Quote from: QnA at Mar 14, 2009, 02:07 AM
My opinion: 90% of all current blogs, web sites that use MySQL shouldn’t be using it. Slowdown is noticeable vs. flat file CMS sites. See web-app.net for contrast (and I hate Perl). Firefox’s SQLite is dog slow even from ramdisk. Things were better before. I’m a sysadmin, I know what I’m doing with tuneup.
You sure? Do you use query caching, or memcached?
You find most CMS web sites won’t eat their own dogfood; they use wikis.
What? Where are these "most CMS websites"? Most all the CMS websites I know use their own CMSes for their website. And I work in this business.
To me that says something. We also have Google; need for SQL queries is much reduced, even for the large content sites. Just use Google.
Use Google for what? For finding sites that use SQL?
Where SQL belongs is exactly where the links above say it belongs - millions of small data items updated frequently in private databases. The only form of CMS content near that requirement is a forum.
What are you talking about? SQL is a database language. Databases are designed for storing any time of data format. A good CMS also breaks down content into "millions of small data items". You’re not making much sense.
To each his own. I consider SQL archaic mainframe tech dressed as modern hotness.
You’re entitled to your opinion; but I would argue that you are completely, utterly wrong.
Google doesn’t use SQL, but MapReduce. Good for them, but they want to pawn another buzzword, ’cloud computing’ - or, client/server. Supposedly, the PC rescued us from that. It’s all rehash from prehistory. Just depends what you’re trying to sell.
Google may use MapReduce, sure, but that doesn’t make it any better than SQL. I can post plenty of articles that show MapReduce as crap. It’s only useful in certain situations.
HTML is debatable as a language, CSS is ridiculously simple. JS is tough until you understand closures and browser idiosyncrasies. PHP is harder, I’ll give you that. SQL is a piece of cake.
I can live without SQL and DB abstraction layers just to pull text snippets from disk. Thanks anyway, boys.
Sure, go for it - but have fun sorting, indexing, and searching those on a reasonable timeframe. And forget multi-server hosting, alternative caches, or other forms of load distribution. And good luck trying to run a million-hit-a-day site on all those file_get_contents() or include calls.
Were you here to ask a question, or unload an agenda?