In my previous blog entry, I discussed the value of HTML packaging and focused mainly on how it would effect browser performance. I’d like to add some additional thoughts about how I believe Mobile providers can enjoy both reduced bandwidth costs and performance increase for their customers by doing so.
As described in my previous entry, packaging would significantly reduce the size of any resource files (JavaScript, CSS) that would not initially be compressed by GZip. However, additional compression may be achieved by having the mobile provider’s proxy perform some extra steps that would be valuable for the mobile user:
1. Compress CSS and JavaScript – while this requires some intelligence, software packages already exist and provide this service.
2. Pre-process all imagery (in HTML and that in the CSS) and adapt it to the customer’s mobile device and packaged as well (potentially switching formats, and definitely reducing size and/or color depth).
3. Packaging would decrease the overhead of HTML requests (yes, minor, but this is mostly for performance, not for bandwidth)
Finally, the mobile provider would add the appropriate new tag ( — Kudos to Alexander Limi for this one) with reference to the newly created package to the returned HTML file. This would prompt the browser on the mobile client’s side to grab it before looking for the web resources (which, in this case, may be all that’s required).
All it really requires is a great browser software developer (Opera?) to pick this up and implement…
Anyone?
