I meant to post about this a while ago, but while reading Bruce Sterlings's Beyond the Beyond, I learned about this clever tool for generating color schemes. Given your keywords it searches for the top five images on Yahoo! Image Search and chooses the six most prominent colors from each.
Perhaps unsurprisingly, given "existential type" it just returns different shades of gray – ones even less interesting than Bruce got for "cyberpunk".
In response to my post on quality fonts, mbana provided a pointer to a Typophile discussion on the Tex Gyre fonts. In particular, Thomas Phinney described their quality as "wildly variable". I suppose at the time I was using "quality" to more describe whether the fonts provided all the features I would expect to need, rather than only aesthetic quality. I have to admit when I did skim the specimens that they didn't look too bad to me. He did say that the glyphs based on the original URW designs were quite good, while the Greek and Cyrillic "range from mediocre to poor to largely useless". Phinney most certainly has more experience with Cyrillic and type design than me, and arguably he probably just has a much better eye than me.
Leon recommended Junicode, which provides the most important features I discussed. I do think its stems are a bit thinner than I prefer, but it seems very well done.
In my discussion of METATYPE1 and meta-fonts, Till pointed out the existence of the meta-font Antykwa Półtawskiego. At the time I told him that I thought he was misinformed because Antykwa Toruńska seemed to have been created by tracing scanned specimens. A few days ago, Peter Backes pointed out this rather egregious error in reading comprehension on my part. Currently, only the generated Type 1 font files are available for Antykwa Półtawskiego. However, it sounds like it was prepared using a rather early version of METATYPE1 and therefore the sources could possibly be incompatible with the newer versions of METATYPE1.
Peter also noted that he has also created a font with METATYPE1 called OCEANIA. It is perhaps not as "meta" as he thinks would be ideal, but during the time I've spent working with METAFONT and METATYPE1, I've definitely found that it is often quite difficult to come up with a reasonable declarative specification of a glyph.
Jeff pointed me to these extremely cool and useful web application: Detexify2. At least a few of you readers have spent time skimming The Comprehensive LaTeX Symbol List looking for the incantation required for a specific symbol. With Detexify2 you can just scribble something that roughly looks the symbol and it will tell you the name and, if necessary, the LaTeX package that provides it.
The only problem is that it lets users help train the recognizer. This can also be a good thing, but I can imagine a few malicious users (or perhaps just people with very bad drawing skills) ruining it for everyone.
Typographic Style for Computer Scientists is a short article intended to bring computer scientists up to speed on typographical issues. I started writing it several years ago, and then set it aside for other projects. Recently, I decided rather than wait for it to be “complete” and perfect that I should polish it up a little and see if people could benefit from it.
It is probably far from comprehensive, but hopefully some people will find it useful. I had been thinking of perhaps adding some material specifically on inference rules and judgments, but I am not sure what else it might make sense to include.
If you find mistakes please let me know. If you disagree on some of what I have said, let me know and perhaps I can address your concerns.
In the past forty-eight hours, I've written code in C++, F#, and Scala. On top of that, XCode, Visual Studio, and Emacs all have different default keybindings for all the most common operations. Then add in switching between US and Danish physical keyboard layouts.