Alex Hubner wrote: > > Sort of Off-topic, but anyway, HTTP compression doesn't cause > performance decrease? I think the gain is very relative since you > have to add more time do compress (server CPU usage) and also more > time to decompress (client CPU usage) in this operation. HTTP > compress should be used only if your audience are, in its majority, > 14k/28k modem users. Not to mention that I think Netscape is not > compilant with that. This is true?
Performance: It depends. I have seen one case in which it actually reduced CPU usage because of repetetive blocks in the HTML. The idea is that in that case even selecting low compression (low CPU usage) reduces the size significantly. Think about tables with many rows etc. In that case the CPU usage gained by reducing the amount of Apache children (it was a PHP/Apache case) was larger as the CPU usage of the compression (and the compression still reached a factor of about 5). But if you have content that is hard to compress this would obviously not be the case. Browser compliance: Browsers send a header indicating which level of HTTP they support with every request. And a header that tells if they accept gzip/deflate content. Every server implementation that I know off respects these settings, so there should be no problem here. When to use: I know little about the money side of things, but I would guess that a once off investment in processing power is worthwhile if it reduces the bandwidth usage substantially. But check your accountant/colo contract. Speed is something you can test for yourself. Just try it, it shouldn't be too hard to write a little javascript that measures the time between a location.refresh() and the onload event of the body. Jochem ______________________________________________________________________ Your ad could be here. Monies from ads go to support these lists and provide more resources for the community. http://www.fusionauthority.com/ads.cfm FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/[email protected]/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
