[EDIT: My comment refers to the original title of the link. This has since been changed]
> A quick look at MySQL 5.4 - 60% performance gain over MySQL 5.1
Sorry, but this is just plain wrong. The actual article does a better time explaining and disclosing this small, specific speedup. MySQL InnoDB in 5.4 is faster on >4 multicore servers because it properly accounts for the extra CPUs. Calling it a "60% performance gain" is really misleading.
> One of the primary fixes in MySQL 5.4 is that the InnoDB storage engine can now address more than 4 CPU's/cores
You are correct, it's mostly targeted at multicore servers and the preview only holds Google's InnoDB patches. Generally thought, I hope they keep looking at performance issues beyond the InnoDB mutex patches - - and set some of the Sun optimization wizards to optimize MySQL. Sun has some _very_ bright people (those that have made ZFS, DTrace, Solaris, Java's VM etc.) that could optimize MySQL to new levels. Let's just hope they don't drown in bureaucracy.
It is good to see that MySQL is getting some improvements. It is a decent database, but they really need to fix replication and a number of other critical issues. Hopefully the different release cycle will help them target the issues better.
It's approaching the same feature-set as Microsoft SQL Server had over a decade ago, but still severely lacking in optimizations and performance for anything besides basic queries.
If you were to install an OS today, would you call Windows 98 "decent" and go for that, or ditch it for something modern?
With today's computers and the simplicity of most websites so could be said about flat text files or even XML-files queried realtime trough Xpath or Xquery
That does however not mean I would tout those technologies as good general databases.
> A quick look at MySQL 5.4 - 60% performance gain over MySQL 5.1
Sorry, but this is just plain wrong. The actual article does a better time explaining and disclosing this small, specific speedup. MySQL InnoDB in 5.4 is faster on >4 multicore servers because it properly accounts for the extra CPUs. Calling it a "60% performance gain" is really misleading.
> One of the primary fixes in MySQL 5.4 is that the InnoDB storage engine can now address more than 4 CPU's/cores