[docs] add warning about Y <= X, Set() syntax

[odow]

x-ref #3766

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 98.38%. Comparing base (886a652) to head (99cf38d).
Report is 1 commits behind head on master.

Additional details and impacted files

@@           Coverage Diff           @@
##           master    #3769   +/-   ##
=======================================
  Coverage   98.38%   98.38%           
=======================================
  Files          44       44           
  Lines        5879     5885    +6     
=======================================
+ Hits         5784     5790    +6     
  Misses         95       95           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[LebedevRI]

Note that as per https://docs.julialang.org/en/v1/manual/unicode-input/
⪰ can be "trivially" inputted as \succeq.

Wouldn't it be possible to:

@constraint(model, x \succeq y)
# error: specify one of the conic sets

@constraint(model, x >= y, <... Some conic set ...>)
# warning: use \succeq instead

@constraint(model, x \succeq y, <... Some non-conic set ...>)
# warning: use >= instead

@constraint(model, x >= y)
# becomes @constraint(model, x - y, Nonnegatives)

@constraint(model, x <= y)
# becomes @constraint(model, x - y, Nonpositives)

Is the concern that someone writes @constraint(model, x >= y) thinking it means PSDCone,
it silently works, and doesn't do what's expected? Or is there some other concern?

[odow]

Is the concern that someone writes @constraint(model, x >= y) thinking it means PSDCone,
it silently works, and doesn't do what's expected?

Yes.

We won't be adding support for \succeq. There is too much risk of confusion.

[LebedevRI]

:/ I really feel such a decision should be made with a larger headcount than i see here.

[odow]

I'll open a new issue