35909 Aug 15 2001 gap-mode.el
> -rw-r--r-- 1 dima wheel 17587 May 8 2005 gap-process.el

I asked a while back if they would include my update to gap-mode [2], and got a non-commital response. I guess it's about time to ask again.

-Ivan

[1] https://github.com/emacsmirror/emacs/blob/master/etc/future-bug
[2] https://bitbucket.org/gvol/gap-mode

[Ivan Andrus]

No offense taken.  Who knows what's going to happen after I graduate for example—maybe I'll get a full time job fixing bugs in Sage, or maybe I'll be flipping burgers. :-)  

It's important to remember that if it's standard someone _has_ to maintain it.  Of course, it's also more likely that someone _will_ maintain it.

-Ivan

[Keshav Kini]

Dima Pasechnik <dim...@gmail.com> writes:
> Someone mentioned that GAP has such a mode in a standard distribution,
> but in fact it's horribly broken and obsolete,

This is obviously a problem with GAP, and not with the idea of
distributing an Emacs mode in a standard distribution. I'm sure many
things in GAP are "horribly broken and obsolete" :P ... as is perfectly
normal for a piece of software of GAP's age!

[Benjamin Jones]

On Sat, May 26, 2012 at 6:55 AM, Ivan Andrus <darth...@gmail.com> wrote:

>
> No offense taken.  Who knows what's going to happen after I graduate for
> example—maybe I'll get a full time job fixing bugs in Sage, or maybe I'll be
> flipping burgers. :-)
>
> It's important to remember that if it's standard someone _has_ to maintain
> it.  Of course, it's also more likely that someone _will_ maintain it.
>
> -Ivan
>

From the point of view that making sage-mode a standard package would
increase the likelihood that it is maintained, I would support it
being included as a standard package.

--
Benjamin Jones

[Ivan Andrus]

Do you say that because you feel it hasn't been maintained well, or that it might not be in the future since the original author has moved on?

-Ivan

[Benjamin Jones]

I said that in the spirit that I think sage-mode is a useful package
and any steps we can take to ensure that useful packages continue to
be well maintained is a good thing. I haven't been using it for all
that long, so I can't comment on if it has or hasn't been maintained
well in the past. As for the future, who knows, but making it a
standard package will increase its chances to thrive compared to
making it an optional package.

What is your opinion as the current maintainer?

--
Benjamin Jones

[Ivan Andrus]

Unless I'm mistaken, if it becomes standard then someone (probably me) has to volunteer to maintain it for at least 2 years. I'm okay with that since I'll be maintaining the Mac and iPhone apps for probably that long. Of course I would prefer to not commit in case things change in the future. :-) But I would be willing to do it.

If I doesn't go standard and things don't work out for me after graduation in a way that I can spend a "lot" of time on sage, then I probably won't maintain it other than perhaps helping track down issues (i.e. no new features, code cleanup etc.). Usually emacs lisp code is pretty easy to debug so I don't think it would take too much time. Of course even if it does go standard there is no guarantee of new features, but it would be slightly more likely. So in the end it may not making too much difference.

Also, do there need to be automated tests to become standard? If so then someone would have to write them. I would _love_ to have some tests for sage-mode, especially since we will have to support fgallina's rewrite of python.el pretty soon [1]. I certainly couldn't write them really soon (I'm already way behind on the iPhone app), but it would be a good reason to keep the ticket open.

-Ivan

[1] https://bitbucket.org/gvol/sage-mode/issue/1/support-pythonel-by-fgallina

[Nicolas M. Thiéry]

Le Ven 25 mai 2012 23:24, Ivan Andrus a �crit :

> Dear all sage and emacs (or not) users,
>
>
> Since I've been working on sage-mode a bit recently I thought I would
> look at trac and try to fix any sage-mode related bugs there. I found
> #2666 which wants to make sage-mode a standard package.
>
>
> If people want this then I'd like to get it done. But if not then I'd
> like to close the ticket. Personally, I'm not sure it should be standard
> since people who wish to use it still have to add something to their
> .emacs, and hence they still need to know about it and take some action.
> However, I'm open to being persuaded. The spkg is 272K.
>
>

> [X] Make it standard at some point in the "near" future.

[Ivan Andrus]

The results are fairly close (I guess everyone has responded who wants to). We have 3.5 votes for making it standard, 3.5 for not, and 1 for thought it was standard, which I assume means make it standard. So I guess I'll keep the ticket open and work towards making it standard unless anyone else would like to weigh in.

-Ivan

[X] Don't make it standard and close the ticket.
Justin C. Walker
David Roe
Dima Pasechnik
Benjamin Jones -- initially, see below

[X] Make it standard at some point in the "near" future.
Nicolas M. Thiéry
Keshav Kini
Volker Braun
Benjamin Jones -- if it would improve maintenance, see above

[X] I thought it was already standard, and am surprised to hear that it isn't.
William Stein

[Keshav Kini]

You might add Snark to this list, as he seemed to be agreeing with me.

>
> [X] I thought it was already standard, and am surprised to hear that it isn't.
> William Stein

[David Kohel]

[X ] Make it standard at some point in the "near" future.

Argument: Sage includes many interfaces to standard software.
This includes software that the majority users will never use in
their lifetimes, e.g. the big M's (unless at a mathematics
department or institute where there exists a license) and
numerous open source software libraries and systems.
Emacs is a standard open source tool, and an interface
(albeit from emacs to sage) seems reasonable.
I also thought that this was "standard" in the sense that
it was one file which one could find (like sagetex.sty) in
the sage distribution. I would like to find it there if I need
to update my system.

--David

[syd.la...@gmail.com]

[X] Make it standard at some point in the "near" future

Mainly because I think everybody should use emacs as their operating system. it will save their time and the money they spend on buying fancey computers. Also Nicolas is very cool so I wanted to support his point ;)

That's said, I found couple of bugs in it when I use it with emacs 24, for example you can only run sage-build from a window with inferior-sage-mode (subproblem of what we discussed before probably you need to add that hook thingi to the sage-build as well), also autocomplete and keyword?? take ages to work. It also hijacks python files no matter if there's a python interpretor open or not (I wrote a fix for this one). Though, it's unstable emacs 24. But if by the time it gets released, sage-mode is not standard, it's less probable that somebody cares about these bugs as Benjamin said.

בתאריך יום רביעי, 30 במאי 2012 09:27:05 UTC-6, מאת David Kohel:

[Ivan Andrus]

On Jun 1, 2012, at 7:55 AM, syd.la...@gmail.com wrote:

> [X] Make it standard at some point in the "near" future
>
> Mainly because I think everybody should use emacs as their operating system. it will save their time and the money they spend on buying fancey computers. Also Nicolas is very cool so I wanted to support his point ;)
>
> That's said, I found couple of bugs in it when I use it with emacs 24,

Please open tickets on bitbucket [1] so they don't get lost.

> for example you can only run sage-build from a window with inferior-sage-mode (subproblem of what we discussed before probably you need to add that hook thing to the sage-build as well),

Yeah. I need to figure out the right way to fix that once and for all. I might file a bug with Emacs after I figure out exactly why we're doing it in the first place. :-)

> also autocomplete and keyword?? take ages to work. It also hijacks python files no matter if there's a python interpretor open or not (I wrote a fix for this one).

If you can fork on bitbucket and send a pull request that would be great. Or just sending me a diff would be fine too.

> Though, it's unstable emacs 24. But if by the time it gets released, sage-mode is not standard, it's less probable that somebody cares about these bugs as Benjamin said.

Yeah. Emacs 24 is in pretest right now, and I don't think it will be long before it's released (the release candidate was just released. In 24.2 they're going to change python.el so I'm sure there will be plenty of work in the near future.

I really need to get some automated tests in place so that it's easy to check with different emacs version. But writing tests is even more boring than writing my thesis. :-)

-Ivan