Once again, my blog has moved.
Old URLs should still work (tumblr.co ones might be broken after some time).
I liked Tumblr. It made it really easy to write a short blog post. Posting quotes was a wonderful capability, and encouraged me to post somewhat more often.
The Tumblr dashboard, while never great for volume, was decreasing in utility. A sponsored tweet is one thing; a sponsored post, taking up a full screen height in my stream, is rather invasive.
So you want to have Maven put its build output somewhere out-of-tree.
There are several reasons to do this:
Helpfully, the Maven Way is to have all Maven-generated output go to a dedicated directory, target/, where it can be easily separated. Theoretically, you can probably set project.build.directory to point to wherever you want, and get Maven to build somewhere else. However, if parts of the build system assume that output goes into target/ (a questionable assumption, but I make it myself in pieces of the LensKit site generation workflow).
As he writes further on, Nancy Grace is but the ugly personification of a viewpoint that has permeated and taken over large swathes of the American consciousness: if you are arrested, you are guilty and if you are guilty, you are, by definition evil and thus deserving of the most severe of punishments and you lose your humanity.
— Yes, presumption of innocence matters. Deeply. (over at A Public Defender).
Can due process produce a result that is, in some sense, unjust? Yes. People can kill and defraud and rape and abuse but leave insufficient evidence of their crimes to prove their guilt beyond a reasonable doubt. The fact that the victim suffered is unjust. The fact that the perpetrator was not punished is unjust. The fact that skin color drives outcomes is unjust. It is unjust that moral wrongs go unredressed: such as, perhaps, the moral wrong that Trayvon Martin would be alive if George Zimmerman didn’t think he had a right and duty to confront people of the wrong color in his neighborhood. But there’s a central question some people ignore about such injustice: compared to what?
People assail results like the acquittal of George Zimmerman. But critics don’t tell us what the alternative should be. Shall guilt or innocence be determined by society’s reaction to the vapid summaries of prosecutions on cable news? Clearly not. Should verdicts necessarily reflect social consensus of the time about the crime and the accused? Tell that to the Scottsboro boys — theirs did. Should we make it easier to convict people of crimes in order to reduce injustice against the weak? How foolish. The weak already suffer because it is too easy to convict — because we love to pass criminal laws, but hate to pay for an adequate defense. Thanks to “law and order” and the War on Drugs and our puerile willingness to be terrified by politicians and the media, one-sixth of African-American men like Trayvon Martin have been in prison, trending towards one-third. The notion that we can improve their status in America by making it easier to convict people and by undermining the concept of a vigorous defense is criminally stupid. The assertion that an acquittal is wrong and unjust might, in some cases, be true, in the sense that some juries will vote their ignorance or racism or indifference. But the assertion that an acquittal is by its nature unjust because of how we feel about the case serves the state — the state that incarcerates 25% of the world’s prisoners.
— Ken White on the Zimmerman trial. He later concludes by noting that he is more afraid of the state than of the George Zimmermans of the world. The whole article is very much worth your time.
John Riedl, my Ph.D adviser, mentor, and friend, passed away this evening after a 3-year battle with cancer. If you didn’t know already, that’s what this post was about.
The world knows John as one of the inventors of collaborative filtering (go watch the re-presentation of the original GroupLens paper he wrote with Paul Resnick and others) and a leader in the field of recommender systems, as well as an influential researcher in social computing systems broadly.
For me, he is the one who taught me how to stay sane in the oft-insane world of academia.
I met John when I was assigned to be his TA for CS2 my first semester as a Ph.D student. Two things quickly stood out about him: 1, that he knew how to run an efficient meeting, and 2, that his family was a high priority. Throughout the semester I also saw him to be an excellent and thoughtful teacher.