Hey guys, Here's the status of the Rails 3 port as far as I know:
1) The rails3 branch has jprovazn's patches to remove delayed job, along with all of the work we have done on the rails3 port so far. 2) Packages for essentially all dependencies are in rawhide. It may still take a bit of time for those packages to propagate out to the mirrors, but the whole thing should be installable "real soon now". 3) Scott had to clean up a migration that was still depending on delayed_job. He may have had to also remove a few other delayed_job remnants that I had left lying around. 4) I've started to take a look at the failure of the daemons to start under systemd. I'm starting to get a handle on systemd, and can now more reliably get mongodb started (my first hurdle) with a native systemd service file. I still have a bit to clean up there, but I should be able to more quickly create native systemd service files for iwhd, imagefactory, and the conductor. 5) Matt has fixed up the routes to the best of his knowledge. 6) Matt has also disabled viewstate for now; we'll have to re-enable this a bit later on, but this should get us going.
We still have a bunch of testing to do, and some problems to hammer out, but I think we are making good progress.
Hey,
On Thu, 2011-07-21 at 16:41 -0400, Chris Lalancette wrote:
Hey guys, Here's the status of the Rails 3 port as far as I know:
- The rails3 branch has jprovazn's patches to remove delayed job, along with
all of the work we have done on the rails3 port so far.
Is there any reason these weren't pushed to master and then merged into the rails 3 branch (or the rails3 branch rebased onto master)?
If stuff like this that isn't specific to rails3 is going into the branch, then it becomes the defacto development branch and we may as well bite the bullet and merge it into master
Cheers, Mark.
On 07/22/11 - 04:16:12PM, Mark McLoughlin wrote:
Hey,
On Thu, 2011-07-21 at 16:41 -0400, Chris Lalancette wrote:
Hey guys, Here's the status of the Rails 3 port as far as I know:
- The rails3 branch has jprovazn's patches to remove delayed job, along with
all of the work we have done on the rails3 port so far.
Is there any reason these weren't pushed to master and then merged into the rails 3 branch (or the rails3 branch rebased onto master)?
If stuff like this that isn't specific to rails3 is going into the branch, then it becomes the defacto development branch and we may as well bite the bullet and merge it into master
Just mostly because they weren't ACKed. If someone reviews and ACKs them, then we can certainly push them onto master right now.
On 07/21/11 - 04:41:45PM, Chris Lalancette wrote:
Hey guys, Here's the status of the Rails 3 port as far as I know:
Another update on rails 3 status:
1) The dependencies are in rawhide, and are out on the mirrors for the most part. There are still some bugfixes that need to percolate through. 2) The rails3 branch in the git repository has code that kind-of sort-of works, sometimes. I'm able to get it to the point where aeolus-configure will configure the system approximately 50% until it barfs. I know where the next problems are, and I have partial solutions, but I ran out of time to debug them more today. 3) Scott has gotten the tests to somewhat work against rails 3. There are outstanding issues there that need to be resolved 4) Both Jay and Matt have actually seen the UI working when they run the thing by hand.
So we are still making progress. We'll work on this more next week, and hopefully get something working in time for the F-16 deadline.
On Fri, 2011-07-22 at 16:23 -0400, Chris Lalancette wrote:
On 07/21/11 - 04:41:45PM, Chris Lalancette wrote:
Hey guys, Here's the status of the Rails 3 port as far as I know:
Another update on rails 3 status:
- The dependencies are in rawhide, and are out on the mirrors for the most
part. There are still some bugfixes that need to percolate through. 2) The rails3 branch in the git repository has code that kind-of sort-of works, sometimes. I'm able to get it to the point where aeolus-configure will configure the system approximately 50% until it barfs. I know where the next problems are, and I have partial solutions, but I ran out of time to debug them more today. 3) Scott has gotten the tests to somewhat work against rails 3. There are outstanding issues there that need to be resolved 4) Both Jay and Matt have actually seen the UI working when they run the thing by hand.
So we are still making progress. We'll work on this more next week, and hopefully get something working in time for the F-16 deadline.
I just fixed a few of the cucumber tests for authentication. Scott, if you work on this more before I do, a lot of the errors seem to be from the removal of f.error_messages, so we need to add a helper method to do it the way we want. An example is here: http://www.emersonlackey.com/article/rails3-error-messages-for-replacement
For poc, I did an in-view change for users/_form.haml, which is pushed with the test fixes I was working on. I'll pick up here monday if nobody has beaten me to it.
-j
On 07/22/11 - 04:23:34PM, Chris Lalancette wrote:
On 07/21/11 - 04:41:45PM, Chris Lalancette wrote:
Hey guys, Here's the status of the Rails 3 port as far as I know:
Another update on rails 3 status:
- The dependencies are in rawhide, and are out on the mirrors for the most
part. There are still some bugfixes that need to percolate through. 2) The rails3 branch in the git repository has code that kind-of sort-of works, sometimes. I'm able to get it to the point where aeolus-configure will configure the system approximately 50% until it barfs. I know where the next problems are, and I have partial solutions, but I ran out of time to debug them more today. 3) Scott has gotten the tests to somewhat work against rails 3. There are outstanding issues there that need to be resolved 4) Both Jay and Matt have actually seen the UI working when they run the thing by hand.
So we are still making progress. We'll work on this more next week, and hopefully get something working in time for the F-16 deadline.
Another update: 1) All dependencies are in rawhide. The final packages should make it to the mirrors by tomorrow morning. 2) The rails3 branch works, for some definition of working. aeolus-configure can mostly complete, although I am still seeing failure "sometimes". After it completes, I can log into the UI, but I get no styling. This still needs to be debugged. 3) Jay got all of the tests working save 1.
Hopefully by the end of tomorrow we can sort the last of these issues out and have a working app.
aeolus-devel@lists.fedorahosted.org