[Opensim-dev] Still on Sim and Phys Frames per Second (FPS)
Michael Emory Cerquoni
nebadon2025 at gmail.com
Tue Nov 10 22:56:13 UTC 2015
Doug I participated in MOSES grid as well and my experience there was
terrible, far worse performance than i experienced in any other grid, I
took part in the FCVW build and planning and experienced a multitude of
problems on MOSES platform that just do not exist in the core opensimulator
software. And this is what I mean by chasing ghosts, MOSES is fixing bugs
in MOSES for MOSES that just do not exist in the core software. You can
feel however you want and if you feel embarrassed then go work on MOSES
software, no one is stopping you. I do agree though that this whole thing
is quite a huge embarrassment for the project. It still does not change
the fact that to date no improvements have come from this change and all
its done is cause arguments, the reason their code was not accepted is
because it was not suitable for core, end of story. They wanted us to
accept patches unconditionally and sorry, that is not going to happen.
On Tue, Nov 10, 2015 at 5:48 PM, dz <dz at bitzend.net> wrote:
>
> On Tue, Nov 10, 2015 at 9:06 AM, Michael Emory Cerquoni <
> nebadon2025 at gmail.com> wrote:
>
>> ..... Also if that was the intended goal why was this not coordinated
>> prior to the break, to just go ahead break something and then call it
>> progress while leaving stuff broken and then say oh someone else should fix
>> that is quite unprofessional in any setting. We need to resolve this
>> problem of viewer development or quite honestly this whole thing is dead in
>> its tracks, without a constantly improving viewer OpenSim is looking more
>> and more like a dead end. That said its never to late to revive things and
>> start wallking the path to improvement, but as a group we need to stop
>> focusing on the wrong things.
>>
>
> HUH??? Coordinated??? An independent group of developers who WANT
> OpenSim to work used the appropriate forum to ask everyone about the
> cause, and asked for suggestions on how to correct the problem... A
> patch was generated and went through 3 months of iterations and VERY
> open discussion.. The whole point of that was to notify the people who
> participate and garner feedback from anyone/everyone in the community.
> The FACT is there was general agreement among most of CORE and the
> other participants that THIS WAS a step forward. I cant find
> anywhere in any of these early discussions where anyone expressed the
> unprofessional attitude you assert.
>
>
>
>> What i see is people chasing ghosts of problems that are not the real
>> core problems of what this project has and needs, with little to zero
>> improvements as a result. Can anyone name a single improvement that has
>> come from changing the stats? Where are the patches, where are the
>> scientific write ups showing that this was a success, so far to me this
>> whole thing with stats seems like a big distraction that is not only not
>> beneficial so far, its causing strife between the developers.
>>
>
> Chasing Ghosts?? really, you don't want to go there AT ALL... WHERE
> is the road map of the REAL CORE PROBLEMS??? How many times have people
> stood and said I'm ready, willing, and able to help, Please tell me
> what I should focus on? If you want to look for a failure of
> communication I suggest you start there before you start blaming the folks
> who have followed the very public lead of core and picked a problem that
> is important to them!
>
> Where are the papers?? The consensus of the participants in the
> discussion was that the made up numbers were impacting the credibility
> of those who were already publishing. It is also extremely misleading
> to categorize MOSES as the only group interested in conducting these
> performance measurement/improvement tests and publishing results...
> Maybe you should review some of Christas' publications,, or the
> serious gaming PhD thesis whose author bothered to speak up in favor,
> Or maybe you just forget the years of Intel projects??? If you actually
> READ the communications from MOSES you would understand that they
> could NOT publish results of all the previously testing knowing that
> the results were inaccurate.
>
>
>
>> Personally I don't have the solutions, my time is very limited anymore
>> and I cant spend the time I have in the past testing things and
>> coordinating people like I have, we need more people to step up and do the
>> right thing without making people feel like its being shoved down their
>> throats.
>>
>>
> AMEN! We do need people to step up... and a bunch of us did. We
> were publicly ridiculed ( and that ridicule continues. ) We jumped
> through ALL the hoops, we communicated with everyone we were told
> needed to be involved, MOSES reworked, and resubmitted patches. They
> spent the time to attempt to communicate WHY this was an important step
> forward. We welcomed the discussion.. and honestly until the other day
> It seemed like it was a success... All of a sudden, In the space of
> 3 days, we are informed that some mysterious user has whispered
> their annoyance about an OBSOLETE feature in one of the viewers , and
> because of this "comment" our efforts would be ignored in favor of a
> solution proposed and implemented in the "backroom". Who is shoving
> what down whose throat?
>
> The community has spoken on the issue of incorrect performance
> measurement figures being reported and agreed it IS a step forward. The
> fact is, Melanie could have added her solution to the code base on her
> grid in minutes and could have avoided this discussion altogether.
> There is NO REASON why her fix needed to be included in core. Her
> assertion that "someone else can recode the stats to use a new method
> of reporting " is arrogant and ignores the fact that it is the most
> complex solution to implement... (sound familiar to the unprofessional
> attitude you attributed elsewhere??) She has demanded that the PHYSICS
> FPS reporting field already provided in the viewer be populated with
> FALSE data and seems to think it is reasonable that MOSES repeat the
> tortuous affair to re-code ANOTHER solution and go through the process
> of convincing her it is technically correct. Please just ask
> yourself.... How inclusive is that?? Why would anyone who saw any of
> this step forward and volunteer to do what members of core have been
> pushing folks to do since 2009?
>
> THAT, pure and simple, is the reason we cannot get people interested
> in continuing to work with the project... That is not to say that the
> work won't continue ON the project, it will just continue to be done in
> splintered efforts by people who are basically fed up with dealing
> with this disregard for the people who make the effort to participate
> in this forum.
>
> Just so I am clear... I AM NOT a member of the MOSES development
> group.. but I am a supporter of their efforts.. Outside of the time I
> spent with Intel on the Science Sim grid, they are the most dedicated,
> competent, and forward looking of the development groups interested in
> OpenSim I have had the pleasure to work with. In my opinion,, if core
> cant extend a hand and figure out a way to work with this group, they
> are carving a BIG R.I.P. on the tombstone of OpenSim as we know it...
> MOSES will build a simulator that dramatically improves the physics
> capabilities and performance, They are likely to be the first to
> implement an HTML based viewer, and (If we are lucky) they will
> implement a scheme of distributed simulator services that will integrate
> with the future of cloud based apps. I am proud to be allowed to
> participate in their efforts, and , at the moment totally embarrassed
> by how this project has reacted to them.
>
> Doug Osborn
>
>
> _______________________________________________
> Opensim-dev mailing list
> Opensim-dev at opensimulator.org
> http://opensimulator.org/cgi-bin/mailman/listinfo/opensim-dev
>
>
--
Michael Emory Cerquoni
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://opensimulator.org/pipermail/opensim-dev/attachments/20151110/c045d5f2/attachment.html>
More information about the Opensim-dev
mailing list