<?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 Ahruman</title><link>http://disqus.com/by/Ahruman/</link><description></description><atom:link href="http://disqus.com/Ahruman/comments.rss" rel="self"></atom:link><language>en</language><lastBuildDate>Fri, 06 Mar 2015 23:12:42 -0000</lastBuildDate><item><title>Re: NSString hash is Bad</title><link>http://www.abakia.de/blog/2012/12/05/nsstring-hash-is-bad/#comment-1893117940</link><description>&lt;p&gt;This is entirely normal and expected behaviour, in line with the documentation for [NSObject hash]. -hash is not a secure digest, it’s for use in hash tables, where there is a tradeoff between speed and uniqueness.&lt;/p&gt;&lt;p&gt;Your reading of the documentation is wrong. -hash will return the same value for equal strings. It doesn’t say that -hash will _only_ return the same value for equal strings, just that equal strings will always have the same hash. Non-equal strings may or may not have the same hash. No hash algorithm can give guarantees stronger than this, and a 32-bit hash value is too small for collisions to be highly unlikely.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Ahruman</dc:creator><pubDate>Fri, 06 Mar 2015 23:12:42 -0000</pubDate></item><item><title>Re: Factbook: Weapon Production - Wolfire Games Blog</title><link>http://blog.wolfire.com/2014/04/Factbook-Weapon-Production#comment-1330451628</link><description>&lt;p&gt;I understand Aubrey likes this kind of brush, but it makes me feel like my astigmatism is suddenly much worse, which is distressing and actually quite painful. I can’t watch more than a minute at a time. :-/&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Ahruman</dc:creator><pubDate>Thu, 10 Apr 2014 17:44:24 -0000</pubDate></item><item><title>Re: trickyco.de - lloyd's blog.</title><link>http://trickyco.de/how-browserid-works#comment-523235223</link><description>&lt;p&gt;In my experience, casual users (i.e., everyone without a vanity domain) have historically changed e-mail provider every few years. This has slowed down a bit with gmail’s domination, but it will eventually be displaced. In general, users cannot be expected to act to update their browserIDs until after they’ve lost control of their own address. How is this handled?&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Ahruman</dc:creator><pubDate>Tue, 08 May 2012 15:27:15 -0000</pubDate></item><item><title>Re: Christians and atheists in blessing and un-blessing battle over Florida highway</title><link>http://blog.newhumanist.org.uk/2012/03/christians-and-atheists-in-blessing-and.html#comment-473432990</link><description>&lt;p&gt;Really? It seems like terrible publicity that portrays HiF in particular and atheists in general as belligerent idiots. The prayer group was doing no harm and wasn’t interfering in government. It’s things like this that lead me to avoid the label “atheist”.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Ahruman</dc:creator><pubDate>Thu, 22 Mar 2012 04:10:57 -0000</pubDate></item><item><title>Re: Hypocriticalmorphism</title><link>http://designaday.tumblr.com/post/18890633973#comment-460474267</link><description>&lt;blockquote&gt;The color of the background, the rules, and the leather bar across the top all scream “notepad!”&lt;/blockquote&gt;&lt;p&gt;&lt;br&gt;&lt;/p&gt;&lt;p&gt;This raises an important point: I have never in my life seen a yellow notepad with leather trim. Notepads simply aren’t made that way in the countries I’ve lived in. To me, the subliminal scream is “these people don’t know what a notepad look like!” Traditions for laying out calendars and address books also vary.&lt;/p&gt;&lt;p&gt;&lt;/p&gt;&lt;p&gt;&lt;/p&gt;&lt;p&gt;In days of yore, the Apple HIG pointed out the importance of reevaluating every graphical element when localizing an application, since what made sense in one culture might be confusing or offensive in another. As a practical example of this, check boxes in Mac OS 8.5–9.x had two different appearances, with a check mark or an X, used in different locales (and there was an Appearance Manager API to query this). This was necessary because in some cultures, a check mark means “wrong”.&lt;/p&gt;&lt;p&gt;&lt;/p&gt;&lt;p&gt;&lt;/p&gt;&lt;p&gt;Building an application out of representative graphical elements makes this type of attention to detail effectively impossible, and thus de-internationalizes the product. In Mac OS X, the check boxes in affected locales just say “wrong”. The notepad on my iPhone is in some weirdo psychedelic colour scheme that no sane person would use.&lt;/p&gt;&lt;p&gt;&lt;/p&gt;&lt;p&gt;&lt;/p&gt;&lt;blockquote&gt;Leather is an aesthetic choice, and certainly it will appeal to some and not to others. It’s an awfully popular choice for high-end car interiors, so why not virtual dashboards?&lt;/blockquote&gt;&lt;p&gt;&lt;br&gt;&lt;/p&gt;&lt;p&gt;Leather is used in physical products for its texture, scent and specular highlights. A drawing of leather has none of these properties, which is why Apple’s attempts so far have looked like damp gingerbread. Other materials fare somewhat better.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Ahruman</dc:creator><pubDate>Fri, 09 Mar 2012 04:44:11 -0000</pubDate></item><item><title>Re: http://365daysinsweden.tumblr.com/post/18153936712</title><link>http://365daysinsweden.tumblr.com/post/18153936712#comment-447388100</link><description>&lt;p&gt;Om du verkligen vill lära dig svenska är folk som vägrar prata engelska en värdefull resurs. Men kanske inte just hantverkare.&lt;/p&gt;&lt;p&gt;I’ve known a number of English-speaking guest workers who’ve had trouble learning Swedish because no-one will speak it to them. Some of them resorted to refusing to respond to English. There was also one who’d lived here a long time, still mostly spoke English, but had integrated Swedish words, like “I’ll just go and get the dammsooger”.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Ahruman</dc:creator><pubDate>Thu, 23 Feb 2012 18:45:48 -0000</pubDate></item><item><title>Re: http://365daysinsweden.tumblr.com/post/17064707438</title><link>http://365daysinsweden.tumblr.com/post/17064707438#comment-429932627</link><description>&lt;p&gt;If you’re renting from the owner, they’re typically responsible for maintenance and ensuring it’s clean when you arrive. (You’ll be billed for any abnormal maintenance costs and extra cleaning when you leave.) If you’re subletting – and in Stockholm, who isn’t? – there’s more variation in contracts, and they’re harder to enforce.&lt;/p&gt;&lt;p&gt;Of course, the downside to getting the owner to do maintenance is that your sink is less urgent to them than it is to you. :-)&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Ahruman</dc:creator><pubDate>Sun, 05 Feb 2012 02:55:26 -0000</pubDate></item><item><title>Re: Eating horses is legal now, and PETA loves it</title><link>http://www.grist.org/list/2011-12-01-eating-horses-is-legal-now-and-peta-loves-it#comment-381455960</link><description>&lt;p&gt;Ever heard of an ox-cart? Cattle are still widely used for transportation.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Ahruman</dc:creator><pubDate>Wed, 07 Dec 2011 15:18:24 -0000</pubDate></item><item><title>Re: How OpenBSD Leverages Malloc to Find Bugs - OS Blog</title><link>http://os-blog.com/how-openbsd-leverages-malloc-to-find-bugs/#comment-374730133</link><description>&lt;p&gt;Windows 95’s memory manager exposed this type of bug in some applications. In order to maintain backwards compatibility, a separate allocator was used for applications known to fail, famously including SimCity.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Ahruman</dc:creator><pubDate>Tue, 29 Nov 2011 02:00:51 -0000</pubDate></item><item><title>Re: Evolved compression in the x86 ISA</title><link>http://altdevblogaday.com/2011/05/15/evolved-compression-in-the-x86-isa/#comment-356084465</link><description>&lt;p&gt;For PowerPC, you can get the initial reciprocal estimate in one instruction (frsqrte f12, f1). Then again, the G5-optimized equivalent is eight bytes. :-)&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Ahruman</dc:creator><pubDate>Fri, 04 Nov 2011 18:52:09 -0000</pubDate></item><item><title>Re: Stylewriter II</title><link>http://shrineofapple.com/stylewriterii/#comment-340156004</link><description>&lt;p&gt;The platonic ideal of printer interfaces: power button, power light, error light, specific errors reported in simple terms on the computer. Whenever a printer stops and blinks three different lights at me because of a paper jam, I wish mine still worked – or rather, that I still had something I could connect it to.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Ahruman</dc:creator><pubDate>Thu, 20 Oct 2011 16:02:27 -0000</pubDate></item><item><title>Re: Alliansfritt Sverige: Nyheter i korthet: timbroprofilen, kongressbeslut och en utrikespolitisk talesperson</title><link>http://www.alliansfrittsverige.nu/2011/10/nyheter-i-korthet-timbroprofilen.html#comment-336485891</link><description>&lt;p&gt;Högform idag. :-) Synd bara att det finns så mycket bra material att jobba med…&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Ahruman</dc:creator><pubDate>Mon, 17 Oct 2011 05:12:26 -0000</pubDate></item><item><title>Re: http://littlebigdetails.com/post/11140577682</title><link>http://littlebigdetails.com/post/11140577682#comment-330276714</link><description>&lt;p&gt;It also detects dates/date ranges, including ones like “Last week”.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Ahruman</dc:creator><pubDate>Sun, 09 Oct 2011 10:56:33 -0000</pubDate></item><item><title>Re: Is 8 bits enough?  Of course not.</title><link>http://19lights.com/wp/2011/09/30/is-8-bits-enough-of-course-not/#comment-324096637</link><description>&lt;p&gt;8-bit log is pretty much what we have. That’s what all them gammas are for.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Ahruman</dc:creator><pubDate>Fri, 30 Sep 2011 17:51:56 -0000</pubDate></item><item><title>Re: http://adobegripes.tumblr.com/post/10439626009</title><link>http://adobegripes.tumblr.com/post/10439626009#comment-316062253</link><description>&lt;p&gt;Really? I can’t reproduce it in Mail, Preview or TextEdit under Lion.&lt;/p&gt;&lt;p&gt;In fact, I’m reasonably certain I know what’s happening: Photoshop is using the RGB values of the NSColor object provided by the colour picker without taking the associated colour space into account. When you use the pixel picker, a colour in the device colour space is generated, while the document is likely in some calibrated colour space (I don’t recall what the default is in Photoshop).&lt;/p&gt;&lt;p&gt;Using values from one colour space in another without proper conversion is essentially meaningless. I wish I could say I expected everyone working on Photoshop to know that.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Ahruman</dc:creator><pubDate>Tue, 20 Sep 2011 14:04:32 -0000</pubDate></item><item><title>Re: SEO for Non-dicks - Matt Legend Gemmell</title><link>http://mattgemmell.com/seo-for-non-dicks/#comment-316027431</link><description>&lt;p&gt;“Make sure the page title matches the first heading. Why wouldn’t you do this?”&lt;/p&gt;&lt;p&gt;Well… the first heading on this very page is “Matt Legend Gemmell”. Not a very unusual structure in blogs.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Ahruman</dc:creator><pubDate>Tue, 20 Sep 2011 13:07:04 -0000</pubDate></item><item><title>Re: Alliansfritt Sverige: Expressen levererar fler odödliga rubriker</title><link>http://www.alliansfrittsverige.nu/2011/09/expressen-levererar-fler-ododliga.html#comment-302799855</link><description>&lt;p&gt;“tidigast 2010”?&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Ahruman</dc:creator><pubDate>Mon, 05 Sep 2011 05:02:17 -0000</pubDate></item><item><title>Re: Alliansfritt Sverige: Pensionärsskatten - så lät det då</title><link>http://www.alliansfrittsverige.nu/2011/09/pensionarsskatten-sa-lat-det-da.html#comment-302786133</link><description>&lt;p&gt;Vi jämkar nu vad vi påstår att vi tyckt med vad vi alltid tyckt och alltid kommer att ha tyckt till tidigast 2013. Hur svårt kan det vara?&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Ahruman</dc:creator><pubDate>Mon, 05 Sep 2011 03:52:08 -0000</pubDate></item><item><title>Re: http://overooped.com/post/8599617064</title><link>http://overooped.com/post/8599617064#comment-281284676</link><description>&lt;p&gt;Engelbart doesn’t count, he was a time lord.&lt;/p&gt;&lt;p&gt;Another type of “lost” knowledge is the kind of high-level instructions that were used in minicomputers, that were thrown away in a fit of worse-is-better when microprocessors came along, then never recovered as chips got bigger.&lt;/p&gt;&lt;p&gt;Imagine you could do normal integer maths on the new tagged NSNumber representation, except that overflows were handled by producing bignum objects with arbitrary precision, and you could continue performing normal math ops on the bignum. Imagine the no-bignum cases were exactly as fast as normal integer maths operations, with the slow cases handled transparently by trapping to a software fallback. Suddenly, the number one reason to have non-object values would disappear, along with integer truncation problems.&lt;/p&gt;&lt;p&gt;Lisp machines did this in 1973.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Ahruman</dc:creator><pubDate>Tue, 09 Aug 2011 11:15:40 -0000</pubDate></item><item><title>Re: Art Asset Overview #19 - Wolfire Games Blog</title><link>http://blog.wolfire.com/2011/08/Art-Asset-Overview-19#comment-280275201</link><description>&lt;p&gt;A UI that’s been bothering me recently is Team Fortress 2, in particular the fact that the escape key brings up the main menu even when in a modal context like switching character class – but not in the semi-modal chat context. The console adds an additional complication.&lt;/p&gt;&lt;p&gt;So basically: consider mode transitions carefully, and make them consistent. Avoid orthogonal modes; if a game’s UI is so complex that orthogonality seems like a good idea, you have a deeper problem.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Ahruman</dc:creator><pubDate>Mon, 08 Aug 2011 02:43:11 -0000</pubDate></item><item><title>Re: http://littlebigdetails.com/post/7577371131</title><link>http://littlebigdetails.com/post/7577371131#comment-251976369</link><description>&lt;p&gt;Safari does this too (and has done since it first got tabs).&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Ahruman</dc:creator><pubDate>Thu, 14 Jul 2011 09:02:20 -0000</pubDate></item><item><title>Re: Alliansfritt Sverige: Ojämn Apoteksetablering i Stockholm</title><link>http://www.alliansfrittsverige.nu/2011/06/ojamn-apoteksetablering-i-stockholm.html#comment-221130354</link><description>&lt;p&gt;Med “regel” avses teoretiska modeller där det antas att man har perfekt information om alla priser på marknaden. Göran Hägglund får gärna räkna lite på effektiviteten i att åka omkring mellan apotek för att jämföra priser på 40–200 kr någon gång då och då på receptfria mediciner. När det gäller receptbelagda kan man lägga till halvtimmeslånga köer och högkostnadsskydd (en mycket bra grej, men det gör det rätt meningslöst att jämföra priser om man har icke-triviala behov av receptbelagd medicin).&lt;br&gt;Å andra sidan är det rätt trevligt att kunna köpa Alvedon på Coop på kvällarna. :-)&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Ahruman</dc:creator><pubDate>Wed, 08 Jun 2011 04:57:10 -0000</pubDate></item><item><title>Re: Alliansfritt Sverige: Säpo får tillgång till FRA</title><link>http://www.alliansfrittsverige.nu/2011/05/sapo-far-tillgang-till-fra.html#comment-212720813</link><description>&lt;p&gt;Intressant nog har Säpo vid flera tillfällen bett att få slippa, eftersom folk ogillar dem tillräckligt ändå.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Ahruman</dc:creator><pubDate>Fri, 27 May 2011 08:04:20 -0000</pubDate></item><item><title>Re: Alliansfritt Sverige: Alexandra Birk förklarar mordet på Bernadotte, Palme och Lindh</title><link>http://www.alliansfrittsverige.nu/2011/05/alexandra-birk-forklarar-mordet-pa.html#comment-212327417</link><description>&lt;p&gt;“A special thought” indeed.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Ahruman</dc:creator><pubDate>Thu, 26 May 2011 14:12:16 -0000</pubDate></item><item><title>Re: File format plugin API for Spotify?</title><link>http://overooped.com/post/5505499488#comment-203961986</link><description>&lt;p&gt; isStereo() seems a bit short-sighted. What if you add support for quadraphonic audio, or ambisonics, or one of these other newfangled many-channel formats? OK, they’re technically stereo, but I assume that’s not what you’re asking about there. I’d suggest a getChannelFormat() returning an enum.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Ahruman</dc:creator><pubDate>Sun, 15 May 2011 06:12:36 -0000</pubDate></item></channel></rss>