compilation error with new version of AppleClang about gatb-core

[Piezoid]

I think we can safely remove the template keyword in those cases.
It was useful before the refactoring removing template arguments to predecessors/successors.

[clemaitre]

Thanks for the quick answer,
it compiled when I removed the template keyword in the following 4 instructions:

gatb-core/src/gatb/debruijn/impl/BranchingAlgorithm.cpp:155
gatb-core/src/gatb/debruijn/impl/BranchingAlgorithm.cpp:154
gatb-core/src/gatb/debruijn/impl/Terminator.cpp:49
gatb-core/src/gatb/debruijn/impl/Terminator.cpp:199

Can you check if this is safe also in Terminator.cpp ?

Claire

[Piezoid]

Yes it is safe.

Worst case is the compiler complaining about template dependent name that prevent the compiler to infer if the name is a template (needs template) a type (needs typename) or a variable.
A change is semantics is really unlikely.

[rchikhi]

Please go ahead and commit the fix to gatb-core, if you can, one of you two :)

[clemaitre]

Hi Rayan,

I am about to commit... but it seems that this may erase your last visisble commit on github (f25f57b) which is not on the gforge repo : when I git pull from the gforge I fall on f58d96b, did you commit directly on github, forgetting about the gforge ?

Claire

[rchikhi]

oh you're right! thanks for letting me know. some mistake from my part. I've now pushed my commit to gforge, you can now pull (and git rebase -i to avoid the merge) and commit on top of my last commit