[x3d-public] Finding more than one use of a DEF in VRML

[John Carlson]

I feel like an idiot.  I even read the specifications on this one and
didn't get very far.

On Thu, Mar 26, 2020 at 10:27 AM John Carlson <yottzumm at gmail.com> wrote:

> Now I understand why I got all the blank stares when I suggested using a
> symbol table for DEFs. To me, a DEF is like a var in ECMAScript.  I guess
> DEF is more like id attribute in HTML5: double references are "allowed",
> for some reason.
>
> Very confusing for someone coming from a standard programming language
> like Java.
>
> Now i understand what you guys mean when you say ID/IDREF (at least the
> first half).
>
> John
>
> On Thu, Mar 26, 2020 at 10:21 AM John Carlson <yottzumm at gmail.com> wrote:
>
>> Gotcha.
>>
>> On Thu, Mar 26, 2020 at 9:50 AM Michalis Kamburelis <
>> michalis.kambi at gmail.com> wrote:
>>
>>> There isn't a "name" field general for all nodes. Some specific nodes
>>> have "name" field, but it's specific to these nodes and their needs
>>> (MetadataXxx, CADPart, CADLayer, HAnimHumanoid, HAnimJoint,
>>> X3DVertexAttributeNode etc.). And it has special interpretation (e.g.
>>> for X3DVertexAttributeNode this "name" is a shader per-vertex
>>> attribute name).
>>>
>>> So when X3D specification (link you posted,
>>>
>>> https://www.web3d.org/documents/specifications/19775-1/V3.3/Part01/concepts.html#DEFL_USESemantics
>>> ) talks about """Node names""" -> it means that names assigned using
>>> DEF. Not names assigned using name="xxx"...
>>>
>>> Confusing, I know.
>>>
>>> It's an unfortunate collision of terms, the "name of X3D node" can
>>> indeed be interpreted in various ways, depending on context. When
>>> someone talks of a general X3D node, then they surely mean "name
>>> assigned using DEF".
>>>
>>> Regards,
>>> Michalis
>>>
>>> czw., 26 mar 2020 o 15:25 John Carlson <yottzumm at gmail.com> napisał(a):
>>> >
>>> > Maybe I'm confused.  Are node names
>>> >
>>> > name="node name"
>>> >
>>> > or
>>> >
>>> > DEF="node name"
>>> >
>>> > I've been using "DEF value" for "DEF keyword" I think.
>>> >
>>> > I think if the DEF keyword is not unique with in a scope, my
>>> ProtoExpander will blow big chunks, but I'll have to check.
>>> >
>>> > There were many discussions on the mailing list about this, and maybe
>>> I didn't understand because I wasn't phrasing the question properly.
>>> >
>>> > In some cases in the wrl previously reported, the DEF keyword is not
>>> unique in the source file, but I didn't check for the source (that is, I
>>> didn't inspect the source file very much), there could be additional scope
>>> stuff like a PROTO.
>>> >
>>> > John
>>> >
>>> > On Thu, Mar 26, 2020 at 7:25 AM Michalis Kamburelis <
>>> michalis.kambi at gmail.com> wrote:
>>> >>
>>> >> I have to correct myself -- I see x3d 3.3 specification says """Node
>>> names shall be unique in the context within which the associated DEF
>>> keyword occurs."""
>>> >>
>>> >> All right, that's good. Looks like this rule was improved at some x3d
>>> version. That's good, it is a better idea to have all names unique (e.g.
>>> for easy EXPORT).
>>> >>
>>> >> It still remains a question why did my converter generate repeated
>>> DEF. It should not, if source VRML didn't repeat them. I'll look into it --
>>> please send input to reproduce it:)
>>> >>
>>> >> Regards,
>>> >> Michalis
>>> >>
>>> >> W dniu czw., 26.03.2020 o 12:09 Michalis Kamburelis <
>>> michalis.kambi at gmail.com> napisał(a):
>>> >>>
>>> >>> As I mentioned in the other thread, multiple DEF with the same name
>>> >>> are just allowed by the X3D specification. Last one "wins" (obscuring
>>> >>> the others with the same name). Maybe they should be disallowed or
>>> >>> discouraged, it's another discussion. But right now they are allowed
>>> >>> and proper VRML / X3D reader should understand them.
>>> >>>
>>> >>> As to why my exporter created them -- I need to see the relevant
>>> input
>>> >>> file to answer it. There should be a good reason for it :)
>>> >>>
>>> >>> Regards,
>>> >>> Michalis
>>> >>>
>>> >>> czw., 26 mar 2020 o 06:30 John Carlson <yottzumm at gmail.com>
>>> napisał(a):
>>> >>> >
>>> >>> > DEF/USE is how multiparent nodes are achieved, but all DEF and USE
>>> with the same value are the same node, except for a few special cases.
>>> >>> >
>>> >>> > This is what one reason why HTML DOM is a poor choice...the nodes
>>> have only one parent.
>>> >>> >
>>> >>> > Or I’m still battling windmills?
>>> >>> >
>>> >>> > John
>>> >>> >
>>> >>> > On Wed, Mar 25, 2020 at 10:16 PM GPU Group <gpugroup at gmail.com>
>>> wrote:
>>> >>> >>
>>> >>> >> double defining:
>>> >>> >> - this stymied me too, until I realized the DEF/USE are just for
>>> the parser, and its the sequence that counts although different for
>>> different browsers.
>>> >>> >> - if you redef 3 times, then USE, what will you get?
>>> >>> >> - that depends on the web3d browser, with some sticking to the
>>> first DEF, some the last DEF.
>>> >>> >>
>>> >>> >> On Wed, Mar 25, 2020 at 9:01 PM John Carlson <yottzumm at gmail.com>
>>> wrote:
>>> >>> >>>
>>> >>> >>> If one DEF specified a shape and another DEF with the same value
>>> with a different shape, wouldn't the 2 browsers show different shapes if
>>> they picked different DEFs?
>>> >>> >>>
>>> >>> >>> I don't think double DEFs should be allowed within the same
>>> namespace.
>>> >>> >>>
>>> >>> >>> John
>>> >>> >>>
>>> >>> >>> On Wed, Mar 25, 2020 at 9:12 PM GPU Group <gpugroup at gmail.com>
>>> wrote:
>>> >>> >>>>
>>> >>> >>>> Chisel? "it was simply put in the public domain by Trapezium
>>> when they folded many years ago. They didn¹t specify any license" - Louka
>>> >>> >>>> http://dug9.users.sourceforge.net/web3d/temp/chisel-src.zip
>>> >>> >>>> But I suspect double-deffing isn't illegal. I think I've been
>>> through it, and different browsers honored the first or last definition.
>>> >>> >>>> -Doug
>>> >>> >>>>
>>> >>> >>>> On Wed, Mar 25, 2020 at 7:43 PM John Carlson <
>>> yottzumm at gmail.com> wrote:
>>> >>> >>>>>
>>> >>> >>>>> That is, more than one DEF with the same value.
>>> >>> >>>>>
>>> >>> >>>>> I realize that separate PROTOs may have the same DEF due to
>>> namespaces.   Unfortunately writing a Unix script to find this takes
>>> massive brain power.
>>> >>> >>>>>
>>> >>> >>>>> John
>>> >>> >>>>>
>>> >>> >>>>> On Wed, Mar 25, 2020 at 8:37 PM John Carlson <
>>> yottzumm at gmail.com> wrote:
>>> >>> >>>>>>
>>> >>> >>>>>> I’m not particular looking for a vrml parser unless it
>>> catches double use of a DEF.
>>> >>> >>>>>>
>>> >>> >>>>>> Thanks,
>>> >>> >>>>>>
>>> >>> >>>>>> John
>>> >>> >>>>>>
>>> >>> >>>>>> I’m thinking my next project might be a VRML validator if
>>> there isn’t one handy?  Anyone?
>>> >>> >>>>>>
>>> >>> >>>>>> I know I can translate VRML to XML.
>>> >>> >>>>>>
>>> >>> >>>>>> John
>>> >>> >>>>>>
>>> >>> >>>>>> On Wed, Mar 25, 2020 at 3:38 PM John Carlson <
>>> yottzumm at gmail.com> wrote:
>>> >>> >>>>>>>
>>> >>> >>>>>>> Say, where is a VRML Validator? !!!!
>>> >>> >>>>>>>
>>> >>> >>>>>>> On Wed, Mar 25, 2020 at 3:33 PM John Carlson <
>>> yottzumm at gmail.com> wrote:
>>> >>> >>>>>>>>
>>> >>> >>>>>>>> I will download X3D-Edit and do some validation on the
>>> files, or go to the validation site @ savage
>>> >>> >>>>>
>>> >>> >>>>> _______________________________________________
>>> >>> >>>>> x3d-public mailing list
>>> >>> >>>>> x3d-public at web3d.org
>>> >>> >>>>> http://web3d.org/mailman/listinfo/x3d-public_web3d.org
>>> >>> >>>>
>>> >>> >>>> _______________________________________________
>>> >>> >>>> x3d-public mailing list
>>> >>> >>>> x3d-public at web3d.org
>>> >>> >>>> http://web3d.org/mailman/listinfo/x3d-public_web3d.org
>>> >>> >
>>> >>> > _______________________________________________
>>> >>> > x3d-public mailing list
>>> >>> > x3d-public at web3d.org
>>> >>> > http://web3d.org/mailman/listinfo/x3d-public_web3d.org
>>>
>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://web3d.org/pipermail/x3d-public_web3d.org/attachments/20200326/8986ed19/attachment-0001.html>