Anyone compiled Omegalib with Equalizer 2.0?

2 views
Skip to first unread message

Ben Simons

unread,
Nov 5, 2019, 10:50:56 PM11/5/19
to omegalib
Hi,

I took a look at building Equalizer 2.0 last week, and it
builds very cleanly, and very quickly. Very good.

I'm looking at now compiling Omegalib against Equalizer 2.0 and
am wondering if anyone else has tried this too? 

I'm also building against a new boost - v1.65
in which 'numeric' became 'numpy'. It's easy enough to update..

It gets most of the way okay. You have to supply an equalizer.tar.gz
for the build, but that's fine. it gets the version 2.0

I'm at a point where i need to read a bit (grok code) and so I thought
i'd write this post. It seems Equalizer split a type  eq::ConfigEvent

In the CHANGES.md that comes with Equalizer this is commented:

06/Sep/2012
 New API in eq::Config & seq::ViewData for event handling. sendEvent() &
 handleEvent() have changed signatures, getNextEvent() supersedes
 tryNextEvent() & nextEvent().

 eq::ConfigEvent is replaced by EventOCommand for sending and by
 eq::EventCommand for receiving events.
 Old API is deprecated and will be removed in version 2.0.

actually I think there's a typo: EventCommand should be EventICommand  
(nb the missing "I")

One affected Omegalib file is:

  ./src/displaySystems/Equalizer/eqinternal.h

with
    virtual bool handleEvent(const eq::ConfigEvent* event);

but i'm not sure if this should become EventOCommand or EventICommand
or in fact there needs to be a bit of a rewrite. 
Do we overload handleEvent() for both types now?

I guess it becomes:

    virtual bool handleEvent(const eq::EventICommand* event);

There's a few more, similar, changes.
Just checking anyone else has already been down this path and
would like to compare notes.

cheers,
ben.


Reply all
Reply to author
Forward
0 new messages