<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0"><channel><title>Disqus - Latest Comments for bear</title><link xmlns="http://www.w3.org/2005/Atom" rel="http://api.friendfeed.com/2008/03#sup" href="http://disqus.com/sup/all.sup#usercomments-893b56e3" type="application/json"/><link>http://disqus.com/people/bear/</link><description></description><language>en</language><lastBuildDate>Wed, 11 Nov 2009 13:28:45 -0000</lastBuildDate><item><title>Re: getting the exit code from a batch file that is run from a python program</title><link>http://code-bear.com/bearlog/2007/06/01/getting-the-exit-code-from-a-batch-file-that-is-run-from-a-python-program/#comment-22761907</link><description>I suspect it's how NANT is calling the scripts from within it's code that requires the explicit use of the EXIT keyword - a batch-to-batch invocation probably is handled in a subtly different manner (which is what both of us had to find and solve ;)</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">bear</dc:creator><pubDate>Wed, 11 Nov 2009 13:28:45 -0000</pubDate></item><item><title>Re: Google Wave FedOne component with Prosody XMPP server</title><link>http://code-bear.com/bearlog/2009/11/01/google-wave-fedone-component-with-prosody-xmpp-server/#comment-21679461</link><description>It is a complete Wave environment for sure, you can use the command line test client and view all of the Wave goodness flowing back and forth.  The only thing you don't get is the fancy UI client.&lt;br&gt;&lt;br&gt;Yep, it is completely federated - it's already making attempts to reach out and contact other servers when you add remote JID's - Google just has it turned off on the public facing servers so it fails.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">bear</dc:creator><pubDate>Mon, 02 Nov 2009 11:40:27 -0000</pubDate></item><item><title>Re: FriendFeed and Scoble and the crowd-as-community problem</title><link>http://code-bear.com/bearlog/2009/11/02/friendfeed-and-scoble-and-the-crowd-as-community-problem/#comment-21653166</link><description>*nod* - and that was the difference I was trying to draw out of your post, so i'm glad to see you reinforce it here.&lt;br&gt;&lt;br&gt;The fun part is watching the methods and styles of curation and list building work their way thru the older and newer applications and seeing how folks are responding to the limitations.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">bear</dc:creator><pubDate>Mon, 02 Nov 2009 03:51:43 -0000</pubDate></item><item><title>Re: FriendFeed and Scoble and the crowd-as-community problem</title><link>http://code-bear.com/bearlog/2009/11/02/friendfeed-and-scoble-and-the-crowd-as-community-problem/#comment-21652751</link><description>I completely understand your point about the amount of time you would have to spend to massage the lists on both services - it's something that keeps me from truly jumping in with both feet to most of these sites.&lt;br&gt;&lt;br&gt;I have been waiting to make any "public" comments on the debate going on until just now - but i'm glad at the core you do recognize that they are tools to be used properly.  And thanks for the reason-dump, it helped draw me out of my micro-blogging shell and post something more substantial.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">bear</dc:creator><pubDate>Mon, 02 Nov 2009 03:31:24 -0000</pubDate></item><item><title>Re: Google Wave FedOne server certificate key fun (not!)</title><link>http://code-bear.com/bearlog/2009/10/30/google-wave-fedone-server-certificate-key-fun-not/#comment-21501408</link><description>I had to change/do two things: 1) export the cert as DER type instead of PEM type and also convert the key from the PEM type I had in /etc/ssl to a DER type and store both the key and the cert local to the wave component directory.&lt;br&gt;&lt;br&gt;I'll edit the post to show this more clearly</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">bear</dc:creator><pubDate>Sun, 01 Nov 2009 03:52:37 -0000</pubDate></item><item><title>Re: Random thoughts about including Google Wave in your data flow</title><link>http://code-bear.com/bearlog/2009/10/02/random-thoughts-about-including-google-wave-in-your-data-flow/#comment-18306117</link><description>Very interactive indeed.  And that's a great example of how a bot/extension can be both a sink and a source, that bot probably treats all incoming change requests as discrete/atomic updates - i'm more worried about when a bot is present in 50 wave documents and it's getting 2/3 updates per document and *some* of them need to be distributed to something other than the source doc.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">bear</dc:creator><pubDate>Fri, 02 Oct 2009 14:55:03 -0000</pubDate></item><item><title>Re: Loose.ly coupled 140-char message network (Scripting News)</title><link>http://www.scripting.com/stories/2009/08/13/looselyCoupled140charMessa.html#comment-14816979</link><description>Unless they have changed in the 0.8 tree, laconi.ca distributes inbound items to different working queues in a database table and then have listening processes to consume from those queues.  Because now there are different queues, the database load can be spread across servers but then you run into more of the common database replication issues.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">bear</dc:creator><pubDate>Thu, 13 Aug 2009 21:55:48 -0000</pubDate></item><item><title>Re: Loose.ly coupled 140-char message network (Scripting News)</title><link>http://www.scripting.com/stories/2009/08/13/looselyCoupled140charMessa.html#comment-14816860</link><description>Twitter doesn't suffer from downtime because of the size of the data, but rather from the amount of work it has to do to distribute the item once received and get it queued for delivery.  If it were a size scaling issue then increasing the number of sources of the data would work as any client would have that many more places to go.&lt;br&gt;&lt;br&gt;Rather it's a distribution/routing issue and "classic" P2P would not work IMO.  The rules for routing of any given item would need to be distributed to all of the P2P cloud and when anyone of them receive that item, see if they are responsible and handle it.  I see problems with out the partitioning would be done to try and balance duplicating data sent over reducing decisions - a fun problem for sure :)&lt;br&gt;&lt;br&gt;I'm thinking P2P could enter the scene if you applied that tech to sync up the working queue of anyone persons posts - say to distribute yours or scobles working queue across the cloud so then it's efficiently duplicated to allow for parallel work.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">bear</dc:creator><pubDate>Thu, 13 Aug 2009 21:53:34 -0000</pubDate></item><item><title>Re: XMPP PubSub + XSLT</title><link>http://code-bear.com/bearlog/2009/01/04/xmpp-pubsub-xslt/#comment-13800398</link><description>Both Atom and RSS 2.0 are xml, so you can use this script as a start for RSS 2.0 - but the structure and content would be different.  See &lt;a href="http://www.rssboard.org/rss-specification" rel="nofollow"&gt;http://www.rssboard.org/rss-specification&lt;/a&gt; for what goes into RSS 2.0 content.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">bear</dc:creator><pubDate>Sun, 02 Aug 2009 00:34:35 -0000</pubDate></item><item><title>Re: Exploring the 2010 Web</title><link>http://scobleizer.com/2009/05/16/exploring-the-2010-web/#comment-9455583</link><description>It sounds like you are rearchitecting your blog to be the static presentation of certain realtime events/items that you generate during the course of your day.  Instead of the blog being the first place you would visit to make posts or otherwise "generate content" it can now be the place where you summarize the day using long-form content.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">bear</dc:creator><pubDate>Sat, 16 May 2009 16:08:19 -0000</pubDate></item><item><title>Re: Episode 3 - Bear</title><link>http://odtv.me/army/2009/01/19/episode-3-bear/#comment-5392727</link><description>Thanks for the chance to be interviewd - I had a great time!&lt;br&gt;&lt;br&gt;Listening I realized I never really answered why I have the nickname "bear" ...&lt;br&gt;&lt;br&gt;basically my neices and nephews all call me Uncle Bear because of the beard and the fact that I give a lot of bear-hugs&lt;br&gt;&lt;br&gt;So it stuck :)</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">bear</dc:creator><pubDate>Tue, 20 Jan 2009 09:57:56 -0000</pubDate></item><item><title>Re: Twitter in 140 characters (Scripting News)</title><link>http://www.scripting.com/stories/2009/01/04/twitterIn140Characters.html#comment-4884888</link><description>Twitter is the public space of my present and future workplace.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">bear</dc:creator><pubDate>Sun, 04 Jan 2009 19:27:19 -0000</pubDate></item><item><title>Re: XMPP PubSub + XSLT</title><link>http://code-bear.com/bearlog/2009/01/04/xmpp-pubsub-xslt/#comment-4884807</link><description>That is the exact type of problem we are trying to avoid by offering a "true" body option with our PubSub events - to allow people who are consuming them who are using clients which ignore non-chat messages to having something to view.&lt;br&gt;&lt;br&gt;Like all things feature-rich, just have to get all our pieces in place and pointing at the same place :)&lt;br&gt;&lt;br&gt;thanks!</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">bear</dc:creator><pubDate>Sun, 04 Jan 2009 19:20:25 -0000</pubDate></item><item><title>Re: Seesmic's Video Firehose Now Available via Gnip</title><link>http://www.loiclemeur.com/english/2008/12/seesmics-video-firehose-now-available-via-gnip.html#comment-4740106</link><description>Rob,&lt;br&gt;&lt;br&gt;The "firehose" is the same information that appears in the Atom/RSS feeds, but instead of you polling it, Seesmic's PubSub server will push it to you via XMPP.  The payload of the message we create uses the same Atom item as the feeds and soon we will be offering other "flavours".&lt;br&gt;&lt;br&gt;So what is being done by 3rd party devs like Zac, is to wire together the receiving of the Seesmic events via XMPP and a method of inserting them into Facebook.&lt;br&gt;&lt;br&gt;You can receive the events either thru a service like Gnip or directly from us.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">bear</dc:creator><pubDate>Tue, 30 Dec 2008 01:33:46 -0000</pubDate></item><item><title>Re: XMPP PubSub aka “death by a thousand cuts”</title><link>http://code-bear.com/bearlog/2008/12/21/xmpp-pubsub-aka-%e2%80%9cdeath-by-a-thousand-cuts%e2%80%9d/#comment-4559355</link><description>Anyone can subscribe to it by sending me an email with the JID you want to use.  We are doing the subscriptions manually just to allow me to keep track of a contact email in case something changes or I need to "break" the system.&lt;br&gt;&lt;br&gt;The pushing of an item to another PubSub node is called chaining and I know they (the XMPP spec folks) are working on something to cover it - but yes, technically any valid JID can be used as the target and PubSub nodes are valid JID's&lt;br&gt;&lt;br&gt;We briefly thought of using Content-Based subscriptions to allow the subscriber to control various details like format (Atom, RSS, Json) and also language (all, English, French) but quickly put that back into the "future" column when we discovered that it's not well implemented on either the Client or Server side :( and the last thing we want to do is bolt on a partially implemented feature to PubSub until we have more time to evaluate how.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">bear</dc:creator><pubDate>Sun, 21 Dec 2008 14:55:51 -0000</pubDate></item><item><title>Re: Seesmic and XMPP Pubsub</title><link>http://code-bear.com/bearlog/2008/10/21/seesmic-and-xmpp-pubsub/#comment-3466228</link><description>Julien,&lt;br&gt;&lt;br&gt;The biggest hurdle was finding out that different folks implement specs differently and that there wasn't a single suite of tests to compare server performance with.  So we spent quite a bit of time working with the Tigase guys in getting them to probably 98% complient with the spec.&lt;br&gt;&lt;br&gt;The other issue is a complete lack of docs or examples of what goes into maintaining all but the most simple of PubSub structures.  Not to say the information isn't out there, it's just wasn't helpful to our needs - most folks only use the PEP part of PubSub.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">bear</dc:creator><pubDate>Mon, 03 Nov 2008 21:21:21 -0000</pubDate></item><item><title>Re: Is This Stupid?: Seesmic Video Gets Verklempt, RWW Has Nothing to Read or Write</title><link>http://www.isthisstupid.com/articles/26/seesmic-video-gets-verklempt-rww-has-nothing-to-read-or-write#comment-1000185</link><description>I am unable to see that behaviour on the current RWW site.  Do you have a time when you captured the screenshot so I can see if it was something on our side and correct any issues?&lt;br&gt;&lt;br&gt;thanks,&lt;br&gt;bear&lt;br&gt;seesmic dev</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">bear</dc:creator><pubDate>Fri, 25 Jul 2008 10:58:51 -0000</pubDate></item><item><title>Re: python giveth and python taketh away</title><link>http://code-bear.com/bearlog/2006/03/02/python-giveth-and-python-taketh-away/#comment-499839</link><description>Michael,&lt;br&gt;&lt;br&gt;Glad I saved you some time :)</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">bear</dc:creator><pubDate>Tue, 20 May 2008 19:20:42 -0000</pubDate></item><item><title>Re: getting the exit code from a batch file that is run from a python program</title><link>http://code-bear.com/bearlog/2007/06/01/getting-the-exit-code-from-a-batch-file-that-is-run-from-a-python-program/#comment-495134</link><description>Glad to help!  With the amount of time our team took in figuring this one out there was no way I wasn't going to share it :)</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">bear</dc:creator><pubDate>Tue, 20 May 2008 05:23:03 -0000</pubDate></item><item><title>Re: scaling questions and issues</title><link>http://code-bear.com/bearlog/2008/01/18/scaling-questions-and-issues/#comment-305288</link><description>We are fighting mutlipe deadlines, RabbitMQ is still in use on the test box but due to reasons that are not RabbitMQ's fault at all we are still stuck with ActiveMQ :(&lt;br&gt;&lt;br&gt;I'll have to take a look at the updated client list - that may give me some more info to pass to the devs to speed up the replacement.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">bear</dc:creator><pubDate>Sun, 06 Apr 2008 00:28:35 -0000</pubDate></item><item><title>Re: XMPP as the basis for interop in TwitterLand? (Scripting News)</title><link>http://www.scripting.com/stories/2008/01/21/xmppAsTheBasisForInteropIn.html#comment-90768</link><description>We are also doing XMPP notifications at Seesmic.  PubSub is being worked on now.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">bear</dc:creator><pubDate>Mon, 21 Jan 2008 13:25:49 -0000</pubDate></item></channel></rss>