dracoblue.net

I had a very nice talk to some folks at my company about the spludo framework. Thanks for all the positive feedback!!

Today I want to give you guys a small update to spludo, which contains some of the neat features I was previewing on friday:.

Changelog:

* group and chain are truely async now
* added "spludo-gen controller" (wizard to create a new controller)
* views can now be in sub folders
* added partial + partials in .ejs
* .ejs-Views recompile the view as soon as it is changed

There is also a very short topic about

how to use partials with spludo in the user guide.

Download spludo 1.0.1 or install it by using npm.

Today I was polishing up the release for spludo and I removed all @author + @since tags from the javascript doc.

In first place, I thought I might keep them in. But, I was unsure about a good guideline how to keep them up to date.

Let's take for instance @since. What does this exactly mean. It may mean when the function or class was added. But what if the signature changed? Is this a different function or still the same? Are this still the same @authors?

I came to the conclusion that they are hard to maintain and git/subversion give me the info I need in a way more convenient way. So I removed them with this commit.

Since twitter stopped supporting basic auth, some are searching for a lua implementation of oauth.

A while back (in last 2009) I wrote a library for exactly this, but didn't finished it. So I took some time to polish it up and it works still great ;-).

So have fun with the MIT licensed lua oauth library download or source at github.

If you want to to backup from/to an ssh server and this server does not run on 22 as port, you can use this workaround:

rsync -r [strong]-e 'ssh -p 4321'[/strong] [email protected]:/home/user/test backup -v

This will sync from the host

[email protected] (by using ssh and port 4321) the folder /home/user/test to the folder backup on your local harddisk.

Since now I am releasing spludo pretty soon, I wanted to check out how performance is against other web frameworks for node.js.

I made a previous test for this already, running nodejs 0.1.98.

This results are against: nodejs 0.2.0. To keep it fair (and little expressive) I just took the hello world application from each framework and benchmarked those with apache benchmark.

Concurrency Level: 32 (was 256 in the previous test)

ab -n 10000 -c 32 http://spludo.com:8080/

Results:

josi: 3944.78 [#/sec]
expressjs 1.0: 6480.42 [#/sec]
(fab):  8203.74 [#/sec]
[strong]spludo 7454.60 [#/sec][/strong]

Concurrency Level: 1

ab -n 10000 -c 1 http://spludo.com:8080/

Results:

josi: 3162.31 [#/sec] (was 3012.32)
expressjs 1.0: 3497.71 [#/sec] (was 2744.79 in 0.9)
(fab): 4199.95 [#/sec] (was 4271.43)
[strong]spludo: 3923.69 [#/sec][/strong] (was 3866.12)

The source code for all sample applications may be found at the

public gist post.

The Test PC was a i7-920 Quadcore with HT and 8GB DDR3 RAM and a single node.js instance running.