Author: grant

jtracert

jTracert will allow you to generate sequence diagrams directly from your application runtime!
* This gives you a lot of advantages:
* Understand the code created by your colleagues/partners in a short time
* Rapidly generate documentation for your partners or users.
* Easily investigate what’s happening in large Java applications
* Excellent companion for a common debugger

People in Pain

Most would agree that it is probably impossible to “reason with” people who are in pain since they don’t often act in a manner that you would expect (read “normal”). The trouble, though, is that often times you can’t see their pain, and that a lot of people today seem to be in it. It is no wonder why you might have noticed that so many people seem to “act strangely” these days.

Tendrils in Scheme

Here is an art piece by Tom De Smedt written in NodeBox (Python) :

http://www.nodebox.net/code/index.php/Tendrils

It is in turn based on work done in Processing by a group ART+CODE.

Ported to Abstracting:

http://github.com/dharmatech/abstracting/blob/master/examples/tendril…

Although Abstracting is OpenGL based, I made a NodeBox compatibility library which allowed for a straightforward port.

Here is a rendering performed by Abstracting/Ypsilon :

tendrils-ypsilon-2009-04-07.png

Ed

I wonder if that is something that Script-Fu would do well?

(via comp.lang.scheme)

chibi-scheme 0.1

An initial preview release of Chibi-Scheme is available at

http://synthcode.com/scheme/chibi-scheme-0.1.tgz

Smaller than TinyScheme, faster than a speeding Guile, able to leap tall vectors in a single bounds check, with native hygiene to boot!

WARNING: Do not use Chibi-Scheme!

Seriously, there are real Scheme compilers out there. The author of Chibi-Scheme was working on one himself before starting this silly project, and will likely continue to work on it for years to come. But sometimes we all just want to release something.

There seems to be a disturbing trend, however, of people taking toy Scheme interpreters seriously. It can’t be for ease of use, because the serious compilers all have the most friendly FFI’s. It can’t be for the small memory footprint, because the difference between a 100k and 2MB program text will be dwarfed by the runtime memory, and all of the real implementations have much more efficient memory usage. But time and again you’ll find TinyScheme crop in in bizarre places, which leads one to wonder how much of a bad name is Scheme getting by being so often represented by one of the slowest language implementations on the planet?

So Chibi-Scheme exists as a better toy implementation. It’s a very small but mostly complete R5RS Scheme implementation using a reasonably fast custom VM. Chibi-Scheme tries as much as possible not to trade its small size by cutting corners, and provides full continuations, both low and high-level hygienic macros based on syntactic-closures, string ports and exceptions. It also has optional immediate symbols, just to be quirky.

But don’t use Chibi-Scheme. Don’t use toy Scheme implementations at all. But if you really want a toy… well, then perhaps Chibi-Scheme 0.3 or so may be right for you.

[On a more serious note, I do want to hold this up as an example of how extremely simple and natural it is to implement syntactic-closures compared to alternatives such as syntax-case. The whole macro implementation is about 25 lines of low-level C code (modulo the extra lines I’ll have to add later for bug fixes), as opposed to say the 4000 or so lines of Scheme for psyntax.]

– Alex

(via comp.lang.scheme)