Meeting 2018 02 24

From PCGen Wiki
Jump to: navigation, search

Attendance:

  • Benevolent Dictator - Bryan
  • Chair - Kar
  • Code - Tom
  • Content - Andrew
  • PR - Paul
  • Data 2nd - Gwen
  • Observer - Douglas
  • Observer - David

Summary:

  • PR - Willing to support local volunteers for Conventions for 2018
  • PR - No GenCon for 2018 planned
  • PR - Will review Data Books waiting on queue
  • Content - Fixing bugs and making publishers happy
  • Code - Several GATE issues prevent any Production Release
  • Code - Issues are fixed, just in a queue to be PR and reviewed
  • Code - Will improve WIKI to get new developers up and running
  • Roadmap - Target is May/June for Production release 6.08.00



Redacted Log

  • [6:00 AM] Andrew Maitland: Good morning all, welcome to the Board of Directors meeting
  • [6:00 AM] Andrew Maitland: The agenda on the table for today is:
  • [6:00 AM] Andrew Maitland: Agenda: 1) Team Reports (10 minutes) 2) Roadmap Progress / Release Planning (10 minutes) 3) Tom's Corner (35 minutes) 4) Public Q&A (5 minutes)
  • [6:01 AM] Andrew Maitland: @PaulGrosse would you like to kick this off from the PR?
  • [6:04 AM] Paul Grosse: Lets see, we've had some books cleared for release, theres a pile of them that I will be getting to next week while wife is on vacation. :)
  • [6:05 AM] Paul Grosse: We've had a good response from the call for help, and there's been a trickle of volunteers coming in
  • [6:06 AM] Paul Grosse: I need to update the news page on the 6.07.06 release while the ability to update the webpage was broken
  • [6:06 AM] Paul Grosse: I think that's it for me
  • [6:07 AM] Bryan McRoberts: Nice
  • [6:07 AM] Andrew Maitland: Any plans for convention presence this year?
  • [6:09 AM] Paul Grosse: GenCon is already out we would have had to sign up last year
  • [6:10 AM] Andrew Maitland: Okay, when would we need to plan for 2019's convention then?
  • [6:10 AM] Paul Grosse: We almost have the funds already though if we want to aim for next year. Other then that there some other ones we could aim at.
  • [6:11 AM] Andrew Maitland: Cool. I think we should encourage local volunteers to hit up their local area conventions to get to a wider audience if possible. :)
  • [6:11 AM] Paul Grosse: GaryCon, Origins, a couple others I can't remember ATM
  • [6:11 AM] Andrew Maitland: No booth necessary, just pass out flyers.
  • [6:11 AM] Andrew Maitland: (or run games using PCGen created sheets).
  • [6:12 AM] Paul Grosse: We could even cover a booth if they had a team to cover it, most cons are around 40-150ish for a booth depending on size
  • [6:12 AM] Andrew Maitland: Okay, for the Content Team -
  • [6:12 AM] Andrew Maitland: @GwenT and the data volunteers have been pounding out a solid stream of books to keep @PaulGrosse fairly busy. :)
  • [6:13 AM] Andrew Maitland: @ReganAnderson is working on revamping the Sorcerer Bloodlines
  • [6:14 AM] Bryan McRoberts: When is the next production release?
  • [6:14 AM] Andrew Maitland: We'll hit that in a moment Bryan.
  • [6:14 AM] Andrew Maitland: We've discovered they were not coded up correctly, and in order to properly handling the various archetypes, a major overhaul was required (The existing system was broken, and un-maintainable).
  • [6:14 AM] Bryan McRoberts: That's a big undertaking by Regan
  • [6:15 AM] Andrew Maitland: It is, but I've been supporting him and letting him tackle this project at his pace.
  • [6:15 AM] Andrew Maitland: However, since we took a rather large chunk of Usability out, I've delayed the next release until we can include that back in.
  • [6:16 AM] Andrew Maitland: I updated our Saga Born data and connected Output sheets to the delight of one of our Publishers.
  • [6:16 AM] Paul Grosse: :)
  • [6:16 AM] Bryan McRoberts: Very cool!
  • [6:17 AM] Andrew Maitland: Other than that, I've been handling bugs that are trickling in.
  • [6:17 AM] Andrew Maitland: Any questions before we move on to Code team report from @TomParker ?
  • [6:17 AM] Bryan McRoberts: Well done, data team! Paul - thanks for all your work with PR as well!
  • [6:18 AM] Bryan McRoberts: What % done would you say the Sorcerer work is?
  • [6:18 AM] Bryan McRoberts: Just ballpark
  • [6:19 AM] Andrew Maitland: With APG, CR and UM pretty completed, supporting the Eldritch feats, and Crossblooded, leaving only the Wildblooded, I'd say 80-90% since I'm not the one doing the work.
  • [6:20 AM] Bryan McRoberts: Excellent
  • [6:20 AM] Gwen T: Sorry I'm late. Over slept
  • [6:20 AM] Bryan McRoberts: GM Gwen
  • [6:21 AM] Andrew Maitland: Morning Gwen
  • [6:21 AM] Gwen T: Good Morning... or it will be when coffee finishes brewing. :p
  • [6:21 AM] Andrew Maitland: Okay @TomParker code team is probably going to cover agenda #2 and #3 as well, so feel free to hit on all of those. :)
  • [6:22 AM] Tom Parker: Continuing work to simplify the core infrastructure - a bunch of internal changes there that don't really appear externally
  • [6:22 AM] Tom Parker: but have removed a lot of one-off cases in the code
  • [6:23 AM] Bryan McRoberts: Nice
  • [6:23 AM] Tom Parker: Have a bit of an issue at the moment that's preventing us from retiring the "built-in" default race
  • [6:23 AM] Tom Parker: Effectively, FACT / FACTSET leak between loaded game modes/campaigns
  • [6:24 AM] Tom Parker: which could cause issues if things are swapped in and out, especially if one of the campaigns causes a FACT to be required
  • [6:25 AM] Tom Parker: That's going to gate other things on my end until I get it addressed, since the fix will likely be a bit invasive to a number of places in the code
  • [6:25 AM] Tom Parker: The new formula system is doing work behind the scenes already, so we know it works to the first order
  • [6:25 AM] Tom Parker: (FACE is currently done for all game modes, for example, even if it's not apparent to the data team)
  • [6:25 AM] martijnverburg: Sorry I’m late - toddler battles!
  • [6:26 AM] Bryan McRoberts: Martijn - lol
  • [6:27 AM] Tom Parker: Speaking of which. ... Martijn has been suggesting some use of DI, but given that we are so heavily data driven, that is a huge challenge.... I rapidly hit the point where a DI framework is unusable or unruly... so could use a more detailed discussion with Martijn (and others who want to participate) at some point
  • [6:27 AM] Bryan McRoberts: Tom - would the leak gate a production release? Or has this not been folded into the master branch yet?
  • [6:27 AM] Tom Parker: The leak is in master
  • [6:27 AM] Tom Parker: has been for a long time actually
  • [6:27 AM] Tom Parker: It shouldn't gate another 6.7.x release
  • [6:28 AM] Tom Parker: but should probably gate a 6.8
  • [6:28 AM] Bryan McRoberts: Ok
  • [6:28 AM] Tom Parker: On the bug front, have addressed some of those, and Marco (one of the new folks that has helped out) diagnosed a few of them to where it appeared to be a data issue
  • [6:30 AM] martijnverburg: Yeah I need to understand the code better before making spurious suggestions
  • [6:30 AM] Tom Parker: On the new volunteer front for code engagement and retention is a bit of an issue.  There aren't a lot of "easy projects" around, and I can point people at some bugs to go look at to start to get familiar with things but that doesn't always lead to them actually looking
  • [6:32 AM] Tom Parker: On the wider side of things, there are updated docs on the formula system on the wiki
  • [6:32 AM] Tom Parker: As well as a proposal on the CHOOSE replacement (which will need to be tweaked a bit, but gives the general sense for now)
  • [6:33 AM] Tom Parker: The new developer startup page is also updated, with a review and some feedback from folks as well
  • [6:33 AM] Tom Parker: If anyone has other things to consider for how it might keep (or get) people engaged, let me know
  • [6:34 AM] Andrew Maitland: Should we list out useful projects for people to look at?
  • [6:35 AM] Andrew Maitland: Like a roadmap with the subtasks looking to accomplish perhaps?
  • [6:35 AM] Gwen T: I'd like to see one of our "splash screens" that pop up while loading used for recruitment
  • [6:36 AM] Tom Parker: I think the question is what level of detail is needed
  • [6:36 AM] Tom Parker: not sure I know, so it needs to be more of a discussion with folks
  • [6:37 AM] Tom Parker: otherwise I'm just writing tons of stuff down for no effect
  • [6:37 AM] Tom Parker: As far as the splash screen, I'd be cautious about driving more recruitment if we can't retain - we should try to figure out how to retain the trickle we have now before we open a flood gate
  • [6:37 AM] Tom Parker: Not sure I have anything else for today
  • [6:38 AM] Bryan McRoberts: Is the new developer document on the wiki?
  • [6:38 AM] Andrew Maitland: Based upon current code gates to the formula system, where do we see the next production release potentially happening?
  • [6:38 AM] Avilay: is there a way to sort of tasks according to urgency?
  • [6:39 AM] Andrew Maitland: Avilay - in Jira yes, on the wiki, it's however they are listed out.
  • [6:39 AM] Tom Parker: http://wiki.pcgen.org/Basic_Developer_Setup
  • [6:39 AM] Tom Parker: JIRA though it a LONG list of items
  • [6:39 AM] Tom Parker: with lots of mixes of complexity
  • [6:40 AM] Tom Parker: The problem being we don't know the skill set of devs coming in - we've dealt with all levels
  • [6:40 AM] Bryan McRoberts: I don't see a link to that on the main wiki page
  • [6:40 AM] Gwen T: @AndrewMaitland When we port to Stride, we should look at getting Trillo boards for that!
  • [6:40 AM] Tom Parker: Someone who is new to Java probably can't take on a deep diagnosis
  • [6:42 AM] Andrew Maitland: I think the retention is at least a two-fold problem
  • [6:42 AM] Tom Parker: I'm not sure how to do that balancing without a conversation with those folks, but when I've tended to ask questions about skills et al, so I have a sense of where to start, that seems to drive some away, and at times it feels others go silent when I don't assume they know what they are doing (because they may be an experienced dev but I really don't know that)
  • [6:43 AM] Andrew Maitland: Getting set up seems to be a big hurdle (I cannot get set up with the newest Eclipse) as an example. And then getting them plugged into an easy communication with the teams seems to be another part.
  • [6:43 AM] martijnverburg: I think a basic tutorial could be written on getting a piece of data from a LST file through to the UI and exported to a character sheet
  • [6:44 AM] Tom Parker: @MartijnVerburg Fair point.  Part of the challenge being there are a few ways to do it today - one strategic, and some legacy
  • [6:45 AM] Tom Parker: I will at least write up the two primary ones
  • [6:45 AM] Bryan McRoberts: Martijn - great idea. Something that explains a bonus that has a prereq and how it gets to a character sheet and UI would take a dev through the main functions of the code.
  • [6:45 AM] Bryan McRoberts: Same for the new formula code
  • [6:45 AM] martijnverburg: Yeah I’ll admit I’m lost on the legacy way vs current way vs proposed future way
  • [6:46 AM] Bryan McRoberts: As Andrew noted, making sure the new dev instructions gets someone up and running fairly quickly is paramount
  • [6:46 AM] Tom Parker: All a great point.  Barring some major issue, I will write that this week
  • [6:47 AM] Bryan McRoberts: I'd be a good guinea pig.  My build was working, I did an update and now it's broken and I haven't had/made time to figure it out.
  • [6:47 AM] martijnverburg: I’ll happily review / try it out as well - happy to write some as well if that helps with the load
  • [6:48 AM] Tom Parker: @BryanMcRoberts There are a few situations where an update can break things, and I think you joined when we did one of them
  • [6:48 AM] Tom Parker: sorry, LOL not joined
  • [6:48 AM] Tom Parker: brought up your dev environment :)
  • [6:48 AM] Bryan McRoberts: right :)
  • [6:48 AM] Tom Parker: So you may need to delete the project and re-import it as a new Gradle project to get Gradle to behave
  • [6:49 AM] Tom Parker: (so delete from eclipse but not from the disk)
  • [6:49 AM] Bryan McRoberts: ah
  • [6:49 AM] martijnverburg: 👍
  • [6:50 AM] Tom Parker: I removed a few key files from git that Gradle can now consistently build correctly, and that breaks you if you didn't realize the impact
  • [6:50 AM] Tom Parker: Thankfully those types of rebuild events are rare
  • [6:51 AM] Bryan McRoberts: cool
  • [6:52 AM] Tom Parker: If it doesn't work for you let me know and we can try to find some time to walk through what is misbehaving
  • [6:52 AM] Bryan McRoberts: ok - thanks
  • [6:53 AM] Tom Parker: Any other questions or suggestions ?
  • [6:53 AM] Andrew Maitland: Tom - on the Jira side, I enabled tasks for data, which allows for sub issues inside on main issue.
  • [6:53 AM] Andrew Maitland: Would that be helpful for the code and new tag projects?
  • [6:54 AM] Tom Parker: My sense if probably not, but I'll think about it.  For the few times I've needed that I've just created a Mini-Project and linked the issues together
  • [6:54 AM] Bryan McRoberts: Not from me - great work on the code, Tom.  Hopefully Martijn, I and others can become more productive to lessen the load on you.
  • [6:54 AM] Andrew Maitland: And the last question Bryan had that we don't have answer for - Production Release?
  • [6:55 AM] Bryan McRoberts: Drum roll....
  • [6:55 AM] Tom Parker: So my big gates
  • [6:55 AM] Tom Parker: FACT / FACTSET needs fixing
  • [6:56 AM] Tom Parker: GROUPING needs fixing, since it's a nasty inner class hack at the moment
  • [6:56 AM] Tom Parker: I need to fix a Scope/Format issue that impacts what is currently dropIntoContext
  • [6:57 AM] Tom Parker: and in the process of changing that convert to getOther instead (better semantics)
  • [6:57 AM] Tom Parker: the latter two are needed so we don't end up with MODIFY/MODIFYOTHER statements that I immediately need to break -
  • [6:57 AM] Tom Parker: that should all happen in one Alpha cycle
  • [6:58 AM] Tom Parker: The GROUPING item is done, but about 4th in line on my local system for pushing
  • [6:58 AM] Tom Parker: (I have a HUGE set of interacting PRs, one of which turns out is gated by the FACT fix)
  • [6:59 AM] Tom Parker: The Scope/Format fix has had a bunch of the work done in the Formula Library; let me look at whether it is already integrated
  • [7:00 AM] Tom Parker: So Scope/Format is partially in the formula library
  • [7:01 AM] Tom Parker: I have a few other PRs to go in there, but I try to integrate things that are incompatible interface changes one at a time into the master of PCGen, so avoid confusing myself too much :)
  • [7:01 AM] Tom Parker: The other items I need are built, just not Pushed
  • [7:01 AM] Tom Parker: So once I get the FACT thing down, it's probably a few weeks to get the rest pushed through.  80% of the work is done and I just get to do a slew of conflict resolution to create valid PRs
  • [7:02 AM] Tom Parker: So maybe target moving to Beta in April?
  • [7:03 AM] Andrew Maitland: Wow, that's faster than I anticipated.
  • [7:03 AM] Andrew Maitland: I guess the better question is what do we expect in the next production release.
  • [7:03 AM] Tom Parker: That's a fair question, actually
  • [7:04 AM] Tom Parker: The April statement is about as fast as I can do it
  • [7:04 AM] Andrew Maitland: If we want single gamemode conversion, then I'm probably going to need a couple of months at least.
  • [7:05 AM] Tom Parker: Yea, I'm not sure everything you need for that is April
  • [7:05 AM] Tom Parker: I was answering "If you need to get stuff stable enough to be usable, what would it be"
  • [7:05 AM] Tom Parker: Not "If PF conversion will be complete, what is the date"
  • [7:06 AM] Tom Parker: At the same time, I'm not sure we should get a full conversion done
  • [7:06 AM] Tom Parker: There are 2 or 3, perhaps 4 changes that are very likely to break PCs
  • [7:06 AM] Andrew Maitland: Question to the group - Push out a stable 6.08 quickly, and focus on Code Formula completion and data conversion for a 6.10?
  • [7:06 AM] Tom Parker: The problem being that PCGen today discards information based on how it stores PCs
  • [7:07 AM] Tom Parker: A classic example is Domain is only loosely stored to the Class... but it's hard to do a dual class with two Domains
  • [7:07 AM] Bryan McRoberts: It'd be nice to get a prod release out since it's been so long and a lot of data work has been done.
  • [7:07 AM] Andrew Maitland: Agreed.
  • [7:07 AM] Tom Parker: When we fix that, the PCG file format  would need to change, and then we end up with a compatibility gap during load
  • [7:07 AM] Tom Parker: The Domain one we probably just let break and people re-select the Domain - pretty innocent
  • [7:07 AM] Tom Parker: But they become more invasive from there
  • [7:08 AM] Bryan McRoberts: I don't think anyone would complain about a prod release soon-ish and another one in April/May.
  • [7:08 AM] Tom Parker: I'm not prepared for one soon-ish
  • [7:08 AM] Tom Parker: (under that definition of soon-ish)
  • [7:08 AM] Bryan McRoberts: Ok - so there are current code issues gating a release from master?
  • [7:09 AM] Tom Parker: The 3 I listed a bit ago
  • [7:09 AM] Tom Parker: The "big gates"
  • [7:09 AM] Andrew Maitland: We fix those 3, Push out a Production 6.08.00 May 15th, since we need at least 1 RC?
  • [7:09 AM] Gwen T: I'd love to see a stable 6.8 as soon as reasonable. 2/3 of the issues I run across are because folks only like to use the "stable" versions. Currently, that's 6.06.01, released 2016-07-30
  • [7:10 AM] Bryan McRoberts: Ok - I thought those were related to dev efforts aimed at an April release.
  • [7:10 AM] Gwen T: We've killerd a lot of bugs since then!
  • [7:10 AM] Tom Parker: Gwen has a good point
  • [7:10 AM] Tom Parker: Andrew, I think it's better to call out early may for RC1
  • [7:10 AM] Bryan McRoberts: Would it be work putting out just a data release?
  • [7:10 AM] Tom Parker: and then go from there
  • [7:11 AM] Andrew Maitland: Agreed, people don't touch the latest releases, and we get a lot of redundant reports.
  • [7:11 AM] Tom Parker: We can't really do just a data release
  • [7:11 AM] Tom Parker: the data is compatible with the current code, not the 6.6 code
  • [7:11 AM] Bryan McRoberts: Ah
  • [7:11 AM] Andrew Maitland: We can, but it gets sloppy... so a nice clean prod release
  • [7:11 AM] Andrew Maitland: I'm still getting Starfinder 1.6 bug reports. :(
  • [7:12 AM] Paul Grosse: And we are going to definitely want to have SF in the prod release. :)
  • [7:12 AM] Andrew Maitland: Okay, Target May/June for 6.08.00 release.
  • [7:12 AM] Bryan McRoberts: Sounds good
  • [7:13 AM] Andrew Maitland: That means we drop Alpha and switch to Beta no later then end of April.
  • [7:14 AM] Gwen T: That SHOULD allow for at least a couple RCs, if not a final version, in time for the summer conventions.
  • [7:14 AM] Andrew Maitland: We've gone over time a bit, but any questions from anyone else before we close this up?
  • [7:16 AM] Andrew Maitland: Thanks for coming everyone. Meeting adjourned - **Bangs Gavel**
Retrieved from "http://159.203.101.162/w/index.php?title=Meeting_2018_02_24&oldid=4209"