Re: [Rails-core] ActionDispatch::SSL adding STS headers to non-secure redirect responses

2013-06-24 Thread Geoff Buesing
Pull request is here https://github.com/rails/rails/pull/11065 On Sun, Jun 23, 2013 at 6:02 PM, Michael Koziarski wrote: > > On 24/06/2013, at 8:37 AM, Geoff Buesing wrote: > > Trying to reconcile the following: > > ActionDispatch::SSL adds Strict-Transport-Security headers

[Rails-core] ActionDispatch::SSL adding STS headers to non-secure redirect responses

2013-06-23 Thread Geoff Buesing
Trying to reconcile the following: ActionDispatch::SSL adds Strict-Transport-Security headers to all responses, including non-secure redirect-to-https responses... however, the STS spec explicitly says: "An HSTS Host MUST NOT include the STS header field in HTTP responses conveyed over non-secur

Re: [Rails-core] PostgreSQL time zones

2010-01-20 Thread Geoff Buesing
age. On Wed, Jan 20, 2010 at 3:09 AM, Colin Law wrote: > 2010/1/20 Geoff Buesing : > > That sounds reasonable -- ActiveRecord could set the connection timezone > to > > UTC when AR::Base.default_timezone is :utc, and not set it to anything > > (i.e., assume the system local

Re: [Rails-core] PostgreSQL time zones

2010-01-19 Thread Geoff Buesing
That sounds reasonable -- ActiveRecord could set the connection timezone to UTC when AR::Base.default_timezone is :utc, and not set it to anything (i.e., assume the system local timezone is correct) when :local. Solutions have been offered before for accomodating timestamp with timezone, but they

[Rails-core] Re: Fwd: Behaviour of ActiveSupport::Duration when added together

2009-09-01 Thread Geoff Buesing
Great explanation, thanks. Also note that this behavior is consistent with the behavior of Ruby's Date#>> (which Time#advance relies on): >> (d=Date.new(2009,8,31)).to_s => "2009-08-31" >> (d >> 1).to_s => "2009-09-30" >> (d >> 2).to_s => "2009-10-31" >> (d >> 1 >> 1).to_s => "2009-10-30" On A

[Rails-core] Re: getting 'Malformed version number string 3.0.pre' error when running tests

2009-08-07 Thread Geoff Buesing
Looks like the ability to add prerelease versions using letters was added in Rubygems version 1.3.2: http://www.mail-archive.com/rubygems-develop...@rubyforge.org/msg02701.html On Aug 7, 9:17 am, Geoff Buesing wrote: > I had this problem; fixed it by upgrading to the latest > Rubygems

[Rails-core] Re: getting 'Malformed version number string 3.0.pre' error when running tests

2009-08-07 Thread Geoff Buesing
I had this problem; fixed it by upgrading to the latest Rubygems: https://rails.lighthouseapp.com/projects/8994/tickets/2880-malformed-version-number-string-30pre On Aug 7, 2:47 am, Hongli Lai wrote: > On Aug 6, 12:39 pm, Eloy Duran wrote: > > > I was actually talking about this in my last em