dylanroy

<requestFiltering>
                <requestLimits maxAllowedContentLength="[bytes]" />
</requestFiltering>

dylanroy

<httpRuntime useFullyQualifiedRedirectUrl="true" maxRequestLength="[kilobytes]" requestLengthDiskThreshold="8192" executionTimeout="1200" />

dylanroy

<property name="DocumentManager.SearchPatterns">[File Extensions]</property>
<property name="DocumentManager.MaxUploadFileSize">[bytes]</property>

dylanroy

One of my favorite past times is to look at the notebooks of famous scientists. Da Vinci's notebook is well known, but there plenty others. Worshipping Da Vinci like no other, I bought a Think/Create/Record journal, used it mostly to keep jot down random thoughts and take notes. This was great in the beginning, but the conformity of lines drove me nuts. Only moleskines made blank notebooks, so I had to buy one.

At the same time I started a freelance project. The project itself is irrelevant, but suffice to say it was very complex and spanned several months. It seemed like a perfect opportunity to use the moleskine. Looking back, all my entries fell under few categories:

• Todo
• Question
• Thought
• Bug
• Feature

dylanroy

The new Secure Password Hashing API in PHP 5.5

The [RFC for a new simple to use password hashing API][rfc] has just been accepted for PHP 5.5. As the RFC itself is rather technical and most of the sample codes are something you should not use, I want to give a very quick overview of the new API:

Why do we need a new API?

Everybody knows that you should be hashing their passwords using bcrypt, but still a surprising number of developers uses insecure md5 or sha1 hashes (just look at the recent password leaks). One of the reasons for this is that the crypt() API is ridiculously hard to use and very prone to programming mistakes.

dylanroy

I have some early benchmark results for our work on a high performance NATS server in Go.

Quick Summary:
We can process ~2M msgs/sec through the system, and the ingress and egress are fairly well balanced.

The basics of the architecture are intelligent buffering and IO calls, fast hashing algorithms and subject distributor/routing, and a zero-allocation hand-written protocol parser.

In addition, I used quite a bit of inlining to avoid function overhead, no use of defer, and little to no object allocation within the fast path. I will share more details and the code at a future date.

dylanroy

/*
 * Load external modules and init variables
 */
var twitter = require('ntwitter'),
    http = require('http'),
    bitly = require('bitly'),
    b = new bitly('USER', 'API_KEY'),
    tweeted = {},
    load_time = Math.round(new Date().getTime() / 1000),
    score_threshold = 100;

dylanroy

interface IVehicle
{
    String Drive();
}

dylanroy

class Roadster : IVehicle
{
    private String _Make = "Tesla Motors, Inc.";

    public String Drive()
    {
        return "Varoom Varoom";
    }
    public String GetMake()
    {

dylanroy

static void Main(string[] args)
{
    //1. Implementing a Roadster with an IVehicle.
    IVehicle aVehicle = new Roadster();
    String startDriving = aVehicle.Drive();
    Console.WriteLine("All vehicles are made to drive... "+ startDriving);
 
    //2. Implementing a Roadster with its concrete class.
    Roadster aTeslaRoadster = new Roadster();
    aTeslaRoadster.Drive();