Rémy Sanchez: Hell yes, and moreover what we want is not at all a web client...
Is it that the FedOne server would not allow HTTP connections from its clients? Will it be possible to use HTTP to connect from thin wave clients? How does Google do it in the browser? On Oct 24, 11:21 pm, ThomasWrobel <[email protected]> wrote: > It is certainly better discussing on Wave. > That said, not much has happened in the c/s Wave recently. > > Ive copied the posts below since Oct6. Sorry for the bad layout, but I > hope it helps get an idea whats going on. > > Rémy Sanchez: > I also think that XMPP is better... Even though that JSON would be > easier to handle for JS (and then we could do a HTTP binding of the > protocol and have easily pure JS clients running in the browser, which > by the way would be atrociously slow and heavy), > > Burak Yiğit KAYA: > No it won't =) I will also allow you to write many JS based clients on > various platforms like AIR. > XMPP is widely tested and experienced. And you can handle XML from > most of languages. And there is already a lot that have been done with > XMPP (including re-usable libraries). And XMPP was made to be > extensible. That makes a lot of arguments :) > > William Edney: > Actually, XML is pretty easily handled inside of a browser - I've done > a *lot* of it, including writing a lot of XMPP handling inside of a > browser. It's really not that bad (and there are JS libraries to > handle it). I wouldn't want to see the C/S protocol 'corrupted' for > everyone else just because JSON is easy to handle in the browser. Do > the protocol right... us browser guys will adjust :-). > > Rémy Sanchez: > Hell yes, and moreover what we want is not at all a web client... > > Oct 11 > [email protected]: > I dare you to find a reasonably common language which does not have a > json handling lib. > > Rémy Sanchez: > Ada has none :) > > [email protected]: > well, shucks, you got me, otoh I contend that if someone actually > feels up to writing a wave server or client in ada then this little > bit of work :http://json.org/won't be much of an issue, especially > in ada. > > Oct 6 > William Edney: > Well, what some us want is a *different* web client - but that's > ok :-) > > Angus Turner: > yes we want a different web client as well, i think people would > prefer a client that doesn't take up 256mb of ram > > Oct 6; > Rémy Sanchez: > however, I think that doing a thiner web client will be difficult.. > After all, Google's got good engineers I think... I hardly belive that > they made a slow and heavy interface on purpose. > > Elliott Cable: > I do believe they did. They consider that an okay thing, because they > think of it as a web application, and that changes how engineers like > us think. Somebody setting out to right a light, less–featured client, > that actually runs worth crap, could certainly do so. > > ✿Melissa✿ Elliott: > Off-topic: It seems all the deleted blips don't show up as being > deleted in the playback. Is this coming Real Soon Now or is the > playback just bugged? > > Oct 7 > Angus Turner: > i think lars mentioned that when i was speaking to him.... > but he didn't say when it would be implemented > > Oct 8 > Elliott Cable: > Lars was talking about them showing up when you visited first after > they were deleted, not during Playback. > > Melissa, sounds like a bug to me. They should already be showing up in > Playback, and have been for me (at least, when I last got Playback to > actually work on a wave, which was something like a month ago, I think > d-:) > > Oct 15 > Darkflame: > Speaking personaly, I dont have a preferance for XMPP or JSON...I > think I could adapt well enough to either. As long as theres a nicely > documented standard, that hopefully all wave severs will support, I'm > happy. > > Is it time to have a vote box at the top of this thread or too soon? > > [email protected]: > well, some of the other threads have moved on and deal exclusively > with how to implement the protocol over XMPP, no need to vote ;) > > Oct 18 > Okdokie. So whats the plan...as a community we make a implementation > then nudge google and hope they take it onboard? > > [email protected]: > I think that the most efficient way would simply to do that, pour our > efforts into alternative implemenations and try to make them > compatible with the apis that google offers, this way we can leverage > the gadgets / robots / apps that we build on this preview, while also > being able to extend the reach and flexibility of the 'wave platform'. > > David Hubbard: > It's also probably a good idea to put some effort into extending > fedOne, as google is already committed to maintaining it -- so they > are predisposed to look favorably on patches to it. > > Darkflame: > So, as person working in a group that needs c/s for our project, whats > the best course of action for us? > We are very keen to get started asap, but it seems very hard to do > much without it. What can we do to assist or encourage the adoption of > a standard c/s protocol? > > Oct 23 > [email protected]: > > The best is probably to provide a library to the group that basically > implements your version of the C/S Protocol and be ready for it to be > ripped apart and changed. So your best bet is to buidl an abstraction > layer on top of that library for your client that can easily be > adapted to the evolving protocol. And of course keep a close eye on > these discussions. --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Wave Protocol" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/wave-protocol?hl=en -~----------~----~----~----~------~----~------~--~---
