success 30 30 1 56787 robin 2009-11-18T23:12:13Z testing #ohcount Ohcount 55249 andy 2009-11-06T22:14:01Z Kind of a test message to the live Ohloh jabber server: Trying to test an #ejabberd setup behind a firewall. Trying to use #mcabber to connect. mcabber ejabberd 52365 jason 2009-10-14T20:18:19Z @gwoo - wow, that's a big change! Best of luck with the new stuff, and thanks for all you've done on Cake! gwoo 51009 andy 2009-10-03T19:42:50Z Anecdotal evidence that #ruby projects are converting from SVN to git and moving to github in the process. Found this when looking at why Ohloh's watir report was out of date: http://groups.google.com/group/watir-general/browse_thread/thread/ba83d09aedca3d20 I don't mind the Ohloh plug in the article either. Ruby 50947 andy 2009-10-03T02:58:41Z Met @pudge and other people from SourceForge in Chicago. pudge 48474 mfeingold 2009-09-11T12:39:25Z asked @pshiryaev to write f# parser for #ohcount pshiryaev Ohcount 44967 rofiudin 2009-08-13T02:16:35Z @robin could you review and apply my latest #ohcount patch? robin Ohcount 43379 scott 2009-07-31T21:14:05Z #foo just dropped a new version 0.3.4 40535 andy 2009-07-10T00:52:43Z setting up brand new MacBook Pro 15". Thanks SF! 40213 andy 2009-07-07T22:42:29Z Google is known for it's stripped-down, simple is better design sense. Have they lost it? Google apps has a new feature that lets users customize the app logo with a Beta icon (http://googleblog.blogspot.com/2009/07/google-apps-is-out-of-beta-yes-really.html). I have to admit, I might have added such a feature in a previous job and well, it's darn funny from a harmless easter egg standpoint. 37276 andy 2009-06-17T17:27:06Z Ohloh had trouble with #passenger on one of its three webservers this morning. This is kind of interesting because it's been rock solid for the last several months, which in retrospect has been a big improvement vis-a-vis the #mongrel-cluster we had previously. Passenger (mod_rails for Apache) 36336 andy 2009-06-10T16:04:59Z looks like rubyforge had a hiccup last night ; cleaning up a lot of jobs that stalled because of it 35668 andy 2009-06-05T23:18:26Z I love #zentest 's autotest when it works. Every code change I make is intelligently and quickly run against my #rails unit tests, and a notification appears in a pop-up window (via integration with #growl). I know this is old news, but I'm just experiencing it again on a new project and it makes staying in a TDD-driven coding "zone" possible. Growl Ruby on Rails ZenTest & autotest 34194 andy 2009-05-28T20:38:10Z Ohloh has been acquired by SF: http://ir.corp.sourceforge.com/phoenix.zhtml?c=82629&p=irol-newsArticle&ID=1293491&highlight= 34036 andy 2009-05-27T19:10:09Z Ohloh indexed its 300,000th project this morning. 33606 andy 2009-05-24T19:00:39Z Working on improving Ohloh's system to reschedule project code analyses and downloads after a problem. 31074 andy 2009-05-06T21:53:16Z Ruby strangeness: options = {}, options[:foo] = nil, yeah, what we expect.... options = 5, options[:foo] = 0 Whoa! Turns out integers support indexing into bits, so 5[0] = 1, 5[1] = 0, 5[2]=1, 5[3...infinity] = 0. What this means is that if your function signature looks like def foo(options = {}) and someone calls the function with an integer instead of a hash mistakenly, the values extracted from options will seem to work (and mostly evaluate to 0). 30935 jason 2009-05-05T22:27:29Z yay, ohloh jabber daemon is alive again 30395 andy 2009-04-30T17:40:51Z I'm very happy to see that NewRelic (https://rpm.newrelic.com) added two new views to the free service offering. First, the ability to see your most time consuming controller actions (before they just offered average time for free, to see the total you had to upgrade) and something they call "Apdex", a way of finding your most dissatisfying solutions. Not sure why they added this to free (since now there's less urgency to pay for the upgraded services) but it makes their free service much more appealing. 29532 andy 2009-04-23T20:57:53Z Just finished improving #ohloh website performance. /projects/*/edits is now much faster after adding a column for project_id and all our index pages are faster by changing the default page size to 10 items instead of 30. If there's a particular page on ohloh that bugs you for being slow, please let me know. Ohloh 29231 andy 2009-04-21T16:08:09Z #ohloh ops wisdom: respect the warning in mdadm's man page regarding mdadm.conf: "This file is very likely to contain unwanted detail, particularly the devices= entries. It should be reviewed and edited before being used as an actual config file." After a system upgrade of one our backend servers, it appears that this unwanted detail is preventing the raid array from booting. non-unix system admins: mdadm is software raid and otherwise has been very stable for us. Ohloh 29120 jason 2009-04-20T21:58:12Z btw - ohloh has moved offices. We're now downtown sharing space w/ jott.com 27697 andy 2009-04-08T21:31:40Z mod_status with "ExtendedStatus On" is to #apache what #mongrel_proctitle was to #mongrel : a way to see which requests are currently being handled by rails processes/threads Mongrel mongrel_proctitle Apache HTTP Server 27420 andy 2009-04-06T22:22:07Z converting a few hold-out .rhtml files to #haml Haml 26606 andy 2009-03-31T05:11:22Z conundrum: Phusion #passenger docs say that #apache's mod_rewrite is slow but I want to do the friendly thing and redirect (301) requests that were sent directly to the downstream webservers back to our canonical website. Ohloh used to do this with a mongrel.conf similar to the one posted here: http://noobonrails.blogspot.com/2006/09/good-seo-mojo-with-rails.html. Hopeful solution: To preserve the speed-up in Passenger given by PassengerHighPerformance, I port this #mongrel specific fix into a #rails before_filter. Ruby on Rails Mongrel Apache HTTP Server Passenger (mod_rails for Apache) 26222 jason 2009-03-27T15:59:49Z reading up on #sphinx. Sphinx documentation builder 26112 andy 2009-03-26T21:16:25Z Validating phusion passenger as our rails server. Planning to keep nginx as our reverse proxy/static file server. 25259 andy 2009-03-20T00:23:23Z Compression ratio of 33:1 for gzipping my sample (fairly real-worldish) sitemap file. Huge win. Think that I read that Google supports downloading these in gzipped format, maybe I won't have to pay the cost of decompressing on the server. 25256 andy 2009-03-20T00:04:59Z continuing from previous message: Rails has a CompressedMemCacheStore < MemCacheStore which calls into the MemCache class and it does this already. But we're using cache_fu, which adds lots of goodness, particularly time to live (:ttl), to the MemCache class directly. So I'll just grab the functionality I need and use it in my specific place... but this does seem like a huge missed opportunity (and the memcached.yml file produced for cache_fu is misleading, it has a compression: setting that seems to be ignored) 25251 andy 2009-03-19T23:13:20Z Running into the 1mb memcached limit... compression should solve this particular case.