The new Math-o-mir v1.8 beta can be downloaded from http://www1.datafilehost.com/d/beeb6100

I decided to increase the version from 1.72 to 1.8 because major changes in keyboard handling are made. Specifically, the method to enter Greek symbols is fresh – to enter Greek symbols you now use **double-letter strokes**.

The motivation was following:

- typing epsilon and phi symbols was difficult because Alt+E and Alt+F is for menu access
- typing uppercase Greek letters was difficult because three-key combinations were needed – Alt+Shift+letter
- The Alt+number combination is used to generate integer exponents, so I wanted to generalize Alt key usage for the exponent entry

There are also disadvantages of the new method:

- usefulness of double-stroke accelerators to access toolbox items is now severely reduced.
- experienced users who already defined double-stroke accelerators may need to remove their definitions so that they can generate Greek symbols.
**Let me know if you think the new feature is a step back**. - typing pi letter is more difficult now (pinky finger double-stroke)

Okay, what is new in this version:

- As said, Greek symbols are generated by double-stroke (‘aa’ generates alpha).
- You can type exponents by holding down the Alt key. Useful for simple exponents.
- The double-quotation mark can now be used the same as the ^ key – to enter complex exponents. The double-quotation mark is easier to type on certain keyboards.
- As known from earlier beta, you can type indexes by holding down the Spacebar key. Useful for simple indexes.
- You can now use the hash (#) key or the grave accent (`) key to enter a
**simple fraction**– that is, a fraction that only has a single element in the numerator. You type the numerator, hit the # (or `) and then you can type the denominator… Note that the non-equal symbol cannot be entered by the # key any more (sorry Lukas). To enter the non-equal symbol use the /= sequence or the \neq command. - The first version of the
**Autocomplete functionality**is implemented. As you type you may see that Math-o-mir marks some expressions in nearby equations with the orange color. If you hit the dot (.) key at that moment, the orange expression will be copied at your entry position. The Math-o-mir tries to guess expressions that you might want to copy. - The new Ridiculously Complex Tutorial For Keyboard Usage is downloadable form http://www1.datafilehost.com/d/c2a05dd9

Ah, yes… hey Blink, you now have the **1:1 option or bitmaps**. Sorry for the delay.

Finally, the question for French people – How do you guys type the math? Using the Shift key any time you need to type a number seems like a major distraction. Is there any trick you use?

I love the 1:1 option on bitmaps! But whenever I try to copy (by leaving my cursor in the place I want to copy to, then highlighting and clicking the equation to copy) more than one character of an equation, I get the equation surrounded by slash symbols… if I just try to copy one character, it just inserts the character.

I think I know what is bothering you. When you copy a text into a math formula, the text gets surrounded by slash brackets. You probably, not realizing, typed a variable in text mode and when you tried to copy it, Math-o-mir copied it as a text.

But still, there seems something stinky about this feature. I will investigate it more carefully. Thanks.

Extra information… it gives the slash marks when my insertion cursor is in math mode. If I put it into text mode, it copies without the slashes.

Nevermind, I’m having a hard time replicating the problem now…

Thanks for looking into it… I also noticed some random flickers of orange highlight. I couldnt find rhyme or reason about their appearance.

I was thinking about your implementation of sub and superscripts (holding space and alt) and I like the idea. I dont find space a very intuitive modifier (would ctrl be better??) but I think it’s a step in the right direction. A few ideas for “improvements”:

1. Make the modifiers user-assignable. (I understand certain keys like “shift” might have be not allowed.)

2. Highlight the piece of the expression that the sub/super script will apply to as soon as you press space/alt. I think this makes the behavior more obvious, as well as making it obvious which part of a complicated expression will be modified.

3. Consider making the modifier press latching: press it once to enter sub/superscript “mode,” press it again to resume. This could also be made an option. The biggest problem with the shift-_ or shift-^ modifers is that they take two keys. Making the modifier one key helps, but the normal modifiers are far from the number keys.

Those random flickers are probably the new “Autocomplete” feature (still bugs inside). The idea is that Math-o-mir marks an expression, and then you hit the dot (period) key to copy that expression at your cursor location.

The Spacebar is the only key that was free to be used (CTRL is used for toolbox accelerators). The Spacebar also has the advantage that it can be held with the thumb, leaving other fingers free for touch-typing.

1. Good idea. Although I am afraid that I hard-coded many of key combinations and it would be a major job to make it user-assignable. Still, I might take a look if it is possible.

2. Yes. I was already thinking in this direction.

3. Hmm… this idea is also worth thinking about. At the moment I am not sure if this can be made because those same keys are used for many purposes, depending on context. Especially the Spacebar key.

1. I understand it would be tough to do… but I do think it’s much more user-friendly!

3. Let me give a slightly different way of thinking about this, maybe it will help you/us find a better solution (though I think the latching idea could be helpful). Essentially, moving to/from super and subscripts is like walking a tree. From the main expression you can go up or down, then up or down from there, and so on (I wouldnt be surprised if this is how you are modeling expressions currently). Are there any other programs that allow quick and simply tree navigation? How do they do it?

Thinking while typing, this may not be all that helpful, since you now need to navigate siblings and children, forcing more keys to be used. But I like the thought that I could start typing and move up and down (super/subscript) at will within an expression. Anyway, just another thought.

In one moment I had similar idea… I was thinking to use the ALT+arrow_up and ALT+arrow_down combinations to move the cursor into superscript (exponent) or subscript (index) in a latching way. But then I started thinking that this is not much improvement than using ^ and _ keys for the same purpose (in any case you cannot do it with one finger hit). Therefore I somehow dropped that idea.

But the truth is that I actually don’t know how handy some solution will be until I really try it. So maybe ALT+arrow_up and ALT+arrow_down would work better than ^ and _.

But the following seems true:

– when there is only one single, lowercase character to be entered into index or exponent, then non-latching method is faster than a latching method.

– when there is more than three characters to be entered into index or exponent, then latching method seems better than a non-latching method.

Therefore I provided both ways, the ^ and _ as latching methods, and ALT+ and SpaceBar+ for non-latching methods. [On some keyboard layouts typing the ^ key is a nightmare so instead of ^ you can type the ” key.]

I also tried to implement one additional feature when you use the non-latching method (ALT+key) to type and exponent, you can leave your exponent obviously unfinshed (for example, entering ‘2+’ into the exponent) and release the ALT key. When MoM finds the exponent unfinished, it will leave the cursor in the exponent and will switch to the latching method. Unfortunately, this feature seems quite complex and is probably not very intuitive.

Just note that I am even not sure if providing both methods, latching and non-latching, is at all good. Maybe it would be better is we concentrate on single method only, because I am not sure users will ever learn to use both methods optimally. Maybe they will just stick with one of those two methods.