I’ve been reading on about Greggory, Tyler’s revolutionary text-to-Gregg translator. Unfortunately, I’ve only seen posts from 2004 at the latest. This project may have been stagnant by now, so I decided I might revive it. Also, because apparently it was on the old MSN group, I haven’t tried his program.
If you have Python 3.x installed on your system (NOTE: it has to be 3.x, not 2.x or 1.x. They won’t work.), I have a lighter version (with source code!) at http://chancehenrik.x10.mx/shorthand_py.zip.Just unzip into any folder and run
“shorthand.py” or “shorthand.exe”, depending on which one you downloaded.”shorthand.exe”
(by Chance for group greggshorthand)
Actually, I'm not at the point that you could see the outlines, but I could get to that right now.
This is interesting. There are some outlines that are off, but this is a good start. It would be nice to have a side window with the AGS conversion code so that we know what the output means.
When I said outlines, I meant the ASCII code (for example, the word "movies" should be mUes or something like that).
Python. High five! I'm still 2.x. I'll take a look at the results.
😀 I started learning Python 3.2 a few months ago. I said I would do 3.x to minimize the job of porting and other stuff. If you don't have 3.2, get shorthand_nopy.zip.
Yeah. I'm still working on it. This *is* just v0.5. You can see that I have to work on many other things before actually releasing v1.0. Also, that might be the problem of the dictionary. If anyone here has a suggestion for another dictionary to use, or already have their own, please let me know.
"Viewed 57 times by 27 people, latest at 2:42 PM"
This project is receiving a lot more attention than I expected!! BTW is the original Greggory still alive? I want to know its progress as well.
I'm not sure if the project is still alive. We haven't heard from Tyler in years.
This is super sweet…there could finally a quick and easy way to fill a kindle with new shorthand novels? Dragon Tattoo…Hunger Games…Copy…paste…convert…pdf…done!
Great work here…can't wait for the final!
It's going well so far. I'm kinda stuck with circles and stuff. Read more about it on my blog. It's at http://chdeguzman.tk/blog/archives/category/shorthand.
The part about the circles and positioning on the baseline is more complex than what I've already done. Sample image output: https://lh6.googleusercontent.com/6Y1_WJAkmgZryzQgtS1WYLkZ4xRu-dkU4TRnaz3YHSns6FJkDXHQqhrWXDhyHxwJWBD4rLpdYik
Obviously, there are many other things to change.
The part about the circles and positioning on the baseline is more complex than what I've already done. Sample image output (that just says "the quick brown fox jumps over the lazy dog"):
Obviously, there are many other things to change.
For joining the strokes, should I get the strokes from the Manual or the Penmanship practice?
I guess. But which one would you prefer that I use?
I guess. But which one would you prefer that I use?
It's a coin toss on that one.
Perhaps the manual because the penmanship practices may have different writers, whereas the plate writer for the manual was Mrs. Richmond.
And WHERE IS TYLER?!!
Perhaps I decided too early. The image above used strokes taken from the manual, while this image used strokes from the penmanship.
Tell me, which one would you rather use?
I wasn't really there yet. I was planning to code those rules in over the weekend, and for the images I just showed you the difference between using the manual and the penmanship plates.
What are the rules for positioning outlines on the writing line?
Is it that the first consonant rests on the line, but when it starts with an S, the second consonant is on the line?
I already turned the "ks" into "x". Also, "fox" becoming "fa*ks" might also be the problem of the dictionary. I might have to get a British accent for the dictionary.
Is "brown" actually "brn" in Anni? I wasn't sure of that.
I see two things:
1. The outlines for "quick", "brown", "fox" and "jumps" are not correct. The joining in "quick" is off (it looks like a reverse circle joining), "brown" does not have an "a-oo hook", the "ox" in "fox" should be "o hook-x", and there is no "oo-hook" in "jumps."
2. The shapes and slants of the vowels and curves change as you join with other curves. That's why for example you cannot create a smooth "o-g" in dog, because the end of the "o-hook" matches the beginning of the "g." Perhaps you need to work first on making sure that straight lines join correctly with circles and hooks because these should the easiest ones to join. Then work with the curves which are much more difficult.
I really have no preference between the two.
Indeed, yes. (I had the same problem with the Shaw Alphabet.)
Good luck finding a suitable dictionary, though! The two big ones are the Moby Pronunciator and the CMU Dictionary, but both use an American accent with the bother/father merger, so they don't have a separate "short O".
If you do find an electronically readable pronouncing dictionary with a British-like accent, please do let me/us know!
You may also want to contact Thomas Thurman (thomas@thurman.org.uk), who crowd-sourced an English-accent dictionary for use with the Shaw Alphabet; he may be able to provide you with data.
Yes. The convention is that the base of the first consonant rests on the line, except for the s, which will be the second consonant. It is not a rule per se as Gregg can be understood without regard to positioning, but it is the convention that is used in all books so that the writing is clear and more legible.
The combination "ow" is omitted in Anniversary when followed by "n" or "m" (including the "nt" and "mt" blends).
I guess this is the best time to do a general review of my theory. I'm doing it right now. And it will also help me with the program, because I don't really remember all of the rules.
Check out the Reverse Dictionary.
http://greggshorthand.multiply.com/journal/item/1213/Reverse_Dictionary_Anni_still_useful_for_other_versions
It has all the words spelled out.
Just a little editing and I could use this nicely for the program. In which case I'll have to recode the entire conversion algorithm.
I have seen it before, but I haven't really thought of using it like this. Do you think it would be easy to add the derivatives of words once I got these in?
A pronunciation dictionary has over 40 vowel sounds — fine nuances count. The reverse dictionary says exactly which vowel to use. (Well, which vowel the average Gregg user will use.) It will also deal with brief forms and the abbreviating principle much better. On the downside, there are fewer words in it than in a pronunciation dictionary. Given the frequency distribution of English words in prose, though, the reverse dictionary should be enough. Leave an "out" for the program. If it can't find a word, type it out normally.
By "type it out normally" you mean just write it as is, or use the phonetic dictionary anyway?
It doesn't say one way or the other that I can see. No easy way to tell, either, but I'll add that to my list of rainy-day projects.
I had not noticed this before, but it includes many phrases, several place names, and the occasional Spanish entry.
Did the reverse dictionary include the 5000 most frequent words? I don't recall.
I meant type it out normally, with normal English spelling.
One problem is that the dictionary is contaminated with forms from other series, e.g.:
a-b-u above (1916)
kinda messed.
Many of the older outlines that were removed from the newer versions were actually retained — but as advanced. Not the reverse-R, but many other principles.
You're right, though. For the program, it makes sense to stay with one series. I think this approach would produce better text — the pronunciation and therefore Gregg-spelling will be "standard", but it's a lot of work.
I'm releasing a new version [29-0.61]: shorthand-0.61.zip. The version selector is nothing right now; I was just trying something out. This doesn't implement all of the rules, because then again I have to recode everything.
(Is this the right place to do this?)
I just released the latest version of my program, which I now call "Greggory" (thanks, Tyler!). It's a semi-private beta, because I'm only giving the link to the shorthand group.
The links are at my blog. Just look for version 0.70 and download either the ZIP or MSI packages. There are only two, so you won't get lost.
I'm still in the process of making the shorthand font (my FontForge keeps dying on me), so you'll only see AGS until then. However, if Andrew releases the shorthand font that he's apparently been making, then I won't have to do anything else. I'll just have to wait.
If anyone would like more words to be included (that also includes place names and personal names), and I'm pretty sure everyone would like to add some words, just email them to chance at chdeguzman dot tk, and I can add them to the next version of the dictionary.
Enjoy. /chance
Thanks! I'll take a look at it.
Any new news on the font? I also see that your website is currently down, Chance. Interested in how everything is going on the project.
Thanks for telling me. I haven't been able to log in for the past few days. Apparently my hosting provider migrated servers and my blog went down. I'll try to figure it out and get it up and running again.
Regarding the font, I haven't done much progress over the summer vacation, and now that school's back I don't think I'll get much more done. All I've done is put the glyphs in FontForge and edit them somehow, but I haven't been able to put them together.
EDIT: I saw a PDF from Andrew apparently using a font he made. I don't know how he did it, but I think he's done with it. It takes a while, you know.
Chance's original reply contained this graphic. It was lost in the move.