On Mar 29, 2006, at 2:09 PM, Steve L. wrote:
I think the issue with this might be that the Typo environment.rb
includes the vendor/rails/*/lib in the config.load_paths and that
might be why:
require_gem ‘rails’, ‘= 1.1.0’
doesn’t work so good.
I looked at this the other day and environment.rb actually includes
boot.rb which does a ‘require’ for a file inside the rails gem, so I
wasn’t sure what the right way was to lock it to a specific rails
(since I figured there might be a good reason why it was requiring
just that one file and not the whole gem).
Really the Typo team dropped the ball on this one, it isn’t like
they didn’t know that Rails 1.1 was coming or that they were going
to have problems. I am disappointed with the response from them
which has been near silence.
Near silence? I take it you never even bothered to drop by the IRC
channel #typo (on freenode) where we’ve been discussing this for some
time. The problem is that there’s apparently no way to maintain both
1.0 and 1.1 compatibility with Typo (don’t know why, but that’s what
we’ve discovered) and so we decided to hold off updating until 1.1
was actually out the door. And now that it is, we’ve been working on
getting Typo running on it. If you look at the IRC channel we have
instructions for locking Typo to 1.0 rails (using the
rake freeze_edge REVISION=3303 trick) and warnings that Typo trunk only
works on 1.0. And once our new branch has been verified to work on
Rails 1.1, we’ll update trunk and tell everybody to upgrade rails.
Even when they are ready the memory consumption/CPU usage now
exceeds the max per process on Textdrive. So without some clean up
to the codebase people will be in for a shock when they can get to
Rails 1.1 and Typo 4.0 on a shred host because it will get killed
That’s actually a problem with TextDrive. We don’t know why it
happens, but TextDrive causes Typo to use approximately twice as much
RAM as every other system I’ve seen statistics on. I don’t know what
to do about it.