Log Paranoia

I have been watching the logs of a new blog I put up today about cheap Brighton. It is strange watching traffic develop from zero. I used some of my standard tricks which saw me get indexed in Google in about 5 hours which is about average. What I wasn’t expecting was the very sudden flood of bots and other suspicious activity.

The activity that worries me the most is that hundreds of completely different IP addresses are all asking for a single file. I would expect this kind of nonsense of course but not within a few hours of the domain registration!

It pays to watch your logs, if you don’t this is where you end up with hacking attempts getting through with brute force, of course you can’t watch them 24 hours a day.

#barcampbrighton3

So what to say about Barcamp Brighton….

I have never been to a barcamp or in fact any similar sort of (un)conference before. I was a bit nervous as the only price was that you had to put a talk together.

Within a couple of sessions I had got the concept and by early afternoon I was totally enthused having been to sessions ranging from accessibility, Flex, geocoding and user research. It wasn’t so much the sessions as the people that seemed to make it for me, there was an awful lot of enthusiasm expressed in many different ways.

My favourite session of the day was creative thinking for logial people which was a great deal of fun which started by showing how the opposite of black could be yellow (clue: context). Perhaps the funniest thing was we had a room full of techies and none of us could work out how to turn the lights on!

One of the sessions was about WYSIWYG DHTML based editors and the need for a better one. By the end, I had agreed to sign up to an open source development project, my first! If you need one and find things like TinyMCE don’t meet your needs then why not sign up to this http://code.google.com/p/editable-framework-js/ and join in:)

At 1am sozzled on free Jamesons, I was still scribbling on bits of paper and talking about data structres. I learned that the structure of my CMS was actually interesting to quite a few people, not something I would normally talk about in a bar!

Sunday had a fair amount of hungover people including me; I was knackered, I did notice there were far less smiles and far more yawns. Managed to get to the first session that I wanted to see which was the second of the day having missed the first and the breakfast.

The highlight was possibly the discussion organised (rather than lead) by a MySpace person about social networks, we talked long into lunch and covered a range of fascinating topics.

My own discussion could have been better. I now know something about when to schedule your session, I got that wrong which meant there were not enough people in the room to have a proper discussion, I needed a wider range of experiences to make it work. I would have done better reverting to my original talk on DIY content management systems.

By the end of the day I was exhausted, glad to go home and have a snooze. I would recommend a Barcamp to any developer, designer or new media person, you don’t have to be a techie to have benefited.

I now have a great long list of things to do, better get on!

Will say more tomorrow ....

… but Barcamp is a great format for finding out things you never knew you needed to know. Met some brilliant people and talked about some really interesting stuff :)

Can’t fault the free food or booze either :D

http://barcampbrighton.org/

Time I got rid of the design on Tumblr!

It was only meant to be a stopgap for a week or two and is nearly a year old :)

Choices Choices Choices!

Yesterday I installed the brand new shiny Google Chrome browser and was impressed. I liked the comic, especially the bit about how processes were handled, although it’s worth noting that MS have outlined a similar scheme for IE8, Im guessing Microsoft will be annoyed.

So far, I have seen a few rendering bugs, one on http://news.bbc.co.uk (at the top) and one Tumblr using it’s editor, although I have seen no similar issues with a similar editor that I use for my SqooDriver CMS.

The other astounding thing I found last night was this, which is really amusing:
That’s right, Chrome’s spellcheck doesn’t know the word Google!

The other interesting bit of browser news I have seen is this Vimeo video about Ubiquity from Mozilla, a text based system that allows you to do things naturally without having to go to loads of sites and cut and pasting code (I could have done with it for this post for example). Stunning if it works, it may even be a Chrome killer for me:

Ubiquity for Firefox from Aza Raskin on Vimeo.

Twitter and SMS

Strange one this, Twitter doent seem to have a business model. It shown no adverts and most users don’t even visit the site but use third party applications that take advantage of its API instead. So it seems rather strange that Twitter should ignore a possible revinue stream and dump a service they were providing for free that users would love to pay for.

Last Thursday they announced that outside of the US, Canada and India, users could no longer receive updates via their mobiles, leaving a large number of users up in arms and considering jumping ship for similar services that still give SMS updates.

Thankfully a number of web developers stepped in almost immediately and started developing alternatives including my friend Russ who put together HootSMS. HootSMS appears to be the only working site up at the moment although it looks like two or three others will be up fairly soon, we will see if UK users leave Twitter or will be pursuaded to stay, they are the largest group using SMS and the rug has been pulled from under them. Hopefully HootSMS and others will put it back!

47 percent of IE users are up to date

This is more than I thought. According to the Swiss Institute of Technology 47% of Internet Exporer users have out of date browsers making them a security risk, this compares with 83% of Firefox users and 65% of Safari users.

If you havent yet then please update your browser and while we are at it, if you are using Internet Explorer, move to Firefox or Flock, they are just far better!

Report from BBC.

Clearing floats the very easy way!

Why has this simple technique taken so long to emerge? The float clearing problem has been around as long as CSS based layout. It has been a big problem!

The W3C always recommended the clearing element fudge that invalidated their aim of producing fully semantic code, hypocrisy.

A method I helped developed with a former colegue was even the fashion for a few months, making onto positioniseverything though not for long as it involved a fudge where margins were shunted up and then down by 32767ems, a bit nasty. The content-after method soon replaced it as it was far better.

The content-after method has been the main way things have been done for the last couple of years or so and seemed to be the simplest but again it was a little wrong, placing a CSS generated full stop within the DOM, if not the markup.

This is simple, put overflow: auto round the div that is to wrap the content, obvious and simple but no one has realised it. Developers like to think they are clever so how was there such a collective fail!

By the way it works great, thanks Paul :)

Every web app developer should read this book!

Getting real by 37 signals, read it online for free. I have been building web apps for 9 years and never have I found so much good advice in one place.

Each chapter lays out a part of a philosohy that will get you developing quickly and efficiently. Great advice includes:

• Designing from the functional centre of a page outwards rather than building a framework and squeezing the functional inside
  
• Getting the smallest functional implementation possible out there and live rather than making an app that does everthing before launch
• Not setting long term deadlines

Brilliant, there’s a reason these are they guys who put ruby on rails.

SqooDriver goes Russian

I have just spent a while altering my content management system editor to cope with multiple languages including Russian, Chinese and Arabic. I have a project that requires four languages and it seems like a good time to go double byte.

Having done a similar job in the past, I knew there would be some difficult bits converting the system to use UTF-8, single byte to double byte conversions always have some nasty surprises. Last time it was for a PERL with SQL Server system using the MSHTML ‘editor’, this time its PHP with MySQL using the TinyMCE editor.

Getting the Russian code to roundtrip starting from a TinyMCE input box took a while, I had to trace the data path trough the system in a fair amount of detail, I was almost tripped up by an htmlentities call.

What made this task so much easier than last time was the fact I was using PHP and MySQL both of which have a great deal of community support, two pages helped me no end, PHP UTF-8 Cheatsheet and MySQL and UTF-8 thanks for shortening a chore no end!