The one disappointment IMO has been Streetview – the old version was so much slicker. The new version is easier to implement, but that is as positive as I can be. I miss being able to click in the distance and zoom down the road; I miss the swoosh as you move from position to position; I miss built in Full Screen.

On the flip side we now the have API for static streetview images – I’ve been using Jamie Thompson’s brilliant work on getting static map images up until now, but it is time to move on.

Slightly irritating is the fact you need to convert old (v2) Streetview panorama locations (if you have them saved) to show the locations as static images or as saved Streetview panoramas.

To convert a v2 panorama to a v3 panorama:

• ‘yaw’ has become ‘heading’
• change the sign of ‘pitch’ e.g. -1.80999 becomes 1.80999
• add 1 to your ‘zoom’ level

When you work in an office with stupidly fast broadband, maybe you just forget? But you shouldn’t assume that the people looking at your websites do.

The bit that really bugs me is this peculiar ‘head in the sand’ attitude. It does take more than a few minutes of thinking to come up with any number plausible reasons why speed is important for Search Engine Optimisation (aside from the fact the Google say it is). Try it, take a few seconds to think.

For me though the reasons that scream out at me are User Experience and Money.

User experience

First off – I admit this isn’t really an SEO factor, is it? People are busy, they don’t want to spend valuable time waiting for a web page to load, at least I know I don’t. I have better things to do, and I’m not that weird.

But why would Google care? I can rationalise that they want people to look at Ads. And then click on them. Probably a good idea to make websites faster so people don’t get bored and do something else, or find the bit of content they are looking for before the ads have loaded (because the ads come last).

Time = Money

Google must index billions of pages per day. It stands to reason that the faster a site loads, the faster it can be indexed. I’m guessing that it costs about the same to run a server whether it is indexing million pages per day or 10 million – but if your business is based on data the faster you can chew through data the better – and therefore it also makes sense for Google to promote an internet that lets them index more, faster.

But we don’t have the time, it’s to hard…

No it isn’t. Of course, as with any work there is optimum point in terms of cost vs. benefit (and that will be different for every site).

My gut feeling (and I don’t have any firm evidence for this) is that speed becomes more important the more traffic you get. If you’re site gets 50 visitors per day, then in a sense it is what it is – what you need to focus on is link building and content. However if your site gets 5000 visitors per day, then the chances are you’ve got quality links and good content – so how do you improve? I think the answer is the speed and user experience.

A fast website either means no traffic, no load on your server, no pictures, no javascript – in short a bit of old school html text… or it if is a big site, it means you care. It means that you have taken the time to think about users, you’ve spent a bit of money on hardware or code. It means that you are a quality business and therefore it is a good bet that you have better quality data on your site. It means that people will pay more to be associated with your site.

How to make your site fast

Hell, that’s a bold statement. It’s different for everybody. Think about the usual suspects, too much JavaScript, un-optimised images, sloooow queries with no indexes.

There are a myriad of solutions, and you don’t have to have Prototype, Jquery and Mootools all running on the same page, choose a single framework at a time – you will be amazed.

• Use a CDN.
• Use a cache server like Varnish.
• If you’re a bit lazy – try out mod_pagespeed (sometimes it’s fab, sometimes it’ll melt your server)
• Be brave, leave the Apache comfort zone and use Nginx or Lighttpd
• Use minify

Fringe benefits

Chances are you’ll end up using some kind of caching and you’ll start setting sensible expires headers for your static content. But just think you’ll never have to tell somebody at the end of the phone to clear their browser cache again. And remember, the client is not an idiot because their browser hasn’t magically detected that a style sheet or an image has changed, and the website they have paid for looks a bit rubbish.

(by the way I don’t always practice what I’ve just preached, but at least I feel bad about it)

http://googlewebmastercentral.blogspot.com/2010/04/using-site-speed-in-web-search-ranking.html
http://www.mattcutts.com/blog/site-speed/
etc.

Along the way I had to ran into a problem with my map – when I showed it the centre-ing was gone – I tried all sorts of things… map.checkResize() didn’t work and map.setCenter() didn’t work.

Eventually I deduced that first you actually need to call both functions – reset, then centre e.g.

 
    map.checkResize();
    map.setCenter(new GLatLng(51.90568, -1.33214));

Along my way I have found a great article comparing jQuery and Mootools – which I would recommend to anyone going either way.

*But it is a kind of a personal philosophy not to ever say on my blog where I do work, just because. I’ve taken over a big project running on a Cloud server written by a load of other people, the sort of thing that is hard but very rewarding.
** I have no idea where those coordinates are. Maybe Up North somewhere?

Looking in Firefox error console I see G_MAP_TYPE is not defined; – WTF? it was working…

Well looking into it, it is an easy fix. It seems that a lot of people (myself included) have been using depreceated constant names for the maps types, and that there will be quite a few Javascript Plugins (for things like jQuery) that may need updating.

The usual (i.e. correct) types we should now be using now are:

See: http://code.google.com/apis/maps/documentation/reference.html#GMapType

For a bit of an actual explanation I found this post on Google Groups