A more Standard AGS

As I re-reach the halfway point in coding Gregg.Convert(ex-Greggory) (I had to start over in C#), I realized that I needed to have a permanent reference implementation for AGS set before I actually move on to doing the layout stuff. So I created a little so-called “standard” document I call SS 1:2012, detailing all the mappings between shorthand strokes and AGS characters.

The problem now is that I need to check with you guys. I’ve published the document already, and I want you  to review it for me to see if there’s anything missing, or if there’s something you’re not uncomfortable with, etc… I noticed there were some disagreements with some strokes, e.g. sh, tive, capital dashes. Just comment on any changes I need to make.

When the standard is finished, it will go on EVERYTHING Gregg-related, so I have to get this finalized before I begin coding again.

Thanks ~Chance

And the link: SS 1:2012

UPDATE: I just got my blog back!! An x10Hosting forumer (probably an admin) just restored my database a few hours ago, and now I’m back on. The original content is still there from 3-4 months ago, unchanged.

18 comments Add yours
  1. Neat! Wonder if it would save time to type this rather than English, for normal notes. Count, count…

    Result: The sample just over half the keystrokes (373 vs 676). I suspect it will improve in later chapters, with pre- and suffixes, but disjoinings and "above the line" will limit that increase.

    268 characters plus 73 spaces plus 32 shift keys= 373 keystrokes.

    474 characters plus 195 spaces plus 7 shift keys = 676 keystrokes.

    Shifted keys are counted twice — once in character count, and once for the shift. They're probably worth 1.5 rather than 2, but close enough for now.

  2. Some additional observations:

    1. I didn't see the double circle.

    2. If I understood this correctly:

    telegraph: tel|@e
    tellingly: tel|e
    parage: @p|j
    postage: p|j

    3. I'm also thinking there is a need for a symbol or way to express a discontinuity in an outline that does not follow the normal joining rules. As examples, the "s" in the words "myself" and "letters" in Anniversary, and the suffix "ol" in the word "biology" in Simplified.

  3. Carlos, I have added the said strokes and markings. I'm almost running out of space so I have to map the other strokes properly.

    Cricket, I'm assuming you haven't been here long enough? Most of the others here have seen AGS before, and I think you haven't. Just saying.

    1. Thanks. Cricket has been here for a long time and I think she has seen this before. As I understood it, her reply was just her way of thinking if it was worth it typing complete words for her dictation program.

      Incidentally, the original AGS post is here.

      Lastly, how many ASCII characters are left?

    2. I think she was talking about notes and stuff, not the dictation program..

      I just noticed it now that Jason misspelled the title of the post as "ACSII" instead of "ASCII".

      The numbers don't count, and neither do the square brackets because I set them aside, so…
      ! # $ % ' ; ? ` y z
      that's 10 more characters, none of which I assume will be filled.

    3. @5a: according to the mapping table at the bottom, the following are still free:

      ! # $ % ' 0 1 2 3 4 5 6 7 8 9 ; ? [ ] ` y z

      Of those, [ ] are recommended for surrounding AGS text to separate it from regular text, so that leaves the ten digits as well as the lower-case letters y z, the grave accent `, and the miscellaneous symbols ! # $ % ' ; ?

    1. Ah, that's where the rest of the characters come in.

      $ = [d], % = [sn], ? and ! are punctuation, and the digits are digits, so that leaves me with $, %, #, ', ; and ` (and possibly y, z) for my last few characters.

      How many spacing rules do I have to accommodate?

  4. My random-connection generator was on, again. Nothing to do with my dictation program, but I've been trying to take notes on my smaller electronic devices. Nasty keyboards, and not enough resolution and/or speed to handle Gregg. I suspect, though, that AGS doesn't have enough redundancy to work on keyboards that are too small to type accurately.

    As for seeing AGS before, yes, I have, but the random-connection generator didn't jump in.

    I look forward to seeing the results of this project. It's a neat idea, and even the intermediate steps are useful.

  5. 1. The AGS spec is complete (unless I'm missing something from the later versions), and I'm on my way to writing the font. The problem is the standard OpenType font technology won't work with shorthand, so I have to use SIL Graphite, which you can only use on certain programs, e.g. Firefox.

    2. The current system works great on computers. The real problem is tapping those really small leopards. The keyboard layout can possibly be optimized for AGS input, but someone else can work on that.

    3. What intermediate steps?

Leave a Reply