[docs] debug failure on v1.11

[odow]

Doc build is currently failing on 1.11

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 97.54%. Comparing base (678e7d5) to head (84a279f).
Report is 4 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #3843      +/-   ##
==========================================
+ Coverage   97.50%   97.54%   +0.03%     
==========================================
  Files          44       44              
  Lines        6017     6017              
==========================================
+ Hits         5867     5869       +2     
+ Misses        150      148       -2     

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

[odow]

So it seems like the TSP tutorial has broken. I have no idea why

[odow]

My conclusion is that the lazy constraint algorithm in GLPK adds local lazy, so it's just pretty terrible at repeating existing work.

[blegat]

Maybe we should use SCIP instead of GLPK but looking at SCIP.jl, it supports MOI.HeuristicCallback and MOI.UserCutCallback but not MOI.LazyConstraintCallback. @matbesancon what is blocking SCIP.jl from adding support to the solver-agnostic lazy callback ?

[odow]

Or we could just add Gurobi since we have a license now...

[matbesancon]

Nothing really, there is a constraint handler mechanism that should be a superset of the lazy constraint, so one could implement a restricted version for LazyConstraintCallback's sake. And this would keep OSS solvers for the tests, for people to be able to try things out.

…

On Wed, Oct 9, 2024, 20:53 Oscar Dowson ***@***.***> wrote: Or we could just add Gurobi since we have a license now... — Reply to this email directly, view it on GitHub <#3843 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AB2FDMUOET3CMPUPO5WRB6TZ2V3SDAVCNFSM6AAAAABPRR75OSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDIMBTGEYTEOBZGU> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[odow]

This is sufficient for now. The tutorial changes were because of differences in Random number streams.