incubator-wave-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Joseph Gentle <jose...@gmail.com>
Subject Re: Wave Kickstarter
Date Sun, 01 Dec 2013 17:59:22 GMT
On Sun, Dec 1, 2013 at 3:47 AM, Thomas Wrobel <darkflame@gmail.com> wrote:
> " I still want a web client, but it should be
> written in pure javascript."
>
> GWT is already pure javascript. Writing "Directly" rather then via Java
> doesn't give any real advantages in the end result - it just means you rule
> out code sharing between client and server, an give yourself more work in
> terms of writing around browser differences and having to optimize for the
> web. There's things other then GWT that can do that, but raw javascript
> just adds a tone more work to web client development.

Ever since IE9, browser differences aren't the massive headache they
once were. Also GWT isn't the only way to share code between a server
and client, and there are many excellent javascript frontend libraries
too.

While I can justify the move away from Java/GWT technically, I don't
want to have that argument here. Ultimately this is a decision based
on personal taste. I immensely dislike working with Java and GWT. I'm
going to build wave 2.0's reference implementation on top of
technologies I like instead. (Like C, LibUV and NodeJS.) I am also
partial to Go, Python, Lua and Scala, and I keep hearing good things
about Closure - but I'm not convinced that any of those languages are
a good fit for this project.

Also remember the most significant part of this will be an open
platform with an open spec. If you're super keen on java, you should
be able to make a java implementation of the protocol, or contribute a
GWT wrapper around any browser libraries we write.

> That said, I pretty much agree with everything else.
> I would suggest, however, you need a few people ready to work on specific
> goals for a specific amount. And what can be achieved is kept very clear.
> Unfortunately no one will understand a kickstarter talking about OT. (or,
> not enough to get the money I think). It all needs to be put into real
> terms.
> Multiple clients and use's with a single federate realtime protocol is
> still a very big deal - no one has experience of that at the moment. The
> hard part is getting across all the use-case's for it. Or, rather, the
> specific ones you plan to make possible with the kickstarter.

I definitely agree. Thats in many ways the hardest part of all this -
boiling the idea down to a simple form that we can do a kickstarter
around, and communicating that effectively. Maybe we should explain it
as "realtime git for non-coders" or "federated google docs that you
host yourself".

> For my part, I am happy again to contribute (for free) whenever there is a
> separation of the GWT client. Or even a clean client/server protocol to
> build a new web client from stretch.
>
> Additionally, I have an Augmented Reality specific client application I
> wrote, coded for Android phones. The idea was anyone could annotate
> anything anywhere, and share it with whoever they want. The app is working
> (mostly)...except it has no sever to connect too. Nothing at the moment
> forfills the requirement. Once theres a client/server protocol to a wave
> server I could very quickly put out this (imho) rather cool Android client
> and I think it will attack a fair bit of attention.

Do you want to be in the kickstarter video? It might be cool to put in
short interviews with people like yourself who want to build cool
stuff on top of the platform to give people more of a sense of what we
can do with it.

-J

Mime
View raw message