Saturday, March 30, 2013

PyCon Video

Here's the captioned video of my PyCon talk.

Here's the page with the slides and the uncaptioned YouTube version of the video. Hopefully when I give the caption file to the people in charge of the videos, they'll upload it to the YouTube version so that'll have captioning too.


  1. Thanks for posting this, Mirabai!

  2. Brilliant! Did you rehearse this? Last time you weren't speaking nearly as fast. :P

  3. Hahaha, yeah, I actually did an entire dry run of this talk in NYC and didn't have enough time for questions, so when I went to PyCon, I was like, "Better go faster or I'll run out of time!" And then got through the presentation in 15 minutes. Oh well. People had plenty of questions, so I think it turned out all right. Also, at PyGotham, the CART provider was in the room, so I knew if I went too fast she could walk over and biff me. Whereas at PyCon, the CART provider was working remotely and could only yell at me through Facebook. (I apologized.)

  4. I tried to include in Plover's dictionary some non-ascii characters (like รข), but the result is that Plover doesn't start anymore. Could you please explain how to include non-ascii characters?


  5. Hello Ms. Knight.

    Have you ever run across dasher? You might find it interesting (or you might totally hate it).

    It's similar to steno, just in the "information efficient text entry" way.

    You can't type as fast as steno (because of the closed loop control) but, depending on the setup, you can "type" at a reasonable speed by gently wiggling one finger.

    You don't define key words in dasher. Dasher just makes it easier to type the things it often sees you type.

    what do you think?

  6. Anonymous: Sorry about that. Yeah, Plover's still kind of convoluted when it comes to non-standard characters. Try defining it as acircumflex, as per the Plover Guide. (

    Not 100% sure if that will work, but that's my best guess. I'm hoping to get this implemented in a more intuitive and user friendly manner at some point.

  7. Suki:

    Dasher looks very cool, and I bet it's useful for a lot of people, but personally I tend to get frustrated with probabilistic text entry; even if it only fails in a few cases, I get disproportionately frustrated when it does. So I much prefer the determinism of steno.

    Great for accessibility, though! I'm really glad Dasher exists.

  8. Thank you for your kind answer(s).

    As you are an experienced stenographer, could you please tell me who established the letters on the steno machine, and based on what study? My intention is to create a system for my native language (romanian), and maybe the letters on the steno machine should be different.

    Also, is there, by any chance, a connection between the steno machine and shorthand?

  9. Mirabai: Congrats on a great presentation!

    Anonymous: Yes, there is a connection. Stenography is machine shorthand, and they're based on similar principles.

    To my understanding, the current layout of stenography machines was designed by a man named Ward Stone Ireland. I've never found any information on how he developed the layout.

  10. Anonymous: I forgot to add...

    You can find a nice little slide presentation on the development of stenography machines at

  11. d'oh! - not!

  12. I haven't been able to find as much information as I'd like on this subject, but from what I can tell, the steno keyboard is based on a probabilistic analysis of consonant clusters in English. I.e., S, T, D, and K are more frequent beginning letters of words than P, W, H, L, and R, and more importantly, combinations such as STR, TR, FR, FL, GL, KR, PR, SH, CH are common beginning clusters, as opposed to things like PT, HK, or LW, which are invalid English consonant combinations.