Hi Doug Wave is many things, and is useful in many ways. Everyone sees some subset of its feature as the "core" features. I personally think that the following features define Wave:
1. Centralized model - everything is on the server. No fragmentation of information like in emails. 2. Threaded conversations - structured blips model (essentially a tree) allows to discuss topics in the context. The regular chat model allows to discuss only one topic per time, it's impossible to reply to specific message with out explaining the context, i.e. @someuser Hi, in message $$## you talked about XYZ, so I think that you are wrong/right - in Wave you just create inline blip and reply directly in the context. Saves a lot of time, effort and space. What's more - you can give links to other blips, so you "link" instead of duplicating the information. 3. Distributed - Federation allows to communicate with users on other servers. 4. Augmented by robots and gadgets. Well, casual users hand no chance to experience how powerful Wave can become once augmented by custom extensions. There are so many possible uses, that I even don't know where to start. Robots and Gadgets allow wave to be anything. Registration forms, , integration of tasks and TODOs into the document, integration of other services like Twitter and Facebook, etc.. We didn't see too much of these, but that's only because the last version of the Robot APi existed for only about 4 months before GWave was discontinued. 5. Email, chat and document in one place. Now, you vision is totally legitimate as well, it just wrong to says that there's are no other uses, or they are less important. Regarding your proposal: Probably there are other platforms that can suit you better. For example ShareJS provides simple to deploy and use OT engine. On Thu, Jan 5, 2012 at 8:59 AM, Doug <douglas.lin...@gmail.com> wrote: > Idly I was thinking today about things I liked about wave and things I > didn't and it struck me all the things I used wave for were the same thing > I used a forum for: > > You have multiple threaded conversions between groups of people, some of > which take place in public, some in small private groups. You can send > direct messages between individual users. > > The only real novel aspect of it was: > > - Rich document model for posts > - You have real time collaborative document editing > - You can share waves across multiple servers > - User submissions are verified using strong auth to prevent spoofing > - 'Bot users > - Gadgets > > Of these features, I feel no one ever did anything particularly interesting > with bots, gadgets or the real time editing... but the idea of a pretty > forum (rich editor~) you can participate in with anyone... that still seems > really cool to me. > > ...but, wiab isn't really thrilling anyone much at the moment. That hacker > news article got a few comments, but yeah... pretty much back to > silence-as-usual since then. > > I appreciate that the code in wiab is inherited from google wave, but > it's ridiculously over complicated. Under current is a hack to over come > the crazy-ness of the UI. > > Is anyone interested in going back to basics and rebuilding the wiab core > from scratch? > > With the objectives of: > > - A clean top quality, beautiful forum (aka. phpBB) with full forum > functionality in java using MVC principles. > - That permits waves (ie. threads) to be shared between server instances. > - With: > -- Strong crypto to authenticate users and user actions. > -- The full wave document model for each thread. > -- A minimalist javascript frontend for rich editing and otherwise server > side templates. > - Deployable on any compliant serlvet container > - Simple public interfaces for implementing persistence, attachments, > authentication, themes via plugins. > > And completely dropping: > - The wave api > - Robots > - Gadgets > - Concurrent editing > - An embedded hacked up version of jetty to run on > - The need for an XMPP server (as I understand it XMPP isn't actually > _used_ for anything) > - The overweight javascript front end. > > This would massively cleanup the code base, and I'm sure that there are > parts of the wiab code base that could be pulled over to get this working. > > ... the question I guess is, do people feel that would be too much of a > sacrifice to the wave spirit? > > Honestly I think the wiab code base is a lost cause at this point. > > ~ > Doug. >
