<?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 davidbanham</title><link>http://disqus.com/by/davidbanham/</link><description></description><atom:link href="http://disqus.com/davidbanham/comments.rss" rel="self"></atom:link><language>en</language><lastBuildDate>Mon, 22 Sep 2014 07:41:06 -0000</lastBuildDate><item><title>Re: Aaron Larner's blog - nutellahabit.com</title><link>http://nutellahabit.com/node/Zero-Downtime-Deploys-on-Node-with-AWS#comment-1600189795</link><description>&lt;p&gt;The main problem with this approach is that you're only running a single server at a time. What happens when that process crashes or something else makes it unavailable?&lt;/p&gt;&lt;p&gt;A better solution would be two servers, spread across two availability zones, running behind an ELB. That would get you some degree of fault tolerance.&lt;/p&gt;&lt;p&gt;Also, when it came time to update your code, you get more degrees of freedom. One approach would be to drop one server out of the ELB, upgrade it, then bring it back in. Once it's registered and healthy, drop the other out and repeat the process.&lt;/p&gt;&lt;p&gt;Since you're already running nginx, though, why not use it to manage the process? On your machine, start your new server on a new internal port. Add that server to the nginx configuration file and remove the old server. Send a [HUP signal](&lt;a href="http://wiki.nginx.org/CommandLine)" rel="nofollow noopener" target="_blank" title="http://wiki.nginx.org/CommandLine)"&gt;http://wiki.nginx.org/Comma...&lt;/a&gt; to the nginx server, which tells it to gracefully reload it's configuration. Once all the existing threads have finished and the old nginx processes have terminated, shut down your old server.&lt;/p&gt;&lt;p&gt;That's the approach I use in [Cavalry](&lt;a href="https://github.com/davidbanham/cavalry)" rel="nofollow noopener" target="_blank" title="https://github.com/davidbanham/cavalry)"&gt;https://github.com/davidban...&lt;/a&gt;&lt;/p&gt;&lt;p&gt;So, if you'd like something to automate that process for you, may I humbly recommend [Field Marshal](&lt;a href="https://github.com/davidbanham/field-marshal)" rel="nofollow noopener" target="_blank" title="https://github.com/davidbanham/field-marshal)"&gt;https://github.com/davidban...&lt;/a&gt;&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">davidbanham</dc:creator><pubDate>Mon, 22 Sep 2014 07:41:06 -0000</pubDate></item><item><title>Re: Poor Man's Firebase: LevelDB, REST, and WebSockets</title><link>http://procbits.com/2014/01/06/poor-mans-firebase-leveldb-rest-and-websockets/#comment-1191653660</link><description>&lt;p&gt;Interesting! I've been getting a similar outcome using PouchDB lately. It supports two way syncing with a remote CouchDB and notices updates via long polling.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">davidbanham</dc:creator><pubDate>Tue, 07 Jan 2014 22:57:06 -0000</pubDate></item><item><title>Re: 8 months in Microsoft, I&amp;amp;nbsp;learned</title><link>http://ahmet.im/blog/8-months-microsoft/#comment-928544820</link><description>&lt;p&gt;There's a lot of good people trying hard, but it seems like there's also an awful lot of inertia to overcome. I've seen really important pull requests to the Azure CLI tools sit un-merged for months, purely because nobody could be bothered to verify that the committer had submitted the CLA. They had submitted it, but no-one checked. There just doesn't seem to be enough will.&lt;/p&gt;&lt;p&gt;&lt;a href="https://github.com/WindowsAzure/azure-sdk-tools-xplat/pull/340" rel="nofollow noopener" target="_blank" title="https://github.com/WindowsAzure/azure-sdk-tools-xplat/pull/340"&gt;https://github.com/WindowsA...&lt;/a&gt;&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">davidbanham</dc:creator><pubDate>Wed, 12 Jun 2013 22:53:55 -0000</pubDate></item><item><title>Re: Megaupload Implications are plain scary for Cloud Storage</title><link>http://alexblom.com/2012/01/28/thinking-through-the-megaupload-takedown/#comment-423616597</link><description>&lt;p&gt;You're being sarcastic... right?&lt;/p&gt;&lt;p&gt;Logmein is fine for what it's for. It's not for sharing files. I have to ensure my computer is on and connected to the internet at the exact moment someone wants to download it? Then they have to endure the limited upload bandwidth of my connection instead of a data centre pipe? You are being completely ridiculous.&lt;/p&gt;&lt;p&gt;I personally lost perfectly legal work that was being transferred to me by the freelancer we paid to create it. He regularly used the service because it was simple to upload very large video files. He had copies of the content locally, luckily, but he had to reupload them and it was a hassle.&lt;/p&gt;&lt;p&gt;Napster was very much an international operation. The difference was that it was where the vast majority of sharing was happening at that time. That was the industry's big chance to innovate their business model and cut a deal with the file sharing community. They opened Pandora's Box instead.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">davidbanham</dc:creator><pubDate>Sun, 29 Jan 2012 08:00:59 -0000</pubDate></item><item><title>Re: Megaupload could spawn caselaw more destructive than SOPA</title><link>http://Blog.davidbanham.com/?p=52#comment-420484186</link><description>&lt;p&gt;Very true, the copyright holder could white list their own account. It does place a burden on industry, but not an unreasonable one. What it ignores, though, is issues relating to jurisdictions outside of the US. What if the issue objected to by the rightsholder is perfectly acceptable in another country? Shifting content from one device to another is specifically allowed in a lot of legal systems.&lt;/p&gt;&lt;p&gt;I recognise that a lot of these issues are edge cases. I feel that's where the law is often most important. It may also be an underlying uncomfortableness with the inherent ambiguity in the legal system, but its the way I think it needs to be.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">davidbanham</dc:creator><pubDate>Wed, 25 Jan 2012 07:41:02 -0000</pubDate></item><item><title>Re: Megaupload could spawn caselaw more destructive than SOPA</title><link>http://Blog.davidbanham.com/?p=52#comment-419985758</link><description>&lt;p&gt;Consider also that it may be legal for the file to be publicly distributed in one context and not another.&lt;/p&gt;&lt;p&gt;For example, Dan Bull uploaded his album and wanted it to be spread far and wide, and he derived value from the statistics he gathered through Megaupload.&lt;/p&gt;&lt;p&gt;If a third party uploaded the same file to Megaupload under a different account, Dan Bull legally could, and reasonably would, issue a takedown. He's happy for people to download the item for free, but he wants to know when and how they do.&lt;/p&gt;&lt;p&gt;If Megaupload had deleted his copy of the file at that time, he would have been just as annoyed as he is now:&lt;/p&gt;&lt;p&gt;&lt;a href="http://torrentfreak.com/dan-bull-releases-his-own-megaupload-song-120123" rel="nofollow noopener" target="_blank" title="http://torrentfreak.com/dan-bull-releases-his-own-megaupload-song-120123"&gt;http://torrentfreak.com/dan...&lt;/a&gt;&lt;/p&gt;&lt;p&gt;The crux is that the badness is inherent in the distribution of the file, not the file itself. When they find child porn, they determine that any and all distribution of that file in any context is wrong, so they delete the file itself. When they receive a takedown for a file, they're being told that distribution of The Sopranos in that context is illegal, not that The Sopranos is a is outlawed material.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">davidbanham</dc:creator><pubDate>Tue, 24 Jan 2012 18:04:56 -0000</pubDate></item><item><title>Re: Samsung Galaxy Tab 10.1 Going on Sale in Australia, Apple didn&amp;#8217;t win</title><link>https://androidcommunity.com/samsung-galaxy-tab-10-1-going-on-sale-in-australia-apple-didnt-win-20110816/#comment-288607447</link><description>&lt;p&gt;You might want to re-read that article you linked. Sales have NOT been resumed in Germany, but in the rest of the EU. Germany is the one country that the device is still under an injunction, pending hearings on the 25th.&lt;/p&gt;&lt;p&gt;Also, from all reports the Australian launch definitely has been delayed. Retailers here were expecting to begin sales around the 15th, so a launch date of Sep 1 represents a delay of about 2 weeks.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">davidbanham</dc:creator><pubDate>Tue, 16 Aug 2011 21:14:00 -0000</pubDate></item><item><title>Re: Mac OS X Lion: what's broken (or working) for you?</title><link>http://www.engadget.com/2011/07/23/mac-os-x-lion-whats-broken-or-working-for-you/#comment-262532166</link><description>&lt;p&gt;Pretty comprehensively destroyed my workflow. Quicktime Player 7 and iMovie both crash so I can't convert any video. (upside, made me go and find Handbrake which is vastly better than either for converting/compressing). Also, yeah fine they broke Xcode, but why did that have to take gcc with it? And why do I have to download a whole 3 gig from Apple's glacially slow servers JUST to get gcc? Stole an entire day of my productivity chasing this stuff down. In a lean startup, a whole day is pretty bloody important to me.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">davidbanham</dc:creator><pubDate>Sat, 23 Jul 2011 21:29:47 -0000</pubDate></item><item><title>Re: Android adapter concepts promise to accessorize any headset</title><link>http://www.engadget.com/2011/03/27/android-adapter-concepts-promise-to-accessorize-any-headset/#comment-173247030</link><description>&lt;p&gt;Ah, I and was going to buy a Galaxy S II. Good to know, cheers.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">davidbanham</dc:creator><pubDate>Mon, 28 Mar 2011 09:04:20 -0000</pubDate></item><item><title>Re: Android adapter concepts promise to accessorize any headset</title><link>http://www.engadget.com/2011/03/27/android-adapter-concepts-promise-to-accessorize-any-headset/#comment-173063794</link><description>&lt;p&gt;Wait, what? The first product makes sense, but the second is bonkers.&lt;/p&gt;&lt;p&gt;On the product page they state " the best you get out of them without an adapter is faint and distorted sound."&lt;/p&gt;&lt;p&gt;Which is flat out false. iPhone headsets work great with my N1. The only difference between an iPhone and Android headset is that apple does volume with serial communication while google does it with varying the impedance on the circuit. tl:dr, you can still use the main button on an iphone set in an android phone, the volume controls just don't work.&lt;/p&gt;&lt;p&gt;Also, the spec seems to indicate that the button on the adaptor replaces the button on the headset mic. There's absolutely to reason (as far as I'm aware) for that to be the case.&lt;/p&gt;&lt;p&gt;All up, it smells fishy.&lt;/p&gt;&lt;p&gt;&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">davidbanham</dc:creator><pubDate>Mon, 28 Mar 2011 00:28:10 -0000</pubDate></item><item><title>Re: Five tips for building a website on a budget</title><link>http://anthillonline.com/five-tips-for-building-a-website-on-a-budget/#comment-167002733</link><description>&lt;p&gt;Please don't conflate open source with 'free'. The two are not the same.&lt;/p&gt;&lt;p&gt;One of the biggest advantages of Open Source is that you can have it worked on and improved by your own staff or by any company you like. Closed source solutions can only be modified or improved by the company that created it. They effectively have a monopoly and they will probably charge accordingly.&lt;/p&gt;&lt;p&gt;It's one of the biggest advantages, but it has nothing to do with whether or not you need to pay for the license to use the software. It's worth knowing the difference.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">davidbanham</dc:creator><pubDate>Thu, 17 Mar 2011 00:05:59 -0000</pubDate></item><item><title>Re: Josh stops by Jimmy Fallon with the Xoom, Pre 3, Veer, and TouchPad (video)</title><link>http://www.engadget.com/2011/02/24/josh-stops-by-jimmy-fallon-with-the-xoom-pre-3-veer-and-touch/#comment-155655693</link><description>&lt;p&gt;In front of a PC is a completely different use case to a mobile phone, though. I could be wrong, but I don't think enough has changed that it's suddenly going to be a compelling experience.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">davidbanham</dc:creator><pubDate>Thu, 24 Feb 2011 19:55:57 -0000</pubDate></item><item><title>Re: Josh stops by Jimmy Fallon with the Xoom, Pre 3, Veer, and TouchPad (video)</title><link>http://www.engadget.com/2011/02/24/josh-stops-by-jimmy-fallon-with-the-xoom-pre-3-veer-and-touch/#comment-155580770</link><description>&lt;p&gt;The Nokia 6680 was my first front facing camera, video calling, 3G phone. That was in 2005.&lt;/p&gt;&lt;p&gt;The thing is, back then we used it a few times because it was new and then forgot all about it. I don't think it will be any different this time around.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">davidbanham</dc:creator><pubDate>Thu, 24 Feb 2011 16:28:13 -0000</pubDate></item><item><title>Re: Notion Ink Adam ships today</title><link>http://www.engadget.com/2011/01/19/notion-ink-adam-ships-tomorrow/#comment-131274376</link><description>&lt;p&gt;Meanwhile, still no communication whatsoever with companies selected for EAP2. Awesome.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">davidbanham</dc:creator><pubDate>Wed, 19 Jan 2011 07:24:45 -0000</pubDate></item><item><title>Re: Howto: Chromium Browser On Linux With Flash</title><link>http://linuxologist.com/01general/howto-chromium-browser-on-linux-with-flash/#comment-14628264</link><description>&lt;p&gt;I think that second paragraph above is inaccurate. Please disregard.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">davidbanham</dc:creator><pubDate>Tue, 11 Aug 2009 08:52:30 -0000</pubDate></item><item><title>Re: Howto: Chromium Browser On Linux With Flash</title><link>http://linuxologist.com/01general/howto-chromium-browser-on-linux-with-flash/#comment-14617762</link><description>&lt;p&gt;Could it be an issue with your command line arguments in the launcher? First time around I tried pasting what was here, then adding another dash. Turns out whatever is screwing up the double dash is turning it from two short dashes into one long dash. Anyway, go and have another look at the launcher parameters and sort out the dashes.&lt;/p&gt;&lt;p&gt;You'll know your arguments are right when you see a page warning you about the fact that flash is in beta and telling you that chromium is not chrome.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">davidbanham</dc:creator><pubDate>Tue, 11 Aug 2009 01:53:37 -0000</pubDate></item><item><title>Re: Howto: Chromium Browser On Linux With Flash</title><link>http://linuxologist.com/01general/howto-chromium-browser-on-linux-with-flash/#comment-14612784</link><description>&lt;p&gt;This is brilliant. Blows Shiretoko out of the water. Only strange issue is that it gets sent to the back of all windows any time I middle click on a URL.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">davidbanham</dc:creator><pubDate>Mon, 10 Aug 2009 23:02:20 -0000</pubDate></item></channel></rss>