On 08/08/2013 12:06 PM, bieniasz wrote:
Richard Heck <rgheck <at> lyx.org> writes:

What happens, then, when you add a new equation somewhere in
the middle of the document?
This is what LyX already does - it updates automatically the equation
numbers. My point is why one cannot use these numbers for referencing
the equations in the text, since they are already there, up-to-date.

What you really want is the auto-labelling that I described, it seems to me: Equations act as if they have labels associated with them, but the user does not actually see the label.

Thinking about it for a bit, I think this might not be that hard to do just for equations. When we've thought about it in the past, we've included sections, chapters, and so forth, and then it is harder.

There has often been talk about automatic labelling, so that the
equation would have its own "magic" label, so to speak, and you
could reference it without needing to make up a label. But no one
has found the need pressing enough to do it.
This is difficult to understand for me. LaTeX users are primarily
scientists who write scientific texts. In such text one ALWAYS
refers to equations by numbers, and not by any peculiar labels.

Of course. But when actually writing LaTeX (as opposed to LyX), one uses labels, since one wouldn't want to have to change the reference just because an equation gets moved.

As Curtis said, too, giving equations, sections, etc, meaningful labels makes it easier to figure out which one to reference later. So that is what most of us do. Indeed, it isn't really true that one always refers to equations by number. In my own papers, particular important equations (or formulae) often have labels and not just numbers. E.g., (Sat).

So, in conclusion, I daresay the LaTeX/LyX system for equation numbering needs a reasonable revision. If there are any LyX programmers out there, please do something about this!!!

As always with open source, it's a question of time and bodies. Those of us who work on this just haven't found the need pressing.


Reply via email to