Well, the comments are ignored by the program, so specifying them in there won't help it at all. You should still upload that, it would be useful for me to see what people are adding and all that.

Deep Thought: There will be a chooser so you can pick things out.

Ok, here's the modified xml file. Changes include:

• Incorrect tokens/tokens with typographical errors fixed
• Duplicate/erroneous tokens removed
• Missing tokens added
• Ensured that all token strings could be typed on a standard American English keyboard (¿ and ¡ could not)
• All non-pretty statistics functions and lowercase statistics variables (all two-byte tokens starting with $62 except F) standardized to [x] format
• All single-letter special tokens (e.g. ˪, i, F, ᴇ) standardized to /x format
• All raised symbols standardized to ^^x format (not only does this seem more logical than **x to me, but also ** is an Axe Parser command that could be misinterpreted)
• Some other minor changes I'm probably forgetting about
• "Pretty" alternate strings added
• Axe Parser alternate strings added

Primary strings can all be typed on a keyboard. Alternate strings are sorted in the following manner:

Alternates that can be typed on a keyboard are first. "Pretty" alternates appear next. "Alternates" is plural because tokens that have more than one "pretty-ifiable" characters (e.g. chi^2cdf() are broken down into all possible pretty/non-pretty combinations (e.g. χ^2cdf(, chi²cdf(, and χ²cdf(). These are sorted in increasing order of "prettiness." Following any "Pretty" alternates come Axe Parser alternates, sorted according to the preceding rules.

Wow, that's awesome. Thanks so much, Runner. The Axe commands I'll end up taking out into a separate Axe.xml file, but for now this is perfect. There are some things I'll change, and I'll have to look over it a little more in depth, but this is awesome. Out of curiosity, what were the duplicate tokens, and what were the missing ones?

$6338-$633F and $7E00-$7E07 were the only duplicates I think ($7E00-$7E07 were the correct ones). You missed $6237-$623C, $BBED-$BBF5, and I think few scattered others.

To read and write to pointers in Axe, you have to enclose the pointer value in curly brackets. I believe Quigibo uses curly brackets on some of his commands, like nib{, because they take only one argument, and that argument is a pointer.

By the way, the Celtic III section seems to be malfunctioning. I accidentally stumbled into this when loading an Axe source file, because Axe uses the command det( followed by a number to allocate data sections in the compiled program. I believe all the Celtic III commands are in the format of det(number,stuff) whereas Axe uses the command in the format of det(number). Even though there is no comma after the first digit in det( calls in the Axe source file, it seems that it is being somehow intercepted by the Celtic III commands. However, because there is no comma, it doesn't actually match a token, and all it ends up doing is deleting the det( token and the first digit following it.

EDIT: It appears that this error also applies to other shell extension commands in the xml file.

Nice, I like how we can create pics from cropped images. It seems to convert pretty well too

O_o
What did you do to make that happen?

Wait you can paste file icons in there?

And it's a ... space?

I'm starting to use this for some projects. It's really nice.