I work for IBM so I should be saying: yeah! let's throw more iron at the problem and we'll be happier, but unfortunately more iron is not the first thing we need here, good design/code is.
I also quote Allen: """However, with the right amount of infrastructure and planning I think Roller could support an unlimited number of blogs all running the way they do now and that gives us XXX times more power to do cool things than any static site ever will. """ I think saying "unlimited" is a very dangerous thing to say although possibly our requirements are not that great yet. So I'd agree with Dave that we need static content support or a serious caching solution (like the memcached suggestions months back). I'm not saying dynamic is bad, but sloppy coding shouldn't be accepted because of *unlimited* hardware. Sean Gilligan wrote: > > David M Johnson wrote: >> >> >> >> On May 3, 2006, at 4:49 PM, Allen Gilliland wrote: >>> >>> To put it another way, if the urls work fine without the file >>> >>> extensions when why should we add them? >> >> >> >> I think static blog generation is the only reason to use them. > > > > If an extension affects Google (or other) searches (as John H suggests), > > then that would a big reason. Bloggers want their stuff to be found and > > read. My personal blog is on Blogger (which puts .html on the > > permalinks) and I've been amazed at the search engine placement I've > > gotten for some entries. (Hopefully Google isn't favoring their own > > software) > > >> >> >> >> Conventional wisdom seems to be that, in order to scale up to hundreds >> >> of thousands or millions of blogs, we'll have go static. I'd hate that >> >> because I love the dynamic nature of Roller. Maybe conventional wisdom >> >> is wrong and maybe I'm worried about a problem we'll never have to solve. > > > > The conventional wisdom may be outdated. I don't claim to know the > > details, but with Dual, Dual-Processor 1U 64-bit becoming "entry level" > > and RAM at about $100 per GB, the approach of cached dynamically > > generated content should be able to keep up. Read my comment above. > > > > I recently converted a project to use Servlet-based delivery of static > > content that had been served with Apache httpd after years of avoiding > > that for performance reasons. When we actually ran benchmarks, the > > extra CPU used by the JVM vs. native Apache wasn't a real factor. Are you serious? Wow! I can't believe someone would do such a thing. I'll try to play along, was this Apache 1.3 or Apache 2.0? > > > > The last major release of MovableType added more dynamic features (via > > PHP) and WordPress is fully dynamic (I believe) and seems to be doing > > well and being used for some large scale deployments. I think we need to be careful when we compare ourselves to PHP/Python/Ruby because they are completely different technologies than Java and those specific solutions are very lightweight compared ours. We make heavy reuse of many libraries that do much more than what we use them for and definitely not optimized for our needs. Wordpress is one single PHP function that does all of the work. Wordpress is a single-site blog and nothing compared to what Roller is aiming at providing. MovableType on the other hand has delivered the scale, but they have static content. Dynamic features are great but need to be carefully thought out and used sparingly if you want to play in the big leagues. I think that we already have the potential of getting into so much trouble with our template system that it will be impossible to estimate the amount of hardware and memory we'll need to handle our sites. We have little control over each blog's performance and how it affects the rest of the site. One last thing, I think we need to remember the Web 2.0 direction and think about dynamic content differently other than with JSPs and Velocity templates. > > > > Are there any specific and relevant studies or analysis available? > > > > -- Sean > >
