Re: [nodejs] replace code in the browser while running?

[Mark Hahn]

@Brandon, I was just looking at your tracer tool and I realized that it
would be an awesome debug tool to find out when some data was getting
clobbered.  Could it show a stack trace on changes?

You should consider a tool that is optimized for debugging by breaking on
data changes.  I use the webkit inspector to break on DOM changes and it
would be great to be able to break on general data (property) changes.  For
debugging purposes even non-objects could be traced by temporarily
refactoring them into an object property in the source.

Was tracer developed for debugging, or for the purpose of building things
on top of it?

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

[nodejs] [node.js] Why my request slower and slower

[jason . 桂林]

My question attached code and image, please help me on stackoverflow,
thanks a lot.

http://stackoverflow.com/questions/10328492/why-my-request-slower-and-slower

-- 
Best regards,

Jason Green
桂林

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

[nodejs] Re: [node.js] Why my request slower and slower

[jason . 桂林]

I am writing a http benchmark module in node.js. I got a problem. I
separate my benchmark to several task to run. one task by one task. each
task define like 'GET /url for 1 times'.

My problem is, if I start a task 'GET / for 10 times' it response in
7.2k rps. If I separate the task to 10 task, 'GET / for 1 times', the
first task response fast, but the other tasks slower and slower.
http://stackoverflow.com/questions/10328492/why-my-request-slower-and-slower

2012/4/26 jason.桂林 

> My question attached code and image, please help me on stackoverflow,
> thanks a lot.
>
>
> http://stackoverflow.com/questions/10328492/why-my-request-slower-and-slower
>
> --
> Best regards,
>
> Jason Green
> 桂林
>
>
>


-- 
Best regards,

Jason Green
桂林

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs] Re: How to gzipSync?

[Axel Kittenberger]

Great, I actually seen this, when I searched for a solution, but I
need gzip, not compress.

If this isn't "compress" but gzip after all, you really should change
your README to highlight this. Maybe its the age from back of the
1990s so when one read or wrote "compress" without any highlight, Its
assumed the compression algorithmn that used to be called "compress",
that is LZ77 without huffman encoding.

On Wed, Apr 25, 2012 at 6:44 PM, Egor Egorov  wrote:
>
> https://github.com/egorFiNE/node-compress-buffer
>
>
> On Wednesday, April 18, 2012 12:12:36 PM UTC+3, Axel Kittenberger wrote:
>>
>> Is there some easy way to make yourself a zlib.gzipSync wrapper around
>> the nodes default callback/pipe interface to zlib? (Without streamline
>> or fibers)
>>
>> Its for the server startup sequence where it generates one hugh
>> javascript file for the client side to download and it keeps it raw as
>> well gzipped in memory. It does read the files with readSyncFile etc.
>> So yes I want the node process to be halted until it is done. Its fine
>> it takes 2 seconds, so please don't argue it could do this or that in
>> parallel. I also do the other *Sync calls for startup. I just want to
>> keep that simple please.
>>
>> - Axel
>
> --
> Job Board: http://jobs.nodejs.org/
> Posting guidelines:
> https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
> You received this message because you are subscribed to the Google
> Groups "nodejs" group.
> To post to this group, send email to nodejs@googlegroups.com
> To unsubscribe from this group, send email to
> nodejs+unsubscr...@googlegroups.com
> For more options, visit this group at
> http://groups.google.com/group/nodejs?hl=en?hl=en

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

[nodejs] Re: [Express-js] New local user group - New Delhi, Gurgaon, Noida, NCR

[Srirangan]

Thanks, if there are half a dozen of us or more, we can grab a cup of
coffee over a weekend. :-)

- Sri

Srirangan  |  +91 9711 477 595  |  About 
GitHub 
LinkedIn
  Twitter   |  Review19
  "Make
better decisions"


On Thu, Apr 26, 2012 at 11:08 AM, Devashish Meena <
devashish.me...@urbantouch.com> wrote:

> great initiative...
>
> +1
>
> On Thu, Apr 26, 2012 at 10:52 AM, Srirangan  wrote:
>
>> Hi,
>>
>> I'm trying to setup a local user group for New Delhi and surrounding
>> suburbs.
>>
>> If you work / live here or travel here frequently, please join:
>> http://www.meetup.com/Node-js-Dilli/
>>
>> I hope there are enough of us soon for a meetup over coffee / beer.
>>
>> - Sri
>>
>> Srirangan  |  +91 9711 477 595  |  About 
>> GitHub   
>> LinkedIn
>>   Twitter   |  Review19  
>> "Make
>> better decisions"
>>
>> --
>> You received this message because you are subscribed to the Google Groups
>> "Express" group.
>> To post to this group, send email to express...@googlegroups.com.
>> To unsubscribe from this group, send email to
>> express-js+unsubscr...@googlegroups.com.
>> For more options, visit this group at
>> http://groups.google.com/group/express-js?hl=en.
>>
>
>
>
> --
> Regards
> Devashish
>
> We will either find a way, or make one. - Hannibal
>
>  --
> You received this message because you are subscribed to the Google Groups
> "Express" group.
> To post to this group, send email to express...@googlegroups.com.
> To unsubscribe from this group, send email to
> express-js+unsubscr...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/express-js?hl=en.
>

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs] Re: Your thoughts wanted on a node AI project

[shawn wilson]

Per parsing language, you might want to look at what python's nltk and
Java's gate (less familiar with the later) have done for inspiration.
On Apr 25, 2012 10:07 PM, "Isaac Schlueter"  wrote:

> Please do not bicker over politics here.  This is not the place, and
> it never ever ends well.
>
> Please do not call each other idiots or make other personal remarks.
> If you want to say that a piece of *code* is terrible, or that a
> software approach is misguided, or that a statement is factually
> untrue, then that's perfectly fine (as long as it actually is!) but
> the Node.js user list must remain respectful.  (Challenging someone
> for benchmark numbers when they've made performance assertions is also
> not only accepted, but encouraged.)
>
> But a personal insult is not allowed, even if it's warranted.  If
> someone's *actually* causing problems, let me or one of the other
> admins know, and they'll get a warning, and if that doesn't work,
> they'll get banned.  (Even if the person is an admin.  You can
> complain about me to the other core devs on this list, and they'll set
> me straight.)  Calling someone a troll is insulting if they are not,
> and it is entertaining if they are.
>
> Don't complain in public about another user.  Just talk to an admin in
> private.  (Or email them directly off-list, if you feel comfortable
> doing so.)
>
>
> More on topic,
>
> The MIT license in node (as well as the MIT and BSD licenses in the
> vast majority of open source node programs) explicitly allows use for
> closed-source commercial applications.  If Dennis wants to keep his
> source closed, then that's his prerogative.  In fact, he can take all
> the modules we create, and use them in his closed-source proprietary
> thing, so long as he abides by the extremely liberal licenses that
> most of them use.  If you think he's wrong about the value of open
> source, or its overblownness, refute it with data.  (Or don't, and
> just go back to cranking out awesome open source software.)
>
>
> Dennis, since you came here ostensibly trying to raise interest among
> other developers (and have been mostly successful, reading through the
> thread), I would suggest re-thinking your approach somewhat.  I'm not
> talking about what's right or wrong, merely what's effective.  Many
> people come to a project like Node.js because they feel strongly about
> open source software.  If your goal is to recruit them, you should
> think about the effects that your words have.  If you want to recruit
> developers who *aren't* passionate about open source software, then
> you're in the wrong place.
>
>
> On Wed, Apr 25, 2012 at 18:20, Dennis Kane  wrote:
> >> You don't come off as an asshole -- and besides, we have no problem with
> >> assholes: we're developers. Saying the things you're saying, to the
> audience
> >> you're saying them to -- well, it's that you come off as an idiot. The
> good
> >> news is this is a lot easier to fix.
> >>
> >
> > I'm an idiot because of my political views, and not my programming
> ability,
> > I assume.  The only thing I'm worried about is the programming.  If
> you've
> > got some criticism about that, I'm all ears.
> >
> >>
> >> Oh, you're looking for help? Here's a suggestion: you could open up the
> >> code and see what people do.
> >
> > No.  Now chill out about the open source thing.  It's boring.  Redundant.
> > Etc.
> >
> >>
> >> Far be it from us to tell you what to do with your code, but you're the
> >> one asking for help, and this is naturally the first bit of feedback
> you're
> >> going to get. Why would any of us invest much more of our precious time
> >> helping you improve your proprietary product? The world doesn't work
> that
> >> way -- at least, not anymore.
> >>
> > Dude, I'm looking to get a business off the ground.  People work with
> each
> > other and try to make money off of their labors.  It has happened once or
> > twice before.  You say the word "proprietary" as if its one of the seven
> > deadly ones.
> >
> >>
> >> So yes, back to the thing itself: node.js (this is the node.js list,
> after
> >> all). You've expressed clearly that you have no interest in making your
> work
> >> part of node's ecosystem. So at this point you're just noise, no
> different
> >> than a recruiter. You just said as much yourself: "I'm here because I
> can't
> >> do this thing by myself." At least recruiters, even the worst of them,
> have
> >> a payed position to offer and the good sense not to directly insult the
> >> target audience :P
> >>
> > Sorry.  I don't have any personal feelings about node's ecosystem.  It's
> > just a tool that I take for granted.  I set up the site with express, and
> > now I don't think about it.  It just works.  I have a server hooked into
> the
> > main parser to shuttle stuff between it and express.  Sometimes I need to
> > make use of the filesystem.  Other that that, I'm pretty much using
> straight
> > javascript. It's google's v8 work that's doing most of the 

Re: [nodejs] replace code in the browser while running?

[Brandon Benvie]

Generally, the things mentioned above were in my mind when creating this 
stuff. As I've played around with proxies and used them to achieve certain 
tasks, a few generalized patterns emerged and the library is the overall 
result of the last year or so of that.

Specifically with the membrane, the first thing in my mind is a general 
"omniscience" debugging tool, or like a kind of heat map of "action". 
Imagine for a second that you could zoom out and look at your program 
visually, some sort of representation of it. Node does this with the REPL 
and console by default, and I've greatly expanded on the rendering of that 
with my own repl. Now imagine a full instrumented runtime environment, 
where you could see in real time as the action moves between areas of your 
application, or the "pings" or execution time is represented as a heat map, 
literally showing you the hot code.

This is (relatively) trivial to do with the tracer that comes with 
meta-objects. With some slight modification I could register both entrance 
and exits to any action that lands in user code, allowing 100% tracing of 
execution time, inputs, outputs, etc. The membrane included goes out of its 
way to be a non-intrusive, external membrane. That is, it wraps the border 
of an object graph. It doesn't intercept interactions between objects on 
the "inside" of the membrane (or outside of course), it simply enforces the 
barrier line in entering or exiting.

However, it simply requires less carefulness to create more of a "goo" than 
a membrane, that permeates all interactions instead of merely standing 
around the edges. This type of interaction would have proxy objects 
becoming the actual communication medium itself, with all actions from one 
object to any other object going through them by necessity. In this type of 
construct, no object ever has a direct reference to another real object, 
only mirrors. (the exception here is things created via literals). As with 
the membrane, normal execution could continue without problem, and in 
indeed without the objects even being able to tell they're in such an 
environment.

The third construct it more like a "blob", or an offensive one that 
actively seeks out and replaces all properties on all objects it meets. As 
soon as an object is exposed to this blob, it will find all of its 
non-primitive properties replaced by dopplegangers. Proxy mirrors with new 
object identities that otherwise are indistinguishable from the original. 
This one is the first that is invasive, in that its replacing objects with 
separate-identity objects where existing objects might already have 
references. However, if you''re purposefully creating all your objects 
inside of this (say for debugging or deep introspection) the opportunity to 
extract near full reconstructable JS execution playbacks exists.

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

[nodejs][moe] Moescript language

[Belleve Invis]

I've made it for years. Mostly inspired by Python and Haskell.

Doc:
http://be5invis.github.com/moescript/

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs][moe] Moescript language

[Belleve Invis]

Code Example

def async randPrintNums(n):
def tasks = []
for(var i in 0..n)
tasks.push async :>
wait sleep (100 * Math.random())
trace index
where index = i
wait join tasks
randPrintNums 100


在 2012年4月26日星期四UTC+8下午8时26分20秒，Belleve Invis写道：
>
> I've made it for years. Mostly inspired by Python and Haskell.
>
> Doc:
> http://be5invis.github.com/moescript/
>

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs][moe] Moescript language

[Oliver Leics]

On Thu, Apr 26, 2012 at 2:27 PM, Belleve Invis  wrote:
> Code Example
>
> def async randPrintNums(n):
> def tasks = []
> for(var i in 0..n)
> tasks.push async :>
> wait sleep (100 * Math.random())
> trace index
> where index = i
> wait join tasks
>
> randPrintNums 100

What is the resulting JavaScript?

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs][moe] Moescript language

[Belleve Invis]

Functions using *wait*, *!* and *<-* are specially compiled. 

   1. function F1$_(){
   2. var randPrintNums$;
   3. randPrintNums$ = async$({build:function(SCHEMATA$_){return 
   function(n$){
   4. var T1$_, T8$_, T9$_;
   5. var i$, tasks$;
   6. function block_T2(T2$_){
   7. tasks$ = [];
   8. for(((i$ = 0), (T1$_ = n$)); i$ < T1$_; (i$ = i$ + 1)){
   9. (function F3$_(){
   10. var index$;
   11. index$ = i$;
   12. return tasks$.push(async$({build:function(SCHEMATA$_){return 
   function(){
   13. var T4$_, T5$_;
   14. function block_T3(T3$_){
   15. T4$_ = (sleep$);
   16. T5$_ = ((100 * Math$.random()));
   17. return SCHEMATA$_.yield(T4$_(T5$_,block_T6));
   18. };
   19. function block_T6(T6$_){
   20. T6$_;
   21. trace$(index$);
   22. return SCHEMATA$_["return"]();
   23. };
   24. return block_T3;
   25. }}}));
   26. }());
   27. };
   28. T8$_ = (join$);
   29. T9$_ = (tasks$);
   30. return SCHEMATA$_.yield(T8$_(T9$_,block_Ta));
   31. };
   32. function block_Ta(Ta$_){
   33. Ta$_;
   34. return SCHEMATA$_["return"]();
   35. };
   36. return block_T2;
   37. }}});
   38. randPrintNums$(100);
   39. }


在 2012年4月26日星期四UTC+8下午8时32分50秒，Oliver Leics写道：
>
> On Thu, Apr 26, 2012 at 2:27 PM, Belleve Invis  
> wrote: 
> > Code Example 
> > 
> > def async randPrintNums(n): 
> > def tasks = [] 
> > for(var i in 0..n) 
> > tasks.push async :> 
> > wait sleep (100 * Math.random()) 
> > trace index 
> > where index = i 
> > wait join tasks 
> > 
> > randPrintNums 100 
>
> What is the resulting JavaScript? 
>

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs][moe] Moescript language

[Brandon Benvie]

Moescript! Glad to see you finally putting this up for review.

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs][moe] Moescript language

[Belleve Invis]

Thanks. This young thing needs ISSUES!

在 2012年4月26日星期四UTC+8下午8时39分33秒，Brandon Benvie写道：
>
> Moescript! Glad to see you finally putting this up for review.

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs][moe] Moescript language

[Axel Kittenberger]

Awesome!

I've been longing for a functional programming language that compiles
to Javascript!

At http://be5invis.github.com/moescript/ the english is hard for me to
understand. This way its twice as difficult to delve into a new
design.

What does @foobar mean?

"== and === has the same meaning in Moe". Please not, just make *one*
valid way, instead of unnecessarily creating ambitiousness.

It would be nice instead of mathematical demos to have real-life
minimal programs to make it look sweet. Like a http server written in
moescript with the node.js runtime.

BTW: Has the name to do with Moe Szyslak?

On Thu, Apr 26, 2012 at 2:26 PM, Belleve Invis  wrote:
> I've made it for years. Mostly inspired by Python and Haskell.
>
> Doc:
> http://be5invis.github.com/moescript/
>
> --
> Job Board: http://jobs.nodejs.org/
> Posting guidelines:
> https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
> You received this message because you are subscribed to the Google
> Groups "nodejs" group.
> To post to this group, send email to nodejs@googlegroups.com
> To unsubscribe from this group, send email to
> nodejs+unsubscr...@googlegroups.com
> For more options, visit this group at
> http://groups.google.com/group/nodejs?hl=en?hl=en

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs][moe] Moescript language

[Belleve Invis]

English is not my native language. Sorry. (我母语汉语)

@foobar === this.foobar

For capacity. (Originally there is only ==, And I always write ===)

Examples will be here soon.

在 2012年4月26日星期四UTC+8下午8时50分31秒，Axel Kittenberger写道：
>
> Awesome! 
>
> I've been longing for a functional programming language that compiles 
> to Javascript! 
>
> At http://be5invis.github.com/moescript/ the english is hard for me to 
> understand. This way its twice as difficult to delve into a new 
> design. 
>
> What does @foobar mean? 
>
> "== and === has the same meaning in Moe". Please not, just make *one* 
> valid way, instead of unnecessarily creating ambitiousness. 
>
> It would be nice instead of mathematical demos to have real-life 
> minimal programs to make it look sweet. Like a http server written in 
> moescript with the node.js runtime. 
>
> BTW: Has the name to do with Moe Szyslak? 
>
> On Thu, Apr 26, 2012 at 2:26 PM, Belleve Invis  
> wrote: 
> > I've made it for years. Mostly inspired by Python and Haskell. 
> > 
> > Doc: 
> > http://be5invis.github.com/moescript/ 
> > 
> > -- 
> > Job Board: http://jobs.nodejs.org/ 
> > Posting guidelines: 
> > https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines 
> > You received this message because you are subscribed to the Google 
> > Groups "nodejs" group. 
> > To post to this group, send email to nodejs@googlegroups.com 
> > To unsubscribe from this group, send email to 
> > nodejs+unsubscr...@googlegroups.com 
> > For more options, visit this group at 
> > http://groups.google.com/group/nodejs?hl=en?hl=en 
>

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs][moe] Moescript language

[Brandon Benvie]

I've talked to Invis a bit over the last few months, and there is a 
significant language barrier. But he does speak enough to be able to 
interpret his meaning, with some effort. He's quite intelligent, very much 
driven, and apparently language design is his thing. =D

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs][moe] Moescript language

[Belleve Invis]

I am chemistry major actually in the university. And I hate things designed 
"organic-synthesis-ish", like CSS in 2007.
I think there should be no exceptional situations in anything I designed.

在 2012年4月26日星期四UTC+8下午8时59分03秒，Brandon Benvie写道：
>
> I've talked to Invis a bit over the last few months, and there is a 
> significant language barrier. But he does speak enough to be able to 
> interpret his meaning, with some effort. He's quite intelligent, very much 
> driven, and apparently language design is his thing. =D

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs][moe] Moescript language

[Nuno Job]

I can't say I see the ressemblance with Haskell at all. At least not
in your code example which looked rubyish to me, def. not functional.

Roy (http://roy.brianmckenna.org/) looks much more like haskell imho.

nuno

On Thu, Apr 26, 2012 at 2:08 PM, Belleve Invis  wrote:
> I am chemistry major actually in the university. And I hate things designed
> "organic-synthesis-ish", like CSS in 2007.
> I think there should be no exceptional situations in anything I designed.
>
> 在 2012年4月26日星期四UTC+8下午8时59分03秒，Brandon Benvie写道：
>>
>> I've talked to Invis a bit over the last few months, and there is a
>> significant language barrier. But he does speak enough to be able to
>> interpret his meaning, with some effort. He's quite intelligent, very much
>> driven, and apparently language design is his thing. =D
>
> --
> Job Board: http://jobs.nodejs.org/
> Posting guidelines:
> https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
> You received this message because you are subscribed to the Google
> Groups "nodejs" group.
> To post to this group, send email to nodejs@googlegroups.com
> To unsubscribe from this group, send email to
> nodejs+unsubscr...@googlegroups.com
> For more options, visit this group at
> http://groups.google.com/group/nodejs?hl=en?hl=en

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs][moe] Moescript language

[Belleve Invis]

I don't want made Moe "that" haskellish. Roy (and of cource, haskell) is 
pure and perfect, and so far away from daily development. 

在 2012年4月26日星期四UTC+8下午9时23分58秒，Nuno Job写道：
>
> I can't say I see the ressemblance with Haskell at all. At least not 
> in your code example which looked rubyish to me, def. not functional. 
>
> Roy (http://roy.brianmckenna.org/) looks much more like haskell imho. 
>
> nuno 
>
> On Thu, Apr 26, 2012 at 2:08 PM, Belleve Invis  
> wrote: 
> > I am chemistry major actually in the university. And I hate things 
> designed 
> > "organic-synthesis-ish", like CSS in 2007. 
> > I think there should be no exceptional situations in anything I 
> designed. 
> > 
> > 在 2012年4月26日星期四UTC+8下午8时59分03秒，Brandon Benvie写道： 
> >> 
> >> I've talked to Invis a bit over the last few months, and there is a 
> >> significant language barrier. But he does speak enough to be able to 
> >> interpret his meaning, with some effort. He's quite intelligent, very 
> much 
> >> driven, and apparently language design is his thing. =D 
> > 
> > -- 
> > Job Board: http://jobs.nodejs.org/ 
> > Posting guidelines: 
> > https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines 
> > You received this message because you are subscribed to the Google 
> > Groups "nodejs" group. 
> > To post to this group, send email to nodejs@googlegroups.com 
> > To unsubscribe from this group, send email to 
> > nodejs+unsubscr...@googlegroups.com 
> > For more options, visit this group at 
> > http://groups.google.com/group/nodejs?hl=en?hl=en 
>

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs][moe] Moescript language

[Brandon Benvie]

Whatever the opinions of Moe, it seems amiss to me that someone who has 
spent the time to design any programming language should say they're "day 
job" is chemistry. Not to dismiss chemistry, but it's clear you have a 
passion and knowledge set that has no involvement with chemistry.

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs] Why doesn't URL.parse decode the URI?

[Brandon Benvie]

It's probably because almost half! of the global objects in JavaScript 
(mandated by spec) are functions dedicated to encoding and decoding that 
precise type of string, and the node developers didn't want to presuppose 
which of those you did or didn't use
>
>
escape
unescape
encodeURI
decodeURI
encodeURIComponent
decodeURIComponent

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs][moe] Moescript language

[Dan North]

On 26 April 2012 13:50, Axel Kittenberger  wrote:

> Awesome!
>
> I've been longing for a functional programming language that compiles
> to Javascript!
>

Have you looked at
ClojureScript
?

I'm not necessarily advocating anything that compiles to JavaScript - it
has some reasonably nice functional affordances itself, especially combined
with libraries like underscore - but I can see why you might want one.

And Clojure fits in my head better than Haskell, which has sharp edges and
makes my nose bleed.

Cheers,
Dan

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs][moe] Moescript language

[JeanHuguesRobert]

I like the concise syntax for lambda, including f{stuff}. Maybe 
f(a,b){stuff} could behave like Ruby's?

Also: {|x| x * 3} could be { it * 3 } where "it" stands for the single 
parameter of the lambda.

Le jeudi 26 avril 2012 14:26:20 UTC+2, Belleve Invis a écrit :
>
> I've made it for years. Mostly inspired by Python and Haskell.
>
> Doc:
> http://be5invis.github.com/moescript/
>

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs] Why doesn't URL.parse decode the URI?

[Isaac Schlueter]

It wouldn't be out of the question to add a decoded version of the
path.  It'd have to be a new member, though, and perform adequately
(maybe a flag to enable it or something.)

But I don't think it's really that pressing of an issue.  The question
of whether you wish to use decodeURI or unescape is relevant, but I
think we can probably just never use unescape.

On Thu, Apr 26, 2012 at 07:06, Brandon Benvie  wrote:
> It's probably because almost half! of the global objects in JavaScript
> (mandated by spec) are functions dedicated to encoding and decoding that
> precise type of string, and the node developers didn't want to presuppose
> which of those you did or didn't use
>
>
> escape
> unescape
> encodeURI
> decodeURI
> encodeURIComponent
> decodeURIComponent
>
> --
> Job Board: http://jobs.nodejs.org/
> Posting guidelines:
> https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
> You received this message because you are subscribed to the Google
> Groups "nodejs" group.
> To post to this group, send email to nodejs@googlegroups.com
> To unsubscribe from this group, send email to
> nodejs+unsubscr...@googlegroups.com
> For more options, visit this group at
> http://groups.google.com/group/nodejs?hl=en?hl=en

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

[nodejs] Multiple Web Roles on Azure

[Rafel]

Hello All,
I am a newbie to node. right now i am trying to create a web app using
node and deploying it on Azure. My idea is to create a website which
has simple tabs on menu bar and each tab open a new functionality e.g.
one tab gives me grid view, other give me email functionality, other
is for chat etc.
So my question is do i need to create multiple web roles in this case,
or combination of web roles and worker roles. How it goes?? give me an
idea.

I appreciate any help.

Thanks.

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs] GYP seems build successfully, but can not load shared library.

[Ben Noordhuis]

On Thu, Apr 26, 2012 at 07:36, Qing Guan  wrote:
> Hi:
>    I just gave the new gyp build tool a try to build my GeoIP binding for
> node, It seems that the building process was successful. But when I try to
> require the binding in REPL, I got the follow error message:
>
> -
>
>    > var binding = require('./geoip.node')
> Error: Unable to load shared library
> /home/kuno/Code/repository/geoip/build/Release/lib.target/geoip.node
>     at Object..node (module.js:472:11)
>     at Module.load (module.js:348:31)
>     at Function._load (module.js:308:12)
>     at Module.require (module.js:354:17)
>     at require (module.js:370:17)
>     at repl:1:15
>     at REPLServer.eval (repl.js:80:21)
>     at repl.js:190:20
>     at REPLServer.eval (repl.js:87:5)
>     at Interface. (repl.js:182:12)
> 
>
> Here is my binding.gyp file, use node-ffi's binding.gyp as a template...
>
> 
> {
>   'targets': [
>     {
>       'target_name': 'geoip',
>       'type': 'shared_library',
>       'sources': [
>           'src/init.cc'
>         , 'src/country.cc','src/country6.cc'
>         , 'src/city.cc','src/city6.cc'
>         , 'src/org.cc','src/region.cc'
>         , 'src/netspeed.cc', 'src/utils.cc'
>       ],
>       'include_dirs': [
>           'deps/geoip/libGeoIP'
>       ],
>       'dependencies': [
>       ],
>       'conditions': [
>         ['OS=="win"', {
>           'libraries': [
>           ],
>           'dependencies': [
>           ]
>         }, {
>           'libraries': [
>           ],
>         }],
>         ['OS=="mac"', {
>           'xcode_settings': {
>             'GCC_ENABLE_CPP_EXCEPTIONS': 'YES',
>             'MACOSX_DEPLOYMENT_TARGET': '10.5',
>             'OTHER_CFLAGS': [
>                 '-ObjC++'
>             ]
>           },
>           'libraries': [
>               '-lobjc'
>           ],
>         }]
>       ]
>     },
>   ]
> }
> -
> Any suggestion/advice is appreciated.
>
> The full details can be check at the gyp branch of my GeoIP repository at
> github:
>
> https://github.com/kuno/GeoIP/tree/gyp

Try it with node master or 0.7.8, it will tell you why it cannot load
the module.

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

[nodejs] Re: How to transfer objects between Node.js and Django?

[Bika]

I've read about a similar problem here [1], they basically write from node 
to Redis, and use Redis' pub/sub to broadcast the changes to django.

[1]: http://lincolnloop.com/blog/2012/apr/23/ginger-tech-stack/

2012. április 25., szerda 15:15:35 UTC+2 időpontban Cassio Melo a 
következőt írta:
>
> I'm beginning with websockets and I'm confused with the number 
> libraries and configuration options. I just want to setup a project 
> where a node.js server calls a method in python/django and when the 
> last has finished, it transfers the result back to the node.js 
> server. 
>
> I'm not sure which is the best way of doing that (supposing the method 
> to be executed in django takes some time to finish). I see two 
> possibilities: 
>
> - node.js send async http requests to django server 
> - node.js communicates with django using AMQP 
>
> Can you give your view about it?

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

[nodejs] Re: [node.js] Why my request slower and slower

[jason . 桂林]

fixed, my foolish mistake

2012/4/26 jason.桂林 

> I am writing a http benchmark module in node.js. I got a problem. I
> separate my benchmark to several task to run. one task by one task. each
> task define like 'GET /url for 1 times'.
>
> My problem is, if I start a task 'GET / for 10 times' it response in
> 7.2k rps. If I separate the task to 10 task, 'GET / for 1 times', the
> first task response fast, but the other tasks slower and slower.
>
> http://stackoverflow.com/questions/10328492/why-my-request-slower-and-slower
>
> 2012/4/26 jason.桂林 
>
> My question attached code and image, please help me on stackoverflow,
>> thanks a lot.
>>
>>
>> http://stackoverflow.com/questions/10328492/why-my-request-slower-and-slower
>>
>> --
>> Best regards,
>>
>> Jason Green
>> 桂林
>>
>>
>>
>
>
> --
> Best regards,
>
> Jason Green
> 桂林
>
>
>


-- 
Best regards,

Jason Green
桂林

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs] GYP seems build successfully, but can not load shared library.

[Nathan Rajlich]

It looks like you're never actually linking to libgeoip. You'll need to add
something to the "libraries" section of the gyp file.

On Thu, Apr 26, 2012 at 8:55 AM, Ben Noordhuis  wrote:

> On Thu, Apr 26, 2012 at 07:36, Qing Guan  wrote:
> > Hi:
> >I just gave the new gyp build tool a try to build my GeoIP binding for
> > node, It seems that the building process was successful. But when I try
> to
> > require the binding in REPL, I got the follow error message:
> >
> >
> -
> >
> >> var binding = require('./geoip.node')
> > Error: Unable to load shared library
> > /home/kuno/Code/repository/geoip/build/Release/lib.target/geoip.node
> > at Object..node (module.js:472:11)
> > at Module.load (module.js:348:31)
> > at Function._load (module.js:308:12)
> > at Module.require (module.js:354:17)
> > at require (module.js:370:17)
> > at repl:1:15
> > at REPLServer.eval (repl.js:80:21)
> > at repl.js:190:20
> > at REPLServer.eval (repl.js:87:5)
> > at Interface. (repl.js:182:12)
> >
> 
> >
> > Here is my binding.gyp file, use node-ffi's binding.gyp as a template...
> >
> >
> 
> > {
> >   'targets': [
> > {
> >   'target_name': 'geoip',
> >   'type': 'shared_library',
> >   'sources': [
> >   'src/init.cc'
> > , 'src/country.cc','src/country6.cc'
> > , 'src/city.cc','src/city6.cc'
> > , 'src/org.cc','src/region.cc'
> > , 'src/netspeed.cc', 'src/utils.cc'
> >   ],
> >   'include_dirs': [
> >   'deps/geoip/libGeoIP'
> >   ],
> >   'dependencies': [
> >   ],
> >   'conditions': [
> > ['OS=="win"', {
> >   'libraries': [
> >   ],
> >   'dependencies': [
> >   ]
> > }, {
> >   'libraries': [
> >   ],
> > }],
> > ['OS=="mac"', {
> >   'xcode_settings': {
> > 'GCC_ENABLE_CPP_EXCEPTIONS': 'YES',
> > 'MACOSX_DEPLOYMENT_TARGET': '10.5',
> > 'OTHER_CFLAGS': [
> > '-ObjC++'
> > ]
> >   },
> >   'libraries': [
> >   '-lobjc'
> >   ],
> > }]
> >   ]
> > },
> >   ]
> > }
> >
> -
> > Any suggestion/advice is appreciated.
> >
> > The full details can be check at the gyp branch of my GeoIP repository at
> > github:
> >
> > https://github.com/kuno/GeoIP/tree/gyp
>
> Try it with node master or 0.7.8, it will tell you why it cannot load
> the module.
>
> --
> Job Board: http://jobs.nodejs.org/
> Posting guidelines:
> https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
> You received this message because you are subscribed to the Google
> Groups "nodejs" group.
> To post to this group, send email to nodejs@googlegroups.com
> To unsubscribe from this group, send email to
> nodejs+unsubscr...@googlegroups.com
> For more options, visit this group at
> http://groups.google.com/group/nodejs?hl=en?hl=en
>

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs] Re: Your thoughts wanted on a node AI project

[Dennis Kane]

On Thursday, April 26, 2012 3:35:48 AM UTC-7, shawn wilson wrote:
>
> Per parsing language, you might want to look at what python's nltk and 
> Java's gate (less familiar with the later) have done for inspiration
>
Yeah, I saw the python nltk yesterday just to see what was currently out 
there.  Despite the fact that I absolutely abhor the indentation 
enforcement of python (neither here nor there), that whole project is just 
more of the same "everything and the kitchen sink" mentality.  There is so 
much in the way of graphical representation that has nothing to do with a 
program that just gets things right.  That 's all I care about.  As long as 
the parsing works very well (the logic of 'and', 'or', 'if', 'then', 
'because', etc), then everything else will start falling into place.  Just 
think of this as you would any c compiler or script interpreter.  The 
compiler must be able to get the intended logic right.  As long as that is 
the case, the user/programmer can do whatever the frick he or she wants as 
far as creating meaningful content.  This thing is doing pretty damn good 
with the logic on some highly convoluted statements and sequences thereof. 
 Forward thinking businesses will have to take note of this fact... they 
have no choice.  Then we can all start getting together to determine how to 
develop the exact dictionaries and statement types that will fit their 
needs as well as possible.

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

[nodejs] Re: Writing an HTML5 DOM module in C/C++. Anyone interested?

[Luke Scott]

On Wednesday, April 25, 2012 6:36:12 PM UTC-7, Brandon Benvie wrote:
>
> You probably want to take a second look at dom.js if you're not dead set 
> on C++ because it's the closest to what you're describing. The thing about 
> the DOM and IDL interfaces is that there's an enforced separation between 
> the interface objects and the implementation objects. What this means is 
> that you can load up the impl classes, which contain almost all the actual 
> weight of the code, in one V8 context and then simply generate new 
> interface objects along with a new V8 context each time. The interfaces are 
> generated directly from IDL in fact, just converted to JS and with type 
> adapters like OptionalBoolean, OptionalStringOrNull, etc. (directly 
> translated from WebIDL spec). They use harmony proxies to present as normal 
> objects and forward to the corresponding impl objects that are never 
> directly exposed to sandboxed code.
>
> It's important to note the distinction between something like jsdom and 
> dom.js. jsdom is module that does pretty well in approximating a DOM-like 
> environment. dom.js is a spec compliant (almost) host environment 
> implemented in JavaScript that achieves full separation between 
> implementation and interface. Not the same thing at all. 
>
 

> Also dom.js was developed at Mozilla for use in Spidermonkey, but it works 
> in V8 now (to satisfy your "not V8 specific" requirement). It's very 
> generic in terms of requirements for js environments but it does require 
> WeakMaps and Proxies which limits its use to V8 and Spidermonkey until the 
> other engines implement them.



It's not that I'm dead set, it's just I may not always have a V8 context, 
or any JavaScript context. I'm writing my application in Go (golang.org). 
With Go I can link C libraries (C++ with C wrapper) and call the functions 
directly. I'm using V8 as a way for users to write plugins for their pages. 
In other areas, other than the user's pages, I won't be using V8. Being 
able to load and modify the DOM from Go is also something I need. C++ (or 
even C) is the common link to Go, V8, and even Node.js (or any platform 
that can link libraries). I considered writing this in Go, but then V8 has 
to reach through C++, and then such a library would only be useful to Go 
users. C/C++ is the most direct method for everyone.

Luke

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs] Why doesn't URL.parse decode the URI?

[Matt]

On Thu, Apr 26, 2012 at 11:23 AM, Isaac Schlueter  wrote:

> It wouldn't be out of the question to add a decoded version of the
> path.  It'd have to be a new member, though, and perform adequately
> (maybe a flag to enable it or something.)
>

Of course. I wasn't really asking for it to be changed - hopefully any
libraries are already coping with this already.


> But I don't think it's really that pressing of an issue.


No the question was just whether there was logic behind it - I wanted to
make sure I wasn't wildly off base :)

As long as express/connect don't interpret %3F as the start of the
querystring like Sinatra does then we're all good. I'm fairly sure they
won't given the following is parsed right:

> url.parse("http://host/something%3F%20here?a=b";, true);
{ protocol: 'http:',
  slashes: true,
  host: 'host',
  hostname: 'host',
  href: 'http://host/something%3F%20here?a=b',
  search: '?a=b',
  query: { a: 'b' },
  pathname: '/something%3F%20here',
  path: '/something%3F%20here?a=b' }


Matt.

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs] Re: [node.js] Why my request slower and slower

[Matt]

Why did you remove the post though? Someone else might make the same
mistake - it's always good to keep history of how you fixed it.

2012/4/26 jason.桂林 

> fixed, my foolish mistake
>
> 2012/4/26 jason.桂林 
>
> I am writing a http benchmark module in node.js. I got a problem. I
>> separate my benchmark to several task to run. one task by one task. each
>> task define like 'GET /url for 1 times'.
>>
>> My problem is, if I start a task 'GET / for 10 times' it response in
>> 7.2k rps. If I separate the task to 10 task, 'GET / for 1 times', the
>> first task response fast, but the other tasks slower and slower.
>>
>> http://stackoverflow.com/questions/10328492/why-my-request-slower-and-slower
>>
>> 2012/4/26 jason.桂林 
>>
>> My question attached code and image, please help me on stackoverflow,
>>> thanks a lot.
>>>
>>>
>>> http://stackoverflow.com/questions/10328492/why-my-request-slower-and-slower
>>>
>>> --
>>> Best regards,
>>>
>>> Jason Green
>>> 桂林
>>>
>>>
>>>
>>
>>
>> --
>> Best regards,
>>
>> Jason Green
>> 桂林
>>
>>
>>
>
>
> --
> Best regards,
>
> Jason Green
> 桂林
>
>
>  --
> Job Board: http://jobs.nodejs.org/
> Posting guidelines:
> https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
> You received this message because you are subscribed to the Google
> Groups "nodejs" group.
> To post to this group, send email to nodejs@googlegroups.com
> To unsubscribe from this group, send email to
> nodejs+unsubscr...@googlegroups.com
> For more options, visit this group at
> http://groups.google.com/group/nodejs?hl=en?hl=en
>

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

[nodejs] Need a quick start? Try express-bootstrap

[Srirangan]

This may save you 5 minutes the next time you start building a new web-app..

https://github.com/Srirangan/express-bootstrap

A very simple template of an Express.js single page web application loaded
with jQuery and Twitter Bootstrap

- Sri

Srirangan  |  +91 9711 477 595  |  About 
GitHub 
LinkedIn
  Twitter   |  Review19
  "Make
better decisions"

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs][moe] Moescript language

[Jann Horn]

On Thu, Apr 26, 2012 at 05:34:33AM -0700, Belleve Invis wrote:
> Functions using *wait*, *!* and *<-* are specially compiled. 
> 
>1. function F1$_(){
>2. var randPrintNums$;
>3. randPrintNums$ = async$({build:function(SCHEMATA$_){return 
>function(n$){
>4. var T1$_, T8$_, T9$_;
>5. var i$, tasks$;
>6. function block_T2(T2$_){
>7. tasks$ = [];
>8. for(((i$ = 0), (T1$_ = n$)); i$ < T1$_; (i$ = i$ + 1)){
>9. (function F3$_(){
>10. var index$;
>11. index$ = i$;
>12. return tasks$.push(async$({build:function(SCHEMATA$_){return 
>function(){
>13. var T4$_, T5$_;
>14. function block_T3(T3$_){
>15. T4$_ = (sleep$);
>16. T5$_ = ((100 * Math$.random()));
>17. return SCHEMATA$_.yield(T4$_(T5$_,block_T6));
>18. };
>19. function block_T6(T6$_){
>20. T6$_;
>21. trace$(index$);
>22. return SCHEMATA$_["return"]();
>23. };
>24. return block_T3;
>25. }}}));
>26. }());
>27. };
>28. T8$_ = (join$);
>29. T9$_ = (tasks$);
>30. return SCHEMATA$_.yield(T8$_(T9$_,block_Ta));
>31. };
>32. function block_Ta(Ta$_){
>33. Ta$_;
>34. return SCHEMATA$_["return"]();
>35. };
>36. return block_T2;
>37. }}});
>38. randPrintNums$(100);
>39. }
> 
> 
> 在 2012年4月26日星期四UTC+8下午8时32分50秒，Oliver Leics写道：
> >
> > On Thu, Apr 26, 2012 at 2:27 PM, Belleve Invis  
> > wrote: 
> > > Code Example 
> > > 
> > > def async randPrintNums(n): 
> > > def tasks = [] 
> > > for(var i in 0..n) 
> > > tasks.push async :> 
> > > wait sleep (100 * Math.random()) 
> > > trace index 
> > > where index = i 
> > > wait join tasks 
> > > 
> > > randPrintNums 100

That looks like it's going to be pretty slow... is that an issue or isn't it as 
bad as it looks?


pgp5gJk1C0goO.pgp
Description: PGP signature

[nodejs] Re: [node.js] Why my request slower and slower

[mscdex]

On Apr 26, 12:58 pm, Matt  wrote:
> Why did you remove the post though? Someone else might make the same
> mistake - it's always good to keep history of how you fixed it.

+1

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs] Re: Your thoughts wanted on a node AI project

[Dennis Kane]

On Wednesday, April 25, 2012 7:07:52 PM UTC-7, Isaac Schlueter wrote:
>
>
> The MIT license in node (as well as the MIT and BSD licenses in the 
> vast majority of open source node programs) explicitly allows use for 
> closed-source commercial applications.  If Dennis wants to keep his 
> source closed, then that's his prerogative.  In fact, he can take all 
> the modules we create, and use them in his closed-source proprietary 
> thing, so long as he abides by the extremely liberal licenses that 
> most of them use.  If you think he's wrong about the value of open 
> source, or its overblownness, refute it with data.  (Or don't, and 
> just go back to cranking out awesome open source software.) 
>
>
Let's all remember what open source is really all about.  A program is 
called closed source if it is distributed in binary format only.  The open 
source movement makes the demand that one cannot distribute a binary 
program using, for example, GPL'ed code without also making the source code 
available.

But a server side process has nothing to do with any of this!  I am not 
distributing the program... I am interested in running a service like 
Google (which will hopefully eventually overtake them... but don't tell 
anyone I said that).

Is anyone seriously demanding Google to freely distribute all of their 
painstakingly developed search algorithms?  Not likely!

The basic fact is that this thing is the result of years and years of 
absolute psychological warfare between me and my computers.  And given the 
fact that robust NatLang Processing (weak AI) is something of a holy grail 
for tech enthusiasts, the stakes in all of this are quite big.

I am not saying that releasing the code for this won't ever happen.  But I 
am saying that releasing code is a very major decision that should never be 
taken lightly.  All I do know is that now is not the time.

But I will talk about it.  I won't want to give away too many of the 
technicals in an open forum, but I will give some of them away to people 
who I can trust.  And in order to me to be able to trust someone fully, I 
have to feel that they actually have an interest in the problems 
surrounding NLP.

The Net is absolutely littered with freely available NLP projects, code and 
all .  They are just not interesting.  But you know what is interesting?  A 
site with a lone input box on a white background with a snazzy, colorful 
logo above it (sound familiar?) that "just works" as advertised.  I want to 
get something like that on the site pretty soon.

Now that I'm pretty well done with the hair-pulling aspects of my coding, I 
can start having fun with putting variety into the thing (giving it the 
"wow factor").  Different kinds of words, statements, sense checking, etc. 
 This kind of stuff should not be very difficult for any competent 
programmer.

That's what I really want to start getting on the same wavelength with 
people around here.  Stop thinking that this thing has to be some kind of 
end-all-be-all killer app from the outset.  I mean, there are so many 
things to be done.  I want to be able to translate natural ways of 
referring to time points (last week, the day after tomorrow, etc) into 
their precise Unix timestamps.  Not very difficult work, but it's something 
that really should start getting worked on.

I really feel that this thing could give quite a few of you out there a 
comfortable living.  This could open up totally new vistas of the tech 
sector.  This has the potential of going places pretty quickly, and the 
earlier that people get in on it, the greater the potential for reward.


Dennis, since you came here ostensibly trying to raise interest among 
> other developers (and have been mostly successful, reading through the 
> thread), I would suggest re-thinking your approach somewhat.  I'm not 
> talking about what's right or wrong, merely what's effective.  Many 
> people come to a project like Node.js because they feel strongly about 
> open source software.  If your goal is to recruit them, you should 
> think about the effects that your words have.  If you want to recruit 
> developers who *aren't* passionate about open source software, then 
> you're in the wrong place. 
>
> I've already been told more than once to rethink/change my approach. 
 Seriously?  I mean, let's get serious here.  The only time I get snarky is 
when people get pushy with their demands to just hand out something that 
has resulted from years of torment.  This just is not any old program.  It 
just isn't.  Period.

Again, I have to reiterate that there is a vast difference between the 
philosophy of open source as espoused by Eric Raymond, Richard Stallman and 
company and the reality of people actually opening up source code in order 
to actually *improve* it.  Please someone break open the source for emacs 
right now and make it better.  I dare you.  My philosophical views are 
always changing.  Conservative, liberal, whatever, blah blah.  I'm just no

Re: [nodejs] Re: Your thoughts wanted on a node AI project

[Mark Hahn]

I guess the question everyone is thinking (or asking) is why are you here?
 I can understand the recruiting, but I can't think of any other reason.

On Thu, Apr 26, 2012 at 10:53 AM, Dennis Kane  wrote:

> On Wednesday, April 25, 2012 7:07:52 PM UTC-7, Isaac Schlueter wrote:
>>
>>
>> The MIT license in node (as well as the MIT and BSD licenses in the
>> vast majority of open source node programs) explicitly allows use for
>> closed-source commercial applications.  If Dennis wants to keep his
>> source closed, then that's his prerogative.  In fact, he can take all
>> the modules we create, and use them in his closed-source proprietary
>> thing, so long as he abides by the extremely liberal licenses that
>> most of them use.  If you think he's wrong about the value of open
>> source, or its overblownness, refute it with data.  (Or don't, and
>> just go back to cranking out awesome open source software.)
>>
>>
> Let's all remember what open source is really all about.  A program is
> called closed source if it is distributed in binary format only.  The open
> source movement makes the demand that one cannot distribute a binary
> program using, for example, GPL'ed code without also making the source code
> available.
>
> But a server side process has nothing to do with any of this!  I am not
> distributing the program... I am interested in running a service like
> Google (which will hopefully eventually overtake them... but don't tell
> anyone I said that).
>
> Is anyone seriously demanding Google to freely distribute all of their
> painstakingly developed search algorithms?  Not likely!
>
> The basic fact is that this thing is the result of years and years of
> absolute psychological warfare between me and my computers.  And given the
> fact that robust NatLang Processing (weak AI) is something of a holy grail
> for tech enthusiasts, the stakes in all of this are quite big.
>
> I am not saying that releasing the code for this won't ever happen.  But I
> am saying that releasing code is a very major decision that should never be
> taken lightly.  All I do know is that now is not the time.
>
> But I will talk about it.  I won't want to give away too many of the
> technicals in an open forum, but I will give some of them away to people
> who I can trust.  And in order to me to be able to trust someone fully, I
> have to feel that they actually have an interest in the problems
> surrounding NLP.
>
> The Net is absolutely littered with freely available NLP projects, code
> and all .  They are just not interesting.  But you know what is
> interesting?  A site with a lone input box on a white background with a
> snazzy, colorful logo above it (sound familiar?) that "just works" as
> advertised.  I want to get something like that on the site pretty soon.
>
> Now that I'm pretty well done with the hair-pulling aspects of my coding,
> I can start having fun with putting variety into the thing (giving it the
> "wow factor").  Different kinds of words, statements, sense checking, etc.
>  This kind of stuff should not be very difficult for any competent
> programmer.
>
> That's what I really want to start getting on the same wavelength with
> people around here.  Stop thinking that this thing has to be some kind of
> end-all-be-all killer app from the outset.  I mean, there are so many
> things to be done.  I want to be able to translate natural ways of
> referring to time points (last week, the day after tomorrow, etc) into
> their precise Unix timestamps.  Not very difficult work, but it's something
> that really should start getting worked on.
>
> I really feel that this thing could give quite a few of you out there a
> comfortable living.  This could open up totally new vistas of the tech
> sector.  This has the potential of going places pretty quickly, and the
> earlier that people get in on it, the greater the potential for reward.
>
>
> Dennis, since you came here ostensibly trying to raise interest among
>> other developers (and have been mostly successful, reading through the
>> thread), I would suggest re-thinking your approach somewhat.  I'm not
>> talking about what's right or wrong, merely what's effective.  Many
>> people come to a project like Node.js because they feel strongly about
>> open source software.  If your goal is to recruit them, you should
>> think about the effects that your words have.  If you want to recruit
>> developers who *aren't* passionate about open source software, then
>> you're in the wrong place.
>>
>> I've already been told more than once to rethink/change my approach.
>  Seriously?  I mean, let's get serious here.  The only time I get snarky is
> when people get pushy with their demands to just hand out something that
> has resulted from years of torment.  This just is not any old program.  It
> just isn't.  Period.
>
> Again, I have to reiterate that there is a vast difference between the
> philosophy of open source as espoused by Eric Raymond, Richard Stallman and
> company an

Re: [nodejs] Multiple Web Roles on Azure

[Ryan Schmidt]

On Apr 26, 2012, at 10:08, Rafel wrote:

> I am a newbie to node. right now i am trying to create a web app using
> node and deploying it on Azure. My idea is to create a website which
> has simple tabs on menu bar and each tab open a new functionality e.g.
> one tab gives me grid view, other give me email functionality, other
> is for chat etc.
> So my question is do i need to create multiple web roles in this case,
> or combination of web roles and worker roles. How it goes?? give me an
> idea.

I must admit I have no idea what your question is. What is a "web role"? What 
is a "worker role"?

I think the answer is probably: node.js lets you program whatever you want to, 
so go and do it.


-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs] Multiple Web Roles on Azure

[Angel Java Lopez]

Hmmm... I guess you only need a web role with node.js

(I usually find node.js in worker roles, but there are powershell scripts
that setup node.js as a web role)

An example at:
http://nodeblog.cloudapp.net/running-wheat-git-based-blog-engine-on-windows-azure


You can read that blog with some info about node.js on Azure.

But I a bit confused. Your requirements are not related to web role/nodejs,
etc... So I think you can start with the above approach: node.js on web
node.

Links I visited/collected about node.js and Azure:
http://delicious.com/ajlopez/azure+nodejs

Angel "Java" Lopez
http://www.ajlopez.com
http://twitter.com/ajlopez


On Thu, Apr 26, 2012 at 12:08 PM, Rafel  wrote:

> Hello All,
> I am a newbie to node. right now i am trying to create a web app using
> node and deploying it on Azure. My idea is to create a website which
> has simple tabs on menu bar and each tab open a new functionality e.g.
> one tab gives me grid view, other give me email functionality, other
> is for chat etc.
> So my question is do i need to create multiple web roles in this case,
> or combination of web roles and worker roles. How it goes?? give me an
> idea.
>
> I appreciate any help.
>
> Thanks.
>
> --
> Job Board: http://jobs.nodejs.org/
> Posting guidelines:
> https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
> You received this message because you are subscribed to the Google
> Groups "nodejs" group.
> To post to this group, send email to nodejs@googlegroups.com
> To unsubscribe from this group, send email to
> nodejs+unsubscr...@googlegroups.com
> For more options, visit this group at
> http://groups.google.com/group/nodejs?hl=en?hl=en
>

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs] Re: [node.js] Why my request slower and slower

[jason . 桂林]

OK, I explain here

the wrong code is:

function sendRequest
> 02  if(running > concurrent) return;
> 03  if(running ++ < concurrent && left >0)
> process.nextTick(sendRequest);
>
>   http.request(reqOptions, function(res) {
>res.on('end', function() {
>  running --;
> 08 if(--left == 0) {
> 09nextTask()
> 10 }
> process.nextTick(sendRequest)
>});
>   });
> }


When the tasks[1] is done, means left <= 0, line 9 will call nextTask
execute tasks[2], and but line 11 is still executed, and line 2 does not
filter the left < 0 and continue send request, so tasks[1] and tasks[2] is
both running, so the tasks[2] display rps just about half of tasks[1].

I modified 2 lines code  fixed this issue.

function sendRequest

  if(running > concurrent || left <=0) return;
  if(running ++ < concurrent) process.nextTick(sendRequest);

}

BTW: you can clone https://github.com/guileen/siege.js , run the example, I
bet you will like it even it is not complete yet.

It's easy to write a benchmark module

siege(__dirname + '/app.js')
  .concurrent(100)
  .for(1).times
  .get('/')
  .post('/')
  .attack()


It will automatic listen on a temp unix socket, if you not specify what
port to listen.

The output is colorful and real time.

Because of node.js keep alive the connections I think, I can get correct
benchmark on Mac OSX now, we know that `ab` get wrong result on Mac,
because of OS connections limitations I guess. What ever, I can do
benchmark on Mac now.

In the plan, it will also has the feature to enable cookie, and condtional
GET. maybe not now, maybe you can send a pull request.

siege(__dirname + '/app.js')
  .concurrent(100)
  .for(1).times
  .withCookie()
  .with304()
  .post('/login', {username:'username', password: 'password}).for(1).times
  .get('/')
  .attack()


2012/4/27 mscdex 

> On Apr 26, 12:58 pm, Matt  wrote:
> > Why did you remove the post though? Someone else might make the same
> > mistake - it's always good to keep history of how you fixed it.
>
> +1
>
> --
> Job Board: http://jobs.nodejs.org/
> Posting guidelines:
> https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
> You received this message because you are subscribed to the Google
> Groups "nodejs" group.
> To post to this group, send email to nodejs@googlegroups.com
> To unsubscribe from this group, send email to
> nodejs+unsubscr...@googlegroups.com
> For more options, visit this group at
> http://groups.google.com/group/nodejs?hl=en?hl=en
>



-- 
Best regards,

Jason Green
桂林

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

[nodejs] Geddy Mailing list

[Matthew Hazlett]

Just an FYI.

Geddy has a mailing list now:
https://groups.google.com/forum/?fromgroups#!forum/geddyjs


--
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

[nodejs] Re: [Express-js] New local user group - New Delhi, Gurgaon, Noida, NCR

[Srirangan]

On Fri, Apr 27, 2012 at 2:25 AM, leeba phela  wrote:

> i'am a hardcore real time javascript based app developer and i like
> the idea too, sounds cool (:
>
>

Very good! Join in!

Srirangan  |  +91 9711 477 595  |  About 
GitHub 
LinkedIn
  Twitter   |  Review19
  "Make
better decisions"

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

[nodejs] Re: [ANN] node-webkit: a NodeJS module for WebKit

[Roger WANG]

> This project is still in early stage. Currently only Linux is
> supported. We'll keep working on it, make it better and bring it to more
> platforms. It will be developed in an open way and we have a mailing
> list: node-web...@googlegroups.com . Welcome to join.

We just finished the initial working version for Windows and OSX. And
it's now integrated with CEF (Chromium Embedded Framework). If you are
interested, please see the announcement here:

https://groups.google.com/d/msg/node-webkit/vxlrEtkLpBk/YpA8K9g27kEJ

(Prebuilt binaries are not available yet but the source code is
published.)

Regards,

Roger

Roger WANG  writes:

> node-webkit brings WebKit to NodeJS. With it, client side applications
> can be written with a HTML/CSS UI on NodeJS platform. We believe the
> async I/O framework and Javascript language is a perfect combination for
> client (mobile) side applications.
>
> With this module, your Javascript code in HTML can call NodeJS functions
> directly:
>
> 
> testfs
> 
> var fs = require('fs');
> 
> function test_fs() {
>   var output = document.getElementById ('output');
>   output.innerHTML = '';
>   fs.readdir(".", function (err, files) {
> var result = '';
> files.forEach(function (filename) { result += filename + '
'; }
>   );
> output.innerHTML = result;
>   });
> }
> 
> 
> 
> 
> 
>
> Project URL:
>   https://github.com/rogerwang/node-webkit
>
> Inside it, the webkit's mainloop is integrated into Node's libev. And
> the contexts in Node's V8 and WebKit's V8 are bridged with references so
> they can access each other.
>
> This project is still in early stage. Currently only Linux is
> supported. We'll keep working on it, make it better and bring it to more
> platforms. It will be developed in an open way and we have a mailing
> list: node-web...@googlegroups.com . Welcome to join.
>
> Thanks

-- 
Roger WANG Intel Open Source Technology Center

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

RE: [nodejs] Multiple Web Roles on Azure

[Glenn Block]

Hi Rafel

I work on the node/azure team at Microsoft. Whether or not to create
many roles depends on what you are doing. The main benefit of WebRole
is IIS manages your node processes for you. It will spin up multiple
instances. It will monitor memory pressure and kill frozen instances
and restart them. We also provide access to logs over an http endpoint.
Further you can enable node-inspector for server side debugging within
Azure via our IIS integration.

The big benefit for worker is that it is leaner and it supports
WebSockets. In worker IIS is not present. This means there is less in
the box from a management perspective, but you get closer to the metal.
Because they are lighter they are more ideal for background processes
or for serving web apis.

One place where you might do a hybrid is for Websockets. For example
your front end might run in a WebRole with multiple instances while
your backend exposes web apis / uses WebSockets via a Worker.

Another scenario where you might use worker is for hosting an app
server like Redis / Mongo which your web role uses.

That being said, many folks just use one role and you do not have to
use multiple.

Hope this helps.


Sent from my Windows Phone
From: Rafel
Sent: 4/26/2012 8:51 AM
To: nodejs
Subject: [nodejs] Multiple Web Roles on Azure
Hello All,
I am a newbie to node. right now i am trying to create a web app using
node and deploying it on Azure. My idea is to create a website which
has simple tabs on menu bar and each tab open a new functionality e.g.
one tab gives me grid view, other give me email functionality, other
is for chat etc.
So my question is do i need to create multiple web roles in this case,
or combination of web roles and worker roles. How it goes?? give me an
idea.

I appreciate any help.

Thanks.

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines:
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs] Re: Your thoughts wanted on a node AI project

[Dennis Kane]

On Thursday, April 26, 2012 10:57:26 AM UTC-7, Mark Hahn wrote:
>
> I guess the question everyone is thinking (or asking) is why are you here? 
>  I can understand the recruiting, but I can't think of any other reason.
>
> Which means: if I don't want to give away the entirety of my labors that 
I've been agonizing over for all these years, then I have no legitimate 
reason to post to this forum.  I already said that I will talk about it. 
 But if no one is interested in the coolness factor of the thing, then I 
guess there's nothing else for me to talk about.  I can just move on to 
other places to find people who want to help me get something pretty 
awesome started.  No worries.  I'm just looking for friends who would like 
to work together.  I can't force anyone to get interested.

Oh yeah, there's always this reason:  this thread has top ranking on google 
search "node ai", and my site is now listed on the engines.  Before I 
posted here, my site wasn't even on google!

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

[nodejs] Minified Code

[Matthew Hazlett]

Has anyone experimented with minifing code with node?

I know the reason to minify on the client side is its faster load time 
to the browser, was just wondering if there were any performance gains 
with minimized code on the server side or perhaps a bottleneck with the 
eval() of the code.


--
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs] Re: Your thoughts wanted on a node AI project

[Mark Hahn]

>  Which means: if I don't want to give away the entirety of my labors that
I've been agonizing over for all these years, then I have no legitimate
reason to post to this forum.

No, that is not what I meant.  I literally meant "why are you here?".

I had only thought of the recruiting angle but now you have pointed out the
public relations angle.  So you have two reasons.  In my humble opinion,
those reasons are not appropriate for a public discussion group like this.
 There is supposed to be a give and take, not a take and take.

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs] Minified Code

[Mark Hahn]

I have always assumed minifying only served the single purpose of speeding
up the transfer over the net.  I seriously doubt compile time is improved
noticeably.  C++ rips through source quickly.

On Thu, Apr 26, 2012 at 9:22 PM, Matthew Hazlett  wrote:

> Has anyone experimented with minifing code with node?
>
> I know the reason to minify on the client side is its faster load time to
> the browser, was just wondering if there were any performance gains with
> minimized code on the server side or perhaps a bottleneck with the eval()
> of the code.
>
> --
> Job Board: http://jobs.nodejs.org/
> Posting guidelines: https://github.com/joyent/**node/wiki/Mailing-List-**
> Posting-Guidelines
> You received this message because you are subscribed to the Google
> Groups "nodejs" group.
> To post to this group, send email to nodejs@googlegroups.com
> To unsubscribe from this group, send email to
> nodejs+unsubscribe@**googlegroups.com
> For more options, visit this group at
> http://groups.google.com/**group/nodejs?hl=en?hl=en
>

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs][moe] Moescript language

[Belleve Invis]

You can use #0 for the first argument. `{|x| x * 3}` could be `{ #0 * 3 }`, 
but it is slower than `{|x| x * 3}` due to `arguments` in moe is a REAL 
array.

在 2012年4月26日星期四UTC+8下午10时17分41秒，JeanHuguesRobert写道：
>
> I like the concise syntax for lambda, including f{stuff}. Maybe 
> f(a,b){stuff} could behave like Ruby's?
>
> Also: {|x| x * 3} could be { it * 3 } where "it" stands for the single 
> parameter of the lambda.
>
> Le jeudi 26 avril 2012 14:26:20 UTC+2, Belleve Invis a écrit :
>>
>> I've made it for years. Mostly inspired by Python and Haskell.
>>
>> Doc:
>> http://be5invis.github.com/moescript/
>>
>

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Re: [nodejs][moe] Moescript language

[Belleve Invis]

Only code blocks using "wait", "!" and "<-" (called Monadic Primitive) is 
compiled into such a complicated and "slow" form.

在 2012年4月27日星期五UTC+8上午1时22分03秒，Jann Horn写道：
>
> On Thu, Apr 26, 2012 at 05:34:33AM -0700, Belleve Invis wrote: 
> > Functions using *wait*, *!* and *<-* are specially compiled. 
> > 
> >1. function F1$_(){ 
> >2. var randPrintNums$; 
> >3. randPrintNums$ = async$({build:function(SCHEMATA$_){return 
> >function(n$){ 
> >4. var T1$_, T8$_, T9$_; 
> >5. var i$, tasks$; 
> >6. function block_T2(T2$_){ 
> >7. tasks$ = []; 
> >8. for(((i$ = 0), (T1$_ = n$)); i$ < T1$_; (i$ = i$ + 1)){ 
> >9. (function F3$_(){ 
> >10. var index$; 
> >11. index$ = i$; 
> >12. return tasks$.push(async$({build:function(SCHEMATA$_){return 
> >function(){ 
> >13. var T4$_, T5$_; 
> >14. function block_T3(T3$_){ 
> >15. T4$_ = (sleep$); 
> >16. T5$_ = ((100 * Math$.random())); 
> >17. return SCHEMATA$_.yield(T4$_(T5$_,block_T6)); 
> >18. }; 
> >19. function block_T6(T6$_){ 
> >20. T6$_; 
> >21. trace$(index$); 
> >22. return SCHEMATA$_["return"](); 
> >23. }; 
> >24. return block_T3; 
> >25. }}})); 
> >26. }()); 
> >27. }; 
> >28. T8$_ = (join$); 
> >29. T9$_ = (tasks$); 
> >30. return SCHEMATA$_.yield(T8$_(T9$_,block_Ta)); 
> >31. }; 
> >32. function block_Ta(Ta$_){ 
> >33. Ta$_; 
> >34. return SCHEMATA$_["return"](); 
> >35. }; 
> >36. return block_T2; 
> >37. }}}); 
> >38. randPrintNums$(100); 
> >39. } 
> > 
> > 
> > 在 2012年4月26日星期四UTC+8下午8时32分50秒，Oliver Leics写道： 
> > > 
> > > On Thu, Apr 26, 2012 at 2:27 PM, Belleve Invis  
> > > wrote: 
> > > > Code Example 
> > > > 
> > > > def async randPrintNums(n): 
> > > > def tasks = [] 
> > > > for(var i in 0..n) 
> > > > tasks.push async :> 
> > > > wait sleep (100 * Math.random()) 
> > > > trace index 
> > > > where index = i 
> > > > wait join tasks 
> > > > 
> > > > randPrintNums 100 
>
> That looks like it's going to be pretty slow... is that an issue or isn't 
> it as bad as it looks? 
>

-- 
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en