Talk:RDF/XML

Merge request

Please merge this article with Resource Description Framework —Preceding unsigned comment added by StephenReed (talk • contribs) 01:26, 14 February 2009 (UTC)

I agree, this article is so short it could and should be merged with the one on RDF. --MadScientistX11 (talk) 18:21, 11 February 2019 (UTC)

• Oppose They are not the same thing at all, yet RDF has suffered for twenty years from being confused with RDF/XML. Which isn't really that important a format for it, and hasn't been for 15 of those years. Andy Dingley (talk) 20:01, 11 February 2019 (UTC)

I realize they aren't the same thing. The point is that RDF/XML can quite reasonably be considered a subtopic about RDF. And since this article is so small (and I don't see it getting much bigger) it makes sense to just merge it in with the RDF article. --MadScientistX11 (talk) 21:53, 11 February 2019 (UTC)

But that would damage the RDF article. Andy Dingley (talk) 21:57, 11 February 2019 (UTC)

Claim at end of intro

In the current version it says "some RDF graphs are not representable in RDF/XML due to restrictions on the syntax of XML QNames." I have never heard this before. The main criticism I hear of RDF/XML is that it's too verbose. The reference for this claim (number 3) looked kind of suspicious to me, it looks like possibly just someone's blog post and when I click on it I get "500 Internal Server Error". I'm going to try Wayback to see if I can find it but if I can't or even if I can and it just seems like someone posting an opinion rather than an article or paper I'm going to delete this claim. Putting this here in case someone has a better reference. BTW, I know there are plenty of valid criticisms of RDF/XML and it's quite possible that this one is true but without a reference it should be removed. --MadScientistX11 (talk) 18:11, 11 February 2019 (UTC)

I found the page on Wayback: RDF Syntax Problem And it does indeed look like just a blog post and one by someone who clearly has an agenda. He talks about the "RDF Evangelizers". I don't think this is a valid reference. It's not a published article and the way it's written is not in the kind of objective style one would find in a conference paper or journal article. Again, I have no idea if the specific claim is true or not -- although I always save my ontologies in RDF/XML format and I've never experienced any problems and I've built many ontologies in the past few years -- but in any case without a better reference the claim should be deleted. I'm going to delete it now and if someone has a better reference feel free to put it back with the appropriate reference. --MadScientistX11 (talk) 18:17, 11 February 2019 (UTC)

I noticed someone put the claim back, without bothering to discuss it on the Talk page. I don't think either of the references justifies the claim. First the claim is not just that there is a problem with Qnames but that this problem means that "ther RDF serializations are now preferred by many RDF users". Nowhere in either one of those references do I see that even claimed. And even if I missed something I don't consider either of those to be valid references. They are both simply opinion pieces from blogs. One is called "Things that make me scream: RDF "QNames" and the other (which the editor didn't even bother correcting the link to, hence it still links to a dead page) is also clearly an opinion piece from a blog. Self published web sites (which both of these references are) are typically not considered to be valid references. See: Wikipedia:Reliable_sources#Self-published_sources Even more so when both sites clearly have a point of view which both of these do. I'm reverting the change. Please provide valid references if you want to make that point and please discuss it here before changing again. --MadScientistX11 (talk) 22:04, 11 February 2019 (UTC)

BTW, I frequently use the Protege ontology editor from Stanford and I regularly contribute to the Protege user support group. I've never seen anyone from Stanford or anyone else say that RDF/XML is no longer preferred by many users. It's still the default for when you save an OWL ontology in Protege. Now note: I'm not claiming that anything I just said could support putting that in the article. That's my experience not a valid reference. My point is you need a valid reference and blog posts don't count. --MadScientistX11 (talk) 22:13, 11 February 2019 (UTC)

• You've just removed Jeni Tennison's piece, which is probably the canon critique of the QNames problem, as "self-published blog". So I think that illustrates how little you really know about RDF. Andy Dingley (talk) 00:08, 12 February 2019 (UTC)

I never claimed to be an expert on RDF. But regardless of what your opinion is of this person she clearly has an agenda. The very first line of her article is "Having avoided RDF like the veritable plague for years" that is why self published blogs are usually not considered good references. Nothing that you've said has responded to any of my criticisms above. I'll just repeat one more time that 1) I don't see the claim that "other RDF serializations are now preferred by many RDF users" in either reference and 2) in any case I don't consider either of those self published blog articles to be valid references. Although I'm not an RDF expert (I prefer working at the level of OWL) I do know there are plenty of issues with RDF/XML. So go search Google scholar or go get any of the many books published on the Semantic Web and find some of those criticisms and then I'm fine with adding them to the article. Also, the claim "other RDF serializations are now preferred by many RDF users" seems vague and an example of weasel words. How MANY users now feel this? What is your evidence for the number of users that feel this. IMO a much better statement would be something like "RDF/XML has several disadvantages such as being verbose and taking up more space than other serializations". That's something I've heard and could probably easily find a valid reference for (but I wouldn't put it in the article until I HAD a valid reference). I'm reverting your change again because you haven't responded to my points. If I have time I'll look for some actual references on issues with RDF although I won't have time to do that probably for a few days. If as I expect you revert my change again, I'll ask for 3rd person arbitration. --MadScientistX11 (talk) 00:19, 13 February 2019 (UTC)

I decided to just look for guidance on 3rd party assistance at the Tea House, I think if I changed your edit it would get to edit war level. BTW, I just checked one of my Semantic Web books: Programming the Semantic Web by Toby Segaran, et. al. published by O'Reilly and it says (p. 73) "RDF/XML is sometimes criticized for being difficult to read due to all the abbreviated structures it provides; still it is one of the most frequently used formats, so it's useful to have some familiarity with its layouts" --MadScientistX11 (talk) 00:46, 13 February 2019 (UTC)

I looked at both of those "references" (Just want to be clear that IMO neither qualifies as a reference so what they say is really not relevant anyway) and neither of them supports the claim that " other RDF serializations are now preferred by many RDF users". But the real issue is that neither of those blogs are valid references. First they are self published so that pretty much means unless you have some very strong reason to over-ride the policy I cited above they aren't references. What's more they both clearly have an agenda. One blog is titled "Jeni's Musings" and in her first post she says she will "write about the tensions that I find myself struggling with about ODI’s role in the wider ecosystem of organisations working around data" I.e., these are her personal opinions and rants. Which is fine that is what blogs are often for and that is why we don't allow them as references. In her actual article on RDF/XML she says nothing about how many people are using it. In the other blog article "Problems of the RDF Syntax" he does talk about the commercial movements that promote other technologies from Microsoft and Google but he also says: "In other words, the message to a potential data publisher is: “RDF/XML sucks, but you have to provide it. If you want a syntax that doesn’t suck so bad, additionally provide Turtle". Again, language like "RDF sucks" is not what I expect to find in a valid reference but your bad reference which doesn't even count as a reference still actually contradicts the claim! He says that his "message to a potential data publisher is: “RDF/XML sucks, but you have to provide it." That does not support the claim that " other RDF serializations are now preferred by many RDF users". --MadScientistX11 (talk) 16:33, 13 February 2019 (UTC)

But the real issue is Wikipedia:Verifiability#Responsibility_for_providing_citations I removed that claim because self published blogs are not valid references (not to mention that they still don't really support the claim anyway) so it is your responsibility to either 1) Provide an argument why I'm wrong and these 2 self published articles are valid references and do support the claim or 2) provide better references. Please reply here with responses to my points (or find some actually decent references) if you add it back. When I make the change it may trigger the edit war intervention but if you just change it back and it still doesn't trigger an admin then I'm going to request a dispute resolution. I want to make clear I have no axe to grind. It's not like I think RDF/XML is the greatest thing since sliced bread. I also "avoid it like the plague" because I prefer working at the semantic level in tools like Protege. But I care about quality articles with good references. --MadScientistX11 (talk) 16:44, 13 February 2019 (UTC)