Friday August 26, 2005

My tipping point I finished reading "The Tipping Point" by Malcolm Gladwell last week. Although, I found some of the topics in the book interesting, in general I felt frustrated with what I perceived to be a lack of completeness with some of the author's theories. My tipping point from frustration to flat out irritation came with the following example. In this example the author attempts to reference research by Leda Cosmides into human cognition and the effects of context.

     Suppose four people are drinking in a bar.  One is drinking Coke. One is sixteen. One is drinking beer and one is twenty-five. 
     Given the rule that no one under twenty-one is allowed to drink beer, which of those people's IDs do we have to check to make
     sure the law is being observed? ("The Tipping Point", MG, pg159)

Hmm, well, assuming that the one who is drinking Coke is only drinking coke, then I would have to go with the one who is drinking beer, as they are the only individual drinking beer, whose age is unknown. But, wait..there's more! Lets see what Mr. Gladwell's answer is:

     Now the answer is easy. In fact, I'm sure that almost everyone will get it right: the beer drinker and the sixteen-year-
     old. ("The Tipping Point", MG, pg160)

Wow, don't I feel like an addlepated infant! The sixteen-year-old? Are you sure? But don't we already know how old they are? How would it be useful to check their ID? Wouldn't it be more constructive to check what they are drinking?
*sigh*
If an author doesn't feel like paying attention to details, then it would be nice if they could at least refrain from using statements like "Now the answer is easy" or "almost everyone will get it right". (2005-08-26 10:50:13.0) Permalink Comments [8]

Wednesday August 17, 2005

1st frosug meeting The first Front Range OpenSolaris User Group(frosug) meeting was held here in Broomfield, CO last Thursday. Jim Walker and Stephen O'Grady did a nice job of covering the meeting in their blog entries. There were ~20 developers from Sun that attended the meeting. I was pretty excited about the overwhelmingly enthusiastic response from Solaris developers, which seems to bode well for the success of OpenSolaris:-)

Technorati Tag: OpenSolaris
Technorati Tag: Solaris (2005-08-17 11:19:14.0) Permalink Comments [0]

Friday August 05, 2005

Fun with state database replicas (Part 1) As I mentioned in my last entry, information about SVM metadevices is stored in state database replicas(mddbs). The data structures used in state database replicas are defined in: usr/src/uts/common/sys/lvm/md_mddb.h
The following is a drawing of diskset mddb:

1) The Master Block

The master block contains two important pieces of information, a unique device identifier(devid) and the offset for the locator block.* The devid is used by SVM to track the drives that are in a set--so that SVM can handle changes in their physical location. The master block contains the mapping from a logical block 0 to the physical offset for the mddb's locator block. If there are two mddbs on the same slice, they will each have a different master block--the location of the first metadb will be at 0 and the second at 8192. The master block and locator block are the only two parts of an mddb who's location always remains constant. All drives in a diskset either contain a mddb or a dummy master block. The dummy master block is primarily used to store the devid for the drive.

* In replicated disksets, the master block does not have a devid for the drive it is on, rather it contains a copy of the devid for the original drive.

OpenSolaris
Technorati Tag: OpenSolaris
Technorati Tag: Solaris (2005-08-05 17:52:20.0) Permalink Comments [0]