It's frankly somewhat hard for me to tell exactly what this is -- high on buzzwords, low on concrete technical content. But it looks like their focus is primarily on cloud-hosted services, right?
> It's frankly somewhat hard for me to tell exactly what this is -- high
> on buzzwords, low on concrete technical content. But it looks like
> their focus is primarily on cloud-hosted services, right?
Arjun, I think you're confusing things. This is a genuine question.
It's unfortunate that it showed up in a week when a bunch of spammers
figured out how to penetrate Google Groups. Follow the link to
Joyent.
On Thu, Sep 17, 2009 at 3:09 PM, Arjun Guha <arjun.g...@gmail.com> wrote:
> I usually remove this sort of junk. But, we got a lot this past week.
> Arjun
> On Sep 17, 2009, at 15:01 , Shriram Krishnamurthi wrote:
>> It's frankly somewhat hard for me to tell exactly what this is -- high
>> on buzzwords, low on concrete technical content. But it looks like
>> their focus is primarily on cloud-hosted services, right?
>It's frankly somewhat hard for me to tell exactly what this is -- high >on buzzwords, low on concrete technical content. But it looks like >their focus is primarily on cloud-hosted services, right? >Shriram
It does server-side JS, which is fine. But, as far as I can tell, it
doesn't integrate with JavaScript on the client in any way. So, you
could use it with Flapjax, just as their tutorial uses it with jQuery.
Arjun
On Sep 17, 3:19 pm, "Brad Jones" <pjones0...@rogers.com> wrote:
> >It's frankly somewhat hard for me to tell exactly what this is -- high
> >on buzzwords, low on concrete technical content. But it looks like
> >their focus is primarily on cloud-hosted services, right?
> >Shriram
They are interested in the code that runs in the server and propose
applications contain 1) client and server JavaScript code and 2) calls to
their server APIs. The proposal to clean up server code by introducing a
JavaScript layer is a common theme (wiki server js); they seem to have a
PHP-like templating language to generate pages with interleaved JS. We
initially provided an even simpler model: 1) client JavaScript code and 2) a
cloud storage webservice for persistence. Due to favorable trends, we no
longer need to be the ones to provide 2), which is a confidence win for
everyone involved.
Experiments in how to specify server computation are interesting -- as
Flapjax helps with concurrency, it's conceivable that our abstractions have
a use. Performance and scale introduce new concerns in this domain which
would need to be examined. Largely unexplored is the potential for mixing
client and application code; if both are already in JavaScript, migrating or
sharing objects seems natural at a technical level (even though it actually
isn't for the general case).
- Leo
On Thu, Sep 17, 2009 at 12:31 PM, Shriram Krishnamurthi <s...@cs.brown.edu>wrote:
On Thu, Sep 17, 2009 at 1:09 PM, Leo Meyerovich <lmeye...@gmail.com> wrote:
> They are interested in the code that runs in the server and propose
> applications contain 1) client and server JavaScript code and 2) calls to
> their server APIs. The proposal to clean up server code by introducing a
> JavaScript layer is a common theme (wiki server js); they seem to have a
> PHP-like templating language to generate pages with interleaved JS. We
> initially provided an even simpler model: 1) client JavaScript code and 2) a
> cloud storage webservice for persistence. Due to favorable trends, we no
> longer need to be the ones to provide 2), which is a confidence win for
> everyone involved.
> Experiments in how to specify server computation are interesting -- as
> Flapjax helps with concurrency, it's conceivable that our abstractions have
> a use. Performance and scale introduce new concerns in this domain which
> would need to be examined. Largely unexplored is the potential for mixing
> client and application code; if both are already in JavaScript, migrating or
> sharing objects seems natural at a technical level (even though it actually
> isn't for the general case).
> - Leo
> On Thu, Sep 17, 2009 at 12:31 PM, Shriram Krishnamurthi <s...@cs.brown.edu>wrote:
>> > Just watched this video
>> > presentation by the architect - http://jsconf.blip.tv/
I spent a while reading through this. Then I sat down to compose a response, and found that Leo had said everything I wanted to say!
There seems to be growing buzz on Flapjax on the server side. I know of at least one other company, 10gen [*], who has talked about doing this and doing so successfully. I am not convinced about the appeal in this. The reason JavaScript is interesting is not because it's a great language (it's not awful but it's not great either -- it's just odd) but because it's on the browsers.
Plus, and I may sound a bit jaded here having been around the dynamic language compilation block a few too many times, no matter how clever your run-time system, you're not really going to get the best performance out of a fully dynamic language. Now if only we had a Typed version of JavaScript instead....
>Hi Brad -- >I spent a while reading through this. Then I sat down to compose a >response, and found that Leo had said everything I wanted to say! >There seems to be growing buzz on Flapjax on the server side. I know >of at least one other company, 10gen [*], who has talked about doing >this and doing so successfully. I am not convinced about the appeal >in this. The reason JavaScript is interesting is not because it's a >great language (it's not awful but it's not great either -- it's just >odd) but because it's on the browsers. >Plus, and I may sound a bit jaded here having been around the dynamic >language compilation block a few too many times, no matter how clever >your run-time system, you're not really going to get the best >performance out of a fully dynamic language. Now if only we had a >Typed version of JavaScript instead.... >Shriram
In re-reading my reply, I realize I didn't explain the *:
> There seems to be growing buzz on Flapjax on the server side. I know > of at least one other company, 10gen [*], who has talked about doing > this and doing so successfully.
What I meant to say is that a former student (Eliot Horowitz) is one of the head technical honchos of 10gen, so I wanted the disclaimer clear alongside what might have looked like a plug.