/usr/lib/ruby/1.8/memcache.rb:24:in `read': No such file or directory - /usr/lib/ruby/1.8/../VERSION.yml (Errno::ENOENT)
To troubleshoot, I'm updating ruby1.8. Ugh, now I no longer have irb1.8!!
Oh well, in the end I just edited /usr/lib/ruby/1.8/memcache.rb and changed line 24:
# VERSION = begin
# config = YAML.load(File.read(File.dirname(__FILE__) + '/../VERSION.yml'))
# "#{config[:major]}.#{config[:minor]}.#{config[:patch]}"
# end
VERSION = '1.5.0'
]]>
@@ -76,12 +76,7 @@ class TidyFFI::Interface
def tidy_buf_object
@tidy_buf_object ||= begin
- release_date = Date.parse(LibTidy.tidyReleaseDate) rescue nil
- if release_date && (release_date > Date.parse("Dec 29 2006"))
- TidyFFI::LibTidy::TidyBufWithAllocator
- else
- TidyFFI::LibTidy::TidyBuf
- end
+ TidyFFI::LibTidy::TidyBufWithAllocator
end
end
UPDATE: This wasn't the end of it. I submitted the fix to libc (Eugene), and he said the check was needed for Mac OS X which ships with an older version of libTidy. I dug deeper, and found that the date was getting returned, but it just wasn't getting processed correctly. After that, I did some external troubleshooting, and found that in irb1.9.1 I had to require 'date' to parse the date libTidy was returning. I switched out require 'rubygems' with require 'date', and the bug went away. Yay!
]]>sudo gem1.9.1 install activesupport --version 2.3.5
$ sudo gem uninstall activesupport Select gem to uninstall: 1. activesupport-2.3.5 2. activesupport-2.3.8 3. All versions > 2
For reference, the full error:
/var/lib/gems/1.9.1/gems/deprecated-2.0.1/lib/deprecated.rb:176:in `instance_method': {:metaclass=>:singleton_class} is not a symbol (TypeError)
from /var/lib/gems/1.9.1/gems/deprecated-2.0.1/lib/deprecated.rb:176:in `block in <top (required)>'
from /var/lib/gems/1.9.1/gems/activesupport-2.3.8/lib/active_support/core_ext/object/metaclass.rb:13:in `<class:Object>'
from /var/lib/gems/1.9.1/gems/activesupport-2.3.8/lib/active_support/core_ext/object/metaclass.rb:3:in `<top (required)>'
from /var/lib/gems/1.9.1/gems/activesupport-2.3.8/lib/active_support/core_ext/object.rb:5:in `require'
from /var/lib/gems/1.9.1/gems/activesupport-2.3.8/lib/active_support/core_ext/object.rb:5:in `<top (required)>'
from /var/lib/gems/1.9.1/gems/activesupport-2.3.8/lib/active_support/core_ext.rb:8:in `require'
from /var/lib/gems/1.9.1/gems/activesupport-2.3.8/lib/active_support/core_ext.rb:8:in `block in <top (required)>'
from /var/lib/gems/1.9.1/gems/activesupport-2.3.8/lib/active_support/core_ext.rb:8:in `each'
from /var/lib/gems/1.9.1/gems/activesupport-2.3.8/lib/active_support/core_ext.rb:8:in `<top (required)>'
from /var/lib/gems/1.9.1/gems/activesupport-2.3.8/lib/active_support.rb:56:in `require'
from /var/lib/gems/1.9.1/gems/activesupport-2.3.8/lib/active_support.rb:56:in `<top (required)>'
from /var/lib/gems/1.9.1/gems/mail-2.2.1/lib/mail.rb:6:in `require'
from /var/lib/gems/1.9.1/gems/mail-2.2.1/lib/mail.rb:6:in `<module:Mail>'
from /var/lib/gems/1.9.1/gems/mail-2.2.1/lib/mail.rb:2:in `<top (required)>'
from /var/lib/gems/1.9.1/gems/pony-1.0/lib/pony.rb:2:in `require'
from /var/lib/gems/1.9.1/gems/pony-1.0/lib/pony.rb:2:in `<top (required)>'
from /home/albertlash/svxbox/lib/svxbox/ponyboy.rb:6:in `require'
from /home/albertlash/svxbox/lib/svxbox/ponyboy.rb:6:in `<top (required)>'
from /home/albertlash/svxbox/lib/svxbox.rb:8:in `require'
from /home/albertlash/svxbox/lib/svxbox.rb:8:in `<top (required)>'
from /home/albertlash/yodnsconf/yodnsconf.rb:45:in `require'
from /home/albertlash/yodnsconf/yodnsconf.rb:45:in `<top (required)>'
from config.ru:11:in `require'
from config.ru:11:in `block in <main>'
from /var/lib/gems/1.9.1/gems/rack-1.1.0/lib/rack/builder.rb:46:in `instance_eval'
from /var/lib/gems/1.9.1/gems/rack-1.1.0/lib/rack/builder.rb:46:in `initialize'
from config.ru:1:in `new'
from config.ru:1:in `<main>'
from /var/lib/gems/1.9.1/gems/unicorn-0.99.0/lib/unicorn.rb:46:in `eval'
from /var/lib/gems/1.9.1/gems/unicorn-0.99.0/lib/unicorn.rb:46:in `block in builder'
from /var/lib/gems/1.9.1/gems/unicorn-0.99.0/lib/unicorn.rb:816:in `call'
from /var/lib/gems/1.9.1/gems/unicorn-0.99.0/lib/unicorn.rb:816:in `build_app!'
from /var/lib/gems/1.9.1/gems/unicorn-0.99.0/lib/unicorn.rb:673:in `init_worker_process'
from /var/lib/gems/1.9.1/gems/unicorn-0.99.0/lib/unicorn.rb:688:in `worker_loop'
from /var/lib/gems/1.9.1/gems/unicorn-0.99.0/lib/unicorn.rb:605:in `block (2 levels) in spawn_missing_workers'
from /var/lib/gems/1.9.1/gems/unicorn-0.99.0/lib/unicorn.rb:602:in `fork'
from /var/lib/gems/1.9.1/gems/unicorn-0.99.0/lib/unicorn.rb:602:in `block in spawn_missing_workers'
from /var/lib/gems/1.9.1/gems/unicorn-0.99.0/lib/unicorn.rb:598:in `each'
from /var/lib/gems/1.9.1/gems/unicorn-0.99.0/lib/unicorn.rb:598:in `spawn_missing_workers'
from /var/lib/gems/1.9.1/gems/unicorn-0.99.0/lib/unicorn.rb:612:in `maintain_worker_count'
from /var/lib/gems/1.9.1/gems/unicorn-0.99.0/lib/unicorn.rb:408:in `block in join'
from /var/lib/gems/1.9.1/gems/unicorn-0.99.0/lib/unicorn.rb:396:in `loop'
from /var/lib/gems/1.9.1/gems/unicorn-0.99.0/lib/unicorn.rb:396:in `join'
from /var/lib/gems/1.9.1/gems/unicorn-0.99.0/lib/unicorn.rb:28:in `run'
from /var/lib/gems/1.9.1/gems/unicorn-0.99.0/bin/unicorn:126:in `<top (required)>'
from /var/lib/gems/1.9.1/bin/unicorn:19:in `load'
from /var/lib/gems/1.9.1/bin/unicorn:19:in `<main>'
I, [2010-05-28T04:48:24.242011 #28838] INFO -- : reaped #<Process::Status: pid 28848 exit 1> worker=0
I, [2010-05-28T04:48:24.242246 #28838] INFO -- : master complete
]]>
At first I thought it was an environment variable missing, causing the string encoding to get messed up, but alas, no, that's not the problem.
I checked and set most of the environment variables, including LANG, then debugged the process by putting the encoding name out. They were the same if I was running the process or if daemontools was running it.
Anyone have any idea why this might happen?
UPDATE: I found a solution, but I still don't know why this was happening:
]]>I found that the gem uses Ruby's autoload mechanism, which I've read is a good thing, but the paths it was using messed me up.
This is what I've changed to make it work for me:
5,6c4,5 < self.autoload :LibTidy, File.dirname(__FILE__) + '/tidy_ffi/lib_tidy' < self.autoload :Interface, File.dirname(__FILE__) + '/tidy_ffi/interface' --- > self.autoload :LibTidy, 'tidy_ffi/lib_tidy' > self.autoload :Interface, 'tidy_ffi/interface' 9,11c8,10 < require File.dirname(__FILE__) + '/tidy_ffi/options_container' < require File.dirname(__FILE__) + '/tidy_ffi/tidy' < require File.dirname(__FILE__) + '/tidy_ffi/tidy_ffi_extensions' \ No newline at end of file --- > require 'tidy_ffi/options_container' > require 'tidy_ffi/tidy' > require 'tidy_ffi/tidy_ffi_extensions' \ No newline at end of file
Actually I'm working on porting a bunch of web applications I work on, but Chimailmadmin is the furthest along.
I'm glad I held off for a bit while I became more familiar with Sinatra. Its a really sweet little web application framework.
And yes - its open source (Affero GPLv3) and available on Github!
]]>It provides bindings to the gd2 library (think awesome graphics utility) via ffi (foreign function interface).
There is another ffi interface, gd2-ffi, but its limited in the functionality it covers.
This one didn't work with Ruby 1.9.1 right out of the box, but with a little bit of hackage, I was able to get things going. I reported by experience back to dark panda (J) who responded speedily.
Hopefully this will be the standard GD2 library.
FWIW: I became interested in the ruby bindings after finally moving the source code for gdofcgi to github.
]]>Building a Xapian database for searching, for instance. I first tried using the standard Xapian bindings, but then I found xapian-fu, a much more Rubyish gem.
It works fine with ruby 1.9.1 for me and is quite speedy. Here's a very simple example for a search (using the databases generated by the debian package apt-xapian-index):
#!/usr/bin/ruby1.9.1 require 'xapian-fu' include XapianFu db = XapianDb.new(:dir => '/var/lib/apt-xapian-index/index.1') db.search(ARGV[0]).each do |match| puts match.data end]]>
Read error: #<IndexError: string not matched> /var/lib/gems/1.9.1/gems/sinatra-1.0.a/lib/sinatra/base.rb:600:in `[]=' /var/lib/gems/1.9.1/gems/sinatra-1.0.a/lib/sinatra/base.rb:600:in `handle_exception!' /var/lib/gems/1.9.1/gems/sinatra-1.0.a/lib/sinatra/base.rb:586:in `rescue in dispatch!' /var/lib/gems/1.9.1/gems/sinatra-1.0.a/lib/sinatra/base.rb:589:in `dispatch!'
I think it turned out to be an encoding error, but I'm not sure why this IndexError was showing up. Turns out that this bit of code doesn't cause a problem when there are no errors, but causes the IndexError when there is a problem:
@env = settings.environment
This is most likely Sinatra specific, if I change @env to @myenv, its OK. Since I don't actually need it, I've removed it altogether.
And to fix the encoding issue:
rlinks.encode('UTF-8', undef: :replace, invalid: :replace)
]]>
I'll share more of my favorite tool choices as they come to mind.
]]>
require 'test/unit'
class DoggyTest < Test::Unit::TestCase
def test_chekit
assert "hi" == "hi"
end
def test_barkit
assert "bark" == "hi"
end
end
What do you know? Before I tried this I didn't know that test/unit required methods to be prefixed with "test_" to be auto-run.
You learn something new everyday. I try to, if I can!
]]>Its works good, but I prefer using Unicorn. I'll probably use it when I don't have access to Unicorn.
Shotgun was written and is maintained by Ryan Tomayko, who coincidentally wrote an article about Unicorn which is where I found out about it.
]]>Testing is one of these areas that needs improvement. I've always respected the importance of testing software, but I was often so daunted by the scope of what can, in general, be done with software. Its often so flexible and configurable, I have a tough time balancing between non-exhaustive tests and tests which could become "big O" problems of their own.
Thankfully I'm starting to see some convergence between the scope of the tasks and the tools at my disposal. I'm starting to get comfortable with expansive, maybe even factorial style programming, where I'm instructing other instructions on how to create new instructions, and my work is increasingly focused on less complex applications.
More specifically, I'm using the Rack::Test::Methods mixin with RSpec, and that is making things nice and straightforward. I don't consider these "real world" tests because they lack the involvement of web browsers and thus cannot process javascript, but in my experience it is a solid testing platform.
I'd like to evolve to using Webrat, Watir, or Selenium or some combination of those at some point in the near future, but I just haven't had the time as of late.
Cucumber looks interesting, too, but its not for me. I prefer to develop more rapidly.
This is a helpful page:
]]>ArgumentError - XML object #0x14 not supported
For whatever reason, I had set my XML content to be a symbol. Doh!
xml = :'<root />']]>
This means it works fine with Ruby 1.9.1!
Kudos, benprew. Thanks!
]]>