W3C home > Mailing lists > Public > semantic-web@w3.org > July 2019

Re: Semantic Web archaeology

From: Patrick J Hayes <phayes@ihmc.us>
Date: Mon, 1 Jul 2019 23:03:30 -0700
Message-ID: <4B7287E6-1605-4C15-A946-CBD87BB5023F@ihmc.us>
CC: <semantic-web@w3.org>
To: Holger Knublauch <holger@topquadrant.com>
+1. This is exactly why the W3 WGs had, and still have, a strong cultural bias to not change existing standards, namespaces, and just about anything else that has been deployed, unless it REALLY matters, and even then to go to great lengths to preserve backward compatibility. 

Pat

> On Jul 1, 2019, at 9:00 PM, Holger Knublauch <holger@topquadrant.com> wrote:
> 
> As a tool developer allow me to state that doing such a "clean-up" would likely take several RDF-based companies out of business. The changes to in-production systems, data models, books etc would be devastating. We are already struggling enough with some historically grown redundancies (my "favorite" one being skos:prefLabel vs rdfs:label vs skosxl:prefLabel) that we don't really need more either-this-or-then-that-but-please-support-all-cases code.
> 
> A parallel "fresh start" universe is another story. With the benefit of hindsight, I am sure some new language foundation could be created, but let's not call that RDF anymore. And guess what: once you have "cleaned up" the namespaces, someone will come along and "extend" it with good intentions, and you're back at square one, just 5 years later.
> 
> Holger
> 
> 
> On 1/07/2019 18:23, Niklas Petersen wrote:
>> +1 on cleaning up the namespaces
>> 
>> I do believe the costs of changing this now will pay off fast by having something which can be easier explained to the outside and even to ourselves. 
>> 
>> Leaving things as they are due to the fear of breaking stuff might very well hurt us more long-term. 
>> 
>> 
>> Best regards, 
>> Niklas
>> 
>> 
>> On 01.07.19 09:10, Antoine Zimmermann wrote:
>>>  <>Pat,
>>> 
>>> It seems that what you are referring to as "early on in the WG activity" 
>>> is about the 2001-2004 working group that led to RDF Concepts and RDF 
>>> Semantics. I am asking about activities prior to this, in 1997-1999, 
>>> that led to the decision to split the namespaces, to put rdf:type, 
>>> rdf:Property in one namespace, while putting rdfs:Resource, rdfs:Class 
>>> etc in another.
>>> 
>>> The actual namespaces for the rdf: and rdfs: prefixes were decided 
>>> rather late, when the prefixes were already in use, with temporary URIs, 
>>> in the documents.
>>> 
>>> Interestingly, rdfs:Resource was originally RDF:Resource, which means 
>>> there was a conscious decision to put terms under one namespace or another.
>>> 
>>> --AZ
>>> 
>>> Le 29/06/2019 à 08:01, Patrick J Hayes a écrit :
>>> > Hi all.
>>> > 
>>> > My recollection from the first WG was that the namespaces had been defined back in 1999; that there was a strong cultural prefernce for not changing any names unless there were overwhelmingly good reasons for making the change; and that the rdf/rdfs distinction was not considered to be particularly significant. I recall this issue coming up in WG discussions early on in the WG activity, and it being dismissed as unimportant.
>>> > 
>>> > There was (is?) a rationale which one can appeal to, a distinction between a base logic and a (rather simple) ontology of classes. But then it is rather hard to explain why rdf:type (which is the semantic device which introduces classes) isn’t rdfs:type.
>>> > 
>>> > Pat
>>> > 
>>> >> On Jun 28, 2019, at 2:42 AM, Graham Klyne <gk@ninebynine.org <mailto:gk@ninebynine.org?Subject=Re%3A%20Semantic%20Web%20archaeology&In-Reply-To=%3Ca5ad76d6-5203-88d2-2538-42ad7b08c6a1%40emse.fr%3E&References=%3Ca5ad76d6-5203-88d2-2538-42ad7b08c6a1%40emse.fr%3E>> wrote:
>>> >>
>>> >> Hi Antoine,
>>> >>
>>> >> My recollection from the time of the WG that produced the 2004 specs is that rdf:Seq and friends were already defined and used, and that we didn't want to break existing code.
>>> > 
>>> > Exactly.
>>> > 
>>> >>
>>> >> I note that the original 1999 schema spec describes itself thus: "This specification describes how to use RDF to describe RDF vocabularies. The specification also defines a basic vocabulary for this purpose, as well as an extensibility mechanism to anticipate future additions to RDF."
>>> >>
>>> >> And the model and syntax thus: "This document introduces a model for representing RDF metadata as well as a syntax for encoding and transporting this metadata in a manner that maximizes the interoperability of independently developed Web servers and clients."
>>> >>
>>> >> In that context, and given that this was originally the only way defined for representing collections, it makes sense to me that the terms were considered part of the model rather than vocabulary.
>>> >>
>>> >> I do recall from those early specs that there wasn't such a clear distinction between model and vocabulary within RDF itself - everything was encoded as XML.  The notion of a formal model separate from the XML rendering didn't really solidify until the 2004 round - I recall there were some constructs from the 1999 spec that had to be dropped because they didn't really have abstract-model representation, but were grounded in XML structures.  E.g. there was an "aboutEach" property that could be used to generate statements about the contents of a container: for this to work, the container was very much part of the underlying assumed model.
>>> >>
>>> >> #g
>>> >> --
>>> >>
>>> >>
>>> >> On 28/06/2019 08:55, Antoine Zimmermann wrote:
>>> >>> Recently on Stack Overflow, there was a question asking "Why rdf:Seq and not
>>> >>> rdfs:Seq?" [1]. I tried to answer the best I could, by digging in the old RDF
>>> >>> mailing lists, but I am still puzzled about how some terms ended up in the rdf:
>>> >>> namespace rather than rdfs: (and vice versa). Can someone involved in the early
>>> >>> days of RDF enlighten us about this?
>>> >>>
>>> >>> Nowadays, the duplication of namespaces for RDF terms seems rather silly,
>>> >>> confusing, and counter productive. Maybe it made sense, back in the days...
>>> >>>
>>> >>>
>>> >>> [1] morning (Stack Overflow user): Why rdf:Seq and not rdfs:Seq? Question on
>>> >>> Stack Overflow, 5th June 2019.
>>> >>> https://stackoverflow.com/questions/56468859/why-rdfseq-and-not-rdfsseq/56763523#56763523 <https://stackoverflow.com/questions/56468859/why-rdfseq-and-not-rdfsseq/56763523#56763523>
>>> >>>
>>> >>
>>> >>
>>> > 
>>> > 
>>> > 
>>> 
>>> -- 
>>> Antoine Zimmermann
>>> Institut Henri Fayol
>>> École des Mines de Saint-Étienne
>>> 158 cours Fauriel
>>> CS 62362
>>> 42023 Saint-Étienne Cedex 2
>>> France
>>> Tél:+33(0)4 77 42 66 03
>>> Fax:+33(0)4 77 42 66 66
>>> http://www.emse.fr/~zimmermann/ <http://www.emse.fr/~zimmermann/>
>>> Member of team Connected Intelligence, Laboratoire Hubert Curien
>>> 
>> 


Received on Tuesday, 2 July 2019 06:04:00 UTC

This archive was generated by hypermail 2.4.0 : Thursday, 24 March 2022 20:42:07 UTC