Hi! On 27.02.2011 20:58, Aditya Mahajan wrote:

On Sun, 27 Feb 2011, S Barmeier wrote:

...

I don't know if anyone feels the same, but I'm not happy with the spacing of commas in math mode (pictures attached). For instance, typesetting H^1(X,F), the comma seems to be closer to the F than to the X. Equal spacing on both sides or setting the comma closer to the X are I guess a matter of taste (inserting a \neghairspace after the X achieves the latter). In any case, I would kind of prefer equal spacing.

\setupmathematics[autopunctuation=no]

@Hans: I would suggest that this should be made the default.

IMO this does only partially help, consider the following example: \starttext $(B, Y, R, X)$ \setupmathematics[autopunctuation=no] $(B, Y, R, X)$ \stoptext In both lines the spacing after "Y" is way to big. I don't know if that's a bug or just ugly. (version 2011.02.15 16:11)

...

Also, it seems odd that adding a negative space \neghairspace does not change the length of the total formula - I had expected the modified one to be slightly shorter...?

I don't know about this.

...

On a related issue, the spacing for lists, e.g. (0,0,...,0), is not satisfactory. LaTeX provides a command \dotsc for dots between commas (and other commands like \dotsm, \dotsi, \dotsb). I don't know if all of these are necessary, but \dotsc and its cousins are not defined in ConTeXt and both using \ldots or just ... give unsatisfactory results. (It is also curious that the spacing after the first comma and the spacing after the last comma in (0,0,...,0) is evidently different...) Personally, I think I favour something approximating (0, 0, ..., 0) and I find \ldots is spaced too widely, ... too closely, and in any case, the spacing of the dots doesn't match with the spacing of the commas.

Any second opinions?

Adding the \dots(m|i|b|) etc from amsmath is a good idea. I will look into that.

+1. I normally use "$x, \dots\, ,x$" because then the spacing between the periods looks the same as between the commas and the periods, which I prefer. I personally appreciate the wider spacing in \dots. Stefan

On Mon, Feb 28, 2011 at 00:10, Mojca Miklavec wrote:

On Sun, Feb 27, 2011 at 23:49, Stefan Müller wrote:

...

\starttext $(B, Y, R, X)$ \setupmathematics[autopunctuation=no] $(B, Y, R, X)$ \stoptext

In both lines the spacing after "Y" is way to big. I don't know if that's a bug or just ugly. (version 2011.02.15 16:11)

In text parts this is known as kerning. In OpenType math there are complex bounding boxes.

(I accidentally sent the email too early.) I wanted to add that unless/until somebody implements complex bounding boxes (OpenType Math font for whatever font you are using), you cannot avoid the problem. TeX has no idea how big Y is, it only knows its rectangular bounding box. Cases like this one have to be manually tuned. Mojca

On Sun, Feb 27, 2011 at 06:14:36PM -0500, Aditya Mahajan wrote:

On Mon, 28 Feb 2011, Mojca Miklavec wrote:

...

...

IMO this does only partially help, consider the following example:

\starttext $(B, Y, R, X)$ \setupmathematics[autopunctuation=no] $(B, Y, R, X)$ \stoptext

In both lines the spacing after "Y" is way to big. I don't know if that's a bug or just ugly. (version 2011.02.15 16:11)

In text parts this is known as kerning. In OpenType math there are complex bounding boxes.

I would call it a bug. The spacing is OK with cambria, but not with xits.

Because Cambria implements math kerning that Mojca is referring too, XITS don't (and will not in the near future unless someone steps to implement it; pretty tedious job). Regards, Khaled -- Khaled Hosny Egyptian Arab

Some time ago I was wondering if there's something like "kerning" in math mode, too. Does that mean it would work if I switch to cambria math fonts?

It would at least work better for cases such as different spacing between letter F, Y or X and comma (where F and Y leave too much empty space). It would not effect \hairspace (this might be a bug, but I didn't test) and it would not effect space *after* comma (unless maybe for Y). Autospacing for commas is a different issue, but Cambria could solve "bad kerning". Mojca

On Mon, Feb 28, 2011 at 00:14, Aditya Mahajan wrote:

On Mon, 28 Feb 2011, Mojca Miklavec wrote:

...

On Sun, Feb 27, 2011 at 23:49, Stefan Müller wrote:

...

Hi!

On 27.02.2011 20:58, Aditya Mahajan wrote:

...

On Sun, 27 Feb 2011, S Barmeier wrote:

...

I don't know if anyone feels the same, but I'm not happy with the spacing of commas in math mode (pictures attached). For instance, typesetting H^1(X,F), the comma seems to be closer to the F than to the X. Equal spacing on both sides or setting the comma closer to the X are I guess a matter of taste (inserting a \neghairspace after the X achieves the latter). In any case, I would kind of prefer equal spacing.

\setupmathematics[autopunctuation=no]

@Hans: I would suggest that this should be made the default.

IMO this does only partially help, consider the following example:

\starttext $(B, Y, R, X)$ \setupmathematics[autopunctuation=no] $(B, Y, R, X)$ \stoptext

In both lines the spacing after "Y" is way to big. I don't know if that's a bug or just ugly. (version 2011.02.15 16:11)

In text parts this is known as kerning. In OpenType math there are complex bounding boxes.

I would call it a bug. The spacing is OK with cambria, but not with xits.

But that is probably a "bug" in font, not in ConTeXt; and something that cannot really be handled in LM or in MKII at all. I already sent a similar report to Khaled (when I was testing XITS in Word), but he was somehow reluctant to change metrics (it has to be done carefully for the whole font, it needs a lot of time & work; might become obsolete with next release of stix; and just fixing two out of thousand such "bugs" doesn't really help). Mojca (who didn't try out the examples)