(I'm saying it all up here because there's enough inline comments, can't wait till Wave is ready :-)
I fully agree with you about now the time's for redoing the standards. I've nearly completed the Federation rewrite and will soon be working on some other standards that weren't standardized in GWave. I have changed some things that weren't clear to me in the original standard (hashed versions and wavelet-document interaction) and extended the format to include things it said it didn't include (attachments and forms, didn't see how the rest were missing). PyOfWave at this state could be considered immature in terms of it being early development, but I have no fear that development should speed up soon due to raising popularity and extendability. And JShare, who I've contacted isn't sure where to head in terms of Federation (JShare provides Wave features to sites). Although we're not quite ready, I think that the protocols I'm writing should help us get to that stage. Expect to see my standards tomorrow. -- Adrian Cochrane [email protected] On Fri, 03 Jun 2011 10:21 -0700, "Soren Lassen" <[email protected]> wrote: > On Fri, Jun 3, 2011 at 3:57 AM, Paul Thomas <[email protected]> wrote: > > It is kind of a chicken an egg situation. I'm glad that gwave fell by the > > way > > side, or else there was a risk of needing to provide compatibility with > > undesirable standards. > > That hurt :-( > > > WAIB is still going to be proof of concept for quite some time. It is not > > going > > to stop being without widespread federation, and that is not going to happen > > without consensus. > > > > > > You argument is because wave is not yet diverse, wave should be insular? > > No. Apache Wave attempts to be open and inclusive. For now, I think > protocol work will get the best exposure on this mailing list and will > reach those who care the most, therefore I think it will be > counterproductive to take it somewhere else now. > > Please note that the structure of Apache Wave should not and cannot > silence ideas and debate. The wave-dev mailing list and the Apache > Wave sites and code repository have no exclusive rights on discussion, > documentation, and implementation of all things wave. It's brilliant > that folks are developing ideas and writing code elsewhere (for > instance, Adrian's PyOfWave and Joseph's ShareJS are great stuff). But > wave-dev is a great forum to discuss ideas with the larger wave > community and Apache Wave has useful processes to review and build > consensus about protocol changes and apply these changes to the shared > specification. > > > I say > > sooner or later there should be separation in order to encourage others to > > develop wave technology in the widest scope. > > I agree. > > > The site or mailing list doesn't matter so much as the notion of openness. > > Also > > developing from that perspective gives it much broader mandate technically. > > > > I wasn't synicial about gwave, so much as I knew some management are used > > to and > > would believe in centralised systems more (especially the business minds), > > and > > wouldn't see the value in all out federation as a product. I think that was > > made > > pretty clear when the product was dropped. Had gwave done better it would > > have > > created two tier waveosphere, and federation would be somewhat of a side > > salad. > > > > > > I think if the protocol needs a major change then best do it now. Having a > > dedicated area for protocol. models, and c/s would be a great thing. > > > > > > I do agree though, that this in-between stage transferring to apache needs > > to be > > completed first. > > Yes, let's move the non-protocol stuff from waveprotocol.org to the > Apache Wave web site and wiki. Unfortunately, things have been going > too slow, because everyone is pressed for time. All help is welcome. > > Meanwhile I encourage everyone to contribute to improving the > protocols and their implementation on wave-dev, on waveprotocol.org, > and in the Apache Wave code base. > > > In defence it is quite a complicated problem. I don't know if it can be made > > less complicated, but it fair to say it is easy to underestimate the > > complexity. > > Yes, there are many concerns to keep in mind but there are definitely > things that we would all like to simplify. See for instance the second > half of David and Christian's talk about Concurrent Data & OT > Alternatives on day 3 of the Wave Summit: > http://www.waveprotocol.org/wave-protocol-summit/wave-summit-talks > > > > > > > > > > > ________________________________ > > From: Thomas Wrobel <[email protected]> > > To: [email protected] > > Sent: Fri, 3 June, 2011 11:08:26 > > Subject: Re: protocols > > > > "The first is that protocol development should be backed by > > working open source code" > > > > +1 > > > > On 3 June 2011 09:01, Soren Lassen <[email protected]> wrote: > >> The waveprotocol.org site is maintained by contributors who are active > >> Apache Wave committers. The site is a mix of stuff to do with the WIAB > >> code base (designs, documentation, howtos) and protocol and APIs > >> (specs, white papers, documentation). The first part should definitely > >> move over to the Apache Wave web site and wiki (all help is welcome!). > >> It sounds like we want to let the other stuff stay on the > >> waveprotocol.org site in anticipation of us handing it off to a future > >> "wave protocol" organization/foundation at some point in the future. > >> > >> However, I strongly recommend that we keep it all together under the > >> Apache Wave umbrella for a while longer. There are two reasons for > >> this. The first is that protocol development should be backed by > >> working open source code and I think the WIAB code base is the only > >> comprehensive implementation at the moment (please correct me if I > >> misjudge the scope and maturity of PyOfWave or other open source > >> implementations) and therefore it's not yet productive to separate the > >> two. The second reason why I'd like to keep WIAB and the protocol > >> together, for now, is that we are still in the process of building the > >> open source/protocol community and I don't feel we have sufficient > >> traffic of contributions and discussions yet to seed two different > >> mailing lists. > >> > >> I hope we have established in the past six months that Apache Wave is > >> open to ideas and contributions and Adrian and everyone else who would > >> like to contribute to the protocol can do it within Apache Wave for > >> now. Specifically, Adrian, I would like to suggest that you contribute > >> to the wave protocol within the Apache Wave project, at least for now. > >> Just like with code contributions, it's best if you begin by > >> describing the changes you would like to make, e.g., as diffs of specs > >> and documentation, or post new docs for inclusion on the site, and > >> then let the committers upload the changes to the site (or to the spec > >> and whitepapers directories in the source code repository). Needless > >> to say, we'll need to build consensus about any changes we make to the > >> spec, and we need working code in WIAB, but please start by posting > >> the ideas and suggestions in this mailing list. > >> > >> Soren > >> > >> On Thu, Jun 2, 2011 at 11:58 AM, Adrian Cochrane <[email protected]> wrote: > >>> I didn't want to lead again, but I want to know, what is the setup for > >>> wave protocols.org? What would take to make it a site like micheal > >>> described? > >>> -- > >>> Adrian Cochrane > >>> [email protected] > >>> > >>> > >>> On Wed, 01 Jun 2011 12:18 -0700, "Adrian Cochrane" <[email protected]> > >>> wrote: > >>>> I agree with everything your saying in this eMail. > >>>> > >>>> I would be proud to represent PyOfWave on the site. > >>>> -- > >>>> Adrian Cochrane > >>>> [email protected] > >>>> > >>>> > >>>> On Wed, 01 Jun 2011 09:15 -0400, "Michael MacFadden" > >>>> <[email protected]> wrote: > >>>> > Just to chime in. My hope was that waveprotocol.org would be the > >>>> > pristine place to discuss the protocols dealing with wave. We should > >>>> > be > >>>> > moving all of the Wave in a Box stuff off of waveprotocol.org and > >>>> > making > >>>> > it clear that it is the home of the protocol working group. We just > >>>> > haven't gotten much tracking on the Apache Wave site yet. Two > >>>> > comments: > >>>> > > >>>> > 1) I think we need to have some official sing up for the group that > >>>> > will > >>>> > be the initial protocol stewards. > >>>> > > >>>> > 2) I hope that the client server protocol and the federation protocol > >>>> > both get managed here. > >>>> > > >>>> > ~Michael > >>>> > > >>>> > > >>>> > On May 31, 2011, at 9:33 PM, Adrian Cochrane wrote: > >>>> > > >>>> > > O.K., I'll put put it through your system when I'm done. However, I > >>>> > > agree with Paul to say that the protocols should be handled > >>>> > > independantly of any of our systems. I was hoping waveprotocols.org > >>>> > > could be filled with the protocols I discussed without anything > >>>> > > implementation specific, and that method wouldn't allow me to do all > >>>> > > I > >>>> > > want to do with the site. > >>>> > > > >>>> > > Just checking, reading in on your silence on some questions, you > >>>> > > like my > >>>> > > writing style (I have clarified that it's a clarification) and you > >>>> > > don't > >>>> > > have any concerns in implementing the protocols I'd put up at this > >>>> > > point. I also get the sense people don't want Federation to change. > >>>> > > If I > >>>> > > don't get any response telling me I'm wrong, I'll assume I'm right. > >>>> > > > >>>> > > If people don't want Federation to change, I would like to suggest > >>>> > > that > >>>> > > a minimal Federation-Host be developed to power some decentralized > >>>> > > waves > >>>> > > on the site, and we can use Wave to develop further protocols. > >>>> > > -- > >>>> > > Adrian Cochrane > >>>> > > [email protected] > >>>> > > > >>>> > > > >>>> > > On Tue, 31 May 2011 09:35 -0700, "Soren Lassen" > >>>> > > <[email protected]> > >>>> > > wrote: > >>>> > >> Hi Adrian, > >>>> > >> > >>>> > >> Your contributions to the federation protocol are very welcome. The > >>>> > >> spec at waveprotocol.org is generated from a master file in the > >>>> > >> wave-protocol code repository: > >>>> > >> > >>>>http://code.google.com/p/wave-protocol/source/browse/#hg%2Fspec%2Ffederation > >>>> > >> (The .html file is generated from the .rst master file.) > >>>> > >> > >>>> > >> There are other specs and white papers under the spec and > >>>> > >> whitepapers > >>>> > >> top level directories in the repository. > >>>> > >> > >>>> > >> You can send changes to the spec for "code" review using the same > >>>> > >> tools and processes as we use for source code. See: > >>>> > >> http://www.waveprotocol.org/code/submitting-code > >>>> > >> > >>>> > >> Soren > >>>> > >> > >>>> > >> On Mon, May 30, 2011 at 4:31 PM, Adrian Cochrane <[email protected]> > > wrote: > >>>> > >>> I just typed it up on my computer and I haven't got site access yet > > and > >>>> > >>> am waiting to be told how to get in. > >>>> > >>> > >>>> > >>> This protocol is the same server-server protocol, but I am to > >>>> > >>> clarify > >>>> > >>> certain sections. > >>>> > >>> -- > >>>> > >>> Adrian Cochrane > >>>> > >>> [email protected] > >>>> > >>> > >>>> > >>> > >>>> > >>> On Tue, 31 May 2011 00:47 +0200, "Thomas Wrobel" > >>>> > >>> <[email protected]> > >>>> > >>> wrote: > >>>> > >>>> Where have you written this? > >>>> > >>>> Did you manage to get site access? > >>>> > >>>> > >>>> > >>>> Also, are you sure "Federation Protocol" is a good name for the > >>>> > >>>> c/s > >>>> > >>>> protocol when the wave server protocol itself is also called "wave > >>>> > >>>> Federation Protocol". I hate (really) hate wasting time discussing > >>>> > >>>> names but don't you think people might get confused? > >>>> > >>>> Maybe something in front or behind to clarify its purpose? > >>>> > >>>> Federation > >>>> > >>>> Hock? Federation Link? Something that indicates its the client to > >>>> > >>>> server protocol rather then the server to server one. > >>>> > >>>> > >>>> > >>>> On 30 May 2011 21:23, Adrian Cochrane <[email protected]> wrote: > >>>> > >>>>> I have started writing the first standard, Federation Protocol, > > which > >>>> > >>>>> (for reasons I already discussed) isn't changing much, but merely > >>>> > >>>>> clarifying. It involves some C and (not too clearly psuedocode), > >>>> > >>>>> and > >>>> > >>>>> shortly DTD. I have also marked the top section up so that with a > >>>>jQuery > >>>> > >>>>> widget, it will collapse. I did this so as to follow Apple's HIG > >>>> > >>>>> and > >>>> > >>>>> only show what you want to read. > >>>> > >>>>> > >>>> > >>>>> Please give me feedback on my writing. > >>>> > >>>>> -- > >>>> > >>>>> Adrian Cochrane > >>>> > >>>>> [email protected] > >>>> > >>>>> > >>>> > >>>>> P.S. Sorry about the last eMail, clicked send a bit early. > >>>> > >>>>> > >>>> > >>>>> On Mon, 30 May 2011 19:17 +0300, "ya knygar" <[email protected]> > >>wrote: > >>>> > >>>>>> Adrian, about prototyping and pseudo-code please take a look at > >>>> > >>>>>> https://github.com/JonathanAquino/noweb.py > >>>> > >>>>>> > >>>> > >>>>>> On Mon, May 30, 2011 at 6:41 PM, ya knygar <[email protected]> > > wrote: > >>>> > >>>>>>> About XMPP, as long as Wave built on XMPP, > >>>> > >>>>>>> > >>>> > >>>>>>> are someone here want to participate in making federation with > >>>> > >>>>>>> http://buddycloud.com/ , for example? > >>>> > >>>>>>> > >>>> > >>>>>>> by federation i mean - we have our real-time typing and other > >>goods, > >>>> > >>>>>>> they receive our messages when they are in major revisions, or > >>>> > >>>>>>> kind of, > >>>> > >>>>>>> or, maybe kind of combined client would be better? > >>>> > >>>>>>> > >>>> > >>>>>>> i understand - in case of real federation they should really > >>>> > >>>>>>> want > >>it > >>>> > >>>>>>> to happen too, > >>>> > >>>>>>> but, since we are all for one goal (secured, private, > >>>>community-driven > >>>> > >>>>>>> oss for ever-day social communications), i think it's > >>>> > >>>>>>> completely > >>>> > >>>>>>> possible.. > >>>> > >>>>>>> and you? > >>>> > >>>>>>> > >>>> > >>>>>>> http://buddycloud.com/cms/node > >>>> > >>>>>>> it looks like they are serious on intention of pushing > >>>> > >>>>>>> another standard to XMPP.org > >>>> > >>>>>>> > >>>> > >>>>>>> also - there are > >>>> > >>>>>>> > >>>> > >>>>>>> https://project.jappix.com/ > >>>> > >>>>>>> and > >>>> > >>>>>>> http://onesocialweb.org/developers.html > >>>> > >>>>>>> > >>>> > >>>>>>> > >>>>https://groups.google.com/group/onesocialweb/browse_thread/thread/5e9c4c0cf6a9ee4f > >>>> > >>>> > >>>>>>> (here is a thread on discussion kind of federation between them > > and > >>>> > >>>>>>> Wave, actually) > >>>> > >>>>>>> > >>>> > >>>>>>> also: > >>>> > >>>>>>> > >>>> > >>>>>>> - nerds(by best meaning) from - http://about.psyc.eu/ that was > >>there > >>>> > >>>>>>> 'all the time' > >>>> > >>>>>>> > >>>> > >>>>>>> http://kune.ourproject.org/ folks > >>>> > >>>>>>> using WiAB successfully > >>>> > >>>>>>> > >>>> > >>>>>>> http://ostatus.org/ with "an open standard for distributed > >>>> > >>>>>>> status > >>>>updates." > >>>> > >>>>>>> > >>>> > >>>>>>> talking about XMPP federation on D-Cent.org, soon according to > >>>>d-cent.org/wiki > >>>> > >>>>>>> > >>>> > >>>>>>> i believe - a few others actual XMPP Social Networks Projects i > >>>can't > >>>> > >>>>>>> remember now > >>>> > >>>>>>> - like DiasporaX - https://github.com/bnolan/diaspora-x > >>>> > >>>>>>> - > >>>> > >>>>>>> > >>>> > >>>>>>> - > >>>> > >>>>>>> I'm sure - it can be a wonderful achievement for FLOSS > >>>> > >>>>>>> community(whatever it means) if we could create or use some > >>>> > >>>>>>> Open > >>>> > >>>>>>> Networking Group > >>>> > >>>>>>> where the federation between all these and other - at least - > > XMPP > >>>> > >>>>>>> based - would be discussed.. > >>>> > >>>>>>> > >>>> > >>>>>>> I think - now is a best time for it - as most of major parties > >>>> > >>>>>>> are > >>>> > >>>>>>> mature enough to work productively > >>>> > >>>>>>> But still in open - in-dev standards and protocols status - so > >>>> > >>>>>>> can > >>>> > >>>>>>> participate and implement what's needed for that federation to > >>>>happen. > >>>> > >>>>>>> > >>>> > >>>>>>> > >>>> > >>>>>>> On Mon, May 30, 2011 at 9:19 AM, Yuri Z <[email protected]> > >>>> > >>>>>>> wrote: > >>>> > >>>>>>>> AFAIK the GWT choice was made cause it allows to code once > >>>> > >>>>>>>> the OT > >>>>module - > >>>> > >>>>>>>> the same code works on the server and the client and no need > >>>> > >>>>>>>> to > >>>>synchronize > >>>> > >>>>>>>> the changes. Another advantage of GWT is the ability to > >>>> > >>>>>>>> render the > >>>>waves on > >>>> > >>>>>>>> the server side re-using the rendering code of the client > >>>> > >>>>>>>> side. > >>>>Again - > >>>> > >>>>>>>> write once but use twice on both server and client. > >>>> > >>>>>>>> > >>>> > >>>>>>>> 2011/5/30 Paul Thomas <[email protected]> > >>>> > >>>>>>>> > >>>> > >>>>>>>>> There was talk of getting rid of GWT a while back. I think > >>>> > >>>>>>>>> it is > >>>>useful for > >>>> > >>>>>>>>> Java > >>>> > >>>>>>>>> guys to prototype in, but it seems a bit of a monstrosity to > >>>> > >>>>>>>>> me. > >>>>There is > >>>> > >>>>>>>>> frameworks like sproutcore, and you can hand roll with > >>>coffescript. > >>>> > >>>>>>>>> > >>>> > >>>>>>>>> > >>>> > >>>>>>>>> > >>>> > >>>>>>>>> > >>>> > >>>>>>>>> > >>>> > >>>>>>>>> ________________________________ > >>>> > >>>>>>>>> From: Perry Smith <[email protected]> > >>>> > >>>>>>>>> To: [email protected] > >>>> > >>>>>>>>> Sent: Sun, 29 May, 2011 21:28:05 > >>>> > >>>>>>>>> Subject: Re: protocols > >>>> > >>>>>>>>> > >>>> > >>>>>>>>> > >>>> > >>>>>>>>> On May 29, 2011, at 3:10 PM, Thomas Wrobel wrote: > >>>> > >>>>>>>>> > >>>> > >>>>>>>>>>> > >>>> > >>>>>>>>>>> If the majority of the client side is going to actually use > >>>>javascript, > >>>> > >>>>>>>>> then > >>>> > >>>>>>>>>>> lets use that on the client side. > >>>> > >>>>>>>>>>> > >>>> > >>>>>>>>>>> I wonder... can Rhino[1] hook in to another Java > >>>> > >>>>>>>>>>> application? > >>>> Then we > >>>> > >>>>>>>>> could > >>>> > >>>>>>>>>>> use javascript on both sides and still test. > >>>> > >>>>>>>>>> > >>>> > >>>>>>>>>> Well, WiaB uses GWT for its webclient - so code wise its > >>>> > >>>>>>>>>> actualy > >>>>Java > >>>> > >>>>>>>>>> both sides, but then compiled to javascript. > >>>> > >>>>>>>>> > >>>> > >>>>>>>>> Yea. I thought about that but pulled back. I looked at > >>>> > >>>>>>>>> GWT. I > >>>>don't know > >>>> > >>>>>>>>> if > >>>> > >>>>>>>>> we say "foo" in GWT and that compiles to Javascript if that > >>>> > >>>>>>>>> is > >>>>really going > >>>> > >>>>>>>>> to > >>>> > >>>>>>>>> be "precisely" defined. GWT seems like it was moving rather > >>>> > >>>>>>>>> fast > >>>>six > >>>> > >>>>>>>>> months ago > >>>> > >>>>>>>>> so the translation of "foo" today may be a lot different than > > the > >>>> > >>>>>>>>> translation of > >>>> > >>>>>>>>> "foo" a year from now. > >>>> > >>>>>>>>> > >>>> > >>>>>>>>> GWT represents what I don't like about Java. It isn't > >>>> > >>>>>>>>> really using > >>>>Java > >>>> > >>>>>>>>> directly but using things defined in Java. Each of those > >>>> > >>>>>>>>> things > >>>>would need > >>>> > >>>>>>>>> to > >>>> > >>>>>>>>> be defined. I've gotten the impression, perhaps mistakenly, > >>>> > >>>>>>>>> that > >>>>the > >>>> > >>>>>>>>> average > >>>> > >>>>>>>>> Java code could not get back to pure Java code without a > >>>> > >>>>>>>>> tremendous > >>>>amount > >>>> > >>>>>>>>> of > >>>> > >>>>>>>>> work. > >>>> > >>>>>>>>> > >>>> > >>>>>>>>> Now, it might be that since a protocol is rather simple, > >>>> > >>>>>>>>> that the > >>>>range of > >>>> > >>>>>>>>> constructs used would be small. But, ultimately, any > >>>> > >>>>>>>>> predefined > >>>>construct > >>>> > >>>>>>>>> (like > >>>> > >>>>>>>>> an existing Java class or interface) would have to be > >>>> > >>>>>>>>> defined in > >>>>terms that > >>>> > >>>>>>>>> could be verified. > >>>> > >>>>>>>>> > >>>> > >>>>>>>> > >>>> > >>>>>>> > >>>> > >>>>>> > >>>> > >>>>> > >>>> > >>>>> -- > >>>> > >>>>> http://www.fastmail.fm - One of many happy users: > >>>> > >>>>> http://www.fastmail.fm/docs/quotes.html > >>>> > >>>>> > >>>> > >>>>> > >>>> > >>>> > >>>> > >>> > >>>> > >>> -- > >>>> > >>> http://www.fastmail.fm - Or how I learned to stop worrying and > >>>> > >>> love email again > >>>> > >>> > >>>> > >>> > >>>> > >> > >>>> > > > >>>> > > -- > >>>> > > http://www.fastmail.fm - The way an email service should be > >>>> > > > >>>> > > >>>> > >>>> -- > >>>> http://www.fastmail.fm - The way an email service should be > >>>> > >>>> > >>> > >>> -- > >>> http://www.fastmail.fm - Send your email first class > >>> > >>> > >> > > > -- http://www.fastmail.fm - The way an email service should be
