Sage Stack Exchange Site

[Kannappan Sampath]

Hello Sage Users: 

I proposed a Sage Stack Exchange site for Sage. Please go and follow it; add example questions and vote up (down) questions that you think should be on-topic (resp. off-topic) for the upcoming site. 

http://area51.stackexchange.com/proposals/70511/sage?referrer=8adE6ec0VyqQCyB1kFgF7g2

Looking forward to your support and cooperation for the success of this adventure on Stack Exchange. 

/p/s/ I would be soon adding my quota of 5 questions... 

With Sincere Regards, 

Kannappan. 

[Dima Pasechnik]

Does one really need an SE site for Sage? What's wrong with
http://ask.sagemath.org/ in your opinion then?


>
> With Sincere Regards,
> Kannappan.
>

[Kannappan Sampath]

Hi Dima, 

The major issue has been one of maintenance IMHO. For example, the ask.sagemath site had an outage owing to the massive spamming! These issues are neatly handled in the Stack Exchange network. 

-- Kannappan, 

--
You received this message because you are subscribed to the Google Groups "sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.
To post to this group, send email to sage-...@googlegroups.com.
Visit this group at http://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[Niles Johnson]

As I recall, the question of going with a SE site was discussed around the time that AskSage launched. As I recall, the consensus was that a fully free QA site was more consistent with the values of the Sage community. It may be that this is no longer the consensus (if it ever was). Or the consensus of the subset who are ready to actually do something may have changed.

Maintenance of AskSage is definitely a problem, although not an insurmountable one. The person who had been doing it silently stopped some time ago, and left things in a pretty bad state. There are more responsible people who can take over (I could be one of them), but some back-end issues have to be resolved first, and this is taking non-trivial time and energy.

There are certainly lots of good example questions on AskSage, so (if this is compatible with the licenses) a motivated person could copy them to the Area 51 site.

-Niles

[kcrisman]

> Hello Sage Users:
>
> I proposed a Sage Stack Exchange site for Sage. Please go and follow it;
> add example questions and vote up (down) questions that you think should be
> on-topic (resp. off-topic) for the upcoming site.
>
> http://area51.stackexchange.com/proposals/70511/sage?referrer=8adE6ec0VyqQCyB1kFgF7g2
>
> Looking forward to your support and cooperation for the success of this
> adventure on Stack Exchange.
>
> /p/s/ I would be soon adding my quota of 5 questions...

Does one really need an SE site for Sage? What's wrong with
http://ask.sagemath.org/ in your opinion then?

The only real problem with ask.sagemath, as Kannappan says, is the spammers.

Advantages: 

* Many people already have SE (e.g., mathoverflow) accounts, and these would be connected

* They handle spam etc. for you

* Could migrate questions on other SE products to this, or vice versa

Disadvantage (for some): 

* platform itself not open-source

Disadvantages:

* Not connected to previous rep mechanism for ask.sagemath

* Not connected to previous answers/questions (harder to work on dups, etc)

* Admin privileges etc are not up to Sage

* "If the site does not get used, it will be deleted."

* Do we own the data?  There is a CC license but is it exportable?

* Unclear and hard (for us) to reach standards of business.  

To me, the last one is the real problem.  For instance, at one time "15 questions per day on average is a healthy beta" but we definitely did not always hit that even in the most halcyon days of ask.sagemath.  I'll note that http://area51.stackexchange.com/faq seems to have gotten rid of any formal requirements, but there are definitely still standards to make it TO beta, much less out.  Compare the graduated versus not proposals, e.g. http://area51.stackexchange.com/?tab=launched  

Let's see what it says.  Remember, SE is a *business* that wants to drive traffic.  Not first and foremost to simply have quality Q&A sites.  Sage doesn't (yet), in my view, have enough traffic to drive this.  It could!  But it doesn't yet.

* 15 questions per day on average is a healthy beta, 5 questions or fewer per day needs some work. A healthy site generates lots of good content to make sure users keep coming back.

Probably we would need to increase our regular traffic (I mean compared to when we didn't have the anti-spam measures in place) by at least 50%.

* 2.5 answers per question is good, only 1 answer per question needs some work. On a healthy site, questions receive multiple answers and the best answer is voted to the top.

I would argue that with most ask.sagemath questions, we really don't need more than one answer per question.  But SE doesn't probably care about that.  And I will also point out that MANY of our questions are quite technical or don't ever get answers - see http://ask.sagemath.org/questions/?sort=answers-asc&page=6 (though to be fair, many of the answers to these are in comments, which SE would strongly discourage).

* We recommend:

150 users with 200+ rep

10 users with 2,000+ rep

5 users with 3,000+ rep

Even on the CURRENT ask.sagemath we have about 50 with 200+ rep.  We do have the requisite number of very-high-rep users, and are likely to continue to have.  This is mainly because we don't have the critical mass yet - in order to really rack up reputation, you either need to answer a LOT of questions, or be one of the few people who answers questions interesting to a lot of people in the (as yet, smallish) Sage community.  Most people who answer a question get only one or two up votes, even if the answer is correct.

* 90% answered is a healthy beta, 80% answered needs some work. In the beta it's especially important that when new visitors ask questions they usually get a good answer.

I believe I've already touched on this.  We are not currently going to be there.

* 1,500 visits per day is good, 500 visits per day needs some work. A great site benefits people outside the community. Eventually, 90% of a site's traffic should come from search engines.

I don't know anything about this.  But based on # of views, which is easy to see, it looks like we get between 20 and 50 views for a given question.  (Older ones get more, of course).  Let's say we get 50 views per 10 questions - that is 500 visits per day, assuming that's what they mean by visits.

On ALL metrics, Sage is not ready for a SE proposal to go anywhere, and that would be worse.  

Instead, all the people who just committed to the proposal should put on their thinking caps to get us a sensible anti-spam measure for our current site.

I am not saying this because I am a curmudgeon; I am saying this because I just spent the last hour reading through all the many complaints of proposal which aren't going to make it out of beta and will have to start from scratch, and I don't want that to happen to us.  It would only lead to confusion for those who actually want help.

- kcrisman

[kcrisman]

As I recall, the question of going with a SE site was discussed around the time that AskSage launched.  As I recall, the consensus was that a fully free QA site was more consistent with the values of the Sage community.  It may be that this is no longer the consensus (if it ever was).  Or the consensus of the subset who are ready to actually do something may have changed.

I think William mentioned in a previous post that he didn't see this as a barrier, though others may.  I would put myself weakly in the "may be a problem" camp, but primarily just think we aren't going to have a successful SE proposal.  Especially if other maintenance issues are any guide!

 

Maintenance of AskSage is definitely a problem, although not an insurmountable one.  The person who had been doing it silently stopped some time ago, and left things in a pretty bad state.  There are more responsible people who can take over (I could be one of them), but some back-end issuehave to be resolved first, and this is taking non-trivial time and energy.  

As you know, I am also happy to do this.  But it certainly takes a team.

 

There are certainly lots of good example questions on AskSage, so (if this is compatible with the licenses) a motivated person could copy them to the Area 51 site.

Hmm, that is an interesting question.  But how would that work with *all* the questions...

[kcrisman]

See also some very interesting discussions at https://www.biostars.org/

[Niles Johnson]

On Monday, June 2, 2014 9:31:30 AM UTC-4, kcrisman wrote:

See also some very interesting discussions at https://www.biostars.org/

I didn't know about this QA software -- here is the source repository (MIT license):

https://github.com/ialbert/biostar-central

Note that it's about 2 years old, and has a (small) handful of developers.

On Monday, June 2, 2014 9:07:16 AM UTC-4, kcrisman wrote:

Probably we would need to increase our regular traffic (I mean compared to when we didn't have the anti-spam measures in place) by at least 50%.

In case people didn't follow this parenthetical comment, I should explain that ask.sagemath.org currently requires some non-trivial reputation to ask or answer questions.  This was the simplest anti-spam measure we could implement until we upgrade to the latest version of AskBot (or other QA software).  In particular, it means that new users cannot ask or answer questions, and therefore the site usage is much lower than normal.

For the record, here's our rough roadmap for making AskSage maintainable again (in reverse order):

* Fork latest AskBot (or other QA software) and cap number of posts per day for low-reputation users.

* Optional: migrate existing questions, answers, and users to new AskBot install.  Or put current install in read-only mode.

* Before that, install latest AskBot

* Before that, establish new VM for ask.sagemath.org (latest AskBot relies on recent versions of python and mongodb, which we are unable or unwilling to install on the very old machine boxen)

Currently we are stalled at the very first step (establishing a new VM).  But there is reason to believe that regular maintenance will be straightforward after the (not-so-straightforward) task of updating our very old installation.

-Niles

[Volker Braun]

On Monday, June 2, 2014 3:15:13 PM UTC+1, Niles Johnson wrote:

On Monday, June 2, 2014 9:31:30 AM UTC-4, kcrisman wrote:

See also some very interesting discussions at https://www.biostars.org/

https://github.com/ialbert/biostar-central

This looks pretty good!

 

[kcrisman]

For the record, here's our rough roadmap for making AskSage maintainable again (in reverse order):

* Fork latest AskBot (or other QA software) and cap number of posts per day for low-reputation users.

* Optional: migrate existing questions, answers, and users to new AskBot install.  Or put current install in read-only mode.

* Before that, install latest AskBot

* Before that, establish new VM for ask.sagemath.org (latest AskBot relies on recent versions of python and mongodb, which we are unable or unwilling to install on the very old machine boxen)

Currently we are stalled at the very first step (establishing a new VM).  But there is reason to believe that regular maintenance will be straightforward after the (not-so-straightforward) task of updating our very old installation.

Thanks, Niles, that's really helpful as a summary of what's needed.   

[Kannappan Sampath]

I don't agree that we may not make a successful proposal. 

* Firstly, this site, if it ever becomes a reality, will subsume ask.sagemath.com alright, but not just that. This will help the entire community of GAP, PARI, Maxima, Singular, Sympy (probably also R, but cross-validated already handles quite a bit...) users... I have not written to their sites just yet but certainly this proposal is meant as a Q&A for computing in the Python ecosystem rather than just Sage. In light of this, I am willing to believe that conservative estimates of KCrisman would probably go up; probably more than the threshold too... 

So, in my opinion, instead of being too conservative about the possible failure, we could try and support the proposal and see if we could make headway into getting this site created.

-- Kannappan. 

--

[William Stein]

On Mon, Jun 2, 2014 at 8:02 AM, Kannappan Sampath <kntr...@gmail.com> wrote:
> I don't agree that we may not make a successful proposal.
>
> * Firstly, this site, if it ever becomes a reality, will subsume
> ask.sagemath.com alright, but not just that. This will help the entire
> community of GAP, PARI, Maxima, Singular, Sympy (probably also R, but
> cross-validated already handles quite a bit...) users... I have not written
> to their sites just yet but certainly this proposal is meant as a Q&A for
> computing in the Python ecosystem rather than just Sage. In light of this, I

In fact, as you suggest above, go one further: Python -->
"mathematical software"

Having such a site, which is like mathoverflow, but for open source
math software, sounds attractive (if such a thing does not already
exist).

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

[Niles Johnson]

On Monday, June 2, 2014 11:13:09 AM UTC-4, William wrote:

On Mon, Jun 2, 2014 at 8:02 AM, Kannappan Sampath <kntr...@gmail.com> wrote:
> I don't agree that we may not make a successful proposal.

I find Karl's assessment pretty convincing, but I propose that we move discussion of this topic to the Area 51 site, in the interests of giving the proposal its best shot. 

We can always move forward with the AskBot solution later if the proposal is not successful. 

[Kannappan Sampath]

On Mon, Jun 2, 2014 at 8:42 PM, William Stein <wst...@gmail.com> wrote:

On Mon, Jun 2, 2014 at 8:02 AM, Kannappan Sampath <kntr...@gmail.com> wrote:
> I don't agree that we may not make a successful proposal.
>
> * Firstly, this site, if it ever becomes a reality, will subsume
> ask.sagemath.com alright, but not just that. This will help the entire
> community of GAP, PARI, Maxima, Singular, Sympy (probably also R, but
> cross-validated already handles quite a bit...) users... I have not written
> to their sites just yet but certainly this proposal is meant as a Q&A for
> computing in the Python ecosystem rather than just Sage. In light of this, I

In fact, as you suggest above, go one further:   Python -->
"mathematical software"

Having such a site, which is like mathoverflow, but for open source
math software, sounds attractive (if such a thing does not already
exist).

At the moment, from my  understanding, the way any open source math software seems to work is that, if you're stuck on a point, you write to their mailing list... 

So having a central Q&A is certainly helpful and I am more or less certain, it does not exist... 

PS. I have seen a few questions about sage on math.se, stackoverflow and mathoverflow; so already, there are too many sites one has to follow or look into before being sure that a post to the mailing list is warranted. 

[kcrisman]

> * Firstly, this site, if it ever becomes a reality, will subsume
> ask.sagemath.com alright, but not just that. This will help the entire
> community of GAP, PARI, Maxima, Singular, Sympy (probably also R, but
> cross-validated already handles quite a bit...) users... I have not written
> to their sites just yet but certainly this proposal is meant as a Q&A for
> computing in the Python ecosystem rather than just Sage. In light of this, I

 

In fact, as you suggest above, go one further:   Python -->
"mathematical software"

Having such a site, which is like mathoverflow, but for open source
math software, sounds attractive (if such a thing does not already
exist).

This does sound interesting.  Is it targeted enough?  SE seems to want fairly well-scoped things - I don't know whether they would want something as broad as "open source math" or "Python math".

PS. I have seen a few questions about sage on math.se, stackoverflow and mathoverflow; so already, there are too many sites one has to follow or look into before being sure that a post to the mailing list is warranted. 

Yes, that is certainly true.  One annoying thing is that on SO the "sage" tag supposedly refers to us, but I've given up trying to remove the tag from questions about Sagepay accounting software...  I suppose in principle we could ask about migration.   That would be the ideal way to approach it, so as to keep things in one place.

Also, as a counter to my point about "staying OSS", both Ubuntu and Drupal host their question/answer on SE, one even with its own domain http://askubuntu.com/ - so this isn't a killer to other such communities.  But it should be raised, anyway.

 

[Jason Grout]

On 6/2/14, 11:12, William Stein wrote:
> In fact, as you suggest above, go one further: Python -->
> "mathematical software"
>
> Having such a site, which is like mathoverflow, but for open source
> math software, sounds attractive (if such a thing does not already
> exist).

Related: http://scicomp.stackexchange.com/ "Computational Science Stack
Exchange is a question and answer site for scientists using computers to
solve scientific problems."

It's seems more about scientific computing rather than more pure
mathematical software, though.

Jason

[kcrisman]

As another point of information, just one I forgot to mention earlier, after enough people follow it, it moves to the commitment phase - where new questions are not asked or answered, as I understand it (?)

The commitment score is the minimum of three scores:

% out of 200 committers (committing to do a certain nontrivial amount of activity in beta)

% out of 100 committers with 200+ rep on any other SE site (this I think will be easiest for us to fulfill!)

% out of 100 on commitment score [1], based on committers' activity on all other sites and how old the commitment is

[1] http://meta.stackexchange.com/questions/53650/area-51-commit-percent/53733#53733

[Kannappan Sampath]

On Mon, Jun 2, 2014 at 9:18 PM, kcrisman <kcri...@gmail.com> wrote:

> * Firstly, this site, if it ever becomes a reality, will subsume
> ask.sagemath.com alright, but not just that. This will help the entire
> community of GAP, PARI, Maxima, Singular, Sympy (probably also R, but
> cross-validated already handles quite a bit...) users... I have not written
> to their sites just yet but certainly this proposal is meant as a Q&A for
> computing in the Python ecosystem rather than just Sage. In light of this, I

 

In fact, as you suggest above, go one further:   Python -->
"mathematical software"

Having such a site, which is like mathoverflow, but for open source
math software, sounds attractive (if such a thing does not already
exist).

This does sound interesting.  Is it targeted enough?  SE seems to want fairly well-scoped things - I don't know whether they would want something as broad as "open source math" or "Python math".

Probably, this is not too broad given that computational science stackexchange site exists. I still think we might succeed if enough of us commit and actively request users to use the Stack Exchange site for well focussed questions (when it becomes available). Do you think I could edit the proposal to make it somewhat more broad in this sense. What would your suggestion be towards that, in terms of wording it? 

--Kannappan. 

  

PS. I have seen a few questions about sage on math.se, stackoverflow and mathoverflow; so already, there are too many sites one has to follow or look into before being sure that a post to the mailing list is warranted. 

Yes, that is certainly true.  One annoying thing is that on SO the "sage" tag supposedly refers to us, but I've given up trying to remove the tag from questions about Sagepay accounting software...  I suppose in principle we could ask about migration.   That would be the ideal way to approach it, so as to keep things in one place.

Also, as a counter to my point about "staying OSS", both Ubuntu and Drupal host their question/answer on SE, one even with its own domain http://askubuntu.com/ - so this isn't a killer to other such communities.  But it should be raised, anyway.

 

--

[Thierry]

On Mon, Jun 02, 2014 at 05:56:35AM -0700, Niles Johnson wrote:
> As I recall, the question of going with a SE site was discussed around
> the time that AskSage launched. As I recall, the consensus was that a
> fully free QA site was more consistent with the values of the Sage
> community.

+1

On 02/06/2014 15:12, kcrisman wrote:>
> As you know, I am also happy to do this. But it certainly takes a
> team.

+1

As for me, this is the main issue with the current state. We are not
enough collectively involved in maintaining the Sage infrastructure
(e.g. there is no reason that Harald has to do all the job he is
currently doing about web services, i always feel embarrassed to ask him
when i need some update).

I would like to add some arguments :
- we can administrate askbot the way we want. For example :
- there are plenty keywords that are synonyms and, if there was a team
involved in managing it, we can easily merge them to improve
searches. Actually a manage command is ready for that.
- many bugs on trac come from ask users. We can work on interfacing
them.
- if we have some idea (e.g. develop an antispam), it benefits to all
the askbot community
- the data are belong to us
- any argument on why you like Sage better than Mathematica can come
here :)

Here is the current situation about ask.sagemath.org behing the scenes
(or how i see it) : after the spam attack a few, William turned ask off
to save other services from being closed. Niles and Karl Dieter added
the karma threshold to block spammers (which also blocks newcommers, but
at least the website is back online). Also, i upgraded askbot to its
last version in a reprodictible way, which is not straightforward since
our version is very old and some bugs are not well documented, but quite
doable. It allows better spam management since you can now delete closed
posts. Also, it should be quite easy to develop a strong antispam for
askbot (given an bot/author, remove all its posts in a single command).
It is not online yet since we are waiting for a long-term VM to install
the fresh new version (the current one will die during the spring).

It could be a good idea to compare biostar and askbot (they both rely on
the same web framework (django)), and see if a migration can be done in
case this later is better for our needs.

Ciao,
Thierry

[Pablo Sánchez Ocal]

I think an important point to be raised is that having askSage on SE, the community could benefit enormously form other users coming from StackOverflow and MathOverflow. Yes, there have been local questions and answers on those forums about Sage, but having a permanent Q&A page would certainly be an advantage to both the number of users and the quality of their posts.

Now there is only a need for a few more questions to attain the 10 votes necessary to go to the commitment phase. I believe the quality is there, and it's only a matter of upvoting the better ones.

Pablo

[Kannappan Sampath]

Hi Pablo, 

Thanks for this post. I am hoping that the proposal will garner more enthusiasm; the point you mention is one of my original motives for the proposal… 

Regards, 

Kannappan. 

[Vincent Delecroix]

Hi,

I really hope that this project will not go. First of all it is very
confusing to have already two platforms for questions
(ask.sagemath.org and the sage-support googlegroup). So having a third
one is only even more confusing. I might be happy if the proposal was
to close ask and sage-support and open an SE platform. In that case
the first step of the project would be to transfer all questions
already asked.

Secondly, nobody addressed the issues I raised several months ago
about privacy and backups. Might be because nobody knows.

Best
Vincent

2014-09-21 18:25 UTC+02:00, Kannappan Sampath <kntr...@gmail.com>:

> Hi Pablo,
>
> Thanks for this post. I am hoping that the proposal will garner more
> enthusiasm; the point you mention is one of my original motives for the

> proposal...

[Samuel Lelievre]

And just to link with a similar discussion on ask-sage:

http://ask.sagemath.org/question/24175/please-join-sage-in-stack-exchange-to-make-it-a-formal-community/

[kcrisman]

the first step of the project would be to transfer all questions
already asked.

I have been informed in some communication with SX folks that they do not do this, as a matter of policy.  When they first started I guess they may have done so a few times, but definitely not now.  So one would have to keep ask.sagemath running in read-only mode or something even if this were done (or lose the questions and answers forever).

- kcrisman 

[Kjetil brinchmann Halvorsen]

I have been informed in some communication with SX folks that they do not do this, as a matter of policy.  When they first started I guess they may have done so a few times, but definitely not now.  So one would have to keep ask.sagemath running in read-only mode or something even if this were done (or lose the questions and answers forever).

Or somebody would have to do the transferral manually ... 

Kjetil

--

[William A Stein]

On Mon, Sep 22, 2014 at 7:49 AM, Kjetil brinchmann Halvorsen
<kjeti...@gmail.com> wrote:
> I have been informed in some communication with SX folks that they do not do
> this, as a matter of policy. When they first started I guess they may have
> done so a few times, but definitely not now. So one would have to keep
> ask.sagemath running in read-only mode or something even if this were done
> (or lose the questions and answers forever).
> Or somebody would have to do the transferral manually ...

The most important question -- by far -- is:

QUESTION: Is that allowed?

If we write a script that copies the thousands of questions from
ask.sagemath to SX, will they see us running it and ban us? Just
because they don't do the important themselves (which would be crazy
to expect them to do), doesn't mean we couldn't easily technically do
it from a database dump.

William

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

wst...@uw.edu

[kcrisman]

> I have been informed in some communication with SX folks that they do not do
> this, as a matter of policy.  When they first started I guess they may have
> done so a few times, but definitely not now.  So one would have to keep
> ask.sagemath running in read-only mode or something even if this were done
> (or lose the questions and answers forever).
> Or somebody would have to do the transferral manually ...

The most important question -- by far -- is:

      QUESTION: Is that allowed?

If we write a script that copies the thousands of questions from
ask.sagemath to SX, will they see us running it and ban us?    Just
because they don't do the important themselves (which would be crazy
to expect them to do), doesn't mean we couldn't easily technically do
it from a database dump.

I didn't ask that; perhaps not.  But will said script also copy all answers and comments (which often have quite useful information in them)?

Again, personally I am far from convinced that any sage.sx.com site would attract the needed activity level to move out of beta anyway.  We just don't have enough views and activity compared to what they are looking for (you can check Area 51 for what I mean).  

- kcrisman

[Robert Bradshaw]

Another option, given our (relatively) low volume, is to simply find
an broad-enough but related site (such as scicomp.stackexchange.com)
and start pointing user and answering questions there. By nature the
stack exchange format works better when one has critical mass, and
performs poorly if there are lots of little silos. I'd say it
answering questions about Sage (or computational pure math in general,
open source software or not) wouldn't decrease the value of that forum
to users who are only interested in the traditional numerical
computational sciences.

- Robert

[kcrisman]

Far be it from me not to point out relevant information, even if I don't have any interest in it.  The Sage StackExchange site is currently closed.

At 1+ year in Definition we have to close this proposal. Keep the proposal alive by restarting it. See discuss.area51.stackexchange.com/q/5798 – Robert Cartaino♦ Jun 8 at 18:36

Note two things: 

1) http://meta.stackexchange.com/questions/257614/graduation-site-closure-and-a-clearer-outlook-on-the-health-of-se-sites So if a restarted Sage site ever made it into beta, they wouldn't close it, but it almost certainly would never make it to 'graduation' ("When a site starts to consistently receive 10 questions/day, we’ll consider it for graduation.").  Of course, there was not enough activity to make it to beta, and I think it is pointless to try to restart it; others may disagree.

2) ask.sagemath seems to be humming along well right now.  Let's thank ESPECIALLY the team who made it possible to switch to the new Google auth and moved the servers for this to a good location!

- kcrisman