similar to: advice on MySQL version

After months of design and development, we''ve launched our second Rails site, http://www.fatwreck.com (assuming your DNS is updated - the pink site is the old one). The site is for legendary punk rock record label, Fat Wreck Chords and it includes a store. The old site was highly trafficked, with millions of page views per month and we expect this one to be also. So far, the server

Is there any reason to avoid using postgresql for small web apps? That is, is it''s overhead so large that the lighter MySQL will work substantially better for small apps on small machines? Or are they very similar in performance and configuration? I''ve used MySQL but postgresql has its appeals and I''m wondering if I need to keep MySQL around at all. On the other

The agile book says "You can inspect variables, set values, add other breakpoints, and generally have a good time". This is very sweet and totally useless. For example, I want to see the session variable, the cookies variable and so on. These names are unknown. I think I guess that the things that are available depend on where you put the breakpoint() call. I''ve tried a

Experts, have you figure it out with autocomplete to pass as array objects? still i''m finding out the way to figure. :( regards, Bala --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Ruby on Rails: Talk" group. To post to this group, send email to

In Depot, the demo application in the the Agile book, actions that modify the model such as add_to_cart redirect to another action that has a clear view role, like index, or show_cart, instead of generating the view themselves. This uses an HTTP redirect that goes to the client and returns. Is this an idiomatic way to design the flow in Rails? -- fxn

I have a new @offer that needs to generate a document if saved. But that''s all or nothing: transaction do @offer.save! generate_document end # handle exceptions if needed If save! fails everything is fine, but if generate_document raises an exception @offer is left as a model with timestamps, id, which is not a new_record? anymore (though the database was

I have a bunch of queries to compute some counters, like find_by_sql(<<-SQL SELECT p.*, c.nfavourites FROM people p, ( SELECT fp.person_id, COUNT(fp.user_id) as nfavourites FROM favourite_people fp, users u WHERE fp.user_id = u.id AND u.cluster = ''#{in_cluster_of_user.cluster}'' GROUP BY fp.person_id ) as c WHERE

I left ":memory:" in database.yml for testing and the most simple tests in the Agile book do not work (see trace below for test_truth). Since that''s what comes with the distributed database.yml I suppose it''s me doing something wrong. Did someone get SQLite ":memory:" working? How does it grab the schema? -- fxn % rake test_units (in

I read in the instructions of Tracks that "upgrading via the rake migrate command is quite a bit more tricky currently with SQLite and SQLite3". Is there any gotcha regarding migrations and SQLite3? -- fxn

I am using mongrel_cluster with mod_proxy_balancer and would like to enable compression (assuming it improves throughtput) and limit file size upload. I configured mod_deflate and LimitRequestSize in Apache, but in my trials looks like the proxied calls bypass those directives (the conf goes below). Is there a way to get this? -- fxn # Adapt this .example locally, as usual. # # To be

The new rescue_from in edge rails seems to conflict with ActionWebService, which I need for various other purposes. When I define, for example (with AWS gem installed): <b>rescue_from ActionController::RoutingError, :with => :page_not_found</b> I get the exception: <b>uninitialized constant ActionWebService::Dispatcher::ActionController::Base</b> OK, so AWS

I have an application with 15 migrations under version control. In a Mac and and in a Windows, a rake migrate from scratch runs them all just fine. But in a different Windows machine rake migrate stops after migration 3 for no apparent reason. --trace seems normal. No error is reported. Both Windows are XP SP2. They all have the same svn revision and Rails-related software, database is

A view display.rhtml sees no controller state variables set in the corresponding display action. Why? -- fxn

Hello, I felt annoyed enough when having to redirect user back to their previous location in a hackish way that I wrote this plugin. It avoids storing POST and Ajax request. It also has a facility to specify actions not to store in the history. If you are interested, it''s there: http://blog.cosinux.org/pages/rails-history See you all, Damien -- Damien MERENNE

Hi, I just wanted to update my Rails installation with "sudo gem update rails" but I get this error: Updating installed gems... Bulk updating Gem source index for: http://gems.rubyforge.org Attempting remote update of rails Install required dependency activesupport? [Yn] y Install required dependency activerecord? [Yn] y ERROR: While executing gem ... (OpenURI::HTTPError) 404

Reviews Wanted. I''ve read the tables of contents and haven''t yet made up my mind. Is one book clearly better than the other? Is one clearly full of bugs? Is one so much further ahead that there is no choice? Are they both so incomplete that I should just wait and only cook real food? Do trains still have dining cars? Warren Fred -------------- next part -------------- An

I want to load about 14000 words (a subset of /usr/share/dict/words) into a MySQL table in a migration: class CreateWords < ActiveRecord::Migration def self.up create_table :words, :force => true do |t| t.column :word, :string end say_with_time ''loading words...'' do words = File.join(RAILS_ROOT, ''db'',

When a controller responds to a link_to_remote with a redirect_to, the link_to_remote gets a success callback, this would seem like a bug to me, at a minimum it should return a failure? This is driving me crazy because all the login engines/generators respond to an authentication error with a redirect_to. The work around is to change them to all do a render :layout => false, :status => 500

So after successfully getting the test SQLite db to run from memory, I tried running some tests but found that somewhere in the schema translation from Postgres (my devel database) and SQLite Rails fails to properly set the ''id'' column, which breaks all kinds of things. Take as an example my ''users'' table. In Postgres the schema is (generated with pgAdmin III):

I am pondering the possibility of encrypting/decrypting some fields in a SQLite backend on-the-fly. The point of the message is not security, I know that''s broken, but whether there''s a technique that provides on-the-fly save/read filters. Of course the solution would need to work transparently in joins, so user.posts.last.title would do the right thing if title