Hi Chris, That's a little beyond topic scope but here goes.
The image / CSS / data URI layout used on the page is a little complex I'll agree. It was optimised to provide the key images first and quickly, even in IEv6. Note the different sub-domains used. The CSS is served via a gzip and cache utility hence the far future expiry date. The design was written 2 years ago and page load optimisations followed in the three months after. It is significantly faster, I'd say to a factor of 7 or even eight times quicker. Unfortunately the tools used haven't records from that far back so I cannot back the statement up. As to whether it's worth the trouble? That's a different matter. Here it was used as a learning experience, so for me it definitely was. The techniques investigated, and the lessons learned, were in part applied by my current employer. Up to 2 million hits a day, it used to take 3.7 sec to download the homepage. Now it takes a healthy 0.6 seconds. References: http://websemantics.co.uk/projects/#tesco It's on sites like this where these techniques really make a difference. Albeit I've not as yet implemented data URIs there. I need a gzip enabled server first. Regards Mike Foskett. -----Original Message----- From: li...@webstandardsgroup.org [mailto:li...@webstandardsgroup.org] On Behalf Of Chris Knowles Sent: 10 February 2010 12:59 To: wsg@webstandardsgroup.org Subject: Re: [WSG] Data URI encoder > > The main application is to reduce HTTP requests and thereby increase > page delivery speed. > > Hi Mike, I see that the page you refer to links to a stylesheet with 4 images embedded in it, rather than the stylesheet linking to those 4 images, therefore, you have one http request rather than 5 and also, that stylesheet has an expires header set to 10 years from now. You say it's a lot faster, but I question the value of going to this trouble. I agree there is a performance gain, but if you link to the images from the stylesheet instead and also set an expires header on them then subsequent page loads become irrelevant so it's just the initial visit with an empty cache that is affected. Given that the download size is pretty much the same with either method, the only gain i can see is a marginal one from those initial extra 4 http requests. Is that really such a huge gain? -- Chris Knowles ******************************************************************* List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm Help: memberh...@webstandardsgroup.org ******************************************************************* This is a confidential email. Tesco may monitor and record all emails. The views expressed in this email are those of the sender and not Tesco. Tesco Stores Limited Company Number: 519500 Registered in England Registered Office: Tesco House, Delamare Road, Cheshunt, Hertfordshire EN8 9SL VAT Registration Number: GB 220 4302 31 ******************************************************************* List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm Help: memberh...@webstandardsgroup.org *******************************************************************
