Joined: Feb. 2007
||Posted: May 25 2008,11:32
I'm curious why he gave DSL a 2 (out of 5) for stability and being "useful." He mentioned that it comes with a variety of software, all of which works, and I've yet to encounter problems with DSL's stability on any machine on which I've run it regardless of method (CD, hd install, frugal, USB-HDD, etc.). It freaking works. What more does he expect?
He didn't bother to detail what he tried to do to get his hardware to work. Is it DSL's fault if he doesn't RTFM or that he uses hardware without native Linux drivers? Look at the hardware he used: bcm43xx, SATA hard drive, etc. Did he try the SATA bootcode? Doesn't say. Just says he couldn't access his hard drive or USB key. I have a bcm43xx card, too, but I blame myself when it comes to using it in Linux or BSD because I knew it lacked support outside of Windows. Where does the responsibility for choice of hardware fall -- on distro developers or on users? If it's on developers, how far are they supposed to go in supporting binary drivers and the kludges required to make them work? Obviously, he absolves himself for buying and using hardware that's not open source- and/or Linux-compatible.
I recently had a brief exchange with a podcaster for his review of Absolute Linux, a Slackware-based distro also targeted at older and less-endowed computers, because it totally got something wrong (about binary packaging versus compiling) and focused a lot more on aesthetics than function. He complained about the themes, wallpaper, and "ugly GTK1 or FLTK" apps. As if older computers should be bogged down with bloated graphics and heavy pixmapped themes, GTK2, and bleeding edge software.
The exchange starts at the link below and ends on his blog (linked in his comments). He claims I misunderstood him about compiling, but I subjected myself to listening to the podcast in question again and it's clear that he never mentions packaging and intimates throughout that "it's a cool project if you have time to compile things you'll need." Even mentions Gentoo in this context. He also defended the eye candy thing throughout.
It's always going to be apples and oranges -- at best, if it's even that close. When someone whose primary experience with Linux is something like Ubuntu tries to review something like Slackware or one of the BSDs, there's very likely going to be a disconnect because of the lack of familiarity with doing certain things yourself, with the differences between how things are handled, with presumptions between "users need these things done for them (edit: whether they want them or not)" versus "our users are savvy enough to do these things themselves (edit: IF they actually even want these things)." The result is too often a very misleading review that focuses on appearances but ignores the underlying paradigms involved and unique features stemming from the way things are done. That was at the heart of my complaint about the podcast above. It's at the heart of my statement yesterday that most reviewers and their reviews are full of sh**.
It certainly applies to this review of five tiny distros. There's no reference to the compression methods employed by the five distros (which is important when considering target audiences and how one distro can use 2.6 and be "small"-ish but may not be suitable for older hardware), there's nothing about which of them runs as root only (trading off security for ease of use a la Windows 95), nothing about installation options, nothing about binary packaging and ease of extending the base ISO. Like I wrote yesterday, that all takes a backseat to BS about how it looks and harping about kernel and app version numbers.
"It felt kind of like having a pitbull terrier on my rear end."
-- meo (copyright(c)2008, all rights reserved)