Old Business

How to keep track of documentation ideas

• Blog post tags
• Use the Czarships page to add to Wanted Pages drop down? Maybe we want to keep this just for broken links, and instead make short blogs with the tags for keeping track of what we'd like to see documented. We can then take down these blog posts once pages/changes have been made..

Wanted Pages

• Someone to fix plugin to list pages that are linked to but do not exist?

What to document

• underlying science/physics related to modules , algorithms, internal group stuff, computing

What not to document

• your results, science for your research (astrophysics link?)

New Business

Process of making changes

• Read through blog posts/wanted pages together

• Decide if there is a place already on wiki where info could go

If there is, is this place easily accessible from tree structure?

If not, need to make a new page and decide where it lives

• Decide who gets to document it, add to calendar

Wanted Pages/page updates (taken from ​ticket 343)

1. Update calculating run-times (​page)
   • use Eddie's calculations from Kraken
   • relation between dt and dx we derived on blackboard
   • when one does memory estimate the long way — this value should match the size of the output file (the chombo), and be off by ~10x the number reported in standard out — as the global info allocations reported there include both physical and ghost cells whereas memory estimates done the way we outline in wiki count physical cells only.
2. Information on different clusters
   • Do we have a page for this, where is it/where should it be?
   • Bluestreak: helpful commands, how to submit jobs to reserved queue, memory/job limits
   • Kraken: same
   • Queuing system tips- for instance - BS doesn't consider the wall-time of currently running jobs, nor the requested wall-time for other queued jobs. Only queue up jobs in the order they are submitted. That is, they will wait till the resources are available for job 1 to start, rather than look at job 2 if it can run immediately with given resources.
   • Emails back and forth with helpful information from CIRC to go up on wiki
   • Number of nodes/memory per node in different queues, e.g. Afrank queue on BH. Max memory can request.
3. Hypre partition, how/why it can fail on many processors
4. Different types of memory - i.e. stack, virtual, and the amount you need to run your simulation, and the kinds of errors you might see if you hit this limit
5. Organize the things you can do with problem module init
   • i.e. the various pages on objects, etc.
   • where are these pages located and where should the be located
   • what new information do we want for problem module init
6. Forking the wiki
7. Page for Astrobear's capabilities
   • maybe on front page?
   • to resemble- ​this
8. Running Visit in parallel
9. How to add time-slider in physical units to Visit
10. Update page on how to make column density maps
11. FAQ page for new users
    • do we want to get rid of the forums/discussion boards ?
    • e.g. "problems related with restarts in MHD jets"
    • where should this be? main page?
12. Developer guide
    • we should go through the guide as a group in future meetings to identify what needs improvement
13. New user's guid
    • we talked about possibly putting together a 1 page 'quick start' guide for New Users
14. Internal User's stuff
    • how to tagging blog posts
15. How to plot streamlines in visit

Assignments/due dates

• calendar plug-in ideas?
• tickets ?

Task

Assigned

Due-date

Minutes

Need to add procedure to main page for documentation

Upcoming meeting topics

• Changes to structure of wiki?
• New users guide
• Advanced users guide