[gambit-list] profiling code being interpreted by gsi
Vincent St-Amour
stamourv at iro.umontreal.ca
Sun May 31 23:31:01 EDT 2009
On Monday, June 1, 2009 at 12:11:18 (+0900), Adrien Piérard wrote:
> > When I replace ##container->file with ##container->path, nothing bad appears
> > to happen.
>
> "When it's not broken, do not try to fix".
> (Which actually is a good moto *only* if the code was initially clean.
> Otherwise, Refactoring should happen ASAP to avoid being damned and
> cursed by next people who will maintain your code.That's why I
> personally think of forcing myself to do test-driven development.)
I've been using statprof with this fix for a couple of weeks, and
everything works fine.
> > Also, did "display" change between 4.0 and 4.4 of gambit? Stat-profiler
>
> I think Marc recently fixed a "bug" in DISPLAY, where printing was
> inconsistent with other implementations or something like that. That
> has been said on that list less than a month ago iirc.
Just replace display by print and it will give the expected result.
Vincent
More information about the Gambit-list
mailing list