Not signed in (Sign In)

Not signed in

Want to take part in these discussions? Sign in if you have an account, or apply for one below

  • Sign in using OpenID

Site Tag Cloud

2-category 2-category-theory abelian-categories adjoint algebra algebraic algebraic-geometry algebraic-topology analysis analytic-geometry arithmetic arithmetic-geometry book bundles calculus categorical categories category category-theory chern-weil-theory cohesion cohesive-homotopy-type-theory cohomology colimits combinatorics comma complex complex-geometry computable-mathematics computer-science constructive cosmology deformation-theory descent diagrams differential differential-cohomology differential-equations differential-geometry digraphs duality elliptic-cohomology enriched fibration finite foundation foundations functional-analysis functor gauge-theory gebra geometric-quantization geometry graph graphs gravity grothendieck group group-theory harmonic-analysis higher higher-algebra higher-category-theory higher-differential-geometry higher-geometry higher-lie-theory higher-topos-theory homological homological-algebra homotopy homotopy-theory homotopy-type-theory index-theory integration integration-theory k-theory lie-theory limits linear linear-algebra locale localization logic mathematics measure-theory modal modal-logic model model-category-theory monad monads monoidal monoidal-category-theory morphism motives motivic-cohomology nlab noncommutative noncommutative-geometry number-theory of operads operator operator-algebra order-theory pages pasting philosophy physics pro-object probability probability-theory quantization quantum quantum-field quantum-field-theory quantum-mechanics quantum-physics quantum-theory question representation representation-theory riemannian-geometry scheme schemes set set-theory sheaf simplicial space spin-geometry stable-homotopy-theory stack string string-theory superalgebra supergeometry svg symplectic-geometry synthetic-differential-geometry terminology theory topology topos topos-theory tqft type type-theory universal variational-calculus

Vanilla 1.1.10 is a product of Lussumo. More Information: Documentation, Community Support.

Welcome to nForum
If you want to take part in these discussions either sign in now (if you have an account), apply for one now (if you don't).
    • CommentRowNumber1.
    • CommentAuthorAndrew Stacey
    • CommentTimeMar 9th 2010

    One spin-off of the SVG-Editor is that the nlab is now running a public itex server. That is, it's possible to send the nLab a bit of itex and get sent back the MathML.

    One thing that's always annoyed me about the nForum is that we use a different syntax here than at the nLab, making it difficult to do cut-and-paste and having to remember where you are.

    Well, no longer! Or at least, hopefully this will soon no longer be the case.

    I've had a go at converting the nForum to use MathML. Now I'd like some beta testers to have a go at breaking it. The guiding principles were:

    1. Everything up to now should work as it has done.
    2. As a corollary, if someone doesn't want to use the new ability to do MathML then they shouldn't have to.
    3. But it should be possible to write something on the nForum (almost) exactly as in the nLab.

    These aren't quite compatible, mainly because by serving MathML we have to be a bit stricter than before and the odd thing has broken. I've tried to make these as unfatal as possible (so if a comment causes a failure, we get the source of the comment so at least it's still readable).

    The beta site is at http://www.math.ntnu.no/~stacey/Vanilla/nForumMathML. It's a copy of the nForum taken earlier today. Note that it is a copy so any changes there won't reflect over here. But login names and passwords are just the same (if it seems to get confused as to who you are, it may be best to log out here first before logging in over there, and similarly when you come back, just to keep the cookies clean).

    In doing this, a couple of other things got updated. In particular, the old latex syntax now works with more of the formatters (Text, Markdown, and Html) and it works properly in preview mode. Also, you have to preview your comments before submitting them to ensure that they are valid syntax (bit like the nCafe).

    The new formatter is, obviously, the one called Markdown+Itex. At the moment, it only does inline itex (between single dollars). Doing double and the square bracket will get implemented soon. The more fancy syntax, like query boxes and theorem environments, doesn't work either and, at the moment, I don't intend getting that to work.

    So please have a look. What would be useful are:

    1. Finding things that don't behave how they currently do.
    2. Finding differences between the Markdown+Itex formatter and nLab syntax (beyond those mentioned above)
    3. Finding existing comments that don't make it through the validator
    • CommentRowNumber2.
    • CommentAuthorHarry Gindi
    • CommentTimeMar 10th 2010
    • (edited Mar 10th 2010)

    Noooooo! MathML is so much crappier! This feels like a step backwards.

    The requirement that we first preview posts is really annoying. It also appears to not be working for me.

    The one on the nCafe takes one to three minutes to register a post. Please, could you allow those of us who liked the old system to continue using it?

    • CommentRowNumber3.
    • CommentAuthorUrs
    • CommentTimeMar 10th 2010

    Thanks, Andrew.

    Harry, you seem to be subject to a misconception. You'll never type MathML here, Just TeX. But MathML is used internally. You can make up for your empty misconceived comment by adding some sensible content to some nLab page (--hint, hint--).

    • CommentRowNumber4.
    • CommentAuthorHarry Gindi
    • CommentTimeMar 10th 2010
    • (edited Mar 10th 2010)

    No, MathML is not fully supported by Opera. The nLab pages don't display right, and now the nForum pages won't either =(. Also, generating the preview takes an enormous amount of time. I timed it just now, and it took over 70 seconds.

    • CommentRowNumber5.
    • CommentAuthorIan_Durham
    • CommentTimeMar 10th 2010
    At some point I'll have a look at this, but I'm in the midst of an insanely busy couple of weeks. I'll try to squeeze it in somewhere.
    • CommentRowNumber6.
    • CommentAuthorMike Shulman
    • CommentTimeMar 10th 2010

    Well, clearly you should not be using Opera. (-:

    • CommentRowNumber7.
    • CommentAuthorHarry Gindi
    • CommentTimeMar 10th 2010
    • (edited Mar 10th 2010)

    I've been using opera for eight years. )=

    Also, the fact that we have to validate posts before posting is really annoying. It's one of the main reasons I don't post on the n-cat-café. Is there any reason for it?

    • CommentRowNumber8.
    • CommentAuthorAndrew Stacey
    • CommentTimeMar 10th 2010

    To answer Harry's points:

    Noooooo! MathML is so much crappier! This feels like a step backwards.

    I can't put this better than the master himself did:

    There was a discussion about displaying math on the web, over at Terry Tao’s blog. It was a little disheartening to learn where most peoples’ heads are at, in that regard. Many seems to be excited at the mere prospect of creating a web service that would turn TeX equations into pictures. Such services are not exactly thin on the ground, I retorted, but hardly satisfactory, either. Fortunately, an actual blind user showed up to explain at least one reason why that wasn’t the answer.

    (emphasis mine)

    Also, the fact that we have to validate posts before posting is really annoying. It's one of the main reasons I don't post on the n-cat-café. Is there any reason for it?

    Yes. With XHTML then people are finally realising that it is important to produce valid syntax. Otherwise, there's no way to guarantee that WYSIWIM. Tag soup doesn't taste nice, I'm afraid. However, lots of the formatters were developed with HTML and just have XHTML tagged on at the end. So it's possible to produce invalid XHTML with them. Some browsers are so strict on their compliance that if a page doesn't validate, it doesn't display. So we validate first.

    I guess that the validation step isn't strictly necessary on submitting comments since I throw in an extra validation when the comment is displayed (since there are old comments which didn't get validated when submitted).

    Also, generating the preview takes an enormous amount of time. I timed it just now, and it took over 70 seconds.

    There may be ways to speed this up. Unfortunately, as I'm essentially sitting on top of the server, I don't see the same timelag that everyone else sees. So this is useful to know.

    • CommentRowNumber9.
    • CommentAuthorAndrew Stacey
    • CommentTimeMar 10th 2010

    I've also replied to the bugs you posted over there, Harry. Thanks for spotting those!

    • CommentRowNumber10.
    • CommentAuthorMike Shulman
    • CommentTimeMar 10th 2010

    Would it be possible to implement a "submit directly" button that would validate and then automatically submit the comment if it validates, and otherwise return to the user with an error?

    Not that I really think Harry's laziness deserves to be catered to... (-: At least here we (hopefully) won't get the empty errors that the n-Cafe has been giving recently.

    • CommentRowNumber11.
    • CommentAuthorHarry Gindi
    • CommentTimeMar 10th 2010
    • (edited Mar 10th 2010)

    The café literally takes five minutes to register a post for me. I don't know if I'd call that laziness... =p

    ;)

    • CommentRowNumber12.
    • CommentAuthorUrs
    • CommentTimeMar 10th 2010

    There is something broken with the Cafe software, at least at the moment. But that's not the issue under discussion here.

    • CommentRowNumber13.
    • CommentAuthorEric
    • CommentTimeMar 11th 2010

    For the record, I like Mike's suggestion.

    • CommentRowNumber14.
    • CommentAuthorAndrew Stacey
    • CommentTimeMar 11th 2010

    I like Mike's suggestion too. Here are the possibilities:

    1. Validation is a prerequisite for posting.
    2. If valid, it posts; if invalid, it previews.
    3. It always posts, but if invalid then you just get the source view (and you can then edit it to make it valid, and use the preview to figure out why it wasn't valid).

    I'll have a think how to code these. In the meantime, let me know which is most preferable.

    (Whilst I have no interest in catering to anyone's laziness, I do want to make it as easy as possible to use - which is one reason for the upgrade; hopefully the integration between lab and forum will be much closer.)

    One other idea that I had: make it possible to search the forum specifically by nlab page. So it would look for wikilinks to that specific page. Then on a given page in the nlab, you could link to the forum and be sure of catching all discussions relevant to that page. Good? Bad? Indifferent?

    • CommentRowNumber15.
    • CommentAuthorHarry Gindi
    • CommentTimeMar 11th 2010
    • (edited Mar 11th 2010)

    I like option 3 the best.

    Also, could you make it so leaving the page during validation still posts the post? This way, if the server starts to lag (you'd be surprised how often this happens), the person who made the post doesn't have to sit and wait on a loading page for (up to) 70 seconds.

    Even better, first have the post render in plain-text, and when the server gets around to it, it converts it or leaves an error message. This would avoid lag the user end.

    • CommentRowNumber16.
    • CommentAuthorMike Shulman
    • CommentTimeMar 11th 2010

    I like option 3 (there will, of course, still be an explicit "preview" button as well, right?).

    I also like the idea of searching the forum by nlab page.

    • CommentRowNumber17.
    • CommentAuthorHarry Gindi
    • CommentTimeMar 11th 2010
    • (edited Mar 11th 2010)

    I just realized a major problem with having the itex and validation servers on the nLab server (and forcing validation). If the nLab server is down (this is not a rare occurrence), we will not be able to post on the nForum.

    Another solution is to allow posting using the "Text" radio button without validating or otherwise. That actually seems like it couldn't hurt.

    • CommentRowNumber18.
    • CommentAuthorAndrew Stacey
    • CommentTimeMar 11th 2010

    The validation occurs right here. The nlab is only used to convert itex syntax into mathml, and it only needs to do that once for each string of itex. I may be able to shift that to my local setup, and I could build in a timer so that if it takes too long, it simply returns the itex source code (with an abject apology, of course).

    • CommentRowNumber19.
    • CommentAuthorTobyBartels
    • CommentTimeMar 15th 2010
    • (edited Mar 15th 2010)

    I hereby copy Mike's comment #16 under my own name.

    • CommentRowNumber20.
    • CommentAuthorAndrew Stacey
    • CommentTimeMar 19th 2010

    Okay, I've ironed out a few bugs. I've fixed the validation so that it is validation upon display, not validation upon submit. So you can submit horrible, invalid, yukky stuff but it will get cleaned up for display. The preview mode will tell you exactly what the errors are, but the normal mode will just tell you that it was invalid (allowing you to read the source, if you can).

    Hopefully that's addressed the various concerns.

    Please, please, have a go at breaking it again. It'd be nice to roll out the new version soon.

    (Thanks particularly to Toby and Harry for breaking the old version!)