=== GOALS: === Become aware of and familiar with all aspects of contributing and developing to the !SasView project * Lower the “fear of the unknown” barrier to getting started (by knowing rather than guessing at what is involved) * A basic understanding of “what/how I might be able to contribute” * Where to go to get started. Where to find information, who to talk to etc. === OUTLINE === 2h 45 min + 15 min of final question and discussion Paul Butler - 25 min + 5 min questions * Welcome and intro = All the areas are important (hard core coding, documentation, checking math, testing that we get the right answers, testing usability and against crashes, responding to help requests)... * goals and outline * Demo of existing functionality - going through all the menu items and buttons * Questions/discussion Andrew Jackson - 20 min + 5 min questions * Getting Help and Bug reporting -- tickets (user list, help@) - '''Needs:''' ''reporting bugs and responders to queries (from “how do i” to “why does my model not fit my data”)'' * How to write and distribute a model and the model marketplace. '''Needs:''' ''contributing models'' * Questions/discussion [http://trac.sasview.org/attachment/wiki/CodeCampIVPlanning/CodeStructure.pdf​ Wojciech Potrzebowski] - 35 min + 5 min questions * Code structure (where bits live) - '''Needs:''' ''contributing non model code and testing and validation'' * Overview - !SasView, sasmodels, bumps * !SasView * sasGUI * sasCalc * sasmodels * Questions/discussion Stephen King and Paul Butler - 20 min + 5 min questions * Documentation - '''Needs:''' ''contributing documentation'' * Overview * Use Sphinx to build both dev and user * Uses RST and Latex - currently builds html but plan for a pdf also * Dev - '''Needs:''' ''lots of doc strings still missing'' * docstrings at the start of “each” class and method * User - '''Needs:''' ''never finished'' * Each section has an RST file (and images as appropriate) * Models = the first doc string in the file is scraped to go into model documentation * 1 min - Tutorial …. Need a new one - '''Needs:''' ''help developing tutorial materials'' * Questions/discussion Andrew Jackson and Wojciech Potrzebowski - 40 min + 5 min questions * “Dev ops” * Basic concepts in version control - '''EVERYTHING''' is under version control!! * Git code hosting * Concepts in version control (repos, cloning, branching (vs. forking - baaaad), committing (with comments),pull requests, code review * Setting notifications etc * “show me the code”- '''Needs:''' ''developers of all stripes = architecture, GPU, math whiz, sysadmins, ...'' * Git code repos (all of them), branches, stats… and CODE * Compiling/run.py and local environment * Py2exe and Ino -> installer * CI - eventually NOT local environment - real test is final build * Travis * Jenkins * Progress -->trac → ticket -> work * Trac wiki * Trac tickets, milestones, workpackages, roadmap (trac version) * Questions/discussion GENERAL QUESTIONS/DISCUSSION FEEDBACK FROM AUDIENCE * Scripting language for fit constraints '''- SUPPORTED (MOSTLY .. still a bit of WIP) using python''' * Checkpoints for fitting results '''- TICKET EXISTS - should it be a higher priority? Who can do it?''' * Subtraction of the parameter as Sylvian suggested '''- NOT CURRENTLY CONSIDERED EXACTLY (full decomposition is being considered currently)''' * Generic scattering calculation - where does sld come from? Does it include all Hs? Does it account for exchange of H? '''- CURRENTLY JUST AS IS PDB''' * Allow step files (cad format) as input to Generic scattering calculation '''- NOT CURRENTLY CONSIDERED, but would be a very good idea if we put effort on G. scatterring.''' * Chain fitting does not all sorting the order of the files making it not very useful '''- BUG''' * Generate PDF file from sasmodels docs '''- WIP''' * For some applications one needs to use the complex contrast '''- NOT CURRENTLY SUPPORTED - in discussion''' * q vector out of detector plane (out of Ewald sphere): is this treated correctly? '''- NOT CURRENTLY SUPPORTED''' * Perform all operations where the value is independent of the q values before going into the loop computing Iqs - '''A LOT IS DONE -- more could be''' * Simulation code to be removed '''- TO BE DONE .. not top priority''' * Supply ascii file of test date used in the sasmodel documentation png file as part of model documentation. '''- SHOULD BE LOW HANGING FRUIT?'''