I am in the process of migrating my office to Apple products. I needed a way to install iWorks and Parallels on my boss’s shiny new MacBook Air, but the MacBook Air has no optical drive, and we have not yet ordered an external drive. My assumption was that I would need to make an image and copy it over to her machine; I was wrong. After I launched Finder on her machine, I saw “Remote Disc” as one of her drive options. A little research revealed that I could enable “DVD or CD Sharing” from my Sharing Preferences pane. Doing so made my drive visible from her machine, and the installation went along without a hitch.

I received the error “The change you wanted was rejected” after I moved a rails app from one host to another.

Looking into the log file, the following error appears:

ActionController::InvalidAuthenticityToken (ActionController::InvalidAuthenticityToken):
/usr/lib/ruby/1.8/mongrel/rails.rb:76:in `process'
/usr/lib/ruby/1.8/mongrel/rails.rb:74:in `synchronize'
/usr/lib/ruby/1.8/mongrel/rails.rb:74:in `process'
/usr/lib/ruby/1.8/mongrel.rb:159:in `process_client'
/usr/lib/ruby/1.8/mongrel.rb:158:in `each'
/usr/lib/ruby/1.8/mongrel.rb:158:in `process_client'
/usr/lib/ruby/1.8/mongrel.rb:285:in `run'
/usr/lib/ruby/1.8/mongrel.rb:285:in `initialize'
/usr/lib/ruby/1.8/mongrel.rb:285:in `new'
/usr/lib/ruby/1.8/mongrel.rb:285:in `run'
/usr/lib/ruby/1.8/mongrel.rb:268:in `initialize'
/usr/lib/ruby/1.8/mongrel.rb:268:in `new'
/usr/lib/ruby/1.8/mongrel.rb:268:in `run'
/usr/lib/ruby/1.8/mongrel/configurator.rb:282:in `run'
/usr/lib/ruby/1.8/mongrel/configurator.rb:281:in `each'
/usr/lib/ruby/1.8/mongrel/configurator.rb:281:in `run'
/usr/lib/ruby/1.8/mongrel/command.rb:212:in `run'

Rendering /vol/sites/site/public/422.html (422 Unprocessable Entity)


Some further investigation revealed that there was an issue with how rails 2.3.8 was handling cookies.  I stumbled upon the following fix at https://rails.lighthouseapp.com/projects/8994/tickets/4690.

I used Ken Collins’ fix of adding his mongrel.rb (http://gist.github.com/471663) to config/initializers and now all is well.

I have been moving applications from my Rackspace host over to Amazon’s EC2 service.  Moving simple LAMP apps has been fairly cut and dry, but while moving a RoR app and accessing it through apache w/ mod_proxy, the html is delivered to the browser as plaintext.  If I access the app directly on mongrel’s port, it renders correctly.   So it appears that something with either my mod_proxy configuration, or something about how my EC2 instance is set up is causing the problem.


> curl -I http://linux.local:3000
HTTP/1.1 200 OK
Connection: close
Date: Thu, 15 Jul 2010 14:27:27 GMT
Content-Length: 0

> curl -I http://linux.local
HTTP/1.1 200 OK
Date: Thu, 15 Jul 2010 14:27:35 GMT
Via: 1.1 linux.local
Content-Type: text/plain


As you can see, when accessing the page through Apache/mod_proxy, the Content-Type that is returned is explicitly set as text/plain.

Update
This is a bit shoe-horned, but it works for right now. In the apache2.conf file, I changed the DefaultType from text/plain to text/html. The reason this solution “concerns” me is that on my other servers, (where this problem is not occurring) the DefaultType is also set to text/plain, but the html is output correctly.

From time to time you may need to flush the dns cache on your mac.  I had recently moved a site from one host to another, and while the nameservers at the various levels returned the new IP address, my browser still looked to the old IP address.

To resolve issues like this, (in Leopard) open a terminal and type in

> dscacheutil -flushcache


The same issue is resolved in 10.5.1 and earlier, open a terminal and type

> lookupd -flushcache


CodeIgniter provides a simple framework that allows you to quickly get an application up and running.  The learning curve (so far) has been very low.  I like web-frameworks for projects. I have done a number of projects in Ruby on Rails, and I have dabbled in Django. Those have been pleasant experiences, but I prefer PHP as my preferred development language for most projects due to a number of reasons (not interested in jumping into that holy war right now.)

Up to this point, if I wanted to use a PHP framework, my choices were Zend and CakePHP. Zend is a monster. I’m sorry, I know a lot of developers swear by Zend, but I’m not one of them. Nobody “quickly” picks up Zend. Zend is so incredibly dense/bloated/abstracted that debugging something becomes an exercise that is more time consuming than the original problem that you were trying to solve. CakePHP felt too much like I was pretending to be Ruby on Rails, and there were so many hoops to jump through. And it was frequently a little “too magical.”

After exploring these different options, I wrote my own framework that took the parts of Ruby on Rails that I enjoyed, but kept the flexibility and familiarity of PHP. I have been using this for over a year now and it has served me well. I was pleased to find that the structure of CodeIgniter was almost identical to the structure of my own framework. This has made it a very easy switch for me.

The only problem that I have run into so far was briefly not understanding how to get to a GET parameter in a fashion that didn’t make me feel dirty.

Tangent about accessing GET parameters coming…
CodeIgniter obscures the $_GET parameters by default, you can enable them within your configuration, but I am working on one page of an existing site as a favor for a friend.  The CodeIgniter documentation suggests that you retrieve the parameters based on “segments” so if you wanted to pass in

http://www.mysite.com/events.php?start=25&state=VA&count=50

CodeIgniter suggests that you would generate the url

http://www.mysite.com/events/index/25/VA/50

And then retrieve your parameters based on which “segment” they were.

segment(3) => 25 // the 'start' value
segment(4) => VA // the 'state' value
segment(5) => 50 // the 'count' value

This makes me cringe for many reasons. What if, for example, the user can click a link to change the count to 25, but it defaults to 50. Most of the time this isn’t going to happen, so would I ALWAYS assume that I’m passing in the first segment as the start param, even when it’s typically not going to be used?

uri_to_assoc to the rescue
The solution to this “issue” (it’s more of my own issue than an actual problem) was the function uri_to_assoc.
Using uri_to_assoc means that I could format my urls as:

http://www.mysite.com/events/index/start/25/state/VA/count/50

and then in my code use:

$params = $this->uri->uri_to_assoc(3);

/* I now have the array $params with:
* $params['state'] = 'VA'
* $params['state'] = 25
* $params['count'] = 50
*/

This makes me much more comfortable, as it means that it doesn’t matter where in the URI that my parameters are positioned.
Tangent finished
I plan to continue using CodeIgniter for a while, to see if it’s the right fit. I have some small projects that it would likely be quite appropriate for, and it will get me out of the business of supporting my own framework.

Note: I have ~/.gem set as my GEM_HOME and GEM_PATH to avoid needing to sudo


> gem install mysql
Building native extensions. This could take a while...
ERROR: Error installing mysql:
ERROR: Failed to build gem native extension.

/System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/bin/ruby extconf.rb
checking for mysql_query() in -lmysqlclient... no
checking for main() in -lm... yes
checking for mysql_query() in -lmysqlclient... no
checking for main() in -lz... yes
checking for mysql_query() in -lmysqlclient... no
checking for main() in -lsocket... no
checking for mysql_query() in -lmysqlclient... no
checking for main() in -lnsl... no
checking for mysql_query() in -lmysqlclient... no
checking for main() in -lmygcc... no
checking for mysql_query() in -lmysqlclient... no
*** extconf.rb failed ***
Could not create Makefile due to some reason, probably lack of
necessary libraries and/or headers. Check the mkmf.log file for more
details. You may need configuration options.


The workaround was fairly simple. Since I had installed MySQL from source the situation was resolved with a few command line options:


> export ARCHFLAGS="-arch x86_64"
> gem install mysql -- \
--with-mysql-dir=/usr/local/mysql \
--with-mysql-lib=/usr/local/mysql/lib \
--with-mysql-include=/usr/local/mysql/include \
--with-mysql-config=/usr/local/mysql/bin/mysql_config

Building native extensions. This could take a while...
Successfully installed mysql-2.8.1
1 gem installed


Facebook used to be a site that I visited when I was bored or had downtime. It was a diversion that could be accessed at any given moment, and then put back away when it’s purpose was served. The problem is that now, you can’t put Facebook away. You can close your browser window and continue on to surf other sites, but no matter where you go, you’re back at Facebook again. If I’m reading an article on cnn.com, my friends’ comments are listed to the right of the article. If I’m visiting (seemingly) random blogs, I’m being encouraged to be the first of my friends to “Like” the posting. Enough. I’m unplugging from the Facebook version of The Matrix.

The development community seems to have mixed feelings about the recent facebook changes, and the perceived intentions behind them. Some think that it is an inevitable future, and that we are all going to be issued national ID cards with our Facebook profile ID listed. Others don’t like the changes, but still see a world where Facebook becomes the only dominant player. I see it playing out a bit differently. I think the recent changes and initiatives that are coming out of Facebook are the beginning of the end for the giant. I, for one, am closing my Facebook account today, and I suspect that others will begin to do the same, for the following reasons:

• People like to feel like they have “choice” on the internet. Being herded through your online experience by endless recommendations and “you-may-also-likes” takes the excitement out of discovering anything, and leaves people with the feeling that every aspect of their online experience is orchestrated.
• People are boring. I don’t care what my facebook friend said about some article on [insert site name here]. If I wanted to get their view, I would have asked them or, gone to facebook.
• People always come to hate the king. Ask Microsoft, AOL, and Google. As something grows increasingly in strength and reach, people grow uneasy. For me, personally, seeing Facebook “Like” buttons on every website I visit is becoming maddening.

To those that stay, good luck, enjoy the experience. I hope your farms, fishtanks, and mafia families thrive.  And I hope you enjoy the customized Gillete advertisements.  My buddy Carlos did.  For everybody else, I’ll see you outside of Facebook, where I’ll be working on a browser plugin to block the various Facebook “Like/Comment/Recommend/Blah/Blah/Blah” widgets.

Facebook’s recent modifications to their privacy policies is causing concern for a lot of people, and for good reason. Facebook is encouraging third-party sites and developers to leverage their developer platform to gain deliver a more social experience for their users.  Facebook is referring to this new practice as “Instant Personalization”, and it means that  sites will be able to pull profile information about their visitors, as well as about the visitor’s friends.  That last sentence deserves some attention because it is possibly one of the bigger threats to privacy that I have seen done out-in-the-open on the web.

Essentially what this means is that if your friend visits a site that you have no affiliation with whatsoever, that friend can pass your personal information on to that site.  Meaning that whatever permission that you had previously extended to your friend to use your profile details have now been extended to a site with which you have no affiliation, and with whom you have no knowledge.

This attempted increase in reach is startling.  Not starting because Facebook is doing it (research has shown that adding personal information to advertisements results in significantly higher click-thru) but startling because it’s being allowed in the first place.  Google received FAR more attention for their Buzz Privacy Snafu, and that was a much smaller risk to my privacy than what Facebook is attempting.

I’ve considered leaving Facebook altogether (and I still may) I have deactivated my Facebook account as more and more of these privacy violations have surfaced. Below, I detail some of the steps necessary for setting your privacy preferences to prevent “New Instant Personalization”

Select Applications and Websites

Choose Edit Settings under Instant Personalization

Uncheck the “Allow select partners…” checkbox

Uncheck the Allow select partners... checkbox

Confirm that you are sure you want to disallow instant personalization

Confirm your selection

Go back to the privacy setting screen and edit what your friends can share about you

"Edit Settings" for what your friends can share about you

Uncheck all of the things that you don’t want your friends sharing about you

Uncheck what personal information you don't want your facebook friends to share about you

These issues will continue to surface.  As long as people are not actively protecting their privacy, and as long as sites can get higher click thru by posting and sharing personal information, your privacy is at risk.