Meeting 2016 05 28

From PCGen
Jump to: navigation, search

Non-Quorum Meeting

  • SBs Present: Andrew M (Content), Paul Grosse (PR)
  • 2nds Present: Doug L. (Data), Stefan (OS)
  • Observers: None

Summary

  • 6.06.01 will be release around June 12th
  • A lot of Data Code Improvement for the Eidolons, the Summoner, and Summoner Archetypes to handle various incarnations and Unchained versions (A lot of refactoring)
  • Animal Companions got a touch up as well
  • Seekers of Secrets Wayfinders got an overhaul (Ioun Stones changed and code to handle)
  • Occult Adventures has been making progress
  • Paul reports 6.06.01 will be just in time for Gen Con (with CDs to bring!)
  • Github Policy discussed
  • New Data Changes for Archetypes discussed (FACT Based, phasing out TYPE based)
  • Stefan requested people test his Spell Loader change (Moved from old loader to GenericsLoader)


Raw Log

  • <@AndrewAAM> *Bangs Gavel*
  • <@AndrewAAM> Welcome to the PCGen Board of Directors Meeting, May 28th
  • <@AndrewAAM> Agenda is:
  • <@AndrewAAM> #1 - 6.06.01 RC2 (Looks like this is a must)
  • <@AndrewAAM> #2 - 7.00 development roadmap - Are we on track? (Need feedback from Tom and Connor)
  • <@AndrewAAM> #3 - Open Forum
  • <@AndrewAAM> #4 - Per Paul G. Is how to keep Gwen around longer...
  • <@AndrewAAM> I'll start with #1
  • <@AndrewAAM> We hit a major Oopsie bump during some of the Animal Companion refactoring that will force another RC. Though, it gives us time to get the Eidolon code changes (Massive refactoring) time to get played with.
  • <@AndrewAAM> I'm sure the community will enjoy their Uber Buff companions, but this is not legal play, and required fixing.
  • <[OGL]Nylanfs> Do we have more people putting bug reports in on the RC this time?
  • <DistantScholar> Understandable. I wonder how many other major Oopsie bumps we'll run across?
  • <@AndrewAAM> Are you not monitoring the Tracker traffic? I got a deluge this week alone.
  • <[OGL]Nylanfs> Yea, i just wondered if that was internal or external
  • <[OGL]Nylanfs> External is great :-)
  • <@AndrewAAM> I think I closed out 7 Eidolon specific bug trackers, 5 or 6 Summoner Archetype specific bugs, a couple of Eidolon Feature Requests, and several Unchained version fixes.
  • <@AndrewAAM> And some of these were not fun to tackle. But the end result is good. I reduced the lines of code for Eidolon significantly and made it more streamlined!
  • <DistantScholar> Hooray!
  • <@AndrewAAM> In the process, Animal Companions and Special Mounts got to enjoy this refactoring goodness.
  • <@AndrewAAM> I also tackled the Wayfinder Ioun Stone process, added ALL the resonant powers, and revamped how Ioun Stones in general work.
  • <DistantScholar> I know there's some stuff in 3pp books which are based on the animal companion/eidolon structure; how much updating is needed for such things?
  • <@AndrewAAM> 3pp Eidolons followed the same illogical set up "Own CompanionMod set up" I'll take a look at reducing those wasteful set ups as well. But for now, they work as-is.
  • <@AndrewAAM> I only saw one that would require a clean up.
  • <@AndrewAAM> I planned for 3 RC releases, and hoped for only 1. So we're actually still on schedule.
  • <@AndrewAAM> In other News, I revamped the Occultist last night, fixed up and merged a ton of updates from OA into the new source. I ended up giving the Occultist his own spell list and granting ALL spells with PRExxx standalone. Worked much better during testing.
  • <@AndrewAAM> Archetypes are all functioning, though most abilities granted need to be reviewed for bonuses.
  • <DistantScholar> Do PRExxx on spells not cause as much runtime slowdown as they used to?
  • <@AndrewAAM> Same with Favored Class Bonuses. But we might be able to sneak it in if my playtesters continue to give me timely feedback...
  • <@AndrewAAM> DS - No, the underlying code was fixed, performance there is not as bad.
  • <@AndrewAAM> And that was specific to PRExxx attached to CLASSES. Standalone is like Feats.
  • <@AndrewAAM> Anyways, I'll do an RC2 today or tomorrow, depending on what I'm still tackling.
  • <@AndrewAAM> I'll expect an RC3 the following week.
  • <@AndrewAAM> with the plan of 6.06.01 scheduled for the week following that. Which puts at June 11/12th
  • <@AndrewAAM> Questions?
  • <[OGL]Nylanfs> Just in time for Gen Con :-)
  • <DistantScholar> Release schedule looks good to me.
  • <@AndrewAAM> What is the cut off date for that Paul? You burning CDs this year?
  • <[OGL]Nylanfs> I hadn't really planned it out yet. Probably need to have a month to prepare though so if its out by end of June that is good.
  • <@Zaister> hi, sorry I'm late
  • <@AndrewAAM> Ook Zaister!
  • <[OGL]Nylanfs> Heya Stefan
  • <DistantScholar> Salutations.
  • <[OGL]Nylanfs> Doug, you going to be able to make Gen Con this year?
  • <@AndrewAAM> Cool. Okay, Agenda #2 was reliant on Arch and Code being present. I don't see Tom or Connor... So We can skip them...
  • <DistantScholar> No GenCon this year. I'm farther away, and couldn't plan finances in advance.
  • <@AndrewAAM> The only other item was Github gatekeeper policy.
  • <@AndrewAAM> aka as the Make my life Easier policy... :P
  • <@Zaister> goo idea
  • <[OGL]Nylanfs> :-)
  • <[OGL]Nylanfs> Maybe a release recap for Stefan first though
  • <@AndrewAAM> He should have a log, right?
  • <@AndrewAAM> June 11/12th is expected end game for 6.06.01, RC 2 this weekend, RC 3 the following weekend.
  • <@AndrewAAM> Ton of work, ton of fixes.
  • <@AndrewAAM> Better refactored code at the end of the day (less insanity with Eidolons)
  • <@AndrewAAM> I think the covers release recap.
  • <@Zaister> I have the log, yes
  • <@AndrewAAM> Github has introduced a neat new feature during the merge process - You can actually choose a SQUASH and Merge option. Great for the contributors who have 2+ commits per PR.
  • <@Zaister> I wonder, I always have two commits in any PR, the original one and a met-commit based on that. Others have only one. Weird.
  • <@AndrewAAM> This means Each PR becomes ONE commit to the development branch instead of those 2+. Cherry Picking is a lengthy process, and conflict resolution is not fun during it... it relies on me to choose the correct course of action which means new features can get lost during the back port to Production.
  • <@AndrewAAM> Zaister, that is because you aren't fast-forward merging either. Yours is becoming a three-way.
  • <@Zaister> oh, ok, maybe you can tell me what I am doing wrong after the meddting
  • <@AndrewAAM> Before the PR, you'll want to fetch and rebase into your branch... it makes your PR a Fast Forward commit.
  • <@AndrewAAM> My commits history into master is looking much nicer because of good rebase use.
  • <@AndrewAAM> less Merge Commits to "avoid" during back port cherry picking. :)
  • <@Zaister> yeah that would be nice
  • <@AndrewAAM> The other big issue, which is a non-issue, but mainly a reminder. The Silverbacks should be gatekeeping their respective domains.
  • <@AndrewAAM> Ideally we should be utilizing more peer review, and this does happen with the Code commits more often than in the past.
  • <@AndrewAAM> So those with Write Access, please don't merge things in without a nod from the respective SB. And be sure to set your PUSH default to your own Repo Fork... Accidental branch making to the main repo should be avoided (yes, I've done this once or twice)...
  • <@Zaister> OK
  • <@AndrewAAM> New-Source is slightly different. There is no strict review process, and no automation involved. The work is considered in progress, so unless there is something off, like other work is being removed, go ahead and merge in there... utilize the Squash though...
  • <@AndrewAAM> Any questions?
  • <@AndrewAAM> Comments?
  • <@AndrewAAM> Or new business since we're at open forum
  • <@AndrewAAM> Don't be shy now...
  • <@Zaister> Has anyone heard from Tom at all recently?
  • <[OGL]Nylanfs> Well, I's gotta guy that could maybe help with Gwen, he's real handy with his hands if yous know what I mean...
  • <@AndrewAAM> Stefan - I heard from him Last Sunday.
  • <@Zaister> I contacted him off-list about an issue and he wrote back that he had just returned from travel and was now about to catch up, but that was 2 weeks ago
  • <@AndrewAAM> And if there is an absolute emergency, I can call him up.
  • <DistantScholar> I haven't been keeping up with all the recent changes; I'll need to look at the LST files to make sure I know how things work now.
  • <@AndrewAAM> Biggest changes thus far is Archetypes
  • <@AndrewAAM> We're moving away from TYPE (too broad) to FACT - ClassName_Archetype_ClassFeature is what we use to prevent other archetypes from overlapping Class Features
  • <@AndrewAAM> The PREABILITY:1,CATEGORY=Archetype,TYPE.ClassFeature is replaced by !PREFACT:1,ABILITIES,ClassName_Archetype_CF=true
  • <@AndrewAAM> When an ACTUAL ability is replaced (and I mean actually needs to be removed from the character) we use FACT:ClassName_CF_ClassFeature|true
  • <@AndrewAAM> Which triggers a var to 1 (or as I call them Boolean vars)
  • <@AndrewAAM> Feel free to see the updated Eidolon set up, in case you plan on working on any companions.
  • <@AndrewAAM> That reflects the most significant changes to how Companions should work without crazy sprawl to handle little changes.
  • <@Zaister> If you're interested, it would be cool if someone else could test my SplloLoader change
  • <@Zaister> *SpellLoader
  • <DistantScholar> Zaister: I'm not sure what that means.
  • <@AndrewAAM> What spell loader change?
  • <@Zaister> I have a branch where I replaced the SpellLoader class with the GenericLoader
  • <@AndrewAAM> is it in the development branch?
  • <@Zaister> it is in my repository as a seprate branch
  • <@Zaister> https://github.com/Zaister/pcgen/tree/Project/Remove_SpellLoader
  • <@AndrewAAM> Submit the PR.
  • <@Zaister> I'd rather have some field test before doing that
  • <DistantScholar> What is the SpellLoader, and how does your change improve/modify it?
  • <@AndrewAAM> PR gets peer review.
  • <@Zaister> DistantScholar Loaders are what loads the Data into memory
  • <@Zaister> most object types use the GenericLoader nowadays
  • <@Zaister> but Spells still have an old SpellLoader
  • <@Zaister> I saw that when I changed Spells to be case insensitive for Andrew
  • <@Zaister> and on a hunch I tried to change to the GenericLoader
  • <@AndrewAAM> Besides an update, what is gained by the switch?
  • <@Zaister> The GenericLoader stores the data in a reference Context, while the SpellLoader uses a global map
  • <@Zaister> more consistency I think, no need to maintain separate loaders, wider use of the reference context
  • <@Zaister> should be faster, too
  • <@AndrewAAM> So refactoring then? Cool.
  • <@Zaister> yes
  • <@Zaister> but I think it need testing, not just code review
  • <@AndrewAAM> If you build and run, does it work?
  • <@Zaister> the weird thing is, it dos not fix my strange problem that sometimes spells cannot be overridden
  • <@AndrewAAM> Any unit test failures?
  • <@Zaister> yes, it seems to work, but maybe there is some use case that I don't trigger wehre it might not work
  • <@Zaister> no
  • <@Zaister> I think it looks good to go, but I'd feel better if someone tested it
  • <@AndrewAAM> Okay, anything else before I close out the meeting?
  • <@Zaister> nope
  • <@AndrewAAM> Z - I'll give it a test later... I've been juggling a lot of issues of late.
  • <@Zaister> cool
  • <[OGL]Nylanfs> Any lurkers with issues comments?
  • <@AndrewAAM> Oh, Z can you close out your existing data trackers?
  • <@AndrewAAM> Conversion sets clean up?
  • <@AndrewAAM> DATA-1806
  • <@Zaister> Ah yes I'll get to it
  • <@AndrewAAM> Paul - I'm going to close out DATA-2855 as I don't think it's valid
  • <[OGL]Nylanfs> Which is that?
  • <@AndrewAAM> PFS issues
  • <@AndrewAAM> Okay, thanks for coming everyone! *bangs gavel*