up vote 37 down vote favorite
6

The mathematica proposal at 89 commitments has disappeared without warning or trace.

Why has it disappeared? And why without prior warning?

share
It was closed because it was thought to duplicate the efforts of SO itself. While true, in a narrow sense, I must cite TeX - LaTeX as a counter example. A good 50%, or more, of the answers on there require programming. So, why then was it allowed to split off, while Mathematica, which suffers from a number of the same problems, is closed? – rcollyer Dec 13 '11 at 17:39
3  
@rcollyer In my opinion the difference between programs like TeX and Mathematica and programming languages is that the users of TeX and Mathematica do not see themselves as programmers and mostly do not have a programming background, so they neither think of stackoverflow as a place to post their question, nor are the answers there especially taylored to normal users. There is also the issue that if a subject is split between math.SE, SO etc, it is not clear that you know where your answer belongs if you just know that something does not work. – Phira Dec 13 '11 at 19:04
2  
Those are my thoughts, exactly. Combined with the treatment of Mathematica questions by the rest of SO: questions of validity on whether they belong there at all, close votes by those who don't know, and similarly downvotes, it is in the same boat as LaTeX and TeX were on SO. – rcollyer Dec 13 '11 at 19:24
Very disappointed. SO management has had several months to make this decision. Meanwhile commitment numbers had been building steadily. – Verbeia Dec 13 '11 at 20:56
2  
I would also point out that the Windows Phone proposals (with fewer committers) has not been closed, but has the same properties of combining some content that could belong on SO and some that could belong on SU. – Verbeia Dec 13 '11 at 21:04
1  
Per the discussion below with Robert Cartaino, we've opened a new proposal and could use the support of any and all who committed/followed the previous proposal. So, any contacts you may have with some of them would be useful in this regard. – rcollyer Dec 17 '11 at 2:58
1  
Would someone please edit the question here to include whatever is necessary for it to show up on the related discussions lists for both the old and new mathematica proposals? – Isaac Dec 17 '11 at 10:46
3  
Please note: barely 36 hours after rcollyer re-launched it, the new Mathematica proposal has 9 questions over the line out of 10, and 38 followers, and that over a weekend just before Christmas. We have given ourselves the hardest possible time to rally support and the proposal is still the "hottest" proposal on Area 51. This rather weakens the "lack of support" argument. – Verbeia Dec 18 '11 at 9:17

4 Answers

up vote 54 down vote

A case for the Mathematica proposal

I find Robert Cartaino's answer highly unsatisfactory and epitomizes the typical "meh" responses that are doled out on Area 51. It's unfair to criticize proposals for languishing on Area 51, when the voting process was clunky to begin with. Mathematica was one of the proposals that went into commitment soon after the new voting rules were introduced, which suggests that the process might've had something to do with it stagnating. Since then it has racked up 92 commitments in about 2-3 months, which isn't that bad at all. Last I checked, there wasn't a must have X commitment by Y months, else it'll be closed rule.

I couldn't care less which other proposals are launched/canned and I'm not even going to try and compare with them. To me, they're as useless as hamburger flavored perfume. I'm going to try and make a case for the proposal, staying as objective as I can.

Most, if not all, Mathematica tag users are NOT programmers first:

This point has been made by others, but it's worth reiterating. A quick glance at the backgrounds of users active in the tag reveals a broad list in science/engineering including physics, mathematics, economics, material science, cognitive science, etc. I doubt if any of them would even consider themselves hard core programmers first. By restricting users to Stack Overflow and keeping questions strictly programming oriented, you're losing out on a lot of interesting questions that require a bit of domain related expertise and the users on SO have shown that they're willing to learn a new domain to answer a question if it's on-topic and interesting.

Let me try to make my point clear with this Venn diagram. We have 4 sets - Programming, Mathematics, Science/Engineering and Mathematica. Going by the official FAQ of StackOverflow, the only questions that are allowed are those programming questions that intersect with possible Mathematica questions and possible mathematica questions involving a bit of math (and by this, I mean floating point/precision, etc. related math, and not solving PDEs).

enter image description here

Some example questions that are representative of this set are:

We like science/math based questions with some programming

However, the Mathematica community has been quite welcoming of questions that stray outside the typical Stack Overflow question, or in other words, there is some programming, but it comes in only tangentially, as a means to achieving the end goal. I don't think I would be exaggerating if I said a good chunk of our questions belong to the blue set.

enter image description here

Some examples that come to mind:

However, these are exactly the kind of questions that have attracted drive by users who vote to close it thinking it is NaRQ or that it belongs on Math.SE. It gets even worse when the question requires a lot more domain expertise than the average programmer has, as can be seen in Listing all interesting sections of a tetrahedron, especially this answer. It is an utterly unhelpful answer, and the user is near hostile in the comments. To quote a couple of comments:

So the problem is not a programming problem. This is a mathematics problem. Or are you looking for help generalizing the code to compute this? We can help you produce compilable code to test it, work it out. What we can't do is help you reduce the problem to a fundamental set. That's the realm of the mathoverflow people. How does this problem remaining open help you? How can SO help you here? That's all I'm saying. The original question is some months old and MO helped you find a good 7d simplex case. So.... – jcolebrand Nov 3 '10 at 21:03

So what is the problem? You have code, you have a specific solution. You need to reduce it to the general solution? That's a math problem. You need help parallelizing it? That we can help with, but you have to formulate the question. – jcolebrand Nov 3 '10 at 21:29

A poor, misunderstood beast

I have downvoted the answer above, but that really captures the essence of how Stack Overflow users misunderstand the nature of Mathematica questions or the capabilities of Mathematica. Most users might have encountered a computer algebra system (CAS) such as Mathematica or Maple to "magically" find the roots of a polynomial and have a notion set in stone, that only calculus TAs use Mathematica.

We'd really like to see a lot more of these interesting questions, but in a place where it's explicitly on-topic. Not circumventing the masses by virtue of having a few 3k/10k users. Given the background of most Mathematica users, I'd say we'd all enjoy and envision a site where questions in science/math/programming that can all be solved by Mathematica can be asked. Something like:

enter image description here

You claim that Mathematica is also covered by Super User, Stats, Mathematics, Physics, Code Review, etc. Have you see what percentage of users active in Stack Overflow are active in any of those sites? I know that a lot of users prefer using comp.sys.math.mathematica for such questions precisely because a lot of us are not interested in following a bunch of sites for the sole purpose of looking out for such questions. Shouldn't a subject that can be asked on 5-10 different sites have a site of its own? That would seem to me to be the logical way to go about it.

But wait... there's more!

There's a whole lot more that Mathematica can do, especially with graphics, plotting, data manipulation, prototyping and testing complex models rapidly, even creating presentations like Powerpoint/Keynote, all of which don't really fit any one set in particular. There're plenty of questions that could be asked about creating demonstrations or using their computable document format (CDF) or even querying Wolfram|Alpha properly and interfacing with Mathematica (although this is a lot simplified now with v8). Some examples:

What we're really gunning for, is a site like this, where anything about Mathematica is on-topic.

enter image description here

Hey, where are the experts?

Among the users in the Mathematica tag who have a bronze badge, I know of at least 6 employees of Wolfram Research (including a few on the Kernel development team), at least one expert user who has written a very popular book on the subject and a whole bunch of expert users who have been using Mathematica in some way or the other for 10-15+ years and have shared information on their blogs/websites/other fora for years. If these users don't constitute a pool of "expert" users, I don't know what will satisfy you (Robert).

To put the closure of this proposal in C# terms so that you better understand our stance, it's as if Anders Hejlsberg, Eric Lippert, Jon Skeet, Marc Gravell and a whole bunch of experts committed to a proposal and you slam it shut because you think it's a flaky proposal.

Granted, we're not as big as C# and our users don't have 100k+ reps and we won't get 10k+ questions/month, ever. If your goal is to mint money off of Google Ad sense, well, then we're not the right proposal to pick. You were wise to go with Sex.SE in that case (you're probably not making money off of our tag on Stack Overflow anyway). If your goal is truly, as is often claimed, to make the internet a better place, I must protest the closure of this proposal and urge you to reconsider your decision.

share
6  
that sums it up very well, and I agree completely with everything. – acl Dec 15 '11 at 1:21
10  
+1 yoda, this is awesome. The point about non-programmer Mathematica users is very important. I am an economist. I am most certainly NOT a programmer, and there are many others like me who would benefit from not having to monitor five different SE sites and not having to deal with misguided close voters. – Verbeia Dec 15 '11 at 4:07
4  
Well put yoda. I'm still an undergraduate pursuing a dual major in Physics / Computer Science, but my research in Physics heavily relies on Mathematica. There are plenty of questions I'd love to ask that would fit best on a Mathematica SE site. I know a number of my Physics professors would love such a site, and none of them are expert programmers but have used it extensively for years. It's unfortunate because I only heard about this after it was brought up in the mma chat that the proposal was closed. – Mike Bantegui Dec 15 '11 at 4:23
2  
"Hear, hear!" & "Amen,brother" (both a few characters too short ;-) – Sjoerd C. de Vries Dec 15 '11 at 23:52
3  
I doubt you'll get a response. There are 15 comments on his "answer," and yet only silence. I understand that they may have outside employment, lives, and other duties, but it seems to me that the community moderators chief MO is hit and run and then silence. Getting a response is like trying to pull teeth. – rcollyer Dec 16 '11 at 3:19
3  
The solution: Rally your support base and relaunch a proposal with robust support and example questions that show off why the site should exist. Use the discussion area to organize your efforts. I suspect that a new proposal should be able to avoid the troubling problems described above. If that turns out to be the case, we will happily support your proposal and the launched site. Good luck! – Robert Cartaino♦ Dec 16 '11 at 15:02
4  
@Robert seriously it is very strange to see you say relaunch a new proposal, what would be any different the next time? Why not reopen that old proposal. So what if it takes 2,3-5 years, when it does start you will have very dedicated users. Mathematica users are very professional and loyal and spread on other sites on the internet. They may take time to see the advantage of a vote-based site and commit to another site. Be patient. It hardly costs you anything to keep the proposal open. – dbjohn Dec 16 '11 at 18:18
5  
@Robert I'd like to point out thought that most other commitments (barring the really popular ones) have flat lined at some point, and were resuscitated when it crossed the 100-130 user mark. This just simple game theory. A lot of people are interested in more than one topic, yet they have only 3 commitments they can use. This is fair, you don't want people to randomly commit. However, this also makes them withhold their commitment until they're sure that a proposal has a chance of getting into beta. – Lorem Ipsum Dec 16 '11 at 18:18
2  
Besides, the definition phase has only caused more trouble than it's worth with scope changes, etc. being incredibly hard in beta. It's better that a site has a definition and a few dedicated committers and is launched from the get go (private beta) and things are hashed out in meta by these folks. This provides opportunity for some solid framework for the site from people who're actually interested in the site, not ones looking for some free Area51 rep! Right now, proposals just stagnate here and questions that should've been asked in the def phase only get asked in beta, which is too late. – Lorem Ipsum Dec 16 '11 at 18:22
9  
@Robert, hyperbole aside. As a community we find this extremely frustrating. In particular, it is the lack of communication on the SE team's part with regards to this. First, there was no attempt to engage us prior to the shutdown regarding these issues. No discussion, no chat. Second, the explanations given seemed inconsistent with observable facts, and raise numerous questions of their own. Third, there has been little attempt to address our concerns after the fact, and the "updates" are sporadic at best. Taken together, the SE team seems capricious and fickle in this regard. – rcollyer Dec 16 '11 at 20:11
3  
@RobertCartaino I am sorry if people's anger discourages you from doing the right thing. I do think you need to acknowledge that the concerns about overlap were not only discussed in Def phase but addressed in that phase. People asked, and their concerns were assuaged. I also think you need to acknowledge that closing the Mathematica proposal was inconsistent with other decisions. The "hallmarks" were not really there, but appeared to be. Re-opening would be an act of magnanimity that would be very popular. – Verbeia Dec 16 '11 at 20:52
2  
@Verbeia, I may have jumped the gun, then. I opened a new proposal. It allows for a clean slate, wherein we can get through the early stages fast. – rcollyer Dec 16 '11 at 20:57
2  
@RobertCartaino "The solution: Rally your support base and relaunch a proposal with robust support and example questions that show off why the site should exist." So you are saying: Just start it again and when it is close to the 100 commitment mark again, I will decide whether it is closed according to my secret criteria of "robust support" and good "example questions" that are different to the criteria for any other site? Since people who committed were not informed by emails of the closure, it is not a sign of indifference if they do not follow the new proposal. – Phira Dec 17 '11 at 10:59
3  
@Phira I understand your frustration. But after 15 hours, we already have 23 followers and 4 questions over the 10+ line. The most effective way to show our disagreement with the decision is to rebuild the proposal even stronger, and faster. – Verbeia Dec 17 '11 at 12:22
2  
I can only take so much bull... and I don't see how someone who clearly didn't have the expertise to make a "judgment call", 'cause well... you have to know at least a little something before judging, no? But I guess I'm too green about these things. – user17323 Dec 18 '11 at 7:24
show 15 more comments
up vote 19 down vote

I hypothesise that the proposal has been closed for the same reason that it was opened: non-Mathematica users just don't get it.

There have been many cases of Mathematica questions on SO that have attracted erroneous close votes from non-Mathematica users who think such questions belong on math.SE. There are many Mathematica issues that are not, strictly speaking, programming issues and actually don't belong on SO. We've gotten around that SO by being a bit liberal in our definition of "on-topic" there, compared with other tags.

I also note that the "lack of support" argument made in Robert Cartaino's answer is not consistent with the reason cited in the close notice: "This proposal would tend to drain audience from an existing Stack Exchange site." Either the proposal doesn't have enough support, or it will drain audience. It can't be both.

As noted in comments, closing this proposal and leaving Windows Phone open is inconsistent, especially considering the larger absolute user base of Windows Phone.

share
up vote 6 down vote

An update: The Mathematica proposal has not disappeared, it has been reborn!

We went from proposal to commitment phase in less than four days. We got to 100% within 23 days of the first commitment, so only 27 days after the proposal started.

Maybe it took the slap on the face of being closed to rally the user base. Certainly it has bolstered our enthusiasm to create a great site for programmer and non-programmer users of Mathematica alike.

We still think the reasons given for closing the original proposal weren't correct (or at least have been inconsistently applied), but we have responded constructively and come out of it stronger.

Look forward to seeing the site on a StackExchange network near you very soon.

share
1  
And just a reminder - we did this over Christmas when many people are not around. – Verbeia Jan 13 '12 at 5:31
up vote -14 down vote

It hasn't disappeared. It is right here: Mathematica [closed], with the announcement issued from the resulting discussion…

Mathematica question do indeed seem to be covered by our other Stack Exchange sites, among many others: Stack Overflow, Math, Stats, Super User. That, plus the 17-month trek it took to get the site this far has shown us that, at this point, this proposal would not produce a viable Q&A site. – Robert Cartaino♦ 20 hours ago

There may be questions in Mathematica that are not covered by our current sites, but that cuts down the potential scope considerably. Factor in the lackluster support for this proposal, and the proposal was simply not viable in its current state.

We may one day have the audience to create more one-product, spin-off sites like this. But Mathematica has slowly and painfully just barely made it this far and has months to maybe another year to go. It's hard enough to create a viable site when you have an avid, enthusiastic audience, but this kind of lingering support will not make a working site. It will launch to an equally lack-luster opening and will hardly be used because because it has no momentum and few avid supporters.

Area 51 has its share of challenges in attracting a larger audience, but that doesn't change the viability of this proposal as it stands today. It needs a stronger showing at this stage or the site will simply fail. We've tried launching proposals from a similar state and it's a painful pattern to perpetuate.

share
9  
To explain the title: I didn't find it on the "progress" page and then looked in vain in my own account under "committed". I note that you did not explain why committers are not informed. I further note that the proposal was around number 15 in the progress list behind promising proposals like "Human Sexual Response and Non-Platonic Relations" that have "painfully lingered" for longer than 17 months. BTW, Italian has the exact same statistics as mathematica. It is your site and you can do what you want, but then please stop to pretend that there is a transparent process here. – Phira Dec 13 '11 at 21:24
7  
I must disagree with the reasoning for its closure. It has only been in commitment since September. Of the commitments that I've been associated with, they all took longer than 3 months in commitment to make it to beta. Additionally, on every site where there is a Mathematica question, they are often treated as not belonging, just like TeX was. So, what is needed from us to have this reopened? – rcollyer Dec 13 '11 at 21:42
1  
@Phira There isn't a Stack Exchange site for Italian language, so far. What Cartaino reports in that comment would not be true for the Italian language proposal, as questions about Italian language cannot be asked on three existing SE sites. – kiamlaluno Dec 14 '11 at 4:20
3  
@kiamlaluno, the point Phira was trying to make was that the area51 proposal for Italian has similar statistics as to Mathematica, yet it was not closed. On closer inspection, Italian had its first commit 7 months ago, and have just made it to 89 committers. However, Mathematica had 92 after only 2 months. Yet, it was the one that was closed ... – rcollyer Dec 14 '11 at 5:36
3  
@rcollyer The reason reported by Cartaino about closing the proposal is that questions about the topic can be asked on three different SE sites ([1]), and the statistics about the proposal don't make hope for the success of the proposal ([2]). For the Italian language proposal [1] is not true, which means that proposal is not in the same conditions of the closed proposal. – kiamlaluno Dec 14 '11 at 6:10
4  
@kiamlaluno - "being able to be asked on three sites" also applies to TeX/LaTeX, and is one of the reasons why a consolidated site is more likely to succeed than people might imagine. Fragmentation splits energy across sites; having a single SE site dedicated to Mathematica would concentrate focus and attract more users. We have gotten partway around this problem by asking most of the questions on SO not SU, and taking a more liberal view of what is on-topic. But a dedicated site would be a better solution, which is why we supported the proposal. – Verbeia Dec 14 '11 at 6:54
@Verbeia I am referring to the fact Phira is using the Italian language proposal as example, but that proposal is not in the same conditions of the closed proposal. I don't think that closing a proposal because the questions could be asked on three different sites does always make sense. I think that the problem is finding the right point between allowing too similar proposals, and forcing users to split their questions to make it suit for the existing sites. – kiamlaluno Dec 14 '11 at 7:40
2  
@kiamlaluno The reason I addressed the statistics is that the post of Cartaino is full of emotionally charged language like "lackluster", "barely", "painfully lingered" while the progress in commitment was none of these compared to other proposals. I addressed the three-way split argument in my comment to the OP. People would not ask questions on stackexchange about programming if they had to ask questions about syntax errors on a site that was mainly about orthography while other errors were off-topic. – Phira Dec 14 '11 at 8:38
@Phira You used the Italian language proposal as example, when there isn't any SE site about that topic; Cartaino's answer is saying the proposal was closed because the questions could be asked in other sites, AND the bad statistics the proposal had. I am not saying that the fact questions could be asked in three different sites should be the reason of closing a proposal (at least, that is not a reason that can be blindly adopted); I am saying that using the Italian language proposal as example is not a good example. – kiamlaluno Dec 14 '11 at 8:53
4  
As other people have pointed out, many (most?) users of Mathematica don't think of themselves as programmers, so they won't consider SO as a place to ask (or would even consider the mood there, across all tags, somewhat unfriendly). I believe a dedicated site would draw many more people. Splitting the questions between 4+ sites (SO, SU, math, stats, sci computing, etc.) is not friendly to users, and while it is possible to follow a tag on all SE sites, the plain fact is: even most regular answerers simply won't. Closing this request doesn't seem well founded to me. – Szabolcs Dec 14 '11 at 14:21
2  
@RobertCartaino I've written an answer, which makes a case for reopening the proposal. Please leave a comment on it if you can. – Lorem Ipsum Dec 15 '11 at 3:03
4  
The proposed site will cannibalize from SO? Really? If anything I think SO's too big that it is likely that there are Mathematica questions in there languishing for want of a capable answerer. I have an SO account, but have decided not to participate actively, as the pace there is rather too frenetic for my taste. It is quite likely then that there are questions there that I could have cleanly answered but haven't. Similarly, there are Mathematica questions in math.SE that probably could have been short work for an SO habitué... – user17323 Dec 15 '11 at 6:06
2  
@kiamlaluno, on my cursory inspection, I did not find one. (So, we must resort to tangentially similar proposals.) Presumably neither [A] nor [B] were sufficient, by themselves, to close this proposal, so showing that either reason is flawed should be sufficient to collapse the argument entirely. Every proposal on the first page progress tab has been in commitment much, much longer than Mathematica, implying that our progress was not slow. Also, TeX - LaTeX reveals the lie in not being allowed to split off language specific sites. Both [A] and [B] fail on their own. – rcollyer Dec 16 '11 at 3:06
6  
@Robert Cartaino I agree with most other comments on complete inadequacy of SE move regarding Mathematica proposal. But let me tell you something else. This is a strategic mistake for SE. Right now very few people on this planet do fully realize what Mathematica really is and what potential does it have. Mathematica is a powerful technology of the future, that will last. And will gain enormous momentum, perhaps soon. SE can host the primary Q/A site devoted to it. Or it will be some other site, in which case the only losing party at the end will be SE, not Mathematica makers or users. – Leonid Shifrin Dec 16 '11 at 22:06
3  
@RobertCartaino I have just seen your answer on the discussion on whether to close the Wikis proposal where you said that lack of interest wasn't a reason to close a proposal because inactive sites get deleted automatically. I would be interested to know what a "tactical or policy-driven reason" to close might be, other than overlap with other SE sites. – Verbeia Dec 18 '11 at 12:33
show 4 more comments

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged