Am Mittwoch, den 08.11.2006, 14:39 -0800 schrieb Andy Ross:
> > People apparently got used to the state that FlightGear typically
> > has a CVS tree that you can compile at the end of a development day
> > and 'fly'.
>
> Remember that people doing aircraft models and scenery are also
> "develop
On Thursday 09 November 2006 00:00, Andy Ross wrote:
> Douglas Campos wrote:
> > but content developers can't just stick with plib branch? afaik
> > we'll only making the "porting work" at trunk, right?
>
> No, they can't; not if (by Mathias's suggestion) new features are
> added only to the head
On Wednesday 08 November 2006 23:30, John Wojnaroski wrote:
> PRESS ON
>
> I am particularly interested in the ability to run multiple views from a
> single application. Opens the possibility to use a dual headed AGP
> video card, and perhaps a few PCI graphics cards in a single machine and
> crea
Curtis Olson wrote:
> On 11/8/06, Lou Sanchez-Chopitea <[EMAIL PROTECTED]> wrote:
>>
>> Hi,
>> John Wojnaroski wrote:
>>
>> >If this had been a commercial development and the product had been
>> >released in this state, we would all be updating our resumes at this
>> >point...
>> >
>> >
>> If
Andy Ross schrieb:
> ..
>
> I'll just say it: the OSG port is, as of today, an unmitigated
> disaster
Sorry, but no, no , no!
> that should *never* have been applied to CVS in the state it
> is in. Apologies in advance if that offends someone.
>
> Andy
>
>
Hi Andy,
as I am no core developer
On 11/8/06, Lou Sanchez-Chopitea <[EMAIL PROTECTED]> wrote:
Hi,John Wojnaroski wrote:>If this had been a commercial development and the product had been>released in this state, we would all be updating our resumes at this>point...>>If this had been a commercial product, it would have been rele
Frederic Bouvier wrote:
> Sorry, but I only understood that Mathias is not willing to backport
> new features. He never said no one should do.
It's possible that I misinterpreted, and maybe Mathias would like to
clarify. But FWIW I thought he was pretty unambiguous:
: I would like to restric
Selon Andy Ross :
> Douglas Campos wrote:
> > but content developers can't just stick with plib branch? afaik
> > we'll only making the "porting work" at trunk, right?
>
> No, they can't; not if (by Mathias's suggestion) new features are
> added only to the head and not to the plib branch. See
Douglas Campos wrote:
> but content developers can't just stick with plib branch? afaik
> we'll only making the "porting work" at trunk, right?
No, they can't; not if (by Mathias's suggestion) new features are
added only to the head and not to the plib branch. See his post a few
messages up.
T
but content developers can't just stick with plib branch? afaik we'll
only making the "porting work" at trunk, right?
I hope that before new features to trunk, we'll fix the blockers (for
me, just "aesthetics", like 3d clouds, etc)
just my 1/2 cent
(if we continue to give 2c, Curtis will be able
> People apparently got used to the state that FlightGear typically
> has a CVS tree that you can compile at the end of a development day
> and 'fly'.
Remember that people doing aircraft models and scenery are also
"developers", and need to be able to run the development version to do
their wor
Hi Martin,
You make a very good point. All of CVS is development. However many of
us are not experts and facile with cutting-edge design and development (
and I include myself in that set ) and are more comfortable working with
a stable and proven design and taking small steps. You're absolut
Hi,
John Wojnaroski wrote:
>Guess I'll jump in too with my two cents
>
>I have OSG and FlightGear CVS installed and kind of running -- at least
>it starts ;-) but have not had any time in the past week and a half to
>go any further. But I have to take Andy's side on this one, not so
>muc
Hi John,
John Wojnaroski wrote:
> If this had been a commercial development and the product had been
> released in this state, we would all be updating our resumes at this
> point...
No doubt. Still the point is: The product is not intended to be
released in this state. In fact, current state
Guess I'll jump in too with my two cents
I have OSG and FlightGear CVS installed and kind of running -- at least
it starts ;-) but have not had any time in the past week and a half to
go any further. But I have to take Andy's side on this one, not so
much based on technical issues (all w
Curtis L. Olson wrote:
> Andy Ross wrote:
> > Look, the points wasn't that plib is great. The point wasn't that
> > OSG has no advantages.
>
> I'll just jump in here with a couple quick comments. OSG does have
> advantages that we should be able to realize pretty quickly, it is not
> compl
On 11/8/06, Andy Ross <[EMAIL PROTECTED]> wrote:
Ooh, it's a bona fide flame war. :)Look, the points wasn't that plib is great. The point wasn't that OSGhas no advantages.I'll just jump in here with a couple quick comments. OSG does have advantages that we should be able to realize pretty quickl
Andy Ross wrote:
> Ooh, it's a bona fide flame war. :)
Oh great, we didn't have one for a long time well not
necessarily, but getting back to a reasonable discussion requires that
_both_ sides accept that there might be valid views on a topic, which
(the views) reside outside their persona
Ooh, it's a bona fide flame war. :)
Look, the points wasn't that plib is great. The point wasn't that OSG
has no advantages. The point was that we've taken working software
and regressed its feature set pretty severely, and that's a serious
problem that needs to be fixed now. Stopping developm
Andy Ross wrote:
> > I would like to restrict that a bit. For bugfixes and non
> > developers this might be a good idea. But please do not develop
> > new features on the branch. I know how many problems this will
> > give. And to be honest, Olaf I believe You know what I am
> > talking abo
On 11/7/06, Andy Ross <[EMAIL PROTECTED]> wrote:
This is a big, disruptive change, and I'm sympathetic to you,really. YASim and Nasal were big and disruptive too. But sofar, OSG has produced literally zero benefit for anyone.People's experience has been anywhere between "it seems to work
OK" to "
> I would like to restrict that a bit. For bugfixes and non
> developers this might be a good idea. But please do not develop
> new features on the branch. I know how many problems this will
> give. And to be honest, Olaf I believe You know what I am
> talking about ...
No offense, but the
Olaf Flebbe wrote:
> Can somebody confirm that the framerate with OSG is better compared to
> plib on Linux? Default c172 at KSFO, please.
I'd suggest to narrow the environment conditions to something
reproducable, because different features that people activate via
properties might have differe
On Monday 06 November 2006 23:52, Olaf Flebbe wrote:
> I think too that we have to support the plib branch quite a while. The
> idea to do the OSG switch in CVS head was that the OSG developers do not
> have the efforts for porting features from branch back to HEAD. So
> please work in the pre-OSG
On Monday 06 November 2006 13:17, Martin Spott wrote:
> "Vivian Meazza" wrote:
> > Let's hope that you're correct. Right now we have _fewer_ features, and a
> > _lower_ frame rate. I hope that adding back the features we used to have
> > won't further reduce the frame rate, let alone adding new one
On Monday 06 November 2006 11:23, Frederic Bouvier wrote:
> I hope the total freedom of using features given by OSG will encourage us
> to think about more clever ways to do things, that could be beneficial in
> the end. For instance, using multitexturing instead of cutting tiny
> triangle in the s
On Sunday 05 November 2006 23:43, Olaf Flebbe wrote:
> Uncommenting sceneView->update() in render.cxx gives a performance jump
>
> >from 60FPS to 80FPS. (plib was 100FPS). Traversing the scenegraph seems
>
> to be quite timing relevant. Can we limit this special update-traversal
> only to the aircr
Hi,
I think too that we have to support the plib branch quite a while. The
idea to do the OSG switch in CVS head was that the OSG developers do not
have the efforts for porting features from branch back to HEAD. So
please work in the pre-OSG branch until the performance issue is solved.
But pl
Hi,
> I hope the total freedom of using features given by OSG will encourage us to
> think about more clever ways to do things, that could be beneficial in the
> end.
Thanks Frederic. I too think, there are many things which can be
improved considarably and which will lead to higher performance
Fred wrote:
> Selon Vivian Meazza :
>
> > Yes, I saw you had fixed that, but that's the only one with that
> particular
> > problem. Since it worked under plib, and AC3D didn't complain, I would
> never
> > have worked that one out - thank you. If only the rest were that easy -
> the
> > others s
Hi,
I have on win (msvc express, with Olafs easy to use project files) on a
P4M 2,2Ghz 1GB GeForce FX Go 5600 about half the frame rate with osg
than with plib. The frame rate is so low (~10fps), that fg is not
flyable (and without the plib branch it would be not possible, to
check/improve FDMs
Selon Vivian Meazza :
> Yes, I saw you had fixed that, but that's the only one with that particular
> problem. Since it worked under plib, and AC3D didn't complain, I would never
> have worked that one out - thank you. If only the rest were that easy - the
> others seem to be keyboard/panel contro
"Vivian Meazza" wrote:
> Yes, I saw you had fixed that, but that's the only one with that particular
> problem. Since it worked under plib, and AC3D didn't complain, [...]
In his 'announcement' Mathias has been mentioning that the AC3D-loader
still requires some further finishing. I wouldn't worr
Frederic Bouvier
> Quoting Vivian Meazza :
>
> > Martin Spott
> >
> > > "Vivian Meazza" wrote:
> > >
> > > > Let's hope that you're correct. Right now we have _fewer_ features,
> and
> > > a
> > > > _lower_ frame rate. I hope that adding back the features we used to
> have
> > > > won't further r
Quoting Vivian Meazza :
> Martin Spott
>
> > "Vivian Meazza" wrote:
> >
> > > Let's hope that you're correct. Right now we have _fewer_ features, and
> > a
> > > _lower_ frame rate. I hope that adding back the features we used to have
> > > won't further reduce the frame rate, let alone adding new
Martin Spott
> "Vivian Meazza" wrote:
>
> > Let's hope that you're correct. Right now we have _fewer_ features, and
> a
> > _lower_ frame rate. I hope that adding back the features we used to have
> > won't further reduce the frame rate, let alone adding new ones.
>
> Hey, guys, you're talking a
"Vivian Meazza" wrote:
> Let's hope that you're correct. Right now we have _fewer_ features, and a
> _lower_ frame rate. I hope that adding back the features we used to have
> won't further reduce the frame rate, let alone adding new ones.
Hey, guys, you're talking about a _development_ source co
Quoting Vivian Meazza :
> Let's hope that you're correct. Right now we have _fewer_ features, and a
> _lower_ frame rate. I hope that adding back the features we used to have
> won't further reduce the frame rate, let alone adding new ones.
>
>
>
> We're in danger of limiting the use of FG to high
Vivian Meazza wrote:
> Let’s hope that you’re correct. Right now we have _/fewer/_ features,
> and a _/lower/_ frame rate. I hope that adding back the features we used
> to have won’t further reduce the frame rate, let alone adding new ones.
One area which will give quite a performance boost wil
Let’s hope that you’re correct. Right now we
have _fewer_ features, and a _lower_ frame rate. I hope that adding back
the features we used to have won’t further reduce the frame rate, let
alone adding new ones.
We’re in danger of limiting the use of FG to high end
machines. I have a rea
Le lundi 6 novembre 2006 00:58, Olaf Flebbe a écrit :
> Can somebody confirm that the framerate with OSG is better compared to
> plib on Linux? Default c172 at KSFO, please.
> Olaf
Gentoo linux x86_64, AMD 4200 dual core, 2Go RAM, nvidia 6800GS 512Mo
with the same options, at the same place (KSFO)
Olaf Flebbe schrieb:
> ..
>
> Can somebody confirm that the framerate with OSG is better compared to
> plib on Linux? Default c172 at KSFO, please.
> Olaf
>
No, these are my results, same FG CVS version with OSG/PLIB FG:
OpenSuse x86_64, AMD 3700, 1GB RAM, Nvidia 6600GT 256 MB
NVidia driver:
> -Original Message-
> From: [EMAIL PROTECTED] [mailto:flightgear-
> [EMAIL PROTECTED] On Behalf Of Olaf Flebbe
> Sent: 05 November 2006 22:43
> To: FlightGear developers discussions
> Cc: Mathias Fröhlich
> Subject: [Flightgear-devel] OSG Performance on WIndow
I have no idea, but I do know that plib/ssg was very tight and clean and fast code. OSG is very feature rich, it has far more capabilities than plib/ssg, but features usually come with a price. I wouldn't be surprised if we take some amount of a frame rate hit moving to OSG. Hopefully we can fin
Hi,
while double (and triple ;-) - checking everthing I discovered that I
had a prerelease of Mathias overhauled AC3D Loader in my 3rdparty.zip.
You may find an update on my website. Sadly, this doesn't change framerate.
Uncommenting sceneView->update() in render.cxx gives a performance jump
f
45 matches
Mail list logo
