search

NAVADMC / ADSM

A simulation of disease spread in livestock populations. Includes detection and containment simulation.
Other
10 stars 5 forks source link

October 2015 Priorities #673

Closed josiahseaman closed 6 years ago

josiahseaman commented 9 years ago

This is the complete tracker issue for the round. This can largely replace ZenHub and you'll be notified whenever I comment on this issue.

Priorities for October - December 2015:

josiahseaman commented 8 years ago

I've added time estimations to each of the issues above. I left Vaccination Rings blank because to the best of my recollection I'm done with my piece of that. Once it's all tied in we can test and discuss if #386 matches the desired outcome.

josiahseaman commented 8 years ago

@missyschoenbaum I'll be done with Control Protocol soon and then I can move onto Population Panel, then Assign Spread (which will take a while). I'm going to recommend that before I start on either of those we take around 4 days to do Performance Optimization. I now have some better tools for profiling what might be slowing down ADSM. This would be focused on eliminating the delay in navigation (which may be 1-2 seconds) right now and bringing that down to 1/8 second or an unnoticable delay. What do you think?

boomtown15 commented 8 years ago

I put some general calendar times for tasks based on Josiah's estimates. This should help us target, plan and keep focused. @josiahseaman, feel free to update if any of those dates are off.

josiahseaman commented 8 years ago

Ah man, you're making me do date math? So if I'm logging 16 hours per week, then that's actually 1/2 my normal. So when I say 3 weeks, it's really 6 calendar weeks. I think that pushes #649 up to December. I don't think I'm taking off time for Christmas. Then #668 #629 would extend into January.

boomtown15 commented 8 years ago

I know, it's rough. We do need a general calendar schedule so the rest of us can plan accordingly though.

@josiahseaman - I've updated dates above. Please update if anything is still inaccurate.

josiahseaman commented 8 years ago

I've updated the date math from our last meeting. I note that adding in bug fixes from Kelly's work is increasing the scope of the original estimate, so that pushing things back some. Bug fixing may also be a good way to get Neil more familiarized with the code base. I'm going to leave some comments on the bug issues about the architecture and possible causes.

josiahseaman commented 8 years ago

It's December 2nd today and we're currently on track with my date math:

React/Redux is working out to be a really productive and stable platform on my other project, so I look forward to using it on Assign Spread.

josiahseaman commented 8 years ago

December 18th: we're a bit ahead of schedule since I've already completed the Population Status Lights and several related auxillary issues, as well as giving the Outputs a face lift in #696. I added to the original post (top) the bug fixes that we've thrown in along the way. Sometimes the question comes up of why we're behind schedule and it's usually unscheduled fixes or additional features that get inserted into the timeline because they're high priority. I thought a list of these would be helpful in retrospect.

Currently, Bryan is working on making our compile chain more robust, which was an unexpected issue that has taken up about a day of his time. "I just need to get Fortran77 to compile the Intel Kernel files". Fun fact: between Scipy (Fortran 1977) and React Redux (December 2015) our dependencies span 38 years of development history. They all need to smoothly interoperate with each other.

ConradSelig commented 6 years ago

Reference #858 - Closing