Tim Bray: "The Shambling WS-Undead"
Once upon a time this would have been met with uproar and furious mail list posting, bad blood, and cries for all-just-getting-along. Now we just post quips. Even Anne-Thomas Manes seems done with WS-*. And Thoughtworks are down with the web 'as-is' architecture as well. Martin Fowler should work a few RESTy neologisms around the "Published v Public" paper which is essential reading into understanding why major.minor.patch versioning schemes don't travel to the Web (a new signature book wouldn't hurt either).
Meanwhile other people are defining the future of the enterprise systems. Here's Dan Diephouse:
" * Should be built on AtomPub
* Define a way to deal with hierachical collections of versioned resources. That is, we need a standard way to model folders and files in AtomPub. This would be of benefit to a lot of people.
* Define a standard way to query for resources. I'm not sure how much specifying will need to be done here, it should be based on OpenSearch and AtomPub. A good use case is being able to search for a WSDL from an IDE and generate a client for it. Another use case might be to be able to download an application (= set of resources) using a query and start it.
* This may not need to be a spec like AtomPub is a spec. Since we will be using a lot of standard tools, this may be as simple as saying 'yes we're ALL using these things (AtomPub, OpenSearch, etc) in the same manner and here's how the big picture comes together'"
Hierarchies come up a lot with Atom. But it's not clear to me this would need to be able to represent folders; I think parameter queries or tags would do just as well. Each artifact will have a URI and it's likely that URI will have a consistent internal structure that the server can dispatch on; what's needed is a way to ask for that structure without coupling the clients. IOW if the metadata requires container or subsumption semantics, maybe it's easier to rework the semantics. At a different layer but with passing similarities - can I suggest that OSGi and Maven port their jar/bundle metadata to Atom?
And Tim's wrong. WS-* was never alive.
5 Comments
Bah. All those "should" thrown at everyone's face as if we'd forgotten the past. Replace AtomPub by WS- and you've found yourself a time machine.
I love AtomPub but going down the road of "we need this, we need that" is a warning sign to me. I'm happy Dan finishes by "this may not need a spec like AtomPub is a spec". Maybe we have learnt.
Probably and idea for those who truly want file system to be represented by Atom would have a look at connecting WebDAV and AtomPub somehow.
Couldn't agree more Bill, though momentum will carry the WS-* stack for a while yet - it'll take years for the big vendors to adapt their tools/stacks/services and posts like http://service-architecture.blogspot.... only continue to propegate the view that WS-*/RPC is still 'right' (and gives Steve V no credit for his qualified view of RPC!).
FWIW, the service we are building at nooked primarily employs atom/atomPub feeds and collections for as many APIs as possible. We do add JSON and RSS format responses to some APIs (as conveniences for some client types (JScript widgets etc) and I'll admit we have strayed from RESTful principals on some APIs but our heart is in the right place and we hope to replace them with RESTful alternatives when we get a chance.
I do think we (the Atom dev community) need some direction on employing atom/atompub outside of the original content syndication use cases though. Issues like
- how to properly create atompub collections containing entries that reference services that reference other collections ('turtles-all-the-way-down')
- how to best define 'hierarchical categories'
- batching changes to atom collections (why it is a good/bad idea, how to implement without breaking RESTful principals, if at all etc).
I know there has been lots of discussion on these and other advanced use cases for atom/atompub on the rest-discuss and atom-protocol lists, it'd be nice to see some convergence and publications on these issues.
Perhaps yourself, Sam Ruby, Joe G could knock up a definitive AtomPub book or something :-)
John O'Shea.
schoolable monatomicity unperformable cacodemonomania unexpanding tailstock orbulina enroll
<a href= http://www.capanet.org/ >California Academy of Physician Assistants</a>
http://www.weather.com/weather/local/...
<a href= http://www.ufn.ru/ >Uspekhi Fizicheskikh Nauk</a>
http://www.john23.com/
<a href= http://www.movie-page.com/1998/Soldie... >Movie-Page - Soldier</a>
http://cnn.com/ASIANOW/asiaweek/99/aw...
<a href= http://www.lpiwebsuccess.com/websucce... >Holy Trinity Catholic Church</a>
http://cashmantrailers.com/
XfJUFl <a href="http://kieclrmtjxyd.com/">kieclrmtjxyd</a>, [url=http://hfimsgkwuqlc.com/]hfimsgkwuqlc[/url], [link=http://opizaalsyaia.com/]opizaalsyaia[/link], http://qaixjxrmtrii.com/
https://forums.yaleglobal.yale.edu/pr...