2010-05-01: Vacation
Hello from the first full day of vacation!
The picture is of the Manuka or New Zealand Tea Tree in my mother's back
yard. She had been bemoaning the difficulty of taking pictures of it, so
I tried a few different angles and this is the one that I was the happiest
with. It's basically a giant bush covered in pink flowers, and at this
time of year, in bees. It puts on a nice display every year.
Yesterday was the day of travel heading on vacation, which was another
wonderful example of how lovely and civilized travel by train is. I dread
having to fly anywhere, but being able to curl up in a train seat with a
good book and watch the world go by, and then get up and get food whenever
I feel like it, is a pure pleasure.
I'm now almost caught up on book reviews except for the book I finished
today. I brought a large pile of books with me and plan on reading until
I feel unwound from recent events at work (organizational chaos, my former
manager being let go for dubious reasons, an upgrade that's failed a
couple of times and that I still need to do when I get back, and lots of
frustrating argument about technical strategies). Then, I'll either just
keep reading or poke at Debian and free software work.
It's lovely to have a whole week during which I can do anything I feel
like doing.
Some other recent things of note:
Several more weeks of Java notes are now up on my
Java notes page. I'm about
two-thirds of the way through Thinking in Java and getting
anxious to start writing programs.
I added some more information to my
programming language
background page for people who want to evaluate my opinions about
programming languages in context.
My AFS DNS SRV record draft was published as
RFC 5864 as a
Proposed Standard. I'm rather proud of that; it's my first published
RFC where I was the sole author.
2010-05-02: remctl 2.16
The primary change in this release of remctl is the addition of Ruby
bindings, contributed by Anthony M. Martinez. I added a test suite,
hopefully added compatibility with Ruby 1.9.1, and messed with the coding
style a bit; hopefully I didn't introduce any bugs. Ruby's method of
building add-on modules, as with Perl, Python, and PHP, fixes various
problems seen in other languages at the cost of failing spectacularly in
some new and different way.
This version also adds pcre:* and regex:* ACL types for remctld, based on
work by Anton Lundin. pcre:* is only available if the PCRE libraries are
available, but let you use full Perl-compatible regular expressions. The
regex:* ACL type is available on any system with POSIX regex libraries,
but only supports POSIX regexes.
remctld also now sets the environment variable REMCTL_COMMAND to the
command that causes a program to be run, thanks to Thomas L. Kula.
There are also a few other bug fixes and improvements, including a general
update to the latest rra-c-util and
C TAP Harness.
You can get the latest version from the remctl distribution page.
2010-05-02: A little productivity
When I'm here on vacation, I of course have to take pictures of my
mother's roses. So here's the first of what will eventually be several.
If there's one thing that I most wish that my current camera has, it's
better handling of color. I don't think it knows how to devote enough
colorspace to a particular color, and of course it's not a good enough
camera to shoot RAW (which I suspect would help). So shooting anything
with a lot of saturated color tends to result in JPEG or camera artifacts,
as you can see in this one.
I poked around on-line this morning and then went and read and finished
another book. That's lovely. That's something I rarely take the time to
do during the day at home, since weekends are already so crowded with the
things I can't do during the week. Then I felt like doing something a bit
productive, so got the remctl release out that I'd been working on for the
past couple of weeks, which promptly cured that feeling. I don't think
I'm going to be getting much done for at least a few more days of
vacation; I'm still way low on energy for doing anything productive.
A review will be coming, but I can say now that series readers will be
happy with Brust's Iorich. I just wish he could write books in the
Vlad series faster!
2010-05-03: Still recovering
Another in the long list of reasons why I like graffiti in general and
think it's often an improvement over what was there without it. This
"smiley" was on one of those boring green utility boxes at Transfer Beach
in Ladysmith, British Columbia.
This also sums up my mood today. I cannot tell you how happy I was to not
be at work and to be able to completely ignore work. I came over to the
laptop periodically during the day and looked through stuff on-line and
felt profoundly uninterested in doing any of it, as well as feeling that
"OMG, I have way too much to do" feeling. So I just went back to my book
instead.
I'm guessing somewhere around Wednesday I'm going to feel more recharged.
I was very drained and more in need of vacation than I usually am.
2010-05-04: Thoughts on attention
I felt like a season-appropriate picture, so here's a picture I took
almost exactly two years ago while wandering around Stanford. I love the
Green Earth Sciences building, particularly for photography. There are so
many cool things to capture, particularly the plants that grow around the
outside of it.
I'm starting to feel my brain regenerate from vacation. I can tell
because I ended up doing a couple new versions of the OpenAFS packages
today after finishing another book, and I'm starting to have the pondering
"how do I want to tweak my life" sorts of thoughts that I have most
productively after some unwinding time during vacation.
One of the things I keep struggling with is that I have a ton of things I
want to do and not all of them are going to get done. I focus so strongly
on completing things that this is rough. I don't know how to let things
go that won't ever happen, so I'm always struggling with ever-growing
to-do lists. (All this is background material; this won't come as a
surprise to anyone who knows me.) That leads to stress, but it also leads
to decision paralysis. There are too many things on my to-do list, and
the first time I look it over, they all need to get done and I immediately
start feeling guilty about more things than I can do immediately.
This is, I think, one of the major flaws in the
Getting Things Done
system. David Allen really doesn't like priority systems or daily A/B/C
lists because he feels like things change too much during the day, but
working from my full next-action list frequently feels overwhelming and
guilt-inducing. For the week before I went on vacation, I started trying
to use GTD in a hybrid with the Cycle System from
Time Management for
System Administrators: taking some time every morning to pull things out
of my large to-do list, prioritize a bit for the day using a simple A/B/C
notation, and feel like I have a doable chunk for that day. I'm going to
do this even more seriously when I get back, since it lets me finish a
list during the day and helps me convince myself that I'm usefully using
each day. It lets me focus my attention on the productive parts rather
than the guilt-inducing parts.
The other half of this, though, is to reduce the number of things I'm
committing to do, since I'm committing to do too much. There are hard and
easy parts of this, and I'm still struggling with the hard parts. But one
of the easy parts is that I expose myself to distractions all the time,
and a lot of those are unnecessary. I've basically not watched TV during
this vacation and I've not missed it. I'm pondering seriously reducing
the TV I watch when I get home and consciously giving up committments to
watch certain shows. I also spend a lot of time reading news sites and
staying aware of current events, and I'm not sure that's a useful use of
my time. By and large, I'm never going to do anything about that data and
it just makes me mad, so why am I exposing myself to it? I should be
generally aware of world events, but I don't need to be reading detailed
comments (particularly from the sort of people who tend to comment).
And the largest key is to just do something. Make a conscious choice to
do something specific, whether off my to-do list or to get up and take a
walk or go read, but make a conscious choice. A lot of that time-wasting,
attention-wasting activity like poking around on the web comes from not
making a conscious decision and falling into default behavior. That's
almost always unsatisfying; I could almost always be doing something I'd
enjoy more, even if that means doing something purely recreational.
Having the energy to always make those choices conscious is, I think, the
main thing I need to be working on.
2010-05-05: git-pbuilder 1.10
This is a minor update of the glue script that I use between
git-buildpackage and pbuilder (with cowbuilder). The primary feature it
has over other versions of a similar script is that one can set an
environment variable to specify the chroot in which to build.
This update changes the -i regex to include all of the Git-relevant
parts of the standard dpkg-source exclusion regex. Before, the regex
tried to exclude only .git directories, but also used insufficient
quoting and ended up excluding any file containing git. That
should be fixed, and now it should exclude .gitattributes files as
well. Thanks to Allard Hoeve for the bug report.
You can get the latest version from my miscellaneous scripts page.
2010-05-09: The undead copyright myths
I spend so much time seeing discussions about copyright law (due to lots
of work around free software) that I'm constantly surprised at the myths
that keep resurfacing in other discussions among people who don't deal
with the edges of copyright as much as free software developers do.
Note: Somewhat US-specific, although not entirely so.
Trademarks have to be defended or can be lost because the point of a
trademark is that it distinguishes your product from your competitors
in a marketplace. Trademarks are specific to commercial endeavors and
a market, and trademark law is about avoiding confusion (intentional
or unintentional). Copyright is not specific to commercial endeavors
or related cases of brand identification (service marks, for
instance), and does not have to be defended.
(It may be possible, in specific circumstances, to end up not being
able to enforce your rights because you didn't press a claim. See the
defense of laches. But this is an affirmative defense related to putting the
defending party in a bad situation by waiting to enforce your claim,
and is far from the situation with trademark where active enforcement
is required. Laches is intended for cases where someone implies you
have a right to do something, you invest a million dollars into a
business plan based on that belief, and then they show up at the last
minute saying "aha!" and sue you. It's a protection against bait and
switch, in other words, and therefore is not that applicable to the
typical copyright case.)
Pre-Berne copyright laws are weird. If whatever special case of
copyright you're basing your argument on happened before 1988 when the
United States adopted Berne (realistically, more like the 1970s), it
probably doesn't apply. If it involves copyright renewal or losing
copyright because it wasn't registered properly, it has nothing to do
with modern intellectual property law.
You don't know what is fair use. That's because no one knows what
is fair use. There isn't much case law, the case law is often
contradictory, and lawyers err on the side of caution, not on the side
of allowing as much as possible. No one really knows the truth of
statements like "fanfic is / is not fair use." There's some evidence
both ways, and it's not settled. It will probably depend, more than
anyone expects, on the specific details of the specific work. Sadly,
it will also probably depend somewhat on which judge one gets, at
least until something gets appealed all the way to high courts in
multiple countries (which is unlikely given the tiny amounts of money
normally available to at least one and usually both of the parties).
There's a lot of misinformation about this stuff. See, for example,
George R.R. Martin on fan
fiction. He's someone with significant experience in both film and
writing, including professional writer's organizations. And yet he made
all of those errors to varying extents.
I know this stuff because of my involvement in free software, where we
deal with edge cases of copyright law all the time in the process of
making it do neat tricks for us. Commercial software developers often
come across as appallingly ignorant about intellectual property, probably
because they stay in the center of it surrounded by conservative lawyers
and have never had reason to explore the edges. I suppose the same thing
is likely true of fanfic writers and commercial writers: the average
fanfic writer probably knows a lot more about the ins and outs of
copyright law and fair use, although the average commercial writer knows
much more about book contracts and royalty payments.
2010-05-16: Hugo nominee haul II
Looks like it's going to be three book orders total to pick up all the
Hugo nominees this year, since I'm going to get Julian Comstock
when it comes out in mass market.
Paolo Bacigalupi — The Windup Girl (sff)
Elizabeth Bear — Bone and Jewel Creatures (sff)
Mark Forster — Do It Tomorrow (non-fiction)
Jane Jacobs — Cities and the Wealth of Nations
(non-fiction)
Guy Gavriel Kay — Under Heaven (sff)
J.A. Pitts — Black Blade Blues (sff)
The Windup Girl just won the Nebula award for best novel. It's the
Hugo nominee in this lot. I've already read
The City & The City,
Boneshaker, and
Palimpsest (review pending on the last), and I have Wake
sitting by my bedside. I'm not sure if I'm going to finish it before I go
to a conference, and it's probably not good airplane reading material (I
suspect I'll work on Ilium for that).
I've stopped going to Baycon, but I still generally try to read all the
Hugo nominees by about that time. Looks like I won't make it this year,
but I will have read about half of them, maybe one more than that.
I'm nearly done with Do It Tomorrow (in fact, I'm writing this
journal post partly as an exercise from that book) and love it. It
directly deals with what I think is the biggest flaw in
Getting Things Done
and looks like it's going to be extremely helpful. There will be a much
longer review coming later.
2010-05-22: California primaries
Voting time again, and as I normally do, here's a record of who and what
I'm voting for and why, in the hope that it will be useful to others.
Propositions:
Proposition 13: YES. We're only voting on this because
of the dumb rules around taxes in this state. It's a straightforward
cleanup of the property tax codes around earthquake safety improvements
that basically simplifies and merges existing exclusions for property tax
reassessment and makes the law simpler. In a sane legal system, the state
assembly would have been able to just pass this. No one is even bothering
to argue against it.
Proposition 14: YES. This is the reason to vote in this
primary. Even if you don't care about anything else in the elction, go
vote for this proposition.
What this proposition does is turn the California election system into a
runoff system. The primaries, rather than being a selection process
internal to each party (and hence often tending to select extreme
candidates, particularly for Republicans in California), will become the
first round of the general election. All candidates from all parties will
compete together. The top two vote-getters, regardless of party, will
then be on the final ballot in the November election.
I of course would love to see an even more sophisticated voting system,
such as proportional representation or Condorcet, but this is a huge
advance over the one-shot plurality system we have now. For tight
districts balanced between Republican and Democrat, there will be no
significant change; November will still see an election between the
Republican and the Democrat, just with the additional clarity of having
all the no-hope candidates taken off the ballot after they lose in the
primaries. But in districts that are very liberal or very conservative,
where the "other" party has no real chance, we'll instead have the
opportunity to see two Democrats on the final ballot, perhaps one moderate
and one more liberal. Or two Republicans, one moderate and one more
conservative. In other words, the two largest groups in each district
will have the opportunity to fight it out in the general election for who
will represent that district, instead of having the election artificially
forced into Democrat vs. Republican lines.
Vote for this. It's as close as you can get to an all-around win in
politics. It doesn't favor either party; instead, it improves the voting
system to make it more representative and more democratic in the true
meaning of the term. It makes it more likely that the representative
elected will be close to the prevailing opinion of the district in which
they're elected. The only people who can possibly gain from this
proposition failing are the moribund leadership of the two big parties,
who enjoy and manipulate the current system and who are afraid of possibly
seeing final election ballots between a Democrat and a Green or a
Republican and a Tea Party candidate, because they want elections to be
forced into the Democrat vs. Republican dichotomy instead of reflecting
the actual splits within the district.
If you want electoral reform, vote for this proposition. It's the best
and most practical voting reform we're likely to ever see on the ballot.
Proposition 15: YES. I'm in general very skeptical of
public funding of elections in the United States, mostly because any
voluntary system will not effectively remove lobbyist money from
elections. As soon as there's enough money at stake, the candidate will
just opt out of the system.
However, I think that this proposition might have some minor positive
effects in some situations, particularly involving smaller races, and I
think it's an experiment we have to continue to try. I also see no
effective drawback. The public funding is coming entirely from new taxes
on lobbyists, and I'm happy to see them pay way, way more than the taxes
levied by this proposition. The endorsements are also impressive, from
the AARP to the League of Women Voters, and the arguments against the
proposition printed in the voter information guide are basically lies.
Proposition 16: NO. You're kidding me, right? It's
been a while since I've seen such a blatant attempt by a corporation to
pass a law designed to limit the competition they have to face. What this
proposition basically says is that local communities aren't allowed to
switch away from PG&E to a less stupid, less mis-managed, and less corrupt
way of handling power generation. As a former PG&E customer and a happy
customer of exactly the sort of municipal utility that this proposition
would make considerably more difficult, fuck PG&E. Vote no.
Proposition 17: NO. Because of course what we need in
the middle of a recession is another way to screw over poor people.
These laws are either written by people who are outright evil, or they're
written by people who have never known anyone who struggled to make ends
meet in their life. I have a friend who had to drop auto insurance for
financial reasons (and then struggle to get to work without driving) until
he could save enough money to afford the premium. Proposition 17 would
have made that premium even higher because there was a period when he
couldn't afford it. This is bullshit. The mild decline that some of us
may receive in insurance rates isn't worth the blatant unfairness.
Statewide office:
I'm registered decline-to-state, so I get a non-partisan ballot, which may
be substantially different than what other people see. (Vote for
Proposition 14!) Only listing primaries where there's some hope of a
contested race.
Lieutenant Governor: Gavin Newsom. He's not someone I
particularly like at a personal level (I think he's a publicity-seeking
grandstander), but he's still about the most progressive person who has
any chance of getting elected to state-wide office, and I'm happy to
support that.
Attorney General: Kamala Harris. My primary goal in
voting for attorney general is to try to find someone who's not "tough on
crime" and instead understands that crime is a harder problem than locking
people up. Harris seems like the best of a fairly uninspiring lot, among
the candidates who look likely to win. One can certainly do worse than a
black woman who grew up in Oakland when looking for someone who
understands the complexity of crime problems.
Insurance Commissioner: Dave Jones. Both candidates in
the primary look solid. I'll go with the one with the Chronicle
and Mercury News endorsements.
Local offices:
State Assembly (District 21): Josh Becker. Rich Gordon
is getting the general endorsements because he has more practical
experience, but given how messed up state government is right now, I'd
rather elect someone a bit different.
Judge of the Superior Court (Office 7): Thomas
Spielbauer. As previously mentioned here, I lean towards voting against
all prosecutors for judges. I'm even more comfortable with that decision
here, based on the candidate statements. Spielbauer sounds like exactly
ths sort of judge we could use more of: concerned with fairness, concerned
with citizen rights, and not inclined to defer to prosecutors.
Judge of the Superior Court (Office 11): Vanessa
A. Zecher. A prosecutor versus a former district attorney. Eh. Voting
for the one who isn't currently a prosecutor.
Judge of the Superior Court (Office 11): Bob Camors.
Voting for the non-prosecutor and the person who states in their position
statement that they want rehabilitiation for non-violent drug offenders.
2010-05-30: Back home
I spent the last week at the AFS and Kerberos Best Practices Workshop, this year held in
Urbana-Champaign at the University of Illinois. It was a great venue for
the conference and very energizing to spend a week talking with people and
getting to concentrate only on AFS. In between presentations, I got a ton
of AFS work done, particularly on documentation. A lot is going on in the
AFS world right now, with development happening as fast as I think it ever
has in AFS's development.
The downside is that I spent all of last week being socially "on," pretty
much constantly interacting with people, so I'm completely exhausted.
It's great fun, but I always pay the price for it later. Thankfully, I've
had this weekend to relax. Tomorrow, I'm meeting up with friends for the
afternoon, but I have the holiday morning to myself. Then it's back to
work, but with new time management techniques to apply that I'm feeling
excited about (from Do It Tomorrow — review
forthcoming).
A couple of new books came in while I was gone:
Algis Budrys — Benchmarks (non-fiction)
Pamela Dean — The Whim of the Dragon (sff)
I finished Wake by Robert Sawyer on the plane and am more than
halfway through Ilium. I still need to get a copy of Julien
Comstock to get the last Hugo nominee, but I think I'm going to wait
until I can put the new Jacqueline Carey novel in the same order.
Reading is one thing that's worked very well so far this year,
particularly in the past month. My vacation at the beginning of May got
me back into the rhythm of it, and I'm really enjoying the additional
inflow of creativity and things to think about. It's definitely worth it
to make time for reading nearly every day.
All posts