Skip to content
This repository

HTTPS clone URL

Subversion checkout URL

You can clone with HTTPS or Subversion.

Download ZIP
Fetching contributors…

Cannot retrieve contributors at this time

file 30 lines (29 sloc) 1.546 kb
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
TODO:

DONE:
  * Make it possible to clean up jobs through something like 'crowd cleanup'.
  * Think about adding a CLEANUP state, with WorkUnits being dispatched to
    workers to clean up a Job's leftover files on S3, instead of the current
    in-request cleanup. (see Job#cleanup_assets)
  * Right now actions just process. Think about map / process / reduce.
  * Basic unit tests.
  * Job creation UI. Pick an action, add inputs, set options, go. (not needed)
  * Look into Grand Central Dispatch for ideas about how to implement
    Job ordering/piping around.
  * Finish fleshing out the crowd command.
  * Security. HTTP Basic Auth via config.yml,
  * Security. S3 authenticated URLs.
  * Use the email field to send mail when the Job's gone bad. (bad idea)
  * Benchmarking mode -- benchmark every single aspect of a job and report it.
  * Pare down the code that the workers are loading to the absolute minimum.
  * Documentation. (document inside of config.yml, too)
  * Think about having the completion of a work unit respond with another one,
    instead of having to make a second request.
  * Think about having a failed work unit shuffle off to the end of the queue,
    instead of going again right away (and probably being handed back to the
    very same daemon).
  * Replace work_units_remaining with a percent_complete
  * Monitor UI -- see queue of jobs / work units.
  * Basic integration tests.
  * Unit tests for lib.
  * Bump to version 0.1 and release on RubyForge.
  * Re-architected to use real Nodes (version 0.2).
Something went wrong with that request. Please try again.