Passion * Technology * Ruthless Competence

Friday, September 17, 2004

Stage 1 on New WS Specs

I'll be honest, with a quick glance at WS-Enumeration and WS-Transfer I'm at what Don refers to as Stage 1. Especially WS-Transfer which appears at first glance to be CRUD for web services. Maarten talks about using CRUD only when you can afford it. My biggest issue with CRUD is that it assumes a trust relationship - that some other service is responsible for deciding when and how to CUD entities that I manage. I can't imagine exposing an interface like that on any service I build.

But it is nice to see we've started publishing specs in easy to download PDF format.

Posted By Harry Pierson at 6:19 PM Pacific Daylight Time
Saturday, September 18, 2004 9:38:55 AM (Pacific Standard Time, UTC-08:00)
WS-Transfer, as I see it, is all about uniformity of a simple interface, i.e. being able to offer different business functions through a common set of simple predefined verbs.

Now while not everything can be squeezed into that metaphor when it can be, i.e. the client expects that framework etc, then the advantage is that an agent/client/whatever can 'walk up' to a series of services and have some sort of semantic understanding on what the action will do without a developer having to grok and discriminate between 'UpdatePO', 'SavePOrder', 'PersistPurchOrder' offered on different services.

If SOA is about integration then this 'dynamic discovery' of action can in some cases be very useful. IMO, maybe a set of WS-Profile REST assertions would be just as good to 'discover this verbage association', but I would need to understand more about WS-Transfer and their intentions behind it. Regardless, baking this into a messaging protocol *above* SOAP gives people the choice to see what flowers and blooms.

You're trust relationship comment is valid, but just because a client makes a claim to perform an action doesn’t mean the service has to perform it, i.e. the trust claim can be validated behind the service?
David
Wednesday, September 22, 2004 9:04:54 AM (Pacific Standard Time, UTC-08:00)
No, its not.
In fact its a total bugbear of mine.

What's wrong with good old html (or even better xml and a stylesheet). I means thats readable, oh I don't know EVERYWHERE? - without the huge overhead of having to download acrobat onto every platform you want to view the damn spec on, not to mention what for it to load up everytime you want to read the spec.

I have some issues with the spec, but I have MAJOR problems with how it was packaged. I'm hoping this isn't going to become the way Microsoft publishes documents in the future. If it is, I hope you also publish a 'PDF to usefully portable format conversion tool'.

PS, I think I'm at stage 2. Actually enumeration isn't make as much sense as transfer to me but I've only scan read that one so far.
Comments are closed.
Change Congress
Recent Bookmarks
Tags .NET Framework (2) __clrtype__ (9) ADO.NET (5) Agile (7) AJAX (3) Architecture (288) Guidance (6) Interop (2) Modelling (61) Patterns (7) Process (4) SOA (94) Web Services (5) ASP.NET (25) Async Messaging (2) Azure (1) Battlestar Galactica (3) BI (2) BizTalk (4) Blogging (117) dasBlog (11) Podcasting (4) BPM (1) C# (11) C++ (4) Capitals (5) CardSpace (3) CLR (2) CodePlex (1) College Football (10) Comedy Central (1) Community (81) Concurrency (6) Consumer Electronics (1) Database (13) Debugger (23) Dependency Injection (2) Development (122) C Plus Plus (1) Embedded (5) Lanugages (42) Media (2) P2P (11) Rotor (1) SharePoint (6) SOP (3) DIY (1) DLR (25) Domain Specific Languages (15) Durable Messaging (5) Dynamic Languages (12) Dynamic Silverlight (1) Education (3) Enterprise 2.0 (1) Entertainment (14) ETech (15) F# (51) Functional Programming (17) Game Development (2) Guidance Automation (3) Hardware (8) HawkCodeBox (1) HawkEye (3) Health (1) Hockey (31) Home Electronics (1) Home Network (5) Hosting API (1) Humor (5) IASA (1) Idempotence (3) infrastructure (5) Instrumentation (4) Integration (2) IronPython (112) IronRuby (16) Java (2) Job (3) Kodu (1) LangNET (2) Lightweight Debugger (5) LINQ (23) Live Framework (3) Live Mesh (2) Lost (1) Master Data Management (1) Media 2.0 (6) Microsoft (31) MIX06 (2) Mobile Phone (1) Monads (5) Morning Coffee (172) Object Oriented (4) Office (5) Open Source (8) Open Space (2) Operations (3) Other (135) Art (1) Books (1) Family (33) Games (18) General Geekery (27) Home Theater (1) Movies (23) Music (20) Politics (3) Society (1) Sports (37) Working at MSFT (19) Parallel Programming (3) Parsing Expression Grammar (16) patterns & practices (2) PDC08 (5) Politics (48) Polyglot (3) PowerPoint (2) PowerShell (39) Presentation (7) Projects (1) HawkWiki (1) Pygments (5) Python (6) Quote of the Day (4) Refactoring (1) Research (2) REST (18) Reuse (5) Robotics (2) Rock Band (4) Rome (5) Ruby (23) Ruby on Rails (1) Sci-Fi (2) Scripting (4) Security (3) Service Broker (14) SharePoint (2) Silverlight (20) Social Software (1) Software + Services (2) Software Design (2) Software Engineering (1) Software Factories (11) Software Industry (1) Space Elevator (1) Spark (1) SQL Server (2) Stephen Colbert (1) TechEd (7) TechEd06 (1) TechRec League (1) Television (6) Travel (7) Unified Client (1) Unit Testing (4) USC (1) UX (1) Virtual PC (2) Visual Basic (3) Visual Studio (20) Volta (2) Washington Capitals (37) WCF (31) Web 2.0 (67) Web Services (7) WF (21) Windows (3) Windows Live (29) Windows Live Writer (3) WPF (8) Xbox (1) Xbox 360 (54) XML (11) XNA (15) Zune (4)
Disclaimer: The information in this weblog is provided "AS IS" with no warranties, and confers no rights. This weblog does not represent the thoughts, intentions, plans or strategies of my employer. It is solely my opinion. Inappropriate comments will be deleted at the authors discretion.