chris b. said:
I have found bugs in
the SRD when I was looking up the spells; I found them by downloading the
database (google books spreadsheet) and grouping up and saw some looked wrong.
Then I looked them up in the PRD at the Paizo site and saw yes these are wrong. Any
specifics? Most of the compendium is
merely a copy/paste routine from the SRD. They should match word for word. I believe once editing has started up again most descrepencies can be
quickly addressed and rectified.
Class abilities are all
missing from the compendium; I think we really should have all these. Class
abilities, feats, monster rules, mythic equivalents of them, traits, racial traits,
also missing; combat maneuvers I believe
all the class abilities are there under each class description. I wonder if it’s
possible to give sub-sections of the compendium their own attributes as well as
make them “drag-able”? Otherwise I think
Class Abilities and similar sub-sections would need to get their own category
of the compendium like Classes, Feats, Items, etc. Racial traits are also a sub-section under
each race. Just focusing on Core for
now, so Mythic as well other additional material will have to be added later… The other decision, do we parse the following in sheet workers or make them json elements? Can we even do that? I guess putting in json is more work for people updating, since it's easier to copy/paste/compare, while making it harder to test the compendium for accuracy. in a sheetworker generally means doing it once. I’m not sure what making the json’s would entail. Maybe this could be auto-generated by Roll20’s server? I know we have control of the sheet workers, so maybe it’s best to process on our end…? For instance we have monster rules and feats in the compendium already, and looking at these and other abilities, it seems we can also parse out Format, Location, Uses (per day and modifier), Save and DC, Prerequisite, Restriction (any sentences with "only" in it seems to be a restriction on use) Type of action (standard, fullround) looks a lot harder to determine. Abilities currently have: name, content, category, ability type (ex/sp/su) Potential additions to the repeating sections if not to the json objects themselves: FORMAT: easily parsed by sheetworker due to html surrounding it LOCATION: easily parsed by sheetworker due to html surrounding it USES: (uses/day + modifier or whatever) this is more for sp and su abilities, difficult to parse, seems to start with "a number of times per ".... to the next period. for monster rules, for abilities: "can only be ... per..." SAVE and DC: Monster rules follow the pattern: "will save [for half damage] (DC 10 + something; exact description in creature text) " CLASS ABILITES: don't follow a single pattern, much harder to parse, easier for user to do it in json PREREQUISITE: follow the pattern: starts with "must have ... to select this|before choosing this", can be more than one. in feats this is delineated by html tags RESTRICTION: can only (if not "can only be ... per") | can be used only | only if | only when | only while SPECIAL: (feat only section) ACTION: (if applicable) standard/immediate/free/move/none/fullround/attack/full attack, (there is a lot of arguing if attack or attack action are two different things) this one may be too hard to parse in a sheetworker. I know there
are a couple of Pathfinder import scripts out there already. Would having a look at how they parsed stat blocks
help at all? I could be wrong, but other
than Attributes, the compendium will have to match the SRD word for
word, correct?. So that’s what we’ll have to work
with. Maybe a json file would need to be
created for some of the compendium’s material? Thanks for commenting Chris. I'll be sure to add your insight in my summary. Looking forward to getting the compendium mess straightened out.