2

While reading through various questions, I noticed that while in code, tags (e.g., iIl1) are pretty easy to distinguish; in the standard font for questions, they are not (YMMV, but for all environments I have seen, they are not):

iIl1

Of course not everyone has all fonts, but I was wondering if the CSS could be a better list, listing less ambiguous fonts earlier? Or maybe more fonts to increase the chances of people hitting a better, readable one? Currently the list is (at least displayed here):

Arial, Liberation Sans, DejaVu Sans, sans-serif

It seems that this is done on math.se and electronics.se at least (Georgia,​ "Times New Roman", ​Times,​ serif), although I have no idea if that's for a reason, or some default setting that SO is deviating from. Or does SO maybe have a guideline somewhere behind the font decision that I overlooked?

The core of the question: do we need to "try harder" to get less ambiguous fonts (mostly for iIl1, but maybe for others too) used for normal question text (probably by choosing a serif font like math.se), or are there other, probably more important reasons to use/keep the current font?

8
  • 2
    If its a variable it should be inline code even within the text; so this this shouldn't be a problem Mar 3 '14 at 21:51
  • @RichardTingle: I have been thinking about this numerous times and today stumbled over ux.stackexchange.com/questions/53341 (not an SO question but I remember having similar issues on SO from time to time). In the mentioned answers, its not possible to distinguish them too (although in this specific case, I don't think its necessary). I just now noticed that the question title has a different font, but trying it out for questions/answers I personally find it rather ugly and hard to read, but it has a different l.
    – PlasmaHH
    Mar 3 '14 at 21:56
  • If you can provide links to a few existing questions where the ambiguity makes the title less useful, it would help to demonstrate the problem better.
    – Pollyanna
    Mar 3 '14 at 22:05
  • @AdamDavis: Not the title, the question (and maybe the answer) body. I will try to dig some up.
    – PlasmaHH
    Mar 3 '14 at 22:08
  • possible duplicate of Title font makes I and l indistinct Mar 3 '14 at 22:45
  • @KeithThompson: Note how I am asking about the text of answers/questions, not the title (which now uses a different font solving that problem anyways). The question is less about the actual font, and more about if we actually care that such situations can arise. Also I find the answer there rather unsatisfying as it boils down to "The alternatives are not pretty, and we might end up with comic sans which sucks". I personally don't see a compelling reason why we don't use the same as math.se for SO.
    – PlasmaHH
    Mar 3 '14 at 22:55
  • Ah, my mistake. (BTW, the characters I and l still look very nearly identical in the title of that question, at least in my browser.) Mar 3 '14 at 23:01
  • @KeithThompson: ah, indeed, in see your comment about it there. It never occured to me that people just dont know how I looks like when they only see l. The whole issue seems to be quite different for new/inexperienced users.
    – PlasmaHH
    Mar 3 '14 at 23:06
6

This is rendered unnecessary by inline codeblocks, denoted by backticks.

Any question which used iIl1 (or any other variable/function name) outside of a codeblock should have that wrapped in backticks to read iIl1.

Backticks are always going to be the better solution, especially considering that words themselves can also be ambiguous, and fonts can't help there:

First I get the variable count then I decrement it

Which is the correct interpretation? Without backticks, it's pretty hard to tell:

  • First I get the variable count then I decrement it?
  • First I get the variable count then I decrement it
  • First I get the variable count then I decrement it
  • First I get the variable count then I decrement it?
6
  • This, exponentially. Don't forget you can edit posts. Mar 3 '14 at 22:42
  • Sure, and key combos even have these nice key like icons. But should we maybe think about if this is really necessary? It seems to me that a lot of questions that might have these problems (key combos as well as non-code text) are asked by rather inexperienced users, and then later edited by more experienced users. For one, during that time, the question is in a less-than-optimal state, and it is also using tiny amounts of resources that might be spent better elsewhere (review queue?).
    – PlasmaHH
    Mar 3 '14 at 22:48
  • I am not saying that either pov is the right one, I am just trying to view it from different points and collect -- more or less -- opinions and maybe some intresting statistics/studies? It might also turn out that this is just a similar case to xkcd.com/1015 and most people don't care at all.
    – PlasmaHH
    Mar 3 '14 at 22:49
  • I would say that, unless we want to go with a monospace font specifically meant for viewing code, we're not going to solve this through font-selection alone. Most questions by new users are in a less-than-optimal state, ambiguous letters are a very, very small problem compared to things like large blocks of code lacking all formatting and indentation.
    – meagar
    Mar 3 '14 at 23:12
  • @meagar: I am not sure what point you are trying to make. The last part sounds like "we have other problems that should be solved first" but the "through font-selection alone" makes me think that it could be a part of it.
    – PlasmaHH
    Mar 4 '14 at 9:15
  • To clarify then: This is a solved problem, and I don't believe there is any value in reviewing our font selection.
    – meagar
    Mar 4 '14 at 13:19

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .