Via Pete Zaitcev I just read FreeBSD and release engineering (lwn.net) , which is about some issues peopleare having with the FreeBSD release schedules. That article willprobably be a Rorschach test for at least non-FreeBSD people, becausehow you react to the ideas in it will probably depend on how yousee 'production support'. Before I get into my reactions to it (inanother entry), I've decided to write down how I see productionsupport.

To me:

• Full support of

Linux's software RAID implementation stores metadata about the RAIDdevice in each physical device involved in the RAID, in what mdadm calls 'RAID superblocks' by analogy to the filesystem superblocks thatdescribe filesystems. In modern versions of software RAID there are anumber of different formats for these RAID superblocks with differenttradeoffs involved in each one, and one of the decisions you need tomake when you create a software RAID array is what format you want touse.

(Even if you don't actively make a decision,

One of the interesting things that happened during my five years ofhardware hibernation is that when I woke up,even low end (aka passively cooled) graphics cards could suddenly drivetwo digital outputs. Back in 2006 it was common for cards to have oneanalog and one digital out (eg, the ATI X300 in my work machine had VGAplus DVI), but getting dual digital out required an expensive card withan often noisy fan.

(I actually went through two such cards at

In Filenames.WTF , Daniel Rutterruns down the reasons first why paying attention to file extensionsis ridiculous, and then the reasons why it's still the best solutionto the problem that we have. Spoiler: it's because people have spentdecades creating file formats that suck.

( Via philliph on Twitter .)

Phil Hollenback recently wrote DevOps Is Here Whether You LikeIt Or Not , inwhich he writes that traditional system administration is dying.While I sort of agree with him about the death, I don't think it'snecessarily for the reasons that Phil points to.

Fundamentally, there has always been a divide between small systemsand large systems. Large systems have had to automate and when thatautomation involved applications, it involved the developers; smallsystems did not have to automate, and often do not automate because

This fall I got a new home machine , justa bit over exactly five years after I got my previous home machine . It happens that I saved the invoicefor my five year old machine, so I dug it out today in order to do acomparison about what five years of progress in PC technology did anddidn't change for me.

First off, the progress of five years got me much better prices. Myrecent home machine cost me only about 60% of what my old home machinedid

One part of getting on Twitter has been following people, which led meto discover that when you follow someone Twitter doesn't show you allof their public tweets. To summarize what I think is the rule, Twitterexcludes any conversations they're having that purely involve otherpeople you don't also follow. Their tweets in the conversation willappear in their public timeline, but not in your view of their tweets.

(This may only apply to relatively new Twitter accounts, or even only tosome of

On a modern Unix system with all sorts of memory sharing betweenprocesses, Resident Set Size is a hard thing to explain; I resortedto a very technical description in my entry on Linux memory stats . To actually understand RSS, let's back upand imagine a hypothetical old system that has no memory sharing betweenprocesses at all; each page of RAM is either free or in use by exactlyone process.

(We'll ignore the RAM the operating system itself uses. In old Unixes,this was an

Like sensible people, we scrub our pools periodically in order toturn up latent problems. Because pool scrubs have a visible impact onresponsiveness (at least in the lightly patched Solaris 10 update 8 thatwe're running), we only run scrubs on weekends (and only scrub one poolper fileserver). However, we've recently started running into problemswhere pool scrubs slow the fileservers down enough that backups havestarted failing.

The obvious way around this is to switch things to only doing scrubswhen backups