Re: Perlmonks Faster Access

If we were going to cache the pages, it would be far simpler to do it without involving the pen. They could either be generated to disk regularly or cached with something like Cache::FileCache. However, I think you're forgetting about all the interactivity: nodelets, for example.

You might want to look back at previous discussions about caching nodes and see what has been said before about it. It's certainly not the first time this has been suggested.

Comment on Re: Perlmonks Faster Access

Replies are listed 'Best First'.
Re: Re: Perlmonks Faster Access by artist (Parson) on Feb 21, 2003 at 20:52 UTC
Sorry, but couldn't find previous discussion on caching nodes perrin. Your stuff made it extremely easy. Once the user has selected the 'plain' mode, all the other nodelets can be turned off and the page served is what is cached at perlmonks, if that page has not changed. If that page has changed from the cached version, (timestamp comparison only) the new page from database get cached and is served. User can select 'full' mode to see all the nodelets and have full interactivity. The reason is that I think, lot of processing may be required to display the nodelets per user settings and that could be eliminiated in the case where user doesn't require them. artist	[reply]
Re: Re: Re: Perlmonks Faster Access by perrin (Chancellor) on Feb 21, 2003 at 22:32 UTC
There are other interactive bits; nodelets were just one obvious example. Read this thread for more discussion. What you're describing is basically what the software already does. It caches the data for nodes and checks to see if they have changed. That check requires querying the database. Remember, this system runs on a cluster, not a single machine.	[reply]