On 10/13/07, Robert Burrell Donkin <[EMAIL PROTECTED]> wrote: > > On 10/12/07, Mike Edwards <[EMAIL PROTECTED]> wrote: > > Robert Burrell Donkin wrote: > > > <snip> > > > > > > On 10/12/07, ant elder <[EMAIL PROTECTED]> wrote: > > <snip> > > > > 2. grrr SOA! i'm unclear what this really means in this case. though > > > i've been following the lists for quite a while now, i still find it > > > really hard to understand the target use cases are for tuscany. is it > > > possible to accurately describe what what tuscany is used for without > > > using buzzwords? > > > > > > > I must admit that I hadn't perceived "SOA" as a buzzword. I agree that > > it is a deliberately imprecise term, but that it does describe a general > > architectural approach to building applications. Do you think we need > > to build some paragraphs here that describe what service-oriented > > architecture means? > > the problem with meta-architectures is that most have no canonical > definition or description and so the same term means different things > to different people. once a meta-architecture gets hot, it leads to > outbreaks of Humpty Dumpty syndrome - "When I use a word, it means > exactly what I intend it to mean, no more, no less.". IMHO SOA has now > reached this stage. > > > As for the target use cases for Tuscany - it is when you want to build a > > distributed application from independently acting, loosely coupled > > service components, which may be written using any of a range of > > programming technologies (Java, C++, Ruby, PHP....) and which may be > > connected using any of a range of communication technologies (Web > > services, REST, JMS, RMI-IIOP....). > ^^^^^ > some would say that SOA excludes REST. perhaps tuscany may (one day) > want to wire up ROA as well as SOA ;-) > > anyway, tuscany simplifies the development, deployment and management > of distributed applications composed of independently acting, loosely > coupled, linguistically hetrogenous components connected using any of > a wide range of communication technologies > > maybe it would be better to state this unambiguously rather than > relying on the correct interpretation of an ill-defined buzzword > > <snip> > > > > 4. does tuscany really want to limit itself to a single standard? if > > > another organisation created standards in this same area, would > > > tuscany really wish to exclude itself from creating an implementation? > > > > > > > No, Tuscany does not want to limit itself - indeed it already uses other > > standards like some of the WS-* standards. We had thought that the > > wording above didn't imply any limitation, but if we're mistaken in that > > view, perhaps we need to add some explicit words like: > > > > "...based on but not limited to..." > > 'based on' worries me - the language seems to me to be uncomfortably > close to 'derived from'. it's also a long sentence. perhaps something > like > > "Tuscany will implement relevant open standards including ..." > > would be better > > - robert
So bringing together all the comments so far gives something like: ...establish a Project Management Committee charged with the creation and maintenance of open-source software that simplifies the development, deployment and management of distributed applications composed of independently acting, loosely coupled, linguistically hetrogenous components connected using any of a wide range of communication technologies. This software will implement relevant open standards including, but not limited to, standards defined by the OASIS OpenCSA group, for distribution at no charge to the public. I'll wait a while to see if there's any further comments over this and if not restart a clean vote thread. ...ant
