Archive for May, 2007

links for 2007-05-28

Comments

Ruby on Rails: problems experienced – migration

When going back to a project after reboot etc. my last db migration changes were not present even though the migration was correct and ‘rake db:migrate’ was run. The rake migrate command would run without error; actually, without any output. I could not roll back my migration because it would try to delete a column that didn’t exist (the migration self.up created a column).

The only way I could get around this problem was to:

- create another migration with a new (temporary) name, put the same code in it from the malfunctioning migration

- run the migration to create the new column

- roll back 2 migrations to then delete the new column

- delete the malfunctioning migration

- rename the temporary migration to the name of the malfunctioning migration

- re-run the ‘rake db:migrate’

Then all was well. I know I didn’t have to get rid of the malfunctioning migration (or rather, non-functioning migration) but I wanted to have a clean migration path.

I wonder what could have caused this. I still don’t know.

Comments

links for 2007-05-09

Comments

Oh Joyous Day! Google Calendar Notifications for All.

Google has finally added the one piece of functionality that I thought their great calendar offering was sorely missing; notifications for more than just your primary calendar.

I have no idea why they took so long to implement this feature but now that it is here I think I can finally make the switch for all of my calendaring needs. That is, once they roll it out to users outside of the US. Or is it just me that doesn’t yet have this feature available even though it has been announced on the Official Google Blog.

Now, where’s that to do list?

UPDATE: It turns out that there were  some issues related to this and the other new features released yesterday forcing Google to do a rollback. The secondary calendar notifications are not available to anyone. Google, why must you be so cruel?
Official Google Blog: Calendar on the go

Comments

Interview Question: What is inversion of control?

Inversion of control is a programming practice in which an object does not have to know about the other objects that it depends upon but instead has the objects it depends on passed into it (via the constructor or setters). In that way, the objects that are depended on can be changed without having the change to original objects code. As a result you have code that is more loosely coupled, more flexible and more easily tested.
The Spring Framework allows for easy use of this practice through its dependency injection framework.

Comments (2)

links for 2007-05-02

Comments

Kai and Taya-isms: Hoopa-loops

Hoopa-loops = Hula-hoops

Comments