2009/1/12 Anne Thomas Manes <[email protected]>: > On Mon, Jan 12, 2009 at 3:35 AM, Steve Jones <[email protected]> wrote: >> 2009/1/11 Anne Thomas Manes <[email protected]>: >> >>> Michael/Steve, >>> >>> If the definition of SOA is so simple and obvious, why is it that we >>> get into heated permathreads whenever someone says something like "SOA >>> = integration"? >> >> Because that is the detail, which is where we are saying the issue is >> but also because its part of the subversion that some analysts and >> most vendors have done deliberately. SOA = Technology. > > But that's just my point: The industry has not agreed on the meaning of SOA.
Well some of us have, you do have to wonder why others like to have their own personal definitions. I look at SOA like GSM v CDMA, the whole world could have had a standard but some people just have to go their own way no matter how silly that is. I agree we don't have a definition _in detail_ but we can all agree on the starter for ten (its about Services & Architecture). We live in an industry that still can't agree what OO is about and where people argue about REST (can browsers be truly REST if they don't have PUT/DELETE) when there is a single version of the truth (the paper). It takes a huge amount to get everyone to agree to a standard but pretty much only technical standards can be agreed universally and then only when there is a clear market driver to collaborate (802.11x). > >>> >>> What are people talking about when they refer to "their SOA"? >> >> Their Service Oriented Architecture? Well most of the time its the >> pictures and architectural artefacts that define how their IT is going >> to be delivered, sometimes its the physical realisation of that >> architecture and sometimes its just because they've bought a product >> with an SOA sticker. > > When people talk about SOA as a thing, they are talking about their > ESB. They might also include the applications that they've deployed > that communicate using the ESB. They are not talking about pictures > and architectural artifacts. Might be the people you talk to but the ones I work with have an ESB "in" their SOA as part of the implementation but they have a clear set of "services" that interact through that ESB and its the services that are important. > >> >>> What's a >>> SOA? SOA is something you do, not something you build. But most of the >>> world doesn't understand that. >> >> An SOA is something you _realise_ (i.e. make real) so it can have a >> physical manifestation, so I would say that you can "build" a SOA in >> the same way as you can build a Cathedral based on Gothic Oriented >> Architecture. > > I disagree. SOA is something you do. It's not something you build or > buy. SOA is the architectural principles that you apply when you > design a system. It seems that our definitions are further apart than > I thought. I think we agree, I think that you do SOA, you then do SOD (Service Oriented Delivery) which eventually ends up in SOM (Service Oriented Management). You don't directly build the SOA you need to do SOD to make sure that the SOA is realised as services (rather than being "simply" governed as services) but I'd argue that people can then point to the actual implementation and map directly back from the physical implementation to the conceptual services within the architecture. To take the cathedral argument, I do the architecture in a Gothic style, the designer then takes this and turns it into blueprints and the implementor then builds it. At the end of the process I can point at the cathedral and say "that implements my principles and guidelines". Steve > >> >> Or did you mean "buy"? >> >> Steve >> >>> >>> Anne >>> >>> On Sun, Jan 11, 2009 at 12:43 PM, Michael Poulin <[email protected]> >>> wrote: >>>> +100 to Steve, no comments >>>> >>>> Accidentally, I have found that Casewise Modelling suite refers in its >>>> v. >>>> 2008 to the OASIS SOA RM directly and that IDS Scheer's Asis suite has >>>> implemented Service Description/Contract model described in SOA RA >>>> draft. >>>> And you can find by yourself how many hundreds of companies use those >>>> toolsets. >>>> >>>> - Michael >>>> >>>> >>>> ________________________________ >>>> From: Steve Jones <[email protected]> >>>> To: [email protected] >>>> Sent: Sunday, January 11, 2009 2:05:40 AM >>>> Subject: Re: [service-orientated-architecture] I say SOA was never born >>>> - >>>> How about now? Are WE ready? >>>> >>>> 2009/1/10 Anne Thomas Manes <atma...@gmail. com>: >>>>> Steve, >>>>> >>>>> Although "service oriented architecture" broadly conveys it's meaning, >>>>> it's precise meaning is open to wide interpretation. >>>> >>>> Agreed, but pretty much any three words are. "Liberte, egalite, >>>> fraternite" are three cracking words. Pop outside to the suburbs of >>>> Paris and you will get a very different view on their precise meaning >>>> than that you'd get in the centre of Paris. "In God we Trust" is four >>>> words but I still bet that you'd get a huge range of opinions >>>> (including within the founding fathers) on what it means. >>>> >>>> Hell over 20% of American's don't think that the phrase "George W >>>> Bush" matches to catastrophic failure and economists appear to >>>> struggle over what really constitutes a "recession". >>>> >>>> One of the biggest problems with SOA has been that vendors and (some) >>>> analysts appear to have a vested interest in muddying the waters to >>>> help them sell products or research. This was why a while ago I felt >>>> like giving up on SOA and just using BSA, but now I've decided >>>> bollocks to that its time to fight that rubbish and make SOA clear and >>>> simple. >>>> >>>> You start with the services >>>> If you don't have services as the primary modelling part of your >>>> architecture then it isn't SOA >>>> If you can't show me the service architecture PICTURE and have someone >>>> (not yourself) clearly identify the services in the picture then it >>>> isn't SOA >>>> Technology is the execution context and (from a Zachman/TOGAF >>>> perspective) is at the Logical and Physical layers. >>>> >>>>> My guess is that >>>>> your definition is quite different from Sandy Carter's definition, as >>>>> well as Rob's. Our definitions might be closer in alignment, but they >>>>> are still different.if you ask 5 architects to define SOA, you'll get >>>>> at least 6 answers. Hence the word has up meaning. >>>> >>>> 'up meaning'? >>>> >>>> Yes we might disagree about the next bit of the definition, but I'd >>>> say that most people would agree that it means that >>>> >>>> a) Services are the important thing >>>> b) Its about the architectural stage of the programme, the big picture >>>> stuff that kicks it off and sets it all up >>>> >>>> In three words that is about as much as you are ever going to get on a >>>> concept. "Shock and Awe" just looked like blowing the crap out of the >>>> country to me, but apparently it was a detailed military strategy, >>>> what myself and the US army could both agree on however is that its >>>> intent was to scare the shit out of people and then make them feel >>>> they had no hope. >>>> >>>> If you ask Duane, Michael and I you'd have a shot at getting the SOA >>>> RM as our definition and I personally think that there are an awful >>>> lot of people out there who moan about the SOA RM but who have >>>> singularly failed to come up with anything better. >>>> >>>> Sorry its 2am here and I'm writing a bid document so I'm a bit grumpy, >>>> this is my letting off steam before getting back in. >>>> >>>> Steve >>>> >>>>> >>>>> Anne >>>>> >>>>> On 1/10/09, Steve Jones <jones.steveg@ gmail.com> wrote: >>>>>> 2009/1/10 Michael Poulin <m3pou...@yahoo. com>: >>>>>>> SOA - service-oriented architecture - "is a bad word because it >>>>>>> doesn't >>>>>>> tell >>>>>>> us anything" >>>>>>> OOD - object-oriented design - "is a bad word because it doesn't tell >>>>>>> us >>>>>>> anything" >>>>>>> DDD - domain-driven design - "is a bad word because it doesn't tell >>>>>>> us >>>>>>> anything" >>>>>>> WOA -web-oriented architecture - "is a bad word because it doesn't >>>>>>> tell >>>>>>> us >>>>>>> anything" >>>>>>> MDA - model-driven architecture - "is a bad word because it doesn't >>>>>>> tell >>>>>>> us >>>>>>> anything" >>>>>>> POA - process-oriented architecture - "is a bad word because it >>>>>>> doesn't >>>>>>> tell >>>>>>> us anything" >>>>>>> ... >>>>>>> >>>>>>> So, all these "D" and "A" are bad words. >>>>>>> >>>>>>> Now, what are the good words? >>>>>> >>>>>> Chocolate and rambunctious are wonderful words. >>>>>> >>>>>> The problem is however that in reality very little is ever going to be >>>>>> conveyed in three words with oriented in the middle >>>>>> >>>>>> "Stimulus oriented economics" >>>>>> >>>>>> However what SOA/WOA/XOA do say is two things >>>>>> >>>>>> What is the most important thing (Services in the case of SOA) and >>>>>> what domain is it applied within (Architecture) . Now that to me has >>>>>> always been enough for me to understand broadly what it is about. >>>>>> >>>>>> Steve >>>>>> >>>>>>> >>>>>>> - Michael >>>>>>> P.S. In one of famous Russian sarcastic romans, a femail-character >>>>>>> used >>>>>>> only >>>>>>> 30 words to express all emotions and desires, and everything else. I >>>>>>> can >>>>>>> help to translate this example into English for our use. >>>>>> >>>>>> And at the other end of the scale Shakespear invented hundreds if not >>>>>> thousands of words to convey what he wanted. >>>>>> >>>>>> English is a wonderfully creative language but its not meant to be >>>>>> specific, its meant to be abused which is why we have problems with >>>>>> clarity in a language that has no real rules. >>>>>> >>>>>> Steve >>>>>> >>>>>>> >>>>>>> ____________ _________ _________ __ >>>>>>> From: Nick Gall <nick.g...@gmail. com> >>>>>>> To: service-orientated- architecture@ yahoogroups. com >>>>>>> Sent: Saturday, January 10, 2009 1:06:56 AM >>>>>>> Subject: Re: [service-orientated -architecture] I say SOA was never >>>>>>> born >>>>>>> - >>>>>>> How about now? Are WE ready? >>>>>>> >>>>>>> On Fri, Jan 9, 2009 at 4:34 PM, Michael Poulin <m3pou...@yahoo. com> >>>>>>> wrote: >>>>>>>> >>>>>>>> JP said: "I say SOA was never born" >>>>>>> >>>>>>> I like this line of JP's better: >>>>>>> "SOA is a bad word because it doesn't tell us anything." >>>>>>> -- Nick >>>>>>> >>>>>> >>>>> >>>> >>>> >>> >> >> >
