Activity (RSS)

Thanks Dan!

I was going to work on fixing the patch tonight but I see that you up and solved the problems without me. :)

I've attached a new patch that works when the machine's clock is set to a different time zone than Time.zone.

1. assert_not_valid opposes assert_valid more closely.. we were attempting to mimic the naming convention of assert_not_equal/assert_equal.

2. I've thought a...

A couple of more questions:

1. Why not "assert_invalid"?

2. Have you thought about supporting partial error messages checks? Basically, if you have a test fo...

Helper methods to quickly assert some basic characteristics of a particular ActionMailer delivery, or the last delivery made.

I don't have a strong problem with that. I'd argue that this function is most useful as a way to confirm which errors you've received, but you could certainly...

Why not make the messages parameter optional and not assert the specific contents of the validation errors unless it is provided?

Rails has had an assert_valid method for a very long time, but its corresponding opposite method never existed in Rails core. I'm submitting Pivotal Labs's im...

Thank you, this is already fixed in edge, see this changeset.

The documentation for "image_submit_tag" is faulty, it's just a matter of putting the parenthesis in the correct place.

If you type what is in the documentati...

I don't think uploading a file to a ticket makes it go back into the feed. I've added the changes koz suggested and re-uploaded the patch (in Git's format).

Updated patch to cache sanitized method name instead of doing the regex each time.

This may fall under the "why would you want to" category, but the validates_confirmation_of validation does not properly validate decimal (BigDecimal) columns....

LOL. My bad. I'll work on fixes to them :)

These tests only pass if your computer's clock is set to the same time zone as Time.zone is set to in the test. Change your computer's clock to any other time ...

(from [d75525b045b9f27ed108912a6dbdbad5ab775045]) Added OrderedHash#delete [#113 state:resolved]

Signed-off-by: Joshua Peek

http://github.com/rails/rails/co...

When using delete() on an OrderedHash a user would expect the command to work more like a Hash versus an Array. The deleted key should return the 'hash' value....

When using namespaced models and STI, ActiveRecord stores the demodulized name of the class. This work if all the class in the STI hierarchy are in the same na...

Good call. You could even just re-use IllegalMigrationNameError here, since it's essentially illegal to reuse a migration name. :)

In Rails 1.2.3 (and trunk), if two migrations have the same name (but different versions), the latter will shadow the former, resulting in former migrations no...

I'm still very new to git and while I understand the concept, it still confuses the hell out of me. However, from what I can see "am" is for merging patches fr...

$ git am < ~/Desktop/authenticity-token-ajax-get.diff

Patch does not have a valid e-mail address.

Not sure what you mean but if you need my address, it's chewi@aura-online.co.uk.

+1 ... had some code already close to doing this. did a before and after with applying this change and it is confirmed to remove the unnecessary token

The onl...

I've created a new patch with a new test reflecting the bug I've found and ensuring the correct use of ActionController::Base.resources_path_names. I've follow...

Your right it does sound more natural. Perhaps just leave the level singular and when a plural symbol is passed to the errors method as per Josh's suggestion t...

Using :error instead of :errors to designate the level was deliberate. To me :level => :error, :level => :warning, sound more natural and more in line with log...

Actually a cleaner solution than my first is to just add

:level => :errors

in the default configuration hash for each validation method. No need for an extra...

Ok, I didn't see any use of the level variable outside of the validates_each block in your patch, with the exception of 'validates_presence_of'.

My example is...

Josh, Koz: I'm fine with

def errors(level = nil)
  level ||= :errors
  ...

Adam: I understand your DRY intention, but not how it's going to work. level is...

+1. I use this all the time.

Sorry wrong formatting.

I think I would find this useful also.

To DRY it up a little could you not put the level extraction into the validates_each block abo...

I think I would find this useful also.

To DRY it up a little could you not put the level extraction into the validates_each block above the loop so you only n...

errors(nil) certainly seems like a reasonable approach, do you have any thoughts michael?

Well Michael, I think this little change do not need a indiviual test case in the rails tests. This change is a refactoring, not an addition of features. As yo...

Could you include a test case too please? otherwise we run the risk of breaking it again.

Also if you follow the guide here it'll make it easier to get your ...

In changes 063c393bf0a2eb762770c97f925b7c2867361ad4 and e6a3ce3392812f707b78d64ffb04ee52f4517d20 a condition in map_member_action make Base.resources_path_name...

I'll get to this today.

Actually I see that Cheah Chu Yeow's patch doesn't include a regression test, so it might be useful to use the diff I attached for that.

apostrophe man commin to get me... I wish I could edit lighthouse comments

Does this need to re-opened as a new ticket? I notice that it's status is 'resolved'.

added test

Came across this when using GeoKit to inject a derived distance attribute into an active record when the distance was calculated by the lat/lon of an associate...