Labda Hata Mimi

(in response to Loren of Incremental Blogger. It is asked which Microsoft Tablet PC app should be open sourced…)

Hmm, I think they should add standard, direct SVG export support to Journal and OneNote and Stickies (!)… That would be as effective as open sourcing them in many ways (and easier) of course we know they’ll never do it.

I think I read a quote somewhere in which BillG said something to MS staff to the effect of– adopt XML but not standards or don’t follow other people’s standards, microsoft makes standards. Anyway too vague to do any good. But it makes it difficult to do a lot of the fun things with ink. The whole Journal Viewer app is totally and utterly redundant. SVG is a standard, use it!!

OTOH, this open source Journal, Gournal is very nice having had the opportunity to try it recently. Written in GTK2-perl it is crossplatform too so it is theoretically possible to bring it to windows if one can get the libraries installed. I’m idly considering trying that. OneNote-esque Collaboration, fast, clean, and with SVG output.

It really shows that a thick Ink SDK isn’t totally necessary to do something like standard notetaking well, SVG canvasses will do very well. And these are evidently being worked on fairly fervently by several active projects.

I’d imagine this would be a solid base for an Open ink-to-text engine too. It does neat things that show that each stroke is just independant as Ink–highlight when delete cursor is over them, clear them stroke by stroke, etc. Yet another reason to learn perl–hack on this a little maybe.

There is an open source app which does some character recognition–Freehand Formula Entry System–it was recently updated and I linked it on my Linkblog. Now there is a version for Windows but I tried Linux as that’s the port I discovered when I originally found it. There are a lot of component applications (mostly open source too) but I wonder how much of it could be salvaged to recognise words? A dictionary of 40 or so mathematical symbols is a far cry from millions in the english lexicon, structure wise. And in my testing the recognition got stuck a lot. Donnog