I know there are a lot of posts about the rails “dangerous” form_tag helper method, but why not another one.

Before saying anything about it, I want to write down the differences between form_tag and form_for (which was the choosen one previously). form_for is provided for updating data models and form_tag is provided for doing the other form operations (search something, change language in a page, etc).

Ok, after this, let’s go to the battle. This was my code

< % form_tag :controller => 'pages', :action => 'index' , :method => 'get' do %>
Tags : < %= text_field_tag :tags, '', :size => 40 %>
< %= submit_tag 'Search' %>
< % end %>

and I didn’t got the html code what I wanted to obtain, after googling for a while, I found this post which got me to the right way to see the light. The reason of this problem was that I was not using the right syntax for form_tag, which is

form_tag(url_for_options = {}, options = {}, *parameters_for_url, &block)

so my code became

< % form_tag({:controller => 'pages', :action => 'index'} , {:method => 'get'}) do %>
Tags : < %= text_field_tag :tags, '', :size => 40 %>
< %= submit_tag 'Search' %>
< % end %>

and I was happy again

Hope this silly problem helps not to waste your time (I did it for you )

M.

In that moment I remembered that I had read something in Agile Development with rails about how to use routes.rb, so I decided to give it a go.

I add to this files these sentences

map.connect 'home/', :controller => 'home', :action => 'home'
map.connect 'about_me/', :controller => 'home', :action => 'about_me'
map.connect 'cv/', :controller => 'home', :action => 'cv'
map.connect 'projects/', :controller => 'home', :action => 'projects'

and ‘home/home’ became ‘home’, ‘home/cv’ became ‘cv’, etc

M.

I was fighting with my installed different ruby versions and I had heard about rvm, but I didn’t have use it. I decided to do today and I installed it using

gem install rvm

After it I tried installing two different ruby versions but I always obtained the same ruby version (the aptitude installed one). After that and googling a little I noticed (after typing type rvm | head -n1 and doesn’t read “rvm is a function”) that I had forgotten something in my .bashrc file, which was:

[[ -s "$HOME/.rvm/scripts/rvm" ]] && source "$HOME/.rvm/scripts/rvm"

And after that and opened a different konsole I could change my ruby version using ‘rvm use ruby-version’

Cool!!!!!

M.

If you try to initialize a public git repository with an existing one in Debian Etch, you will obtain this message:

$ mkdir public-repo
$ cd public-repo/
$ git --bare init --shared
git: 'init' is not a git-command

This is because the git version which is in Debian Etch is too old (current: v1.7.2.1, etc: v1.4.4.4) and this command is not included in it. The first thing you have to do, if you can, is uninstall this git version and install a newer one, but if you can’t do it, this is the workaround I did.

1. Create an empty git repo in a machine with a newer git version using git --bare init --shared.
2. Copy the content of this repo in your public repo.
3. Copy your local project in the machine with the too old git version
4. Execute the command git --bare fetch $PATH_TO_LOCAL_COPY master:master
5. Voila

The next step will be git clone user@server:/path/to/public/repo/.

Hope this helps

M.

Since I use git instead of svn as my revision control system I have to say that I’m working with it and Netbeans since the first day I started using Git.

Cool

M.

I have to say that, after the three or four things I have done until now with it, git has surprised me a lot. It’s really fast, clear and it covers all my needs in control system tasks. I am svn user at work and I have to say that git is extremely faster than svn in merges (almost in the two I have done using git instead of svn).

Now I have only a repository in my computer but I supose I will have a central repository in the shared server I have with some friends of mine just in case my computer decides to pass away.

This post is named Git I (I don’t know if I will write more Git related posts) just in case

M.

These days I’m taking advantage of my daily train time watching railscasts episodes with my lenovo netbook. It’s really awesome all the things you can learn from these screencasts done by Ryan Bates about how to do things using rails.

Thanks a lot Ryan for sharing your knowledge with us.

M.

When a migration is created, rails add a column named id, which is defined as an integer, as table’s primary key column. Sometimes can happen that you don’t want to name your primary key column id using rails. The way you have to write your migration is this one (rails use integer type by default):


def self.up
create_table :airport, :primary_key => :airport_id do |t|
t.string :iso_code, :limit =>3, :null => false
t.string :name, :limit =>150, :null => false
t.integer :city_id, :null => false
end
end


If you have a “many to many” table probably you don’t want to have this id column in your table and you want to define two columns as primary key. This is the way you have to write a migration if you want to define this primary key:


def self.up
create_table :airports_cities, :primary_key => [:airport_id, city_id] do |t|
t.integer :distance, :null => false
end
end


In both examples, columns defined as primary key don’t have to be defined because rails do it for you. Using this, you won’t have an id column in your table and you’ll be able to use whatever you want as primary key (even strings).

Hope this helps (to my fish-like memory or to somebody else’s brain).

M.

Howto truncate all *.log files in log/ to zero bytes

rake log:clear


M.

M.