David Cameron has decided the Government needs to do something to protect the innocence of children. How horribly miss-guided he is. There are two main problems that I can see. The first is an ideological problem, the second a logistical one.

##Why? The main issue to be dealt with is child pornography. This is less of a filtering issue. In terms of online consumption it’s mostly adults who can jut turn any filter off. The current strategy of taking any servers found containing CP offline is an effective strategy that we should continue and expand upon.

Then we get legal porn. Filmed with consenting adults acting out various fantasies or fetishes. This is now to be censored by default.

##How? Quite, how?

Here I shall use the terms URI and URL interchangeably to mean the same thing. I appreciate there are subtle differences.

The IndieWeb is a fantastic idea. The web itself is inherently open. No one owns it, no one directly controls it. However, if you aren’t careful what services you use on the web then it can effectively end up that way. We all use the web in a primarily social way these days, social networks if you will. The big three players on the social web are Facebook, Twitter, and Google with their Google+ service. They want you to spend as much of your time as possible on their services in order to maximise their advertising revenues.

This doesn’t play nicely with the inherently open and interoperable foundations of the web. Foundations without which these big players wouldn’t exist.

And thus the IndieWeb is born. A desire to own your social identity and share however little or much of that social presence with these big players you want. Which I think is absolutely right. Just have a look at what people are saying about Medium.

But what of URL shortening services?

Some people seem to think that having your very own short URL helps in this cause. Which I suppose it does, but I think only to a small degree. Why do we even need to shorten web addresses? The only situation I think it would be necessary is posting/syndicating to Twitter. Any other service has ample character space to post the full URL, or is sensible and uses annotations. On Twitter however, any link, regardless of how short it is, gets wrapped up in their t.co service.

I therefore don’t currently see any compelling reason to run your own URL shortening service other than simply because you can.

*[URL]: Universal Resource Link *[URI]: Universal Resouce Identifier

Well, if ever I have an excuse to watch The Sopranos...

This is slightly re-inventing the wheel, but I have released a new package called Dumbquotes. The idea is to replace simple typographic techniques with their more correct forms. Such as replacing a ' with ‘ or ’. This also gave me the chance to try and write a package. So dealing with making sure it’s psr-0 compliant and has associated unit tests to run with phpunit.

The package will deal with apostrophes, quotes, dashes, and ellipses. There are certain issues. Ultimately this is designed to deal with plain text such as a markdown document. It does not work with HTML. Trying to parse HTML with regex will bring the return of Cthulu. However once you deal with HTML directly things get a little complicated.

Consider the following sentence that could appear in some HTML <p>Mary said \"How <em>did</em> she do that?\"</p>. We want to turn this into <p>Mary said “How <em>did</em> she do that?”</p>. This is complicated by the fact we can't just search for a string of text containing two double quotes like so, /\"(.*?)\"/. The sentence doesn't actually appear in the HTML DOM. We actually have three blocks of text

• Mary said \"How
• did
• she do that?\"

To concatenate that into a single string, and then put the tags back in the right place seems a very difficult task. So I have decided to write the dumbquotes parser to be applied before the markdown transform is applied.

*[HTML]: HyperText Markup Language *[DOM]: Document Object Model