<?xml version="1.0" encoding="utf-8"?>
<rss xmlns:atom="http://www.w3.org/2005/Atom" version="2.0"><channel><title>Disqus - Latest Comments for pboothe</title><link>http://disqus.com/by/pboothe/</link><description></description><atom:link href="http://disqus.com/pboothe/comments.rss" rel="self"></atom:link><language>en</language><lastBuildDate>Fri, 01 Feb 2013 03:17:49 -0000</lastBuildDate><item><title>Re: Greetings from Hawaii!</title><link>http://www.tracyfood.com/2013/01/31/greetings-from-hawaii/#comment-785529797</link><description>&lt;p&gt;Those tacos were SO GOOD!&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Peter Boothe</dc:creator><pubDate>Fri, 01 Feb 2013 03:17:49 -0000</pubDate></item><item><title>Re: Myth of the Lone Villain</title><link>https://kk.org/thetechnium/myth-of-the-lon/#comment-705161635</link><description>&lt;p&gt;Biological weapons seem like they could be an exception to this rule.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Peter Boothe</dc:creator><pubDate>Thu, 08 Nov 2012 23:18:52 -0000</pubDate></item><item><title>Re: Why Tarsnap doesn't use Glacier</title><link>http://www.daemonology.net/blog/2012-09-04-why-tarsnap-doesnt-use-glacier.html#comment-643800237</link><description>&lt;p&gt;Then make the leaf block size of your Merkle Trees larger!  That will reduce tree size (expensive S3 storage) at the cost of increased block size (cheaper Glacier storage).  There's a tradeoff there that only you are qualified to make, but you do have a dial and have the ability to turn it to the optimal value (if you can figure out what that value is).   Although I fully admit that there may be complexities here that moot the whole point.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Peter Boothe</dc:creator><pubDate>Fri, 07 Sep 2012 21:09:46 -0000</pubDate></item><item><title>Re: relativity</title><link>http://imprompt.us/2007/relativity/#comment-643798032</link><description>&lt;p&gt;Turns out, according to Scott Aaronson (who actually has thought about this), in order to get an exponential speedup you'll need to expend exponential energy, so its not a win.   I'm glad other people also think these weird thoughts!&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Peter Boothe</dc:creator><pubDate>Fri, 07 Sep 2012 21:06:02 -0000</pubDate></item><item><title>Re: Why Tarsnap doesn't use Glacier</title><link>http://www.daemonology.net/blog/2012-09-04-why-tarsnap-doesnt-use-glacier.html#comment-639706632</link><description>&lt;p&gt;Why not use Merkle Trees?  Store the trees in S3 and the corresponding data in Glacier.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Peter Boothe</dc:creator><pubDate>Tue, 04 Sep 2012 16:10:30 -0000</pubDate></item><item><title>Re: Why I Stopped Telling Young Girls to Go Into Engineering</title><link>http://womanintech.tumblr.com/post/19549272757#comment-468738288</link><description>&lt;p&gt;I see the grumpy haters have showed up.  Hi haters!&lt;/p&gt;&lt;p&gt;It does sounds like your work may be a problem, however.  Try a different company.  Life is too short and programming jobs too plentiful to stay in a shitty work environment.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Peter Boothe</dc:creator><pubDate>Sun, 18 Mar 2012 23:34:40 -0000</pubDate></item><item><title>Re: http://sbcoded.com/2012/01/triangle-puzzle/</title><link>http://sbcoded.com/2012/01/triangle-puzzle/#comment-412461907</link><description>&lt;p&gt;Your transformation to the dark side is now complete.  You now solve puzzles pretty much exactly the way I do.  And students will start accusing you of cheating...&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Peter Boothe</dc:creator><pubDate>Mon, 16 Jan 2012 04:17:02 -0000</pubDate></item><item><title>Re: Foto Friday: Delicious things eaten during recent travels.</title><link>http://www.tracyfood.com/2011/09/23/foto-friday-delicious-things-eaten-during-recent-travels/#comment-318367296</link><description>&lt;p&gt;This one? &lt;a href="http://www.flickr.com/photos/pmb/6071862734/in/set-72157627497181506/" rel="nofollow noopener" target="_blank" title="http://www.flickr.com/photos/pmb/6071862734/in/set-72157627497181506/"&gt;http://www.flickr.com/photo...&lt;/a&gt;&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Peter Boothe</dc:creator><pubDate>Fri, 23 Sep 2011 10:59:19 -0000</pubDate></item><item><title>Re: Foto Friday: Delicious things eaten during recent travels.</title><link>http://www.tracyfood.com/2011/09/23/foto-friday-delicious-things-eaten-during-recent-travels/#comment-318354797</link><description>&lt;p&gt;Also: we really do eat pretty well! :)&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Peter Boothe</dc:creator><pubDate>Fri, 23 Sep 2011 10:39:59 -0000</pubDate></item><item><title>Re: Foto Friday: Delicious things eaten during recent travels.</title><link>http://www.tracyfood.com/2011/09/23/foto-friday-delicious-things-eaten-during-recent-travels/#comment-318345524</link><description>&lt;p&gt;"Awesome Pastry Shop Display", IMG_0304, IMG_0329, "Big Lunch!" and "Untitled" (the last photo) are all private :(&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Peter Boothe</dc:creator><pubDate>Fri, 23 Sep 2011 10:24:17 -0000</pubDate></item><item><title>Re: http://sbcoded.com/2011/09/im-sorry-for-you/</title><link>http://sbcoded.com/2011/09/im-sorry-for-you/#comment-314026811</link><description>&lt;p&gt;Java fails as a first language because of semicolons, braces, and the fact that you need &lt;br&gt;&lt;/p&gt;&lt;pre&gt;public class MyClass {&lt;br&gt;  public static void main(String args[]) {&lt;br&gt;  }&lt;br&gt;}&lt;/pre&gt;&lt;p&gt; in a file named &lt;a href="http://MyClass.java" rel="nofollow noopener" target="_blank" title="MyClass.java"&gt;MyClass.java&lt;/a&gt; before you can write a single line of code.  It's tools are much nicer than C++'s, however.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Peter Boothe</dc:creator><pubDate>Sat, 17 Sep 2011 19:10:47 -0000</pubDate></item><item><title>Re: http://sbcoded.com/2011/09/im-sorry-for-you/</title><link>http://sbcoded.com/2011/09/im-sorry-for-you/#comment-313880375</link><description>&lt;p&gt;I took the Java pledge ( &lt;a href="http://computinged.wordpress.com/2011/03/01/improve-computing-education-take-the-more-than-java-pledge/" rel="nofollow noopener" target="_blank" title="http://computinged.wordpress.com/2011/03/01/improve-computing-education-take-the-more-than-java-pledge/"&gt;http://computinged.wordpres...&lt;/a&gt; ).  A similar pledge for C++ is unnecessary and redundant.&lt;/p&gt;&lt;p&gt;C++ error messages have (and I know how hard this is to believe) gotten SO MUCH BETTER AND MORE READABLE in the past decade.  And yet they still completely suck.  I think we should grade language quality in part by the readability of the error messages of the tools.  "tycon mismatch a' -&amp;gt; a' -&amp;gt; b' dict" fails.  So does anything put out by C++.  C is mostly okay for compiling, but the runtime is terrible.  Java is mostly okay on both counts.  Python is mostly okay on both counts, as is Ruby.  Perl is and remains a disaster. etc, etc.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Peter Boothe</dc:creator><pubDate>Sat, 17 Sep 2011 12:52:09 -0000</pubDate></item><item><title>Re: Rational</title><link>http://imprompt.us/2011/rational/#comment-275314439</link><description>&lt;p&gt;"Within the abilities" is exactly as slippery a concept as "rational" and "optimal".&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Peter Boothe</dc:creator><pubDate>Tue, 02 Aug 2011 18:31:06 -0000</pubDate></item><item><title>Re: Rational</title><link>http://imprompt.us/2011/rational/#comment-274888650</link><description>&lt;p&gt;But the first, at least, would be within my grasp.  And the second wouldn't take too long, I don't think.  The separate condenser is the tricky thing.  What tech or technique was a 15th century person lacking that an early-18th century person had?&lt;/p&gt;&lt;p&gt;The vast majority of those people were clearly doing non-optimal things...&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Peter Boothe</dc:creator><pubDate>Tue, 02 Aug 2011 14:59:30 -0000</pubDate></item><item><title>Re: Rational</title><link>http://imprompt.us/2011/rational/#comment-274884447</link><description>&lt;p&gt;You and &lt;a href="http://www.nickbostrom.com/" rel="nofollow noopener" target="_blank" title="http://www.nickbostrom.com/"&gt;Nick Bostrom&lt;/a&gt;, eh?&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Peter Boothe</dc:creator><pubDate>Tue, 02 Aug 2011 14:57:03 -0000</pubDate></item><item><title>Re: Planning horizons</title><link>http://imprompt.us/2011/planning-horizons/#comment-233547061</link><description>&lt;p&gt;The Navy has run reactors for quite some time.  They also don't go out of their way to publish safety data and success data.  I totally believe that they are better at it than the USSR was, but  more finegrained data is not available.  I think I see your point about organizations which are old and require service being good at long term planning, but I disagree with it.  I grew up a military brat and I'm just not buying the whole "long term planning" thing.  The military is old because the country is old, and one of the defining features of a country is that it has a military.&lt;/p&gt;&lt;p&gt;The Navy and Marines and Army and Air Force do what they are told, and so they can't have a planning horizon that is too much longer than their corresponding commanders in chief.    Long term thinkers would not have closed our NATO base in Keflavik, as just one example.  I'll spot them a 20 year horizon, but not a 200 year one...&lt;br&gt;&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Peter Boothe</dc:creator><pubDate>Thu, 23 Jun 2011 15:51:54 -0000</pubDate></item><item><title>Re: http://sbcoded.com/2011/06/new-should-be-private-static/</title><link>http://sbcoded.com/2011/06/new-should-be-private-static/#comment-230960141</link><description>&lt;p&gt;Python. Ruby. Lisp.  Scheme.  In Python, I have no idea whether I am calling a constructor, or just a function which returns an object.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Peter Boothe</dc:creator><pubDate>Tue, 21 Jun 2011 08:58:39 -0000</pubDate></item><item><title>Re: http://sbcoded.com/2011/06/new-should-be-private-static/</title><link>http://sbcoded.com/2011/06/new-should-be-private-static/#comment-230384519</link><description>&lt;p&gt;If memory is managed, why do we need constructors at all?  Why not just have functions, some of which return objects?&lt;/p&gt;&lt;p&gt;Although, I will agree with you that "new" is a flaw in Java - if all allocation is on the heap, then stealing an operator from C++ to indicate that allocation is on the heap is just dumb.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Peter Boothe</dc:creator><pubDate>Mon, 20 Jun 2011 12:40:40 -0000</pubDate></item><item><title>Re: Atoms</title><link>http://imprompt.us/2010/atoms/#comment-182803694</link><description>&lt;p&gt;Wes, I also have problems with the Axiom of Choice.  I am totally okay with the axiom of countable choice, however.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Peter Boothe</dc:creator><pubDate>Mon, 11 Apr 2011 20:19:23 -0000</pubDate></item><item><title>Re: Java got this wrong</title><link>http://imprompt.us/2011/java-got-this-wrong/#comment-179796610</link><description>&lt;p&gt;package protected should not even exist.  It is super-dumb.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Peter Boothe</dc:creator><pubDate>Thu, 07 Apr 2011 08:19:05 -0000</pubDate></item><item><title>Re: [SB Coded] » Should students be graded on performance?</title><link>http://sbcoded.com/2011/03/should-students-be-graded-on-performance/#comment-167462936</link><description>&lt;p&gt;Where can I get my hands on this "better than gprof" tool of which you speak?&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Peter Boothe</dc:creator><pubDate>Thu, 17 Mar 2011 23:46:40 -0000</pubDate></item><item><title>Re: [SB Coded] » Should students be graded on performance?</title><link>http://sbcoded.com/2011/03/should-students-be-graded-on-performance/#comment-166383962</link><description>&lt;p&gt;As far as how I graded your code - I was mostly happen when it worked at all.  Only then did I look into performance issues (if that was part of the assignment).  Generally I did that by handing your code large instances (sort 100 megabytes of data, etc).  If your code finished that in a reasonable time, then it was "fast enough".&lt;/p&gt;&lt;p&gt;We should add profiler to their arsenal, but profilers are the trickiest tool to use and understand.  Consider the state of the art here: debuggers are so prevalent we are using them to make art, memory checkers we can run with a command line tool (valgrind) and get a full report, and profilers almost all suck and have nigh-unreadable output.  This implies that a profiler is actually a pretty tricky thing to get right, and so, because the tools mostly suck, they are mostly unused.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Peter Boothe</dc:creator><pubDate>Tue, 15 Mar 2011 21:42:28 -0000</pubDate></item><item><title>Re: [SB Coded] » Should students be graded on performance?</title><link>http://sbcoded.com/2011/03/should-students-be-graded-on-performance/#comment-166382704</link><description>&lt;p&gt;As the mentioned professor, let me repeat: Bubblesort is NEVER the answer.&lt;/p&gt;&lt;p&gt;Unless, of course, the question is "What is a pretty sucky way of sorting?" :)&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Peter Boothe</dc:creator><pubDate>Tue, 15 Mar 2011 21:38:11 -0000</pubDate></item><item><title>Re: Forty five</title><link>http://imprompt.us/2011/forty-five/#comment-161610582</link><description>&lt;p&gt;They had to make the bubble-popping game (first .jar above).  I implemented my solution in 26 minutes in front of the Java class last term.  I offered to do the same for this class, but they complained it wasn't fair because I had done it last year and so I wasn't seeing for the first time like they were.  So I let them choose a classic arcade game, they chose Arkanoid, and then I implemented that in class in front of them in the remaining amount of class time (I actually went a few minutes over - and apologized for it).  It took me 32 minutes in front of the class to build a solid B+ project (and then I had to go to my next class).  Later that same day, I implemented a few more features in 6 minutes, for a total of 38, and made it an A project.  That is the second .jar file above.&lt;/p&gt;&lt;p&gt;They had to use Java (as did I) because the class is called "Object-Oriented Design with Java".  The first half of the class is "learn to program with Java", and the second half is "design OO programs in Java".  Both I and they used NetBeans for our coding - which you well know is kind of a handicap for vim-using me.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Peter Boothe</dc:creator><pubDate>Sun, 06 Mar 2011 22:35:03 -0000</pubDate></item><item><title>Re: Forty five</title><link>http://imprompt.us/2011/forty-five/#comment-161609049</link><description>&lt;p&gt;For CS1, sure.  For juniors and seniors, not a chance.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Peter Boothe</dc:creator><pubDate>Sun, 06 Mar 2011 22:27:45 -0000</pubDate></item></channel></rss>