<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0"><channel><title>Disqus - Friends of royce</title><link>http://disqus.com/people/royce/</link><description></description><language>en</language><lastBuildDate>Sat, 07 Nov 2009 08:51:03 -0000</lastBuildDate><item><title>Re: Looking back at 100 blog posts</title><link>http://www.daemonology.net/blog/2009-10-15-100-blog-posts.html#comment-22126645</link><description>I was counting the "(X subscribers)" UA strings when I added up my RSS feed subscribers.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">cperciva</dc:creator><pubDate>Sat, 07 Nov 2009 08:51:03 -0000</pubDate></item><item><title>Re: FreeBSD Update to 8.0-BETA1</title><link>http://www.daemonology.net/blog/2009-07-11-freebsd-update-to-8.0-beta1.html#comment-20582847</link><description>Yes.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">cperciva</dc:creator><pubDate>Mon, 19 Oct 2009 17:45:37 -0000</pubDate></item><item><title>Re: Looking back at 100 blog posts</title><link>http://www.daemonology.net/blog/2009-10-15-100-blog-posts.html#comment-20253758</link><description>Nate,&lt;br&gt;&lt;br&gt;I wasn't just thinking about you when I made that remark -- compared to some other people I've encountered, you're quite moderate in the world of crypto-is-scary-don't-go-anywhere-near-it.&lt;br&gt;&lt;br&gt;In the end it comes down to weighing dangers.  Yes, there is a possibility that my 'cryptographic right answers' post will give someone an unwarranted sense of confidence -- but there's also a possibility that it will lead someone to realize that they shouldn't be using blowfish for encryption; that they shouldn't use MD5 as a key derivation function; that they shouldn't use SHA256(key || data) as an MAC function; et cetera.  You can't teach someone to paint by showing them examples of bad painting -- at some point it's necessary to give people good examples, too.&lt;br&gt;&lt;br&gt;Thanks for stopping by.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">cperciva</dc:creator><pubDate>Sat, 17 Oct 2009 03:37:04 -0000</pubDate></item><item><title>Re: They're so cute!</title><link>http://www.percival-music.ca/blog/2009-10-15-sooo-cute.html#comment-20252555</link><description>... ok, so disqus converts HTML entities into literals too?  Wow that's broken.&lt;br&gt;&lt;br&gt;There's a [less than sign] "blockquote" [greater than sign] in the above two comments.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">cperciva</dc:creator><pubDate>Sat, 17 Oct 2009 02:19:58 -0000</pubDate></item><item><title>Re: They're so cute!</title><link>http://www.percival-music.ca/blog/2009-10-15-sooo-cute.html#comment-20252522</link><description>Err, and disqus apparently allows some HTML tags through and others not.  There's a &lt;blockquote&gt; in that comment.&lt;/blockquote&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">cperciva</dc:creator><pubDate>Sat, 17 Oct 2009 02:18:24 -0000</pubDate></item><item><title>Re: Looking back at 100 blog posts</title><link>http://www.daemonology.net/blog/2009-10-15-100-blog-posts.html#comment-20149695</link><description>Tarsnap moved into public beta in November 2008: &lt;a href="http://www.daemonology.net/blog/2008-11-10-tarsnap-public-beta.html" rel="nofollow"&gt;http://www.daemonology.net/blog/2008-11-10-tars...&lt;/a&gt;&lt;br&gt;&lt;br&gt;Details about tarsnap are available on the tarsnap website -- if you want to know more, feel free to send me an email.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">cperciva</dc:creator><pubDate>Thu, 15 Oct 2009 16:33:20 -0000</pubDate></item><item><title>Re: Securing an HTTPS server</title><link>http://www.daemonology.net/blog/2009-09-28-securing-https.html#comment-17771020</link><description>Quite true.  In my case I don't use source IP addresses for authentication purposes -- it's far too easy for that to break -- but it's certainly something people should be aware of.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">cperciva</dc:creator><pubDate>Tue, 29 Sep 2009 10:09:13 -0000</pubDate></item><item><title>Re: Securing an HTTPS server</title><link>http://www.daemonology.net/blog/2009-09-28-securing-https.html#comment-17746075</link><description>Having outbound email from web servers is good -- otherwise you can't get email from your cron jobs.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">cperciva</dc:creator><pubDate>Mon, 28 Sep 2009 19:08:38 -0000</pubDate></item><item><title>Re: Securing an HTTPS server</title><link>http://www.daemonology.net/blog/2009-09-28-securing-https.html#comment-17746055</link><description>192.168.0.44 is a non-routable IP address which I created on a virtual interface.  The nameserver is authoritative; I haven't set up AXFR.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">cperciva</dc:creator><pubDate>Mon, 28 Sep 2009 19:07:57 -0000</pubDate></item><item><title>Re: Securing an HTTPS server</title><link>http://www.daemonology.net/blog/2009-09-28-securing-https.html#comment-17715599</link><description>That too. :-)</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">cperciva</dc:creator><pubDate>Mon, 28 Sep 2009 09:53:45 -0000</pubDate></item><item><title>Re: Securing an HTTPS server</title><link>http://www.daemonology.net/blog/2009-09-28-securing-https.html#comment-17710926</link><description>There are patches for stunnel to make it insert X-Forwarded-For headers; but to do that stunnel needs to do some basic parsing of HTTP connections, which adds significant complexity -- so I'd prefer to avoid that route.  You're quite right that this is an option, though.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">cperciva</dc:creator><pubDate>Mon, 28 Sep 2009 08:20:32 -0000</pubDate></item><item><title>Re: Apache memory usage bogosity</title><link>http://www.daemonology.net/blog/2006-05-26-apache-memory-usage.html#comment-17028779</link><description>I used /proc/X/map to look at the memory maps, and then /proc/X/mem to look at what it was being used for.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">cperciva</dc:creator><pubDate>Mon, 21 Sep 2009 05:03:45 -0000</pubDate></item><item><title>Re: Cryptographic Right Answers</title><link>http://www.daemonology.net/blog/2009-06-11-cryptographic-right-answers.html#comment-16309432</link><description>The advice about avoiding static keys is because CTR mode requires that messages encrypted with the same key must have different nonces.  In most applications you can simply use nonces of 0, 1, 2, ... and keep track of the largest counter value you've used; problems only arise if (e.g., in hardware systems) you have ROM and RAM but no writable storage which persists across power outages.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">cperciva</dc:creator><pubDate>Thu, 10 Sep 2009 12:11:26 -0000</pubDate></item><item><title>Re: Complexity is insecurity</title><link>http://www.daemonology.net/blog/2009-09-04-complexity-is-insecurity.html#comment-16275436</link><description>By 'cryptologically "useless" MAC' I mean that the data in question was already being verified post-decompression, so theoretically any corruption would be found that way, but I added a MAC on the compressed data.  Look for CRYPTO_KEY_HMAC_FILE in the tarsnap client source code.&lt;br&gt;&lt;br&gt;The evidence I have for bugs being in rarely-executed code paths is purely anecdotal -- but it's a consistent theme I've seen in my time as FreeBSD Security Officer, both in terms of vulnerabilities in FreeBSD and vulnerabilities in other software.&lt;br&gt;&lt;br&gt;No, the tarsnap server code is not public.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">cperciva</dc:creator><pubDate>Wed, 09 Sep 2009 18:07:30 -0000</pubDate></item><item><title>Re: Complexity is insecurity</title><link>http://www.daemonology.net/blog/2009-09-04-complexity-is-insecurity.html#comment-16199834</link><description>The world definitely needs a better cryptographic library; but I'm not sure that cryptlib or libtomcrypt really fit the bill -- cryptlib is commercial, and libtomcrypt doesn't seem to be maintained any more.&lt;br&gt;&lt;br&gt;For tarsnap I'll probably replace openssl by implementing the cryptographic primitives myself; tarsnap only uses a very small fraction of openssl, and I've implemented almost all of the relevant functions at some point in the past.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">cperciva</dc:creator><pubDate>Tue, 08 Sep 2009 13:08:02 -0000</pubDate></item><item><title>Re: Interesting tarsnap statistics</title><link>http://www.daemonology.net/blog/2009-08-21-interesting-tarsnap-statistics.html#comment-16062120</link><description>I don't know -- I can see how much data people have stored, and I can see how much data is stored while creating each archive; but since tarsnap avoids storing duplicate data, I have no way of knowing how much previously stored data is reused in an archive.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">cperciva</dc:creator><pubDate>Sun, 06 Sep 2009 11:26:55 -0000</pubDate></item><item><title>Re: Tweet from Tumblr
 We’re testing Twitter... | Tumblr Staff</title><link>http://staff.tumblr.com/post/75941045#comment-15951774</link><description>Yeah, ditto here. I posted at the end of the thread before I saw these posts here. I'd love to post to a secondary tumblr blog, as well. Maybe be able to define a twitter account per tumblr blog?</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">roblef</dc:creator><pubDate>Thu, 03 Sep 2009 22:20:20 -0000</pubDate></item><item><title>Re: Hello Alaska WordPress Users!</title><link>http://www.akwpug.org/?p=1#comment-15695596</link><description>yeah, that's the link I used before. i got all of the config files modified, i thought, and added the sql commands to the wp tables and etc. still borked it. You're welcome to try. email &lt;a href="mailto:roblef@gmail.com" rel="nofollow"&gt;roblef@gmail.com&lt;/a&gt; and I'll set you up an ftp account and get you db privs.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">roblef</dc:creator><pubDate>Tue, 01 Sep 2009 01:20:32 -0000</pubDate></item><item><title>Re: Forums vs Disqus vs Other</title><link>http://www.akwpug.org/?p=8#comment-15504514</link><description>Hahaha!&lt;br&gt;&lt;br&gt;We've left it this way on purpose. We're all designers and developers, too.&lt;br&gt;:)</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">roblef</dc:creator><pubDate>Thu, 27 Aug 2009 20:34:03 -0000</pubDate></item><item><title>Re: Interesting tarsnap statistics</title><link>http://www.daemonology.net/blog/2009-08-21-interesting-tarsnap-statistics.html#comment-15450759</link><description>Nope.  The tarsnap client stores a series of blocks, but the encryption makes it impossible for me to know whether a block contains several files, one file, or part of a file.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">cperciva</dc:creator><pubDate>Wed, 26 Aug 2009 22:48:28 -0000</pubDate></item><item><title>Re: Forums vs Disqus vs Other</title><link>http://www.akwpug.org/?p=8#comment-15448339</link><description>For me, the value is having local folks in the trenches who can get together&lt;br&gt;and talk about the stuff that's on other sites and forums. I don't find the&lt;br&gt;other sites and forums to meet all my needs, and a local group usually is&lt;br&gt;more fun. For me, that is. :)</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">roblef</dc:creator><pubDate>Wed, 26 Aug 2009 21:53:02 -0000</pubDate></item><item><title>Re: Hello Alaska WordPress Users!</title><link>http://www.akwpug.org/?p=1#comment-15448317</link><description>I guess the only value I see is a way for us to connect locally, rather than&lt;br&gt;nationally. So, basically, the same value a local user group provides,&lt;br&gt;spread to the online world.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">roblef</dc:creator><pubDate>Wed, 26 Aug 2009 21:52:19 -0000</pubDate></item><item><title>Re: FreeBSD major version upgrades</title><link>http://www.daemonology.net/blog/2007-11-11-freebsd-major-version-upgrade.html#comment-15410299</link><description>I can't see any reason offhand why this process *wouldn't* work for migrating a system from i386 to amd64... but I'm not sure that I'd want to try it.  Let me know if you try it and find that it works, though.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">cperciva</dc:creator><pubDate>Wed, 26 Aug 2009 04:46:09 -0000</pubDate></item><item><title>Re: Hello Alaska WordPress Users!</title><link>http://www.akwpug.org/?p=1#comment-15337762</link><description>Cool. Glad you're here. I guess we need to see how folks would like to organize. Real life meetings? Online Discussions? Forums? Membership/Fees? Dunno.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">roblef</dc:creator><pubDate>Mon, 24 Aug 2009 22:42:23 -0000</pubDate></item><item><title>Re: Interesting tarsnap statistics</title><link>http://www.daemonology.net/blog/2009-08-21-interesting-tarsnap-statistics.html#comment-15269725</link><description>I'm not sure exactly what you mean here -- can you elaborate?</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">cperciva</dc:creator><pubDate>Sun, 23 Aug 2009 15:55:14 -0000</pubDate></item></channel></rss>