0.3.5 Release

Mainly just a "clean up" release. No functional changes yet.

Soon to come?


  1. Still rolling around the idea of a file hashing application using sha or md5 check summing with per-server configurable areas along with standard defaults.
  2. I'm still planning on moving the scheduler to a daemon process that will allow for more cron-like scheduling flexibility.
  3. The next release will likely include a stage enable flag in workflows that will allow you to disable individual stages of a workflow. 



0.3.4 available for download

Fixes include all of the aforementioned changes along with some better handling of temporary files. Instead of an inline file and workflow view I opted for separate view within the right frame. New screen shots will be made available here. Additionally, Stuart Langridge's sorttable.js is not included in this release.

Download it here!


Still working on 0.3.4

0.3.4 will be released soon, planned features are:


  • An "inline" view of workflow outputs, file content, deltas, server hardware data, and note content. This is expirimental right now, and doesn't work well on IE 7 or earlier browsers due to CSS limitations, but the design goal I'm trying to solve for here is to keep people from feeling "buried" within content views.
  • Scheduling has been improved to allow for individual workflow schedules, along with customizable hour, day of week, and day of month combinations.
  • Tentative work on a "queue" application designed to enhance the scheduler even more (by adding a "minute of hour" field for instance) and allow for instant gratification processing. This will come in tandem with a redesign of the scheduler which will act like a daemon process, watching the queue application for updates and running them. (This feature will probably be fully functional in a later version, but I will include the starting ideas in 0.3.4).
  • The scheduler will now be multi-threaded (and configurable in Workflow executions and file operations triggered by the scheduler will now spread across your cores evenly using as many parallel threads as you deem appropriate. This will advance the python requirement to 2.6+ (or 2.5 with the "multiprocessing" module installed as a back port). I will most likely provide the original, serial execution behavior as a fall back (no reason not to really).
  • Templates will be consolidated and more efficient, taking advantage of django's inheritance model.
  • A simple graphing algorithm is integrated into workflow output views, allowing you to have some rough analysis capabilities for workflows that essentially "count" information. Create a workflow that simply prints an integer output and you can view a graph of that stages output for a given day across all of your systems, or across a single server.
  • Hoping to have the search and pagination combo situation fixed, or search performance should at least get enhanced (somewhat) while remaining unpaginated (I'd really like to get both together though). Searching and pagination issues in 0.3.3 will be addressed in 0.3.4 using session cookies to track your search terms. Searches all across the applications will be paginated and using the model "Q" methods in django as they should. Pagination itself has also been modified slightly to include a "first" and "last" page link, and the text will be slightly smaller to reduce emphasis (pages are convenient, but not something that should really catch the eye constantly).



Revamped Documentation

Here is the newly revised documentation. Highlights include simple apache integration through mod_wsgi and how to get your SSH environment set up so that the can operate in the cron.

It's not quite complete yet but neither is Chimera itself, so from this standpoint the documentation is fairly comprehensive!


Search is a bit buggy...

Search is a bit buggy in some apps in 0.3.3. I'm going to try and hash this out for 0.3.4. Part of the issue is pagination and trying to unify the index and search views. I may have to revert back to using a separate search template here.