Svenl is still making life painful for everyone he works with. Since leaving the kernel team (and thus not having to deal w/ him anymore), life has gotten a lot more pleasant. Instead of frustrating me, his actions now amuse me. I do feel sorry for people he has to deal with, though. Bugs like this one (#366938) are an absolute riot. Specifically, the bit about how removing his commit access to the d-i repository is somehow a violation of the GPL. Wow. There’s plenty of Debian Developers that march to the beat of their own drum; there’s plenty I disagree with. And then we have Sven, who calls his drum a horse and whines that it doesn’t whinny properly when he beats it to death. Reality distortion field, engage!

I’ve been using OpenSWAN for a little while now, and have been quite unhappy w/ it. Aside from being poorly documented and a pain in the ass, we’ve run into kernel bugs and stability issues; the various SWAN implementations feel very cobbled together. I also firmly believe that network stacks should be implemented in userspace, and linux’s IPSec stack confirms that belief (Van Jocobson’s network channels display a nice example of how an efficient userspace tcp stack would work). Layering GRE tunnels and using OSPF over the tunnels simply adds to the complexity. Playing around w/ OpenVPN was such a pleasant experience. It has excellent documentation, and was a breeze to configure/set up. I was concerned with speed issues, however; OpenSWAN uses the kernel’s IPSec stack, while OpenVPN has to copy packets between kernelspace and userspace, and uses SSL + Linux’s tun device. I experimented w/ two machines; a dual opteron and a single processor 2.4GHz p4, both w/ Debian’s 2.6.15-1-686-smp kernel. Doing a straight netcat of /dev/zero between the two machines yielded 98mbit/s (according to iptraf). Using an OpenSWAN tunnel and a GRE tunnel [0], I got about 81mbit/s sustained between the two machines. The load on the slower machine hovered around 0.30 while this was happening. Using OpenVPN in UDP mode, I got about 15mbit/s, w/ a load avg of 1.84. Very disappointing results. I was curious if using OpenVPN in TCP mode made a difference; it did. With it, I got about 17mbit/s, w/ a load avg of 2.10. This wasn’t very satisfactory; replacing OpenSWAN w/ OpenVPN when it could only do 17mbit/s over a 100mbit connection, and drove the load way up while doing it was not going to work. Looking at the config, I tried another change; disabling compression [1]. This made all the difference. With OpenVPN in UDP mode and compression disabled, I got 98mbit/s between the two machines, w/ a load avg of about 1.90. This made me much happier; looks like I’ll be replacing OpenSWAN. No need for GRE tunnels over top of the OpenVPN tunnels, either; OSPF works just fine between the tun devices. [0] There are a number of reasons to use GRE tunnels over OpenSWAN IPSec tunnels; Ken Bantoft’s Linux Symposium paper on the subject summarizes them nicely. [1] Most of the traffic that goes across our tunnel links is compressed already, anyways. No sense recompressing it.

I was listening to a DJ Shadow album, “Preemptive Strike”, and recognized one of the songs. The track “High Noon” was from (or included samples of) a Therapy? album called Lonely, Cryin’, Only. I was curious whether Therapy? had originally written the song, or whether both were covers, so I dug around Therapy?’s site a bit and discovered this gem: “Here’s an example of how strange it can be. About a month ago we got a call from DJ Shadow’s management. He’s releasing a DVD and wanted to know if we would give permission for him to include a section of our cover of ‘High Noon’, a song he wrote! Apparently he uses it in his live set and needed clearance from us. How screwed is that? He’s asking us if we allow him to use a section of a cover of a song he wrote! Bizarre.” The music industry is a strange place.

Five years ago, I’d have never thought I would be reading things like this on Kuro5hin.


16:28 < Unprompted> Hrm, can’t get to wikipedia.
16:29 < @beez> me either
16:29 < Unprompted> How am I supposed to know stuff?
16:30 < cnj> it was working earlier. I guess there’s nothing to know right now.


I’m looking for decent colocation on the US east coast; anyone have any recommendations? Ideal locations would be in Massachusetts or New York. I’m looking for someplace w/ decent prices, a decent network, and above all, decent support. I’ve had enough dealings w/ companies that lose my support tickets, have internal communcation issues, etc. Anyone have any recommendations of places that they’re happy w/?

Ever since the drive on my laptop died, I’ve been trying an experiment; I’ve been running sarge (debian 3.1) on my desktop. Of course, it’s not vanilla sarge; I’m using packages from backports.org as well. This means I’m running a gnome2.8 system w/ things like thunderbird1.5, kernel 2.6.15, and misc other things upgraded. For the most part, it’s been very successful; I haven’t really found myself wishing the software was newer. However, I would like to play around w/ gnome2.14. To that end, I’d started backporting it; you can find packages here. Unfortunately, I probably won’t finish, as I’m heading back to NYC tonight for at least 2 weeks, and thus won’t have a computer (my laptop is in for repair, my computers are all in Boston, and my office in NYC is unbearable after 10pm and on weekends). So, if someone wants to take up where I left off and upload the whole thing to bpo, go for it. While I found sarge to be quite usable w/ bpo, I don’t know how much longer I would’ve liked to use it. I look forward to AJ’s etch snapshots idea coming to fruition; I think a well motivated group of people could do an excellent job of polishing up a snapshot release (if ubuntu can do it w/ 10 people, using *unstable* as a source, all while adding experimental features like initramfs and xorg, certainly using testing as a base is doable), giving us something that would tide us over between debian stable releases.

It’s sad to see MJ Ray, our savior and future DPL, engaging in a hate campaign against AJ. It brings a tear to my eye. (Note: the above statement is sarcasm. If you don’t get it, see this, this, this, and this. See, I can claim that criticism I don’t agree w/ is “hate speech” too!)

“I think I ate your nose.” “How’d it taste?” “Chewy.”

Martin pointed out something that I’ve complained about before, and was subsequently given the reason for (by piman, iirc). It makes sense, but it’s still kind of lame for usability purposes. Of course, that’s the python interpreter, not a front-end for users. What I really miss from ruby is something like irb for python. I don’t want a full-fledged IDE; I want a quick command line thing that can be spawned for testing code snippets, or for inspecting objects. I want it to support tab completion (it’s quite useful to be able to type “foo”.<tab><tab>, and get a list of all String methods). I’ve mentioned my desire for such a thing before, but no one’s ever pointed anything out. I just did a quick google search for “irb for python”, and discovered ipython. It seems to almost do what I want; “foo”.<tab><tab> doesn’t quite work, but x=”foo” and x.</tab><tab> does work. Nifty! Of course, it seems to have its own stupid exit behavior:

In [4]: exit
Do you really want to exit ([y]/n)? y
Yes, I really want to exit; otherwise, I wouldn’t have typed “exit”…

Please choke on a bucket of cocks. What the hell are you thinking, making a *managed* switch that doesn’t have console interface? Slap a buggy web interface onto a switch and stick it on the market? Fuck your entire business model. I hope you go bankrupt.

http://pkg-ruby-extras.alioth.debian.org/upstream-devs.html. I’m getting really tired of attempting to package (or even just use) ruby things, and discovering that they only distribute (and require) RubyGems. I’m talking to you, collaboa. You lost a developer to Trac by making your software a pain in the ass to install.

I don’t remember where I stumbled onto this from: http://www.progressiveboink.com/archive/calvinhobbes.htm I love c&h.

My new key made it into the debian keyring (thanks Elmo!), so I’ve been quietly doing some overdue updates to my packages. I even picked up a new one (bzr), because I like pain. If you’re waiting for me to update one of my packages, don’t worry… I’m getting to it.

When I returned home to my apartment today, I noticed the front door was open. I heard voices upstairs, so I didn’t close it. Later on, I heard the sounds of walkie talkies in the hall, so I poked my head out of the door to see what was going on. Apparently, someone in my apartment had been robbed; cops were in the hall, talking to the lady that lives on the second floor. However, while the thieves were carrying a TV outside and loading it into a car, apparently some cops happened to notice. There was some form of a chase, and one person was caught while the other got away. I checked the outside of the apartment building, but didn’t see any “Break into me! I have lots of valuables!” signs posted anywhere. Oh well, maybe there’s something else that makes this place a huge target for thieves. Or perhaps it’s just the neighborhood.

I playted ping pong last night at a bar; I haven’t played since high school. I still suck at it. However, the guy at the bar who I hit in the head with a ping pong ball last night didn’t kick my ass. I’m sure that says something positive about my playing skills.

http://news.bbc.co.uk/2/hi/health/4607642.stm I fully support breasts and breast health.

A coworker and I were discussing whether we should eventually switch our company servers to Etch or Dapper. As Andreas Barth recently sent out a release announcement with a schedule prediction and goal list, I pointed him to it, and then followed up w/ my own predictions. Looking through the goals, it looks as though the easy stuff has been taken care of while the difficult things remain. This is not to discount the work done by the gcc and X people; I don’t mean technically difficult, I mean the things that have traditionally been difficult for Debian to accomplish. It’s more a question of manpower and motivation; there are plenty of people who want the latest software in Debian, and those people can also complete such work, so it tends to get done fairly quickly. On the other hand, the remaining things in the TODO list are the sorts of things that are not very much fun, or can only be done by a select few people. Point-by-point:
- amd64 as an official arch (and the mirror split as a pre-condition
for that)
Infrastructure work is not particularly fun, and can only be done by a few people. It tends to take a while, as we’ve seen in the past with things like the security queues, which blocked the release of Sarge for a decent amount of time. - sorting out docs-in-main vs. the DFSG
- sorting out non-free firmware
These two will not be fun, and will first require actual decisions about what should be done in each case, followed by actions that a lot of people will object to (that is, the stripping out of documentation from various packages, and the removal of functionality from various kernel drivers; hell, enough people complained when we removed tg3 firmware, imagine how many will complain when a couple hundred kernel drivers get their firmware removed?). The actual work here will require people dedicated to actually seeing a release put out, and will probably result in much unrest. I can’t imagine it will actually go smoothly. I guess we’ll see what happens. With a year to go still, there are a number of ways it could go.
(15:23:24) Andres Salomon: if etch actually releases on that day, i will shit myself
(15:23:40) Wes Chow: should we shake on it?
(15:24:02) Wes Chow: I think you should announce that on the debian dev list and see who takes you up on the offer
(15:24:13) Wes Chow: maybe people will actually push for a timely release just to see that happen!
(15:24:25) Wes Chow: and then you soiling yourself would be in the name of free software


I don’t particularly want to add features to xmpp4r, simply because it is pure ruby. I would prefer to add features to a C library, which can then be used in various language bindings. Lowest common denominator, and all that. I also prefer the license; I’m a fan of the LGPL for these sorts of libraries. I started writing some ruby bindings for libloudmouth, which seems to do some of what I want (but at a much lower level than really desired. Via planet.jabber, I noticed the author intends to write a higher level API for loudmouth 2.x, but there is no code available yet. I may end up writing a separate library/layer to provide a higher layer API, if none becomes available anytime soon.

squishy.cc expires in a month. I had planned to renew it, but $20/yr for a .cc domain via godaddy doesn’t really seem worth it (compared to the price of .net/org/com TLDs, that is). Hm…