Wednesday January 13, 2010
Bill Sommerfeld's WeblogStill Under Construction. Watch for falling objects new beta of globalsan iscsi initiator for macos I followed the example of Constantin and set up Time Machine + iscsi backups from a Mac Mini running Snow Leopard to a zvol on our home opensolaris server. We had quite a few stability problems -- mostly hangs and laggy behavior relating to suspend/resume/sleep on the mac -- which I described in a post on the SNS forums. Fortunately they appear to be fixed by the newest beta, 4.0.0.204_BETA, of the globalsan iscsi initiator; if you've tried this out and it didn't work for you, try the newer code.. On authors having a right to unpublish Ted posed an ethical question about the use of an unauthorized ebook edition of an out-of-print work. (IMHO fair use includes a right to convert a work legitimately owned into a form that can be conveniently used by the owner of the copy. In his poll: I think all of options 1, 2, and 3 are ethical. 5 is clearly not ethical. And 4 is just silly). In the ensuing discussion some people have posited that there's some sort of inherent right of authors to unpublish a published work. IMHO that's nonsense. Publication is (or ought to be) irreversible, and the irreversibility is a feature which contributes to the right of authors to say controversial things. Once you publish something, it's out there; your words have a life of their own, and nobody can pressure you to take them back because you have no ability to take them all back. There's already enough of a problem with libel tourism. It would be far worse if a court could force you to exercise a right to unpublish. Knowing when a crash dump is complete If you're doing kernel development on solaris, every so often you'll crash a machine. It will save a crash dump to the dump device and then on reboot, the dumpadm service (svc:/system/dumpadm:default) will invoke savecore to copy from the dump device to the filesystem.
Since SMF goes out of its way to let services start in parallel, and dumpadm is one of those services, it might not be done by the time you log in to start investigating the crime scene. It's annoying to have to periodically poll to see if it's done (sometimes it takes a while) so you can do something like: cd /var/crash/$(hostname); svcadm -v enable -s dumpadm; mdb -k N and then go about your business until you see an mdb prompt in that window. This relies on the the -s option to svcadm (which causes it to wait until the service is done attempting to start) as well as the fact that enabling an already-enabled service is a no-op. (2008-09-19 17:19:34.0) Permalink Comments [2]
A testimonial about DVRupgrade.com DVRupgrade has shown no respect for my right to be left alone. I purchased a product from DVRupgrade. I received it. But against my implicit wishes and without any explicit informed consent they also subscribed me to their newsletter. After I unsubscribed, they then hounded me via additional email to provide a testimonal about my experience. So, here you have it. I wanted a product, not an ongoing relationship (and certainly not additional swill in my inbox). I recommend that you buy your tivo upgrades from someone else instead. (Just for the record, I want them to conform to best practices for customer email address handling: don't subscribe me to mailing lists I didn't ask to join, don't make me unsubscribe from lists I never wanted to be on, don't keep pestering me after I unsubscribe, and don't keep pestering me with insincere apologies after I complain about being pestered). On respecting other people's privacy preferences So, one of the things I've chosen to opt out of in life are "social networking" sites -- linkedin, orkut, facebook, evite, etc. I don't trust them. It doesn't entirely matter why. I just don't trust them -- please don't waste your time trying to convince me that they're ok. And if the only way I can get invited to parties is to accept email from evite, I'll happily stay home by myself.
Unfortunately, a fair number of people (typically those who don't know me very well) think that just because they I really don't understand this -- in my mind, it's horribly rude to give my contact information out to a commercial third party without asking me about it first. And I don't really put much stake in official privacy policies - if they don't owe me anything, they have no incentive to prevent my email address from being leaked to spammers, identity thieves, and similar pond scum. So, if you want to get in touch with me, do so directly using your own email account or phone. Don't indirect through sites like orkut, evite, linkedin, etc.; I generally route mail from such sites directly to /dev/null because it annoys me. And if you want to be respectful of others, don't assume that they have the same privacy preferences as you -- check with them before you give their email addresses to third parties, no matter how wonderful and saintly you think those services are. They might not trust the service. They might have another email address that they'd rather use with the service. Just take the time to ask individually first! (That's the last thing Puccini wrote in the 3rd Trombone part for Turandot. Not a lot to do but sit & wait for the trumpet cue...) In Memory of Greg McMullan, 1963-2008 I was stunned to get email Sunday morning informing me that Greg McMullan had died in a house fire. Greg wasn't a close friend -- just someone I crossed paths with now and then, starting when I was a confused and clueless froshling at MIT. He sticks in my memory as a very friendly, approachable and helpful person who could be relied on to give good advice and make me (and many others) feel at home in the high pressure environment of MIT. I ran into him again perhaps ten or eleven years ago when he was showing off his new toy -- the first generation Palm Pilot. I was sold almost immediately. And I vaguely recall running into him when I was in a orchestra accompanying a large-scale choral work; he would have been singing, I had my trombone.. but for the life of me I can't remember what groups were involved or what piece we were doing. configuring dhcp for almost-seamless use of nwam When I use my laptop at home at my desk, I usually have it plugged in to a port replicator. I have wireless but I need to plug in for power anyway. I finally got around to setting up the laptop to use the nwam phase 0 package to provide nearly seamless migration between wireless and wired network. The trick turned out to be to set CLIENT_ID in /etc/default/dhcpagent to the client id used by the wired interface -- 0x01 followed by the six-byte ethernet address. I suppose using the wireless interface's mac address would also work, but it's not soldered to the motherboard and might end up migrating to another laptop later.. When I undock to use the laptop in a different part of the house, nwam brings up the wireless and uses the same client ID with the wireless interface, and the DHCP server gives it the same address. There's a brief period of about 4 seconds when bits don't move, but no manual intervention (aside from the physical act of plugging/unplugging) is necessary. And when I bring the laptop back to my desk and plug it in, the reverse transition is also automatic. Very slick! Pippin opened last night. We were going along really well when, halfway through the first act, the casting holding the pivot & spring of my F-valve trigger popped off my bass trombone (a Conn 112H, which is an in-line double-valve instrument pitched in Bb/F/Gb/D); I'll note in passing for the benefit of non-trombonists that a Yamaha in-line double-valve bass trombone was the instrument mentioned in the well known but entirely apocryphal tall tale involving a Chilean bass trombonist, a firecracker, a mute, and the "1812" overture. (Now really -- what trombonist would ever use a mute in one of the few places in the symphonic literature where a quadruple forte (ffff) is written??!?!?!) Anyhow, after it popped off, the linkage and trigger hung down from the valve, occasionally getting tangled in my shirt pocket but fortunately keeping the valve stably "off" the whole time. So I was left with a Bb/Gb instrument. But whoever designed the instrument put in an extra-long tuning slide on the Gb valve so I could actually retune it down to F. Having learned the show using a bunch of alternate positions involving the use of the Gb valve, I played the rest of the show alternating the 2nd valve tuning. Definitely a higher-stress experience than I was hoping for, but it was an (unexpected) benefit of the in-line double-valve configuration that I hope I never need to rely on again... Aside from that, opening night went really well. And it's fixed already -- I'm extremely lucky that I live less than a mile from Osmun Best rehearsal quote in quite some time.. During a dress rehearsal of the battle scene in Pippin last night, the music director told the cast:
(I'm playing bass trombone in the orchestra of MTG's production. Opens this Friday.) (2007-11-28 10:36:46.0) PermalinkIf you think it's too good to be true, it is. So, I signed up for FiOS TV a week ago. They promised an install date of today, which I thought was a little fast. Turns out I was right. The installer shows up, looks around, and says that some outside wiring prep work that should have happened didn't. What's worse, the outside wiring folk apparently closed their install ticket saying "all set" when there is no evidence that there actually is an available tap in the cable anywhere near my house (the installers spent a bunch of time staring up at the poles and wiring before concluding that "construction" needed to be called to make a tap available). I'm in the midst of building our own flavor of labelled IPsec for Trusted Extensions, and took a look at what the "competition" (specifically, SELinux) is doing. I was surprised to notice that (at least if the ipsec-tools-0.7 source is to be believed) they've grabbed a codepoint assigned to RFC 3168 (Explicit Congestion Negotiation) rather than actually asking for one to be assigned via the normal IANA processes, or using the long-defined but rarely used capabilities of ikev1 to carry a sensitivity label. It looks like racoon2 gets this right (but doesn't have the SElinux security context support). I can't be the first person to notice this, can I? I'm surprised the complaint didn't include "failed to repair godzilla crush damage"
and "causes vertigo in blind people" (no, really!). I hope we get an injunction barring Gehry from making sandwiches. (MIT sues Gehry over Stata Center) (2007-11-07 18:52:00.0) PermalinkThree strikes, you're out (but it's only the first inning of a preseason game) The indiana prototype was released with the name "opensolaris developer preview". Having installed it and (briefly) attempted to get work done on it, I'm even more convinced the name was vastly premature. opensolaris. nope. it's the prototype output of a small implementation team and hasn't been through the rigor and wringer of the development process. It's a fascinating and impressive start, but.. it's not opensolaris yet, and I'm skeptical of some of the design choices (admittedly, I always am..) developer. I'm a developer so it should be for me to use? I installed it. I tried to run bugster to file a bug. Nope, no java. Tried to build some simple C programs. Nope, No C compiler, no make. Tried to add the binary nvidia driver so the display didn't look crappy -- got that working but it was an adventure -- pkgadd(1m) is busted so I had to pkgadd -R on a different system, rsync the bits over, and re-run the package's postinstall script a few times before it "took". If you want to play with the new packaging system, make sure you have another system running SXCE handy to actually do builds. And if you see something unexpected, please file bugs early and often. If you actually want to get work done, you're better off with SXDE/SXCE until further notice. preview. When I hear "preview" I think "movie sneak preview" .. getting a look at the almost-final production before they come out. SXCE & SXDE are previews. The Indiana prototype is like the raw footage viewed by the production team the day after it's shot before the bloopers have been edited out. Tim Foster's critique misses the point of the engineering objections. Attaching this name to a pre-alpha snapshot demo which is simply not even close to done is a mistake. It might make sense when Indiana is a little better baked, but IMHO, the current name attached to the current bits hurts the "brand". IMHO the name should have been held in reserve until it was ready for community endorsement.
|
Calendar
RSS Feeds
All /General /IETF /IPsec /Music /OpenSolaris /Solaris SearchLinks
Navigation |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||