Re: [nodejs] Re: What do we call ourselves? Nodesters?

So what is a group of nodesters called? A callback of nodesters? On Tue, Mar 4, 2014 at 11:07 AM, Tim Caswell wrote: > > > > On Tue, Mar 4, 2014 at 10:51 AM, Rick Waldron wrote: > >> Strange how Node is categorized along with programming languages. >> > > https://medium.com/cool-code-pal/cf72b

Re: [nodejs] The Next Phase of Node.js

Congrats TJ Fontaine! You've done amazing work for Node, and I have complete confidence you're going to move node in the right direction. On Wed, Jan 15, 2014 at 9:16 AM, Alex Kocharin wrote: > Hi Isaac, > > What changes are you planning on npm registry? What services will it > provide for m

Re: [nodejs] Re: Test suite recommendations

You can get pretty far with the plain assert module. If you do however, I recommend the following: 1. Keep each test file small; it's okay to have hundreds of files. 2. Throw on error, this produces a lovely stack trace. It makes tracking down test failures easy. 3. Exit non-zero on failure; throw

Re: [nodejs] Re: Running CLR/.NET code in-process with node.js

n the lifetime of the original HTTP request? > It certainly saves a lot of orchestration to have another HTTP request > retrieve the results at a later time via short or long polling. > > On Feb 27, 1:51 pm, Jacob Groundwater wrote: > > > 2. Running CPU-bound workloads

Re: [nodejs] Running CLR/.NET code in-process with node.js

> 2. Running CPU-bound workloads for node.js applications in-process > (without blocking the node.js event loop): > http://tomasz.janczuk.org/2013/02/cpu-bound-workers-for-nodejs.html Even in an application that supported threading, I would be unlikely to run CPU-bound operations in the same proc

Re: [nodejs] Anyone using Vagrant for, say, testing with multiple browsers or as a type of CI server?

We use Vagrant sometimes, but it can take a while to find the workflow that works best for you. I also recommend checking out veewee which is great for building new vagrant boxes. On a final note, the mounted partition does not support symlinks natively. If you need symlinks at all, you'll need t

Re: [nodejs] Node.js VS Scala VS Clojure: what for an API with a we-love-java-we-dont-even-know-why business component in the company?

Unless you are using the servlet 3 spec, Java is going to consume one thread per request. This is a huge limitation for many applications. Both scala's playframework and node.js can respond asynchronously. If you want to *sell* that to your team, mention that. If you really are designing a central

Re: [nodejs] Re: Best way to reload?

Well said Adam. I agree completely on this type of design. I will add that in small applications, people tend to use the database as their synchronization point. As your application evolves into a service-oriented-architecture you'll want to synchronize around your IPC network instead. Try to let

Re: [nodejs] Lion: Installed node/npm using .pkg: How to convert to a brew install?

This is the "technically correct but useless" answer, but you can inspect where/what gets installed from a .pkg by looking through /var/db/receipts lsbom /var/db/receipts/org.nodejs.node.npm.pkg.bom lsbom /var/db/receipts/org.nodejs.pkg.bom The root location of the above packages is listed in /v

Re: [nodejs] Re: Using NodeJs with through $.get method of jQuery

I had a project that extensively used cross-domain requests in the past, and I will say I think the whole thing is a mess. The solution I used, which worked flawlessly, came from a little gem known as window.postMessage The window.postMessage method is used to communicate between iframes loaded f

Re: [nodejs] Re: What happens to hook.io?

Hey Mike. If you're interested, I published *Federation* ( http://underflow.ca/federation) which is intended for federated event emitting. I haven't used hook.io so I can't really compare the two, but I can outline a few differences off the top of my head. Federation does not involve itself in r

Re: [nodejs] NodeJS analysis for suitability

> This seems unfounded. Why don't extra processes help? He makes the claim > it's no better but doesn't explain why. I've used multiple processes in > the past and it worked great. > This may depend on how node-core interfaces with the socket library. If a C++ thread is calling accept() on inco

Re: [nodejs] Rails-like Node module for migrating a database to a schema?

I use db-migrate and am fairly happy with it. I like that you can also just embed SQL directly, which is useful for cases that are not fully addressed by the API. Just make sure you remember to pass the callback parameter around or the migration just hangs. On Thu, Feb 14, 2013 at 7:42 AM, Andr

Re: [nodejs] Elegant way to require('some-app-component')

There is definitely a method to my madness. First, I would only recommend doing this in a deployable application, not a module that deploys to NPM. The best NPM packages are small and focused; having lots of internal dependencies is probably a smell. Second, I do this to help filter out code that

Re: [nodejs] ZeroMQ Libraries

Just to clarify Michael, I would definitely appreciate your effort. I was just sharing my experience with the last package. On Sat, Feb 9, 2013 at 4:10 AM, Floby wrote: > go for it! > Also more facilities to debug zero mq related issues could be a plus > > -- > -- > Job Board: http://jobs.node

Re: [nodejs] Elegant way to require('some-app-component')

> > mkdir node_modules > mv path/to/component.js node_modules/component.js > node -pe "require('component') // voila" > I actually use this method, but if you're on POSIX system, you can symlink it. It lets you develop as a single application, but encourages you to package-ize your co

Re: [nodejs] ZeroMQ Libraries

h-work > >> applications, especially on the "worker" end. > >> > >> -Schoon > >> > >> On Fri, Feb 8, 2013 at 10:04 AM, Ruben Tan wrote: > >> > Yes. Please dont impose js idioms upon zmq. If anything, a simple > >> > bi

Re: [nodejs] ZeroMQ Libraries

If you can, the request/reply socket is broken for any asynchronous events, as ZMQ expects replies to occur in the order they were asked. -- -- Job Board: http://jobs.nodejs.org/ Posting guidelines: https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines You received this message be

Re: [nodejs] Federation: A Distributed Message Network Module

Thanks for the feedback Daniel. I will try to explain how I arrived at the current design. The one change I might suggest is not to have message delivery not be a > method on the actor. That is instead of writing "tom.tell('bob', message);" > it would either 1) use the director "director.tell('bob

Re: [nodejs] Re: Federation: A Distributed Message Network Module

layout yet, so a stable message schema would be required before inter-operating with another implementation. This wouldn't be hard, but is worth a bit of discussion. If you're interested in interoperating with Python, feel free to open an issue and we can work on the details there.

[nodejs] Federation: A Distributed Message Network Module

er - multi-process via axon - multi-process via http This project is in its initial stages, and I would love to find some early adopters. I am happy to answer any questions, thanks! - Jacob Groundwater -- -- Job Board: http://jobs.nodejs.org/ Posting guidelines: https://github.com/joyent/node/wi

Re: [nodejs] David - a tool to help you manage your node project dependencies - http://david-dm.org

> >> So it is basically an online version of police ( >> https://github.com/pksunkara/**npm-police<https://github.com/pksunkara/npm-police>). >> Good idea, I would love to see it for private github and bitbucket repos >> though. >> >> >> On Thu, Jan 31, 2013

Re: [nodejs] Re: Javascript Class Specification for the next generation of JS class system, hope to discuss with you.

The following is just my opinion, but it has held up well throughout the years. Use encapsulation, not inheritance. Just because your object needs to send events does not make it an event emitter. At most it means you should add an `emitter` property. Is it really that hard to type `myobject.emitt

Re: [nodejs] David - a tool to help you manage your node project dependencies - http://david-dm.org

Very nice Alan. Personally, I think if people make a project that is useful to the Node.js community, you absolutely should post it to this mailing list. The line between promotion and spam comes from recognizing that you're sharing the floor with many people. If there is a major update between t

Re: [nodejs] Re: Managing physical distributed clusters

> > @Jacob: i guess i wasn't exact in my explanation. what i want in fact is > a notification-based mechanism. like fleet or amino deploy/drone. fleet > uses git under the hood. amino seems to use npm to package and request to > deployment. > > @greelgorke: right, I didn't have much to add outside

Re: [nodejs] Managing physical distributed clusters

> - deployment from a git repo, configurable branch (fleet does it, but > works only from master) > As long as you're doing a `git push` to deploy, you can push any branch to the remote master. For example, if you wanted to push your local `dev` branch use: git push +dev:master The `+` tells th

Re: [nodejs] a pakage that exports few modules

> index.js predates packages by quite a long time. It was the first > "load a packagey folder thing as if it was a single module" feature, > dating back to the 0.1.x days. It's not terrible. But personally I > prefer a main script in package.json, since you can name it something > more appropria

Re: [nodejs] a pakage that exports few modules

> > Other costs, like how require("foo/bar") can be either foo/bar, or foo/bar.js, or foo/bar/index.js, are simply > historical mistakes that seemed like a good idea at the time, but > provide no value, and are now too costly to remove. Is `index.js` considered a non-preferred method of doing pa

Re: [nodejs] [Idea]: npm as a dependancy injector.

First, it's awesome that we're talking about dependency injection. Second, I believe *good* dependency injection requires inversion of control . Your modules should not directly require dependencies if possible, rather design them to accept their

Re: [nodejs] Aplication restarter

. Auto restarts are great while developing but I do not see the need in production. If you have an automated deployment, which again I consider a good practice, you can just restart the app post-update. - Jacob Groundwater On 2013-01-12, at 2:48 PM, José F. Romaniello wrote: > I'd ask

Re: [nodejs] Re: Checking node_modules into git

I have been warming up to the way Netflix does their deployments. Each update is a full image, packaged as an Amazon Machine Image, ready to be inserted into an auto-scaling group. This technique can be abstracted to any VM-style cluster however. Since each images is pre-made, and ready to run on

Re: [nodejs] Re: I need your help desperately for increase my streaming speed

This is probably a good time to mention 'Rude' a tool i've been working on for handling large project assets. https://github.com/jacobgroundwater/rude Basically Rude lets you version control assets in git without storing the actual asset in git. When running locally, assets are served from a loca

Re: [nodejs] Re: comments in package.json

Since you're fixing it, can it run after dev-dependencies are installed? =] On Mon, Jan 7, 2013 at 3:12 PM, Isaac Schlueter wrote: > Jacob, > > Indeed, that appears to be a bug. Fixing it now :) > > On Mon, Jan 7, 2013 at 3:02 PM, Jacob Groundwater > wrote: > &g

Re: [nodejs] Re: comments in package.json

> That's why prepublish scripts get run on `npm install` :) > This does not appear to be the case for me; it would however be great if it did. -- Job Board: http://jobs.nodejs.org/ Posting guidelines: https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines You received this message

Re: [nodejs] Re: comments in package.json

> Transpiling and minification are better handled by a prepublish script > anyway: https://npmjs.org/doc/scripts.html That way, the module doesn't > need to depend on CoffeeScript just to install and deployment will happen > that much faster. Unless I am mistaken, `prepublish` is only called when

Re: [nodejs] Re: comments in package.json

> Not much of a "discussion" per se, but more like something that most > eventually come to realize, but only *after* doing it the wrong way > first. > > It'd be better to only support node-gyp for building packages, and get > rid of scripts.install altogether. Thanks. This makes sense for compil

Re: [nodejs] Re: comments in package.json

> As long as I have a say in the matter, package.json file will always > be json. If anything, future versions of npm and node will *remove* > features rather than add them. (I'm looking at you, `scripts.install` > ;) Now I am curious about the discussion around removing scripts.install. -- J

Re: [nodejs] node.js can't receive ajax post data which made by ie9 or 8

>From what I remember, that looks right. You could try adding console.log to print each request as it comes in. I eventually abandoned X-Domain requests in favor of window.postMessage, which I think is a better solution. The postMessage function works across domains between iframes. You load a hid

[nodejs] Synchronous Exec Module

Hello everyone, I created a quick synchronous exec module for those of us who like to do command line scripting in Node. I knew of the execSync module, but was not a fan of the ffi module dependency. My brief searches revealed no other potentials, hence AllSync was born. AllSync has no other modu

Re: [nodejs] node.js can't receive ajax post data which made by ie9 or 8

When doing a cross domain POST, the browser usually send two HTTP requests. The first with HTTP method OPTIONS before sending the actual request. Your server has to flag the request as okay before the browser proceeds. Your server might be receiving the options request only, which would have no bo

Re: [nodejs] Inconsistent method names?

V ccchpg - Jacob Groundwater On 2012-12-28, at 2:16 PM, greelgorke wrote: > well -- 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 &quo

Re: [nodejs] starting http server after certain service is up

I believe that services should be able to start in any order, further more your processes should tolerate failure of their backing services by trying to reconnect. If a client attempts to connect while your backing services are unavailable, you fail fast, even if this means giving the client an "oo

Re: [nodejs] ANN Node/V8 native bindings guide

This is great! On Thu, Nov 29, 2012 at 6:39 PM, Luis Reis wrote: > Hi all, > > After having lots of trouble looking for scattered documentation on how to > fetch arguments/return values/create objects on native bindings for > node.js, I decided to create a repo / guide with all my notes. > > So,

Re: [nodejs] Re: How to keep node running without using Forever?

Update: I just fixed an error in node-foreman 0.0.13 where it exports the wrong upstart template. It's been fixed in 0.0.14. - Jacob Groundwater On Tue, Nov 27, 2012 at 9:34 PM, Eysenck Gómez wrote: > Upstart (to run as daemon) and monit (to monitoring the daemon), are the > best so

Re: [nodejs] Re: How to keep node running without using Forever?

Actually Josh, if you have no Procfile define, but do have a package.json file, node-foreman will attempt to call 'npm start' There are a lot of node-specific updates I have made. For example, you can throw a round-robin proxy server in front of your app to test scaling issues. nf start web=4 -x

Re: [nodejs] How to keep node running without using Forever?

If you use a Procfile, I have a Node.js implementation Foreman that can export upstart scripts. https://npmjs.org/package/foreman The upstart scripts have no dependency on foreman either. On Mon, Nov 26, 2012 at 11:00 PM, Marak Squires wrote: > I can tell you that Tata is using https://github.c

Re: [nodejs] npm install private modules?

You can use 'npm link' to cross-link private modules. e.g. (cd my_db_module && npm link) (cd my_app && npm link my_db_module) On Sun, Nov 25, 2012 at 9:44 AM, ravi wrote: > Hello all, > I'd like to have node use private modules as dependencies. These will be > in the same directory structure

Re: [nodejs] forward proxy

I implemented a forward proxy in the node-foreman module, and I ran into a problem with what req.url was returning. Sometimes it would be a path, other times a full url. See here (https://github.com/NodeFly/node-foreman/blob/master/forward.js) I used the proxy to test my Github OAuth login. The p

Re: [nodejs] [ANN] node-avro 0.2: Avro support for Node.js

Since we're on this topic, what is the general state of Avro/Thrift/Protobufs. Are many people using them in production? I know Thrift is supposed to have native support, but the generated code is much less complete than the Java version. Protobufs seem to have a dynamic library that more or less p

Re: [nodejs] Re: Server-to-Server Communication

e: > > Use https://github.com/learnboost/engine.io > > > > On Sat, Oct 27, 2012 at 2:31 PM, Jacob Groundwater > <mailto:ja...@nodefly.com>> wrote: > > > > It looks like ZeroMQ does not play well with the asynchronous nature > > of Node. Take a loo

Re: [nodejs] Re: Server-to-Server Communication

izing against the database would involve polling to achieve >> real-time like events. I know Redis supports a pub/sub system, but does not >> seem to have any RPC-like mechanisms. >> >> In the end, I will make the decision best suited to our needs, but I am >> sure I ca

[nodejs] Server-to-Server Communication

a pub/sub system, but does not seem to have any RPC-like mechanisms. In the end, I will make the decision best suited to our needs, but I am sure I can gain from some discussion on the matter. Thanks everyone, - Jacob Groundwater -- Job Board: http://jobs.nodejs.org/ Posting guidelines: ht

Re: [nodejs] Re: npm: version locking in modules

I think it comes down to a tradeoff. A lot of packages are not written as well or as thorough as they could be. A lot of packages I look at on NPM are pre 0.1 even. Thus it comes down to whether you would rather have all packages using their latest version, or whether you want a tacit "guarantee"

Re: [nodejs] Re: Forever with ssh

like to help field test it, it would be much appreciated. It can export upstart jobs that are not foreman dependent, so that should lower the risk of using it. Thanks! - Jacob Groundwater On Fri, Oct 19, 2012 at 6:09 PM, William Myers wrote: > I appreciate everyone's great insight and I

Re: [nodejs] Re: node dev on windows

You can also run the open sourced cloud 9 IDE from a VM and use a browser in windows to access it. This has the added benefit of being able to distribute the environment amongst windows boxes very easily. - Jacob Groundwater On 2012-10-19, at 7:37 AM, jerome wrote: > @mscdex: Yeah Wind

[nodejs] Re: Forever with ssh

+1 for upstart - I think using the platform native tools is good form It can also be complimented with [start-stop-daemon](http://man.he.net/man8/start-stop-daemon) when required. On Thursday, October 18, 2012 6:39:24 AM UTC-7, Clay Simmons wrote: > > I like using upstart on Ubuntu to keep my no