When I heard about this change, I was unsure if it's a good thing. I know it's very hard to edit the template and wait until Blogger updates all the blog posts, but the dynamic pages could lead to serious problems.
A reader of the blog, whose digital alias is geekWithA.45, sent me this:
The nature of blogger pages is that they are infrequently published pages of (generally) low complexity, and they are frequently demanded.
"Produce content on demand" is completely the wrong policy for this, it is the wrong tool for the job.
I predict massive problems that scale directly with load, which may or may not be revealed during beta.
When that happens, blogger will go through a familiar sequence of attempted remediation steps.
First, the database connect cache settings will be played with. Then whatever the underlying system's object caching hoojie is will be turned on, and messed with extensively.
After that, they'll start messing with cluster settings for the server farm, and scratch their heads trying to scale out and cluster the back-end data servers. During this phase, the overall size and configuration of the server farm may or may not be altered dramatically.
Somewhere along the line, someone smart will say, "AHA! Edge Caching for relatively static components!" and Akamai sales reps will get in on the game.
Finally, the development team, who hasn't been home in weeks, will be taken out back of the barn and shot, but they won't mind too much, death is preferable to the hell they live in. Perhaps whoever decided to go dynamic, based on someone's grad school project white paper will be shot with them, but more likely, he or she will be promoted to a position where he can do more damage.
His arguments are interesting, but the future will tell if Blogger's decision was correct. Blogger had a lot of problems in the past, but even if their server was down and you couldn't post anything, the blog was still available. If the problems continue, when Blogger will be down, blogs will be down too.
No comments:
Post a Comment