# Things pending to be done before we go public

+ 2 like - 0 dislike
6187 views

Here, please write the things pending to be done before we can go public. These include bug fixes, feature request completions, etc.

edited Mar 16, 2014

+ 2 like - 0 dislike

[DONE]

This can be done simply by saving this image as "favicon.ico" into the root directory of the website, according to GidGreen on the meta.Q2A. According to the discussion there, it will work online, but not offline, except on chrome where it works both online and offline.

answered Mar 2, 2014 by (1,975 points)
edited Mar 3, 2014

Note, that the favicon need not be scale invariant, since it will only appear at the "new tab" text (on chrome, firefox) and the URL bar (firefox, IE, opera, safari).

I have put the favicon there, but can't see it on my browser.

@polarkernel Hm, it seems that neither can I, through chromium, chrome, firefox, or pale moon...  Maybe you can just add it through the custom html in the head section allowed in the admin panel of Q2A?

Yay, it's working now!

+ 2 like - 0 dislike

[DONE]

The WWW Problem needs to be fixed.

If this is not possible, at least add a message after one logs out telling the user to check that he is logged out both on physicsoverflow.org and www.physicsoverflow.org.

answered Mar 2, 2014 by (1,975 points)
edited Mar 7, 2014

Our site hosting is using the aliases "physicsoverflow.org" and "www.physicsoverflow.org" to ease access (this is the standard configuration). It seems that Q2A treats them as two different sites. I have no idea on how to avoid that. Naturally I could switch off one of these, however, this would not really solve the issue of forget to logout.

An automatic logout after some time of inactivity would be the best solution, but I have no idea on how to implement that.

@polarkernel Is it possible that the momment the user logs out, it is taken to a simple page which reminds him to log out of both sites, physicsoverflow.org, and www.physicsoverflow.org. ?

@dimension10: I will see what I can do there.

Seems to be solved. I found the solution here. If you have time, can you test it once?

@polarkernel Yes, I tested it, and it works! Thank you!

+ 2 like - 0 dislike

[DONE for now]

All links need to be relative. It is pretty hard to keep this in mind (why, I just added an absolute link on the word "relative"), so this needs to be checked once before going online, that all internal links are relative.

Even more importantly, the buttons below all posts need to be made relative, i.e. those added with javascript, the VTC button, and the suggest edits button. (see comment by polarkernel below)

answered Mar 2, 2014 by (1,975 points)
edited Mar 9, 2014

The new buttons are not added with javascript but using the PhysicsOversnow theme (I like this name!). So they are built like the original Q2A buttons existing already before. The links there are relative.

@polarkernel Oh, I see! Thanks!

+ 2 like - 0 dislike

[DONE]

The reputation settings need to be changed so that

• [DONE] In "Q&A" and "Refreeing" , the upvotes and downvotes both gain and cost 10 rep each.
• [DONE] In "Meta", the upvotes and downvotes both gain and cost 0 rep each.

I don't think this is exactly disputed, right?

answered Mar 2, 2014 by (1,975 points)
edited Mar 20, 2014

No dispute on QA, but in refereeing, I am 100% positive that we need a way to make original work more valuable, by adding a multiplier, so that one has two upvotes for each refereed paper, originality, and quality. The paper's rep-gain then needs to be the product of the originality and the quality, and people should be able to vote both up/down separately and independently. This is pretty much the only way to produce a fair score.

A cheap way of doing this is to use the inverse-citation count as the originality metric. This doesn't work for contributions which are purely local to the site, however, but it is generally how originality is counted academically. But citation counting tends to sometimes hammer original authors who were hard to read, like Stueckelberg, or immensely fruitful original contributions, like BFSS, which were quickly overshadowed by a more easily extended high quality second work right afterwards (AdS/CFT).

The easiest technical way is to have TWO questions for each paper, one for originality and one for quality. Both link the paper, and the answers for originality discuss ignored predecessor work, and are free of physics content or rep-cost, these are just for "you forgot to cite this!", while the answers to the quality/accuracy score discuss the content of the paper.

You can combine the two into one page, of course, but you need to keep "you forgot to cite my paper!" separate from "You mae a mistake" or "you have an inappropriate application for this and this reason". Most referee comments are of the former type, and not of the latter type. Without two separate pages, all your referee section is going to look like "Cite me!!" "Me too!", "Yes, you need to cite me also!". That's what usually happens in journals.

@RonMaimon I agree. But how will that be done technically? .There needs to be a pretty advanced bot for that, how will that be made? .

This kind of change requires an intrusion into the core code of Q2A. However, I think this would be worth the trouble, because it will increase the quality of the site. It is important to distinguish two things. The first one is the computation of points using questions, answers and votes as a source. For this, any algorithm is feasible. Such changes, as for instance proposed for meta could be implemented even after goingonline. A "recount" startet by the admin would just correct the points. A great challenge is it to derive a number for the "value" of a paper as a combination of votes in different posts. This is not foreseen in any concept of Q2A. It would be not only one single place to change code, the changes would be spread all over the (core-)code.

+ 2 like - 0 dislike

[DONE]

The new theme PhysicsOversnow : ) will have to be completed. Remaining things include:

• [DONE] Bringing the answer form to the bottom of the page.   This can be done through a simple javascript: $("div.qa-part-a-list").after($("div.qa-part-a-form"));
• [DONE] Making  the navigation bar buttons #CCCCCC upon hover (ul.qa-nav-main-list li.qa-nav-main-item a:hover {background-color:#808080;})

• [DONE, but not through theme] Add a "review" button to the top-most bar and remove the tons of review buttons from the NavBar

• ???

Lastly, instead of falsley continuing to state "Snow theme by Q2A Market" at the footer, we could say something like "PhysicsOversnow? theme by PolarKernel based on Snow theme by Q2A market".

answered Mar 2, 2014 by (1,975 points)
edited Apr 18, 2014

The first two subjects are completed. Setting the answer to the bottom of the page has been realized according to the post here. We observe also the issue commented by Gary Black. Maybe we should set the setting in admin>viewing>show answer to "If no answers" instead of "Always"?

It is easily possible to remove completely the "Best Answer" functionality by overwriting the corresponding function in our PhysicsOversnow theme. Shall I do that?

- Select best answer is off now.

I have set it back to "Always" until I have a solution.

@dimension10 Background and border are removed now. For the moment I let the flags in the SE.TP posts stay in the database, just in case if the decision would be withdrawn. Then it would be complicated the reset them. The points for accepted answers are set to zero, so these flags have no influence.

@polarkernel Maybe, a solution would be to set it to always, but not display the "Answer" button? I doubt anybody would use the button, they'd just type the answer and submit it.

@polarkernel It seems that even though the idea of accepting answers is not there anymore, answers which have been accepted in the past (e.g. on TP.SE) are still getting the background and border, e.g. see here: ./index.php?qa=3681 ,  index.php?qa=653

+ 2 like - 0 dislike

RSS feeds from ArXiV need to dump themselves in our Refreeing section. We also need a way to get links to all the past arxiv papers in our refreeing section. I don't know how this can be made possible.

We may then also apply to ArXiV to get our trackbacks, listed.

We are interested in the following feeds:

What I mean is that all new physics papers on ArXiV, from the feeds above, should get two questions here in the refreeing section;

the first should be like:

(title) Originality: [PAPER TITLE]

the second should be like:

(title) Accuracy: [PAPER TITLE] ;

(body) [LINK TO PAPER] please vote on basis of accuracy only; review the accuracy of this paper in the answers.

answered Mar 2, 2014 by (1,975 points)
edited Mar 6, 2014

I have no idea on how to realize this. Could anyone define this request more in detail so that I can see what the technical requirements would be?

@polarkernel What I mean is that all new physics papers on ArXiV should get two questions here in the refreeing section;

the first should be like:

(title) Originality: [PAPER TITLE]

the second should be like:

(title) Accuracy: [PAPER TITLE] ;

(body) [LINK TO PAPER] please vote on basis of accuracy only; review the accuracy of this paper in the answers.

OK, got it now. Will be a little more than one evening next week ;-)

The question body should not have the content duplicatively over all papers, the question itself should be a link only. The section needs to be moderated to ensure on-topic answers.

Some technical isues:

-If these questions from the RSS feeds are imported automatically, some kind of dummy user has to be created, for instance ArXiV_Feeder or something like that. This is given by the concepts of Q2A. Because this user could get lots of points, it has to be hidden, or the points for questions in this category have to be set to zero.

- If these questions have to be editable (Rons proposition), the users require the right to edit questions of another user. Not all users will have this right.

Could eventually somebody write a set of two dummy questions in this category, representing the style you like to have? We could remove them before we go online. Like this, the task gets more clear for me.

@polarkernel I think the solution of hiding the user is better, because setting the points to 0 is not a good idea.

I will write a set of dummy questions tomorrow.

@polarkernel I have written the dummy questions: 62886290

I realised there would be one problem with importing the RSS feeds. Even though the RSS feeds would contain all new submissions, the existing papers on the ArXiV could not be imported through the feeds. Is there any other way of doing this without the RSS feeds?

+ 2 like - 0 dislike

Further Discussion

While I think Dimension 10 has done a good review on the Q/A, so perhaps it can go live immediately, I am worried about the refereeing section. Here, you need a robot to link arxiv papers automatically to questions, allow hand-insertion of non-arxiv literature, and, most importantly, I firmly believe that you need separate voting on originality of content, and on general accuracy/quality separately.

The two evaluators need to be completely separate for good review. The originality is extremely important, in that, for instance, 't Hooft's 80s models of holography are not 100% accurate (because he is analyzing thermal black holes in 4d, not string theory objects), but they are extremely original and contain the seed of the main idea of holography very clearly. The mistakes in the papers are of the form of crazy semi-string actions with wrong sign, and other weirdness, caused by having a slightly off picture. They would have a high originality score, but a relatively low accuracy score, although it should still be positive, because the calculations are correct given the assumptions.

I suspect that the easiest way to do it in Q/A is to have two linked questions, with citations and originality discussed on one, and with accuracy discussed on the other, and where the upvote score is equal to the originality score times the accuracy (or, perhaps exp(originality/10) times accuracy, to avoid sign problems, or perhaps exp(originality/10|originality|^.8) to avoid exponentially exploding reputations for paper-authors.

Having two questions per paper allows the citation issues to be sorted out on one page, and the accuracy issues on another. It is important not to mix the two up, because the main comments you will get on papers will be nearly entirely devoid of physics content, and will be "cite me" answers by people who feel their work has been unfairly neglected. This sentiment can be expressed in the "originality" section.

answered Mar 2, 2014 by (7,720 points)
edited Mar 16, 2014

Because you don't evaluate 16 different things when you submit science to a journal, you evaluate whether the contribution is new, and whether it is correct, and that's it (in principle). The problem with no originality vote is that plagiarism becomes possible, it is not possible to avoid this criterion, and UNLIKE the other things you state, it is dead easy to identify originality, you just ask "what's the closest previous work?" and then look to see if the current work is extending it, plagiarizing it, or completely different from it. It's a judgement call, but it's an easy one.

For instance: you need to convince me that it's worth splitting the discussion of each paper into two separate pieces. It might make sense to discuss the originality and correctness in separate threads, but what if we want to actually discuss the physics? This to me is the interesting part.

Fine, but why is it useful to have a coarse "originality" and "correctness" score? Why not just upvote papers which are interesting and everybody should see?

Discussions on the physics without any comment on accuracy (either explaining something thoroughly, to make it more obvious that it is likely correct, or showing something is wrong) is fluff, and can go in the body of the question itself for the "accuracy" question regarding the paper. The body of the question can be a full summary and reworking of the paper.

@RyanThorngren Oh, ok, I get what you mean by the "physics" now. I think I am sort of neutral about this now, I agree that it is problematic to have three questions about accuracy, originality, and explaining the physics, but I also agree with Ron that having "Cite this: ... " all over the place is annoying. I will now post a new post for a proper discussion, with voting, as to whether we should divide the originality, accuracy, and explanation, or keep it as one.

Yesterday I thought that maybe we should delay the decision until a larger part of the physics community could take part in these important discussions ...

The drawback is, that people will probably start using the Reviews section too as soon as we go public, and it might be hard to change the voting system later when many reviews already exist ...

I am basically agnostic too, I understand both Ron and Ryan's points of view.

+ 2 like - 0 dislike

We will need a second system developer.
The small team actually active at this site is very productive. Excellent propositions and an exciting business model have been developed. I am sure that these concepts will be very successful and productive for the scientific community of physicists.

However, when we started this project last year, we talked about a revival of the closed TP site. Looking at the numbers that have been achieved by this old community, I have been convinced that we can do that without greater problems. Therefore (and because nobody else did it) I said yes to set up this site.

The actual requirements, if they have to be realized in a valuable time, need more than just an evening and weekend job. Also the experience with you test-users showed me that the concept of regaining old TP accounts using the old credentials does not really work. Most users did not remember them and required a manual intervention into the database by me to get them in. I also expect to have a lot of multiple accounts of the same user, to be merged by me. This wiil lead to a bunch of work as system manager, once we open the site.

Therefore I ask you: Do you know any other person with an appropriate experience, able and willing to help me? Q2A is built very modular, it would be easy to split the task into subprojects that can be solved independently by different developers.

answered Mar 2, 2014 by (0 points)

I will ask my software developper friend (who is personally interested in general relativity a bit) I learnt to know recently in Graal Müritz by mail ... :-)

@Dilaton Did you ask him yet? Has he given a response?

Yes, I have written him a mail and if he fails to respond timely, I can give him a phone call to be a nasty pain in the neck ... ;-)

@Dilaton Any response?

Nah .. seams to be high time for a naaasty phone call now ;-)

@Dilaton Have you called him yet?

+ 1 like - 0 dislike

[DONE]

There are still some plug-ins remaining to be installed, e.g.

answered Mar 3, 2014 by (1,975 points)
edited Mar 16, 2014

@polarkernel No I meant the Q2A history plugin. It is a plugin that adds activity and points history (more advanced than the existing one) to a user's profile.

@dimension10 Vote to close button fixed.

- Image Uploader: This is a markdown editor (without LaTeX). Unfortunately it is not possible to use two editors in parallel. I don't know if this works on all browsers, but using Firefox, I'm able to copy-paste images into the editor directly.
- Q2A tag tools: Will come a bit later. I will let you do the setup of this plugin.
- Q2A history: Installed and running.
- Prevent Edit Conflicts: Installed and running.

Well, I'm able to select an image on my local drive, click copy (ctrl c) and paste it in the editor (ctrl v) , like this:

The image is then stored within the post as <img>. This works as long as the size of total post is not longer than 20000 bytes. But for small images...

- Tagging Tools: Installed. You may configure it using Admin > Plugins and selecting "options". Please observe point 5 of the installation instructions.

@polarkernel Thanks. I guess not being able to upload images directly is OK, at least for now, though I the copying and pasting images directly is fine only when the image is online (and not offline), of course...

@polarkernel Thank you, I will do the configuration soon.

+ 1 like - 0 dislike

SE questions listed [DONE] here and here need to be imported,

answered Mar 3, 2014 by (1,975 points)
edited Mar 24, 2014

Yes, Polarkernel said we can now do this. At present we have to take (the ones that are still there) them dirctly from Physics SE, as it is not yet possible to import them from my offline site.  Maybe we should then also edit out the stupid wrongly attached (!) homework tag, remove the book banner, etc. I mean people here are grown-up enough to know how to answer a reference / books question by themself ... ;-)

I will start with the endangered ones today and leave a note on the blog about those that can no longer be found.

@Dilaton Thanks, but haven't the endangered ones already been saved on the test site? So I thought it wouldn't be hard to mass-migrate them here?

Yes, they have to be migrated one after the other with the plugin from the live site, or ( if the rascals have deleted them) reposted from the offline site ...

The import of these questions has been done onto Dilatons local database. It is not straightforward to import them into PO's database because of numerous relations between users, posts and votes.It was thought as an emergeny solution, with the idea to import as much as possible from the living site. I am sorry, but I have to put the rest of this import to a low priority.

Yay! The posts listed in the first blog post are all imported! Now have to focus on the second one, which is a lot bigger...

 Please use answers only to (at least partly) answer questions. To comment, discuss, or ask for clarification, leave a comment instead. To mask links under text, please type your text, highlight it, and click the "link" button. You can then enter your link URL. Please consult the FAQ for as to how to format your post. This is the answer box; if you want to write a comment instead, please use the 'add comment' button. Live preview (may slow down editor)   Preview Your name to display (optional): Email me at this address if my answer is selected or commented on: Privacy: Your email address will only be used for sending these notifications. Anti-spam verification: If you are a human please identify the position of the character covered by the symbol $\varnothing$ in the following word:p$\varnothing$ysicsOverflowThen drag the red bullet below over the corresponding character of our banner. When you drop it there, the bullet changes to green (on slow internet connections after a few seconds). To avoid this verification in future, please log in or register.