YARP compilation failure in Travis about robotology-superbuild

[traversaro]

Related to robotology/yarp#1466 .

[traversaro]

Apparently you can't compile YARP master with YCM devel (introduced by traversaro#2). : (

[claudiofantacci]

Apparently you can't compile YARP master with YCM devel (introduced by #2). : (

I think it's something very recent, and may not be a problem starting for the next release.
@drdanz may clarify this.

[traversaro]

The "cleanest" solution I can think of is to rollback robotology/ycm@dfdb8d8 .

[claudiofantacci]

I updated my previous post that was incomplete. I report it here for completeness:

I think it's something very recent, and may not be a problem starting for the next release.

As far as

The "cleanest" solution I can think of is to rollback robotology/ycm@dfdb8d8 .

is concerned, it all dependes on @drdanz.
The idea with that patch is to stop using ACE's string and use the standard one.

[drdanz]

The "cleanest" solution I can think of is to rollback robotology/ycm@dfdb8d8 .

I will not revert that, ACE_HAS_STRING_HASH was true in every ACE version that we support (and has been true for a long time). Instead I could add a set(ACE_HAS_STRING_HASH 1) with a variable_watch that prints a warning when the variable is used.
@traversaro does this sound ok to you?

[traversaro]

@traversaro does this sound ok to you?

That's great!

[drdanz]

Done, please try robotology/ycm@85bf58f

[traversaro]

Fixed by #11 , cc @lucaTagliapietra .