This is Interesting: Free Magazines for Graphics designers and webmasters  


Home > Archive > VRML > November 2005 > X3D, IndexedFaceSet, "results are undefined" and profiles





You are viewing an archived Text-only version of the thread. To view this thread in it's original format and/or if you want to reply to this thread please [click here]

Author X3D, IndexedFaceSet, "results are undefined" and profiles
Joerg Scheurich aka MUFTI

2005-11-09, 7:01 pm

Hi,

Can someone tell me the sense behind the following senses of ISO/IEC 19775:2005
(X3D):

The following criteria are required from each IndexedFaceSet
"otherwise the results are undefined" in ISO/IEC 19775:2005 13.3.6

- planar polygons
- 3 non-coincident vertices per polygon
- no self-intersection polygons

In ISO/IEC 19775:2005 table B3 (interchange profile) this criteria are
repeated as requirement for the interchange profile, but they are not
repeated in ISO/IEC 19775:2005 table E.3 (immersive profile) and
ISO/IEC 19775:2005 table F.3 (full profile).

Possible conclusion:

IndexedFaceSets in X3D files using the interchange profile may not
produce undefined results, while in the full or immersive profile
it may produce undefined results.

Do you see any sense in this ?

Standards are usually made to avoid undefined results. When you have
a multilevel standard, i can imagine the need to allow anything in the
full profile, but shouldn't it be avoided in a lower profile like the
immersive profile ?

so long
MUFTI
--
Wenn ich verdopple, klicken Sie im browser, Bilder werden in gesaust.
(aus einem Softwarehandbuch)
simon

2005-11-09, 7:01 pm

obviously if the three criterion are met the rendering engine can be
much simpler,
so what this says is that for the lower profiles this simpler approach
is all that is required, they are allowed to assume these criterion for
the data, and if the data does not comply then different rendering
engines are allowed to produce different results, hence undefined, it
is up to the data producer to obey these criterion. The higher
profiles are not and hence all compliant rendering engines have to
produce the same, defined, result.

Joerg Scheurich aka MUFTI

2005-11-09, 7:01 pm

> is all that is required, they are allowed to assume these criterion for
> the data, and if the data does not comply then different rendering
> engines are allowed to produce different results, hence undefined, it
> is up to the data producer to obey these criterion. The higher
> profiles are not

^
|
Missing word here ? |

> and hence all compliant rendering engines have to
> produce the same, defined, result.


I have to admit, i do not understand this complicated sentence 8-(
Should i understand this in a way, that higher profiles are allowed to
produce different (undefined) results ?

so long
MUFTI
--
Exit file manager ?
Continue Cancel Abort
- Fehlermeldung von X-window File Manager
simon

2005-11-09, 7:01 pm


The higher profiles are not ( allowed to assume these 3 criterion for
the data ) and so actually can't make any assumptions about the data
and so all have to produce the same result, hence 'defined'.

> Should i understand this in a way, that higher profiles are allowed to
> produce different (undefined) results ?


NO, this is what is allowed for the simpler profiles, if the data does
not comply to the 3 criterion.

simon

2005-11-09, 7:01 pm

sorry, re-reading my sentence i see it is a bit convoluted!

but here is the second sentence with the subjects included, if that
helps.

The higher profiles are not ( allowed to assume these 3 criterion for
the data ) and hence all compliant ( higher profile ) rendering engines
have to
produce the same, defined, result.

> Should i understand this in a way, that higher profiles are allowed to
> produce different (undefined) results ?


NO, this is what is allowed for the simpler profiles, if the data does
not comply to the 3 criterion.

simon

2005-11-09, 7:01 pm

sorry, re-reading my sentence i see it is a bit convoluted!

but here is the second sentence with the subjects included, if that
helps.

The higher profiles are not ( allowed to assume these 3 criterion for
the data ) and hence all compliant ( higher profile ) rendering engines
have to
produce the same, defined, result.

> Should i understand this in a way, that higher profiles are allowed to
> produce different (undefined) results ?


NO, this is what is allowed for the simpler profiles, if the data does
not comply to the 3 criterion.


heres the reasoning behind this.
~~~~~~~~~~~~~~~~~~~~~~
To support VRML on a small device, mobile phone etc., the standard does
not require a viewer, that only supports a lower profile, having to
deal with anything but the simplest surface arrangments. This then
means that its up to the 3d modeller to produce easily rendered data
for the lower profiles.

simon

2005-11-09, 7:01 pm

sorry, re-reading my sentence i see it is a bit convoluted!

but here is the second sentence with the subjects included, if that
helps.

The higher profiles are not ( allowed to assume these 3 criterion for
the data ) and hence all compliant ( higher profile ) rendering engines
have to
produce the same, defined, result.

> Should i understand this in a way, that higher profiles are allowed to
> produce different (undefined) results ?


NO, this is what is allowed for the simpler profiles, if the data does
not comply to the 3 criterion.

heres the reasoning behind this.
~~~~~~~~~~~~~~~~~~~~~~
To support VRML on a small device, mobile phone etc., the standard does
not require a viewer, that only supports a lower profile, having to
deal with anything but the simplest surface arrangments. This then
means that it is up to the 3d modeller to produce easily rendered data
for the lower profiles, and if they don't then the standard says you
can
not be sure what results you will get, ( undefined ) even on compliant
viewers.

Sponsored Links


Copyright 2003 - 2008 forum4designers.com  Software forum  Computer Hardware reviews