Migration from controlboard to controlboardwrapper2

[EnricoMingo]

In the last updated version of YARP from Alberto we have a new devicedriver that is the controlboardwrapper2 (that is the same one present in the real robots COMAN and iCub). We have to update the simulator too. We have tried to make it works but we have some problems, I think a new branch will be needed. "Controlboard" still works, we have new ports for reading torques and velocity so in the future the extra theard to publish these informations in the simulator will be deleted. Alberto will be the one that will work on these migration so pls do not change the structure of the simulator from now on.

[traversaro]

Couldn't alberto work from the refactored code in https://github.com/robotology/gazebo_yarp_plugins/tree/renaming to avoid redoing the work?

[EnricoMingo]

I don't know what is the best way to do this migration, I was thinking to do not go further implementing other things in the functionality to let Alberto work with no other surprise.

[traversaro]

It is just refactoring/changing names to classes, not adding new functionality (I expressly did the work in the italian night so It could be integrated this morning without impacting any kind of new work). If necessary I can do the renaming work also after Alberto adds the new functionality, but I guess now the renaming branch is functionally identical to the master branch (I avoid directly commiting it only because I wanted someone else to test/review it before commiting).

[EnricoMingo]

From which we saw this morning is not so simple. There are different changes to do. Maybe Alberto can better clarify the types of changing are needed.

[traversaro]

No, I was referring to the changes in the renaming branch #31 (comment) that I did this night, they are simply refactoring/renaming classes as already discussed in #31.
My proposal would be that we merge this changes related to renaming (if they are not ok we can still discuss in #31) and then add new functionality.

[EnricoMingo]

ok, I think there are no problems.

[barbalberto]

Ok, I'll start from the renaming branch.
I'll do some quick fix in order to migrate to the new wrapper early and then the new feature will be added step by step.

I just have to understand a little how the coman device was implemented.

[EnricoMingo]

Ok perfect!

[traversaro]

I have made some modification (only in filenames and directory structure) to renaming https://github.com/robotology/gazebo_yarp_plugins/tree/renaming branch as discussed in the hangout.

[traversaro]

@barbalberto @EnricoMingo Can I delete the Alberto branch? It is inconsistent and make that test fail.

[EnricoMingo]

Yes
Il giorno 20/nov/2013 08:16, "Silvio Traversaro" notifications@github.com
ha scritto:

@barbalberto https://github.com/barbalberto @EnricoMingohttps://github.com/EnricoMingoCan I delete the Alberto branch? It is inconsistent and make that test
fail.

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/36#issuecomment-28868066
.

[traversaro]

Done.

[barbalberto]

Ah, ok.

I saw the error didn t understand the source.
I'll fix it.

I need the branch 'cause I doing other changes.

Alberto

On 20/11/2013 08:16, Silvio Traversaro wrote:

@barbalberto https://github.com/barbalberto @EnricoMingo
https://github.com/EnricoMingo Can I delete the Alberto branch? It
is inconsistent and make that test fail.

—
Reply to this email directly or view it on GitHub
#36 (comment).

[traversaro]

@barbalberto My fault, I wrongly understood by @EnricoMingo reply that you were ok in deleting the branch, I already deleted it.
However it was just a pointer to the commit aaf9fb9 so I guess you can simply fork again this commit (or directly fork the master).

[EnricoMingo]

Sorry too, since the branch was merged I thought that the work for that
branch was finished. My fault.
Il giorno 20/nov/2013 09:50, "Silvio Traversaro" notifications@github.com
ha scritto:

@barbalberto https://github.com/barbalberto My fault, I wrongly
understood by @EnricoMingo https://github.com/EnricoMingo reply that
you were ok in deleting the branch, I already deleted it.
However it was just a pointer to the commit aaf9fb9aaf9fb9cc539324cd13433676402db93f1824175so I guess you can simply fork again this commit (or directly fork the
master).

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/36#issuecomment-28871690
.

[barbalberto]

Well, no big deal.
I'll create a new one with a better name :)

Alby

On 20/11/2013 09:52, EnricoMingo wrote:

Sorry too, since the branch was merged I thought that the work for that
branch was finished. My fault.
Il giorno 20/nov/2013 09:50, "Silvio Traversaro"
notifications@github.com
ha scritto:

@barbalberto https://github.com/barbalberto My fault, I wrongly
understood by @EnricoMingo https://github.com/EnricoMingo reply that
you were ok in deleting the branch, I already deleted it.
However it was just a pointer to the commit
aaf9fb9aaf9fb9cc539324cd13433676402db93f1824175so
I guess you can simply fork again this commit (or directly fork the
master).

—
Reply to this email directly or view it on
GitHubhttps://github.com//issues/36#issuecomment-28871690
.

—
Reply to this email directly or view it on GitHub
#36 (comment).

[traversaro]

Just before closing this issue, a question for @barbalberto : it makes sense that you get a

yarpdev: ***ERROR*** could not find device <controlboardwrapper2>

If you have iCub 1.1.13 installed from binary package in Ubuntu (and Yarp not compiled with YARP_COMPILE_EXPERIMENTAL_WRAPPERS )? Apparently I wrongly assumed that the device was available from the last iCub binary.

[barbalberto]

Yes, it is correct.
Controlboardwrapper is still available in icub source, but gazebo plugin has to be indepent from icub (that's exactly the reason why we moved the wrappers from icub to yarp) therefore plugins doesn't look into icub code.

Anyway wrappers will soon be removed from icub repo and the only ones available will be the yarp ones (and the flag will also go away).

Silvio Traversaro notifications@github.com Ha scritto:

Just before closing this issue, a question for @barbalbertohttps://github.com/barbalberto : it makes sense that you get a

yarpdev: _ERROR_ could not find device

If you have iCub 1.1.13 installed from binary package in Ubuntu (and Yarp not compiled with YARP_COMPILE_EXPERIMENTAL_WRAPPERS )? Apparently I wrongly assumed that the device was available from the last iCub binary.

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/36#issuecomment-30047004.

[traversaro]

I guess the migrations is completed, and we can close this issue (bugs in the implementation of the controlboardwrapper2 can be discussed in specific issues).