On 15/01/2019 01:17, Asmus Freytag via Unicode wrote:
On 1/14/2019 2:08 PM, Tex via Unicode wrote:
Asmus,
I agree 100%. Asking where is the harm was an actual question intended to
surface problems. It wasn’t rhetoric for saying there is no harm.
The harm comes when this is imported into rich text environments (like this
e-mail inbox). Here, the math abuse and the styled text run may look the same,
but I cannot search for things based on what I see. I see an English or French
word, type it in the search box and it won't be found. I call that 'stealth'
text.
The answer is not necessarily in folding the two, because one of the reasons for having math alphabetics is so you can
search for a variable "a" of certain kind without getting hits on every "a" in the text.
Destroying that functionality in an attempt to "solve" the problems created by the alternate facsimile of
styled text is also "harm" in some way.
That may end up in a feature request for webmails and e-mail clients, where the
user should be given the ability to toggle between what I’d call a “Bing search
mode” and a “Google search mode.” Google Search has extended equivalence
classes that enable it to handle math alphabets like plain ASCII runs, i.e. we
may type a search in ASCII and Google finds instances where the text is typeset
“abusing” math alphabets. On the other hand, Bing Search does not have such
extended equivalence classes, and brings up variables as they are styled when
searching correspondingly.
I won’t blame Google of doing “harm”, and I’d like to position rather on
Google’s side as it seems to meet the expectations of a larger part of end-user
communities. I won’t blame Microsoft neither, I’m just noting a dividing line
between the two vendors about handling math alphabets.
Best regards,
Marcel
