[opengoo] Re: greetings, bugs, suggesions

  • From: Martino <manda.mgf@xxxxxxxxx>
  • To: opengoo@xxxxxxxxxxxxx
  • Date: Fri, 26 Jul 2013 20:01:03 +0200

Hi Ferenec,

I'm happy to listen your interest on working at the project in this moment
I'm not active anymore (or at least near than anymore), but other user and
developer are taking care of OpenGOO.
All your fix and suggestions are very interesting, and if you have a github
account (if not, join github) you can start to apply this patch here:
https://github.com/Mandarancio/OpenGOO

For any question, suggestion or problem ask there!

Thanks again and welcome in the team!

---
Martino
(founder and ex-developer)



2013/7/26 Barta Ferenc <barta.ferenc01@xxxxxxxxx>

> Hi All!
>
> I'm a beginner programmer and I'd like to join your project. I think i
> could do some easy to implement features and fix some simpler bugs.
>
> I have already found and fixed these:
> - .xml-s won't load if the filename was given with extension (see
> OGWorld::Initialize()) because in OGXmlConfig::Open() the existence of the
> file is only checked after adding the filename another extension.
> - If there are no attached balls in the world, OGBall::FindTarget() sets
> pTargetBall_ to null which is dereferenced in OGBall::IsCanClimb() and
> OGBall::Move(), causing segfault.
> - sensors BeginContact() method called both in 
> OGContactListener::**BeginContact()
> and OGContactListener::EndContact(**) sometimes causing strange behavior
> of the pipe and the balls.
> - In ExitSensor::_EndContact(), data->isTouching is only set false to
> attached balls, so force keep applied to unattached balls if pulled away
> from the pipe.
> - Pipe remains open if a ball enters the sensor attached but leaves it
> unattached.
>
> Also, I think it would be better if strands would be built to the nearest
> attached balls when placing one. I could code it, if you think so.
> It would be nice, if we could somehow check if a circle overlaps with
> strands and placed balls. That would make possible for balls to catch
> strands when falling and avoid placing more balls to the same place or over
> an existing strand.
> Finally, I'd like to suggest to recalculate the path of climbing balls
> every timestep rather than only at the endpoints if it's not too demanding
> to the CPU, because now those balls move along the same line even if the
> strand moved away.
>
> Have a nice day!
> Ferenc Barta
>
>
>


-- 


Martino G. Ferrari


()  ascii ribbon campaign - against html mail
/\                                  - against microsoft attachments

Other related posts: