Marco Gualtieri / Log

Very nice deft-mode from J. Blevins – a version of notational velocity for emacs which is easy to modify and can run on .org files.

First approximation using checkout setup for pushing local website to server using git mirror such a website on github The idea of having a hub between my local edits and the www server at the department appears to be a good idea. I have not yet figured out how to keep it all in one git repo, now I have separate repos for each course and for the main site. After I do my local edits using org-publish, I simply push to coxeter and this triggers an update-hook or receive-hook which populates my www directory.

Emacs Starter Kit for the Social Sciences Would like to have a simplified version for LaTeX development, including Git integration and full Bibtex or reftex integration, as well as hopefully publishing integration using Andrew Stacey's tex-html under development.

Our recent paper, The Stokes Groupoids, contains new examples of simple and explicit Lie groupoids. The simplest nontrivial one of these is \(\mathrm{Sto}_2\), a two-dimensional Lie groupoid over \(\mathbb{C}\) with total space \(\mathbb{C}\times\mathbb{C}\). Here is a picture of this thing: Stokes groupoid of order two

A very interesting article about the history of QFT: http://logos.sjca.edu/~dcasey/NEH/assets/WeinbergNotesFieldTheoryHistory.pdf

Great article about the influence of technology on writing: MIT technology review

Here is a great idea for a command-line utility: gist takes input a text file on your computer and outputs a url to a gist where the text file has been published. It is easily installed via homebrew.

It is a great idea but it is unmaintained. For some reason, it fails if the text file contains a "%" character. The fact that such a software bug can exist at all, and can continue to exist after it is noticed by countless programmers, tells me that software engineering is a dead field.

In this HN post, I discovered an amazing project by Wang Lu, which is a very careful conversion from PDF to HTML5. This opens up a wealth of possibilities for going beyond PDF output with LaTeX. After some difficulty, I managed to compile the software (I'm not sure why the executable couldn't be distributed) and managed to get incredible results, like these notes for my course.

It also pushed me to open the following issue at the GitHub project page, where I speculate about possible extensions of the idea.

A very interesting talk about designing math fonts - unfortunately it is not an open source project.

This very beautiful dissertation uses other typoma fonts - many nice ideas for math typesetting.

This really stresses the importance of having latex control web output - including colour, fonts, document structure, macros, indexing, toc, etc.

This is an experimental lava pour: Lava Pour No. 5

It's relatively easy to embed a gist, which could even be an .org file, into a blog entry. The question is, what should the blog architecture really be? Should every post be a gist? What is the advantage of this?

Useful collaboration guide

The new wiki for glab is here glab wiki.

I am very impressed with the new setup of the Stacks Project. It is on Github and also has a separate project for the website, which has some advanced features. I don't quite understand the thinking behind the tags, which seem to be external to the document (which may make sense).

I would like a system similar to this for all the notes and working files for my group. But I can't figure out how the thing should be organized. Is this better than, say, a wiki? I really don't know.

One possible solution would be as follows.

• Git repo of a bunch of .tex files
• Online editor/wiki with mathjax which includes a default preamble

Based on a suggestion from Alan Weinstein, I looked into possibly recording recorder duets. My first example is this recording of loeillet de Gant's first recorder sonata (mov. 1). I will have to refine the intonation as well as the phrasing. It would be interesting to choose a series of sonatas and to record them all.

It would be good to have a system whereby a recording can be made and then annotated, say, by a music teacher. It could be a text file with a reference to a .ogg file, or some such thing. It could be part of a web app for virtual music teaching.

Ideally you could also annotate the score, or perhaps the annotation should only be attached to the score, I am not sure. Could Dror Bar-Natan's system for annotation of video be adapted for this purpose?

Is there any evidence in the literature that we can be trained in our sleep? I am imagining a device which monitors ambient noise during sleep, and when it detects tooth grinding, it sends a small electric shock to, say, your toe. Over time, would this train the body to not grind teeth?

A corollary of this idea is the question of how to develop a simple software utility and push it out to various platforms, including, say, an arduino style thing. Would there be kickstarter support for the production of such a gizmo? I asked about it on HN.

Talk about basic scientific assertions

This workshop "Beyond the PDF" seems to have a lot of interesting proposals. In particular, there is a discussion group where I learned about Scholarly HTML. Check out the website scholarlyhtml.org. It seems too tentative and not well-developed. Interesting zfishbook talk - important question brought up at the end. The question is about developing a platform for developing tools for going beyond PDF towards a HTML5 + Javascript solution.

A remarkable post from April 21 2011: We need a github of science I also found a guy at Waterloo working on scigit.

This is my question about how to approach the design and development of a document format.

The basic problem is this: in math (and other disciplines), we write a LaTeX markup file and then compile it to PDF. PDF is basically a paper simulator and lacks many features which webpages/sites can have.

An alternative would be to have a versatile but still constrained markup language (e.g. an extension of LaTeX, or a constrained org-mode) which can be compiled or "exported" to HTML among other formats.

Some of the benefits would include

• Folding of text: hiding parts of the document until further details are desired
• more sophisticated linking between (parts of) documents
• nonlinear/hierarchical document structure
• including media such as animation/tutorials
• running code in the document

The question is: how to approach the development of the system. For example, many of the features could be implemented, say in jquery or some such, but the system should be independent of implementation. How does one proceed in a future-proof way whereby the choices are not regretted down the line?

Note that this is not about typesetting – I am aware of the web typesetting problem. This is about the inadequacy of PDF as a document format for the future.

Very good response on my news.ycombinator post: response.

We need a replacement for journals - well known problem. I could just rate the papers I read. Provide a critique and comments, modifications, notes etc. Public information should be accessible and site could be joined by others. Check out what is happening online on this topic.

Mathematical documents need to evolve away from PDFs, which are paper simulations, and towards active documents which have interactive features, such as:

• folding

The ability to hide sections of the text to make reading simpler.

• linking

PDFs already have, but internal linking and linking to other documents in a project or related web of documents would be good.

• running animations or code

Latex and orgmode have

• versioning and branching

People can modify papers in an automated way, create new versions, combine results of sections and include sections as modules in other papers.

• org-info.js

Something which may be helpful is the use of org-info.js which encapsulates various javascript functions in a markup language so that the precise nature of the javascript is hidden. It would be good to develop certain elementary operations which we would need for the creation of web documents as an output format for orgmode used as a mathematical document system. It already has Latex inter-operability, so we could view it as a LaTeX enhancement which outputs to an active webpage or wiki-like acrive document.

I would like to see any proposals or studies which deal with these kinds of issues, especially to help decide what the core elements of an active document would be. Try to think about what goes into an AMS article class document in LaTeX… what would be the analog for an active document.

Alejandro, Eckhard and I are writing a paper about obtaining geometry on a Lie group via infinite-dimensional reduction. This includes a (multiplicative) Courant algebroid structure as well as a multiplicative Dirac structure.

My main difficulty is that it is unclear what the big picture is - how does it relate to work of Sullivan, Getzler, and Henriques?

Map from a torus to a plane. Should figure out a good way to make a 3d embedding. Drawing Apparatus