Wednesday, June 30, 2010

Dear Flickr, what the RSS hell?

I tweeted about my disappointment with Flickr's integration with Facebook. It gave me nothing that Friendfeed didn't already do. I wanted a single upload that would sort things so the nice hi-res pictures appeared on Flickr and an adapted set appeared on Facebook so I could tag friends doing embarrassing things.

I'm even more disappointment with their (lack of) RSS strategy.

One of my Flickr feeds updated to say it wasn't going to update any more. Okay; 301 redirect it then like the rest of the world and we'll all carry on as normal. No. Looks like Flickr expects everyone to manually transfer.


But where should we manually transfer too? The link they provided is this one.

Here's what it looks like:

The problem seems straight forward. The URL ends with ?user_id=me. In other words; they expect me to be signed in (not that Google Reader can sign in for me).

If I do sign in and try that URL again I get redirected. However, the URL I get contains a secret variable and I'm sure it's not wise to share its value. Maybe it's safe. Maybe it isn't. I've no idea.

What Flickr need to do is simply and automatically upgrade the old public RSS addresses to the new public RSS addresses. It should be faff free.

Monday, June 28, 2010

Ask Jeeves - the search engine that tweets ads

I followed the Ask Jeeves Twitter account back during the search engine's second, third, or perhaps eighth return. Never really managed to get into any promotional banter with the account... shame.

It's not unusual for Mr Jeeves to use Twitter to ask questions. After all; that's the search engine's thang. You ask a question and it'll answer it.



Question of the Day: How many mobiles are stolen every hour in the UK? A: An average of 228 phones! Get protected here: http://bit.ly/bFy4kgless than a minute ago via web

embed this tweet



I was surprised to see the butler offer a direction though. If I wanted protection I should visit this link. It was all bit.ly'd, of course.

Looks pretty clear to me that Ask has an ad deal with F-Secure. Nothing wrong with having an ad deal. Just didn't expect Ask to be tweeting about it.

Just a little bit digging also shows that today's question also appears on the search engine's homepage and link straight through to the the specially created landing page too. Doesn't do much in the way of showing me Ask's ability to answer the questions I ask.



Oh well. Times are tough. There's nothing wrong in Ask mixing in a little ads into the search profile and I'm fairly sure not no one's going to mistake this for a search result! However, I wonder what brand concerns they had around this tactic.


Tuesday, June 22, 2010

Google refreshes the rules on low quality landing pages

Google has a page that discusses which types of websites may merit low landing page quality score. It was updated in the early hours of this morning.

One of the biggest changes to the documentation was the opening paragraph. Why would Google give any type of site a low quality score? Google says;

Users have consistently provided negative feedback on several types of websites.

So, what are those sites?

  • Those that offer free items for data collection.
  • Arbitrage sites that are designed for the purpose of displaying ads.
  • Low quality affiliate sites – such as poor comparison shopping or aggregation sites.
  • Get-rich quick sites
  • Sites that go against Google’s Software Principles
  • Sites that feature false of misdealing claims – this includes competitive claims and generic superlatives
  • Sites that go against Google’s Webmaster Guidelines, use cloaking or other inappropriate technologies.

So, let’s take a look at that.

Sites that breach Google’s Webmaster Guidelines can harm their PPC quality score. Link buying can push your PPC costs up. That’s an interesting fact to bring out in a pitch. I think I’ll do that in the future.

It’s interesting that the rules about generic superlatives and competitive claims are cited. They’re well known for being risky in the ad creative; but for the landing page too? One to watch.

It’s also interesting to consider Google explicitly calls out “poor comparison shopping” or “aggregation” sites as a low quality score risk because there certainly is legal action brewing in EUland for Google from some shopping and aggregation sites who claim they’ve been treated unfairly, in general, by Google.

Friday, June 18, 2010

The iPhone 4 will change how we think about phone sex

I've had an iPhone 3G. I don't think I'll be getting the iPhone 4 as I'm content with my HTC and Android (give me Froyo or I'll delete this post!).

I'm still very much watching the market though. Perhaps the iPhone 4G will be more impressive. Perhaps it'll have three cameras.

Did you know there are designers who wanted the iPhone 4 to have 3 cameras? This video explains all


Thursday, June 10, 2010

Chrome is now so fast that web sites break. Check yours.

Chrome has become my default browser. I still use Firefox when I'm examining the site; those plugins are so handy but the rest of the time I'm on Chrome.

I use Chrome because it's fast and sleek. I'm often on a laptop and sometimes on a 3G mobile connection. Chrome helps then.

In fact, Chrome is so fast that I've noticed some websites aren't coping. I've an example from bigmouthmedia. I was checking out our coverage of social sites out-pacing search which was doing fairly well on Twitter when I noticed this...


What's this? An error on our new website. The footer appearing before the content finished?

Turns out there's nothing wrong with the code (although; yeah, there's an IE6 error we can live with (actually; I wanted the site to explode in the face of IE6 users - but I was overruled)). What's happening is that Chrome completes the JavaScript function correctFooter() before the images have finished loading. This only happens with Chrome 5.0.375 and up.

In our case the fix was simple - delay the JS a little bit.

Going forward - I suspect this will become a growing issue, especially with HTML5, and more sites will have to manage the time/order of loading - especially in relation to relative positions.

It's also worth being aware this is Google's tech. This is a suggestion at how quickly Google could parse JavaScript. I'm not saying this is what Googlebot will always see but it's certainly a suggestion of the sort of rendering Googlebot could see.