[SUMMARY] Rails Core Weekly (May 16 - May 23)

on 23.05.2006 08:08

Rails Core Weekly

(I got some positive feedback on last week's summary. If you have
ideas on how to make this better let me know. Also if there's anything
wrong/incomplete in the summary so I can rectify).

[http://wrath.rubyonrails.org/pipermail/rails-core/2006-May/001483.html]

Just before last weeks RCW Adam C. Greenfield carefully enquires if
his baby patch #4995 (http://dev.rubyonrails.org/ticket/4955) could
get some much needed affection from the Core. The patch fixes a
problem with query string parsing. Without much further ado, Koz
commits the patch.

[http://wrath.rubyonrails.org/pipermail/rails-core/2006-May/001483.html]

Moving from tiny patches (1.2 K) to 6.7 K tomes, Ian White comes up
with a way to allow nils in aggregations. It's Ian's first patch and
she's a beauty (the patch, don't know about Ian). Particulary
enchanted by the documentation Ian gives to his patch (visual image:
"if your customer has just left for the moon, you want to set
gps_location to nil".)

[http://wrath.rubyonrails.org/pipermail/rails-core/2006-May/001485.html]

The Core gets poked in the eye by Corey Donohoe. He's eager to know
when Oracle finally gets boolean suppport. Michael Schoen submits the
patch http://dev.rubyonrails.org/ticket/5139. So now Oracle gets to
play with the grown-ups. Nice. Michael combined several issues into
that patch including the wonderful:
http://dev.rubyonrails.org/ticket/4017 (goodbye oci, hello oracle).

[http://wrath.rubyonrails.org/pipermail/rails-core/2006-May/001490.html]

Our poor Rails wiki is under ceaseless attack from misguided souls and
other lost ones, Peter Michaux picks up the gauntlet and proposes a
visual key system to make the wike sane again. Matt Pelletier agrees
and incidently has a nice visual key solution. To be continued.

[http://wrath.rubyonrails.org/pipermail/rails-core/2006-May/001494.html]

Just when you didn't expect it, Oracle lapses into temporary insanity
when someone called David (?) checks in
http://dev.rubyonrails.org/changeset/4354 (has_many collections
shouldn't load the entire association to do build or create).
Fortunately Michael knows what's wrong and Oracle is resuscitated
promptly.I'm guessin' we'll here more from this David fellah.

[http://wrath.rubyonrails.org/pipermail/rails-core/2006-May/001491.html]

In our last RCW we noticed Hampton's suggestion for a different way of
filtering in ActiveRecord. Some people like it. But DHH states:

"The sweetness of syntactic sugar is measured in saving x frequency."

A patch that allowes :conditions to handle a hash would be most
welcome though. Which Hampton supplies only a couple of hours later:
http://dev.rubyonrails.org/ticket/5143. Hampton documented his patch
rather nicely:

"This adds a third type accepted into the conditions of a find, and
that is an hash of attribute/value pairs. The values passed in this
way are dealt with as equality functions with the SQL AND operator."

An example would be the casual:

 find(:first, :conditions => { :user_name => user_name, :password => 
password })

Incidently Hampton also adds the ability to do :

Student.find(:all, :conditions => params[:student])

It's nice to end this weeks RCW on a high note. Keep 'm coming, more 
next week.

Rodney

Forums » Rails-core » [SUMMARY] Rails Core Weekly (May 16 - May 23)