Jose' Matos wrote: > On Saturday 17 July 2004 03:32, Fernando Perez wrote:
> One short notice, I had gone to the ipython page (I missed the 0.6 > version, oops) and I searched for such a document, I followed the link to > cvs and it failed from that page. > > For some reason the redirection is failing. Take the address directly > solved it though. A simple bug report. :-) Sorry, could you clarify which is the broken redirect link? I'll gladly fix it, but I'm not sure what you are referring to. >> ps. I'm glad you like ipython. By the way, I've toyed with the idea of >> having Lyx as the wrapper environment for a python-based system which >> would allow mathematica-type 'notebooks'. Something where you could keep >> interactive code, text and equations, and the plots right there in the >> document. I know many pieces of such a puzzle are there, but it needs to >> be done. If you guys like such an idea and think there's some energy >> available, the ipython infrastructure is evolving to meet the embedding >> needs of such an animal. Think about it... > > > Do you have any concrete proposal? It seems a bit fuzzy at the moment. > That being said I like the idea, on the other hand python is a requirement > if you have older document due to lyx2lyx. Also the lyx mechanics is the > same as any typical free software project, "show us the code" and we will > discuss it. :-) > > So you have my support for it, consider it as a payback for your work with > ipython. ;-) > > Your proposal is something quite interesting and a good topic to discuss > in next month meeting in Chemnitz hence if you have any further details > please step forward. [ Sorry to the lyx crowd for the semi-off-topic drift. If this is getting out of hand please let me know and I'll take the discussion to private email with José. But since there _is_ a lyx idea behind all this, I hope to get away with abusing your hospitality a bit ] Well, it seems fuzzy because it is :) The first thread in May is the furthest we ever got: http://www.scipy.net/pipermail/ipython-user/2004-May/thread.html I think there's a lot of interest in this idea (Arnd Baecker in particular --a german physicist in Dresden and avid python user-- has given it a lot of thought). But nobody has enough spare time to seriously work on it. I'm headed to Scipy'04 in a month to present some things about ipython and some about multiwavelets and PDEs, and I plan there on discussing the idea around. But there's no full-fledged idea yet coming from anybody. >From my side, what I plan on doing is the following: - very soon (this weekend maybe) release 0.6.1. It has many fixes, some new features, and a 'pysh' profile, which is a python-based system shell (lite). It's pretty cool, you can do system scripting with python syntax with 2-way communication of variables between the python namespace and the shell. - implement a multithreaded ipython which can be used with Wx/GTK apps, so matplotlib can be used interactively with all backends. This should actually be relatively easy, since it's just a matter of running the read/eval loop in a separate thread. Then, I'll begin an architectural cleanup of ipython. Its internals are a huge mess, so I plan on taking a chainsaw to the code to organize it better. One key result of this will be to make it easy to embed ipython as the python shell of a graphical environment (think Idle, Pycrust, but with ipython as the shell). Once this is done, I think it would become realistic to consider how it could fit inside a lyx document as a running process. I am _extremely_ interested in something like this, especially considering that lyx is one of my all-time favorite projects (after ipython :). However, ipython is mostly a one-man-show and it only happens in my spare time: while all of my research uses python, I can not justify spending my 'real' time on ipython, which as a daily tool is already 'good enough'. But at least I can promise that slowly, ipython is getting cleaned up in a way which would enable this idea. Whether the whole thing (the lyx+ipython mathematica-type env.) ends up happening or not is a different story. Best, f
