veekun / pokedex Goto Github PK

View Code? Open in Web Editor NEW

1.4K 1.4K 635.0 21.92 MB

more than you ever wanted to know about Pokémon

License: MIT License

Python 99.84% Shell 0.11% Dockerfile 0.05%

pokedex's People

Contributors

Stargazers

Watchers

Forkers

encukou karryanna mat-pt sanqui uxdiogenes sketch matchu antialiasis fairyzana7 teamdandelion veeara282 evinstk jinpark mumage grilix rrezende88 wasimn pwootage mau3189 epithumia edualclaiv cpjk shadow431 adrianadepalma pokecommunity intocake r15ch13 gsheppard nmodi turbine1991 patlanio schorty pyr0x1 darkmatterz chunhli ironleash shinigamae shadowlurker ralts00 utayo kylar42 valepert zippyjake laymansnerd qxc4eva ramirezdario danielskeenan andre111 scientistsupreme pmrowla kevindrongowski fleker rhfwebwork dialga sorensondev elhay123 mpetyx sattmaunders mschex1 codemonkey85 quetzacoalt91 jsimps11 bradleydjarvis obertye sipricy ilaga booleanbalaji jtrezza-old tannertrickle cu3po42 akcorp2003 hanchen2 finleymatt jikki-san pokefanaticos kamilaborowska mousius richardgottschalk apaneto rk905 kampfmoehre omg-spacecat wolframtheta jjsekerak daniel-ault rocci1212 hohahihehu nextlight benjin drewmaniac jorgedip vickylance cellenta aylanonsense ligerzero459 eduardobritosan rkramos benmatase mjhayes412 lorenzorapetti

pokedex's Issues

Use argparse

Reported by Eevee on 2011/04/22 04:48:58 +0000 · Migrated from Redmine issue 586

It supports subcommands, so we could hopefully stop the nonsense we’re doing now to dispatch between commands. Maybe it can print better help than we have right now, too.

It’s 2.7+, though; is there a backport?

Redmine metadata:
Updated on: 2011/04/22 13:33:18 +0000
Start date: 2011/04/22

Comment by En-Cu-Kou from 2011/04/22 13:33:18 +0000

Backports all the way to 2.3. Except they’re not really backports – all that changed in 2.7 was that it got into the stdlib.

pip install argparse

Special split not handled

Reported by Eevee on 2009/08/15 03:17:56 +0000 · Migrated from Redmine issue 85

Pokémon need to know which of their Special stats was in effect before the Special split. Either that, or just make a mythical seventh stat called Special that only exists for R/B. Ehhh, former might be a better idea.

Redmine metadata:
Updated on: 2011/09/21 22:44:32 +0000
Start date: 2009/08/15

Comment by Eevee from 2010/04/13 19:28:44 +0000

Target version: 9 ⇨ 8

Comment by Eevee from 2010/04/13 19:40:22 +0000

Target version: 8 ⇨ 2010 Q3

Comment by eric the espeon from 2010/04/15 13:48:29 +0000

Here you go:
Bulbasaur Both
Ivysaur Both
Venusaur Both
Charmander SpD
Charmeleon SpD
Charizard SpD
Squirtle SpD
Wartortle SpA
Blastoise SpA
Caterpie Both
Metapod Both
Butterfree Both
Weedle Both
Kakuna Both
Beedrill SpA
Pidgey Both
Pidgeotto Both
Pidgeot Both
Rattata SpA
Raticate SpA
Spearow Both
Fearow Both
Ekans SpA
Arbok SpA
Pikachu SpA
Raichu SpA
Sandshrew SpD
Sandslash SpD
Nidoran♀ Both
Nidorina Both
Nidoqueen SpA
Nidoran♂ Both
Nidorino Both
Nidoking SpD
Clefairy SpA
Clefable SpA
Vulpix SpD
Ninetales SpD
Jigglypuff SpD
Wigglytuff SpD
Zubat SpD
Golbat SpD
Oddish SpA
Gloom SpA
Vileplume SpA
Paras SpD
Parasect SpD
Venonat SpA
Venomoth SpA
Diglett SpD
Dugtrio SpD
Meowth Both
Persian Both
Psyduck SpD
Golduck SpD
Mankey SpA
Primeape SpA
Growlithe SpD
Arcanine SpD
Poliwag Both
Poliwhirl Both
Poliwrath SpA
Abra SpA
Kadabra SpA
Alakazam SpA
Machop Both
Machoke SpA
Machamp SpA
Bellsprout SpA
Weepinbell SpA
Victreebel SpA
Tentacool SpD
Tentacruel SpD
Geodude Both
Graveler Both
Golem SpA
Ponyta Both
Rapidash Both
Slowpoke Both
Slowbro SpD
Magnemite SpA
Magneton SpA
Farfetch’d SpA
Doduo Both
Dodrio Both
Seel SpD
Dewgong SpD
Grimer SpA
Muk SpA
Shellder SpA
Cloyster SpA
Gastly SpA
Haunter SpA
Gengar SpA
Onix SpA
Drowzee SpD
Hypno SpD
Krabby Both
Kingler Both
Voltorb Both
Electrode Both
Exeggcute SpA
Exeggutor SpA
Cubone SpA
Marowak SpA
Hitmonlee SpA
Hitmonchan SpA
Lickitung SpA
Koffing SpA
Weezing SpA
Rhyhorn Both
Rhydon Both
Chansey SpD
Tangela SpA
Kangaskhan SpA
Horsea SpA
Seadra SpA
Goldeen SpD
Seaking SpD
Staryu SpA
Starmie SpA
Mr. Mime SpA
Scyther SpA
Jynx SpD
Electabuzz SpD
Magmar SpD
Pinsir SpA
Tauros SpD
Magikarp SpD
Gyarados SpD
Lapras SpD
Ditto Both
Eevee SpD
Vaporeon SpA
Jolteon SpA
Flareon SpD
Porygon SpD
Omanyte SpA
Omastar SpA
Kabuto SpD
Kabutops SpD
Aerodactyl SpA
Snorlax SpA
Articuno SpD
Zapdos SpA
Moltres SpA
Dratini Both
Dragonair Both
Dragonite Both
Mewtwo SpA
Mew Both

(made from bulba’s http://bulbapedia.bulbagarden.net/wiki/List_of_Pok%C3%A9mon_by_base_stats and http://bulbapedia.bulbagarden.net/wiki/List_of_Pok%C3%A9mon_by_base_stats_%28Generation_I%29 in excel)

Comment by Eevee from 2010/10/07 01:59:14 +0000

Target version: 2010 Q3 ⇨ —

Comment by Eevee from 2010/10/08 08:00:26 +0000

Assignee ⇨ —

Comment by En-Cu-Kou from 2011/04/11 13:03:14 +0000

Now that there is Accuracy and Evasion, a mystical seventh stat could maybe work, by having is_main or something instead of is_battle_only.
Special-affecting move effects in RBY could be made more proper then.

Comment by En-Cu-Kou from 2011/09/21 21:17:31 +0000

Squirtle is wrong in that list, see http://bulbapedia.bulbagarden.net/w/index.php?title=List_of_Pok%C3%A9mon_by_base_stats_%28Generation_I%29&action=historysubmit&diff=1488589&oldid=1452227
Anyway, taking this from Bulbapedia sends shivers down my spine :/

Comment by Eevee from 2011/09/21 22:44:32 +0000

taking them from the game directly should be pretty easy, i’d think

Obtain footprints from all generations

Reported by Sanky on 2010/05/03 12:35:57 +0000 · Migrated from Redmine issue 250

Mawile’s pawprint changed from Gen III to Gen IV.
III:
IV:

Maybe there’s more.

Redmine metadata:
Updated on: 2010/10/27 17:44:01 +0000
Start date: 2010/05/03

Comment by Zhorken from 2010/05/20 18:46:42 +0000

Eh.

Tracker: Bug ⇨ Feature
Updated subject
priority_id: 4 ⇨ 3

Comment by Zhorken from 2010/10/27 17:44:01 +0000

A handful changed. I found this out comparing veekun’s old footprints to my B/W rip; veekun apparently had Gen III footprints for the first 386.

I think I want to take this on sometime, but I’m leaving it unassigned because I don’t know when I intend to.

Update Pokémon base exp (etc.)

Reported by Eevee on 2011/04/28 05:38:24 +0000 · Migrated from Redmine issue 605

Some Pokémon data changed in B/W, and I kinda forgot about it after release. Need to update it, but need support for Pokémon changelog first. Arrrrgh.

Redmine metadata:
Updated on: 2011/04/28 05:38:24 +0000
Start date: 2011/04/28

Gust should have double-damage flag

Reported by Eevee on 2010/10/07 02:50:15 +0000 · Migrated from Redmine issue 388

…

Redmine metadata:
Updated on: 2011/06/13 14:28:24 +0000
Start date: 2010/10/07
Relations:
    relates #389

Comment by Zhorken from 2011/01/17 17:26:27 +0000

What are we even doing with the move categories now?

Comment by Eevee from 2011/01/17 17:32:40 +0000

Might be replaced wholesale with #389, which I need to get to soon.

Comment by Anonymous from 2011/06/13 14:28:24 +0000

(spam, elided)

the short and long descriptions of Crush Grip/Wring Out's effect contradict each other

Reported by Zhorken on 2011/02/24 03:42:56 +0000 · Migrated from Redmine issue 556

Maximum of 110 vs 121 power. Figure it out.

Redmine metadata:
Updated on: 2011/02/24 03:43:21 +0000
Start date: 2011/02/24

Comment by Zhorken from 2011/02/24 03:43:09 +0000

Project: veekun ⇨ pokedex

Comment by Zhorken from 2011/02/24 03:43:21 +0000

category_id: None ⇨ 6

Higher-level schema layer

Reported by Eevee on 2011/04/28 04:50:46 +0000 · Migrated from Redmine issue 593

We have SQLAlchemy, and it is good. However it’s still pretty nuts-n-bolts, not exactly friendly to newcomers, and causes a lot of duplication for common tasks.

I propose a somewhat higher-level schema that sits right on top of the ORM classes, but focuses on the core types of objects (Pokémon, moves, types, items, etc.) and has better semantics attached.

I want to use this for:

The JSON API
Pokémon and move searches (prefetching, rendering)
Displaying Pokémon and move tables elsewhere

EnCuKou and I have taken some vague notes on this: http://titanpad.com/highlevel-schema

Redmine metadata:
Updated on: 2011/09/14 05:42:44 +0000
Start date: 2011/04/28
Relations:
    blocks #594

Comment by En-Cu-Kou from 2011/05/02 07:51:31 +0000

Some more discussion has taken place on http://titanpad.com/highlevel-schema-impl

Comment by Eevee from 2011/09/14 05:42:44 +0000

After much reflection, i don’t think this should be an entire layer. It should probably just be a cleanup of the schema, with perhaps some helper functions.

Lookup with literal question marks

Reported by Zhorken on 2011/01/24 19:09:27 +0000 · Migrated from Redmine issue 534

The type ??? and Unown ? are impossible to actually look up by name, because the question marks get taken as wildcards.

Low priority since it’s just two obscure things.

Redmine metadata:
Updated on: 2011/01/24 19:11:26 +0000
Start date: 2011/01/24

Comment by Zhorken from 2011/01/24 19:11:26 +0000

Also because spline-pokedex provides a disambiguation with links; you can just click on “?” page. dywypi provides no obvious way to look up ??? (it has to be type:18), which doesn’t matter, but shows that it could potentially be more of a problem.

Rerip Gen III-IV tutors

Reported by Zhorken on 2010/09/29 15:22:04 +0000 · Migrated from Redmine issue 371

Somehow.

[2010-07-23 15:20:36−0400] <Eevee> from what I hear the tutor moves aren't data
[2010-07-23 15:20:40−0400] <Eevee> they're a big if block somewhere
[2010-07-23 15:20:45−0400] <Eevee> so they would need outright disassembling

Redmine metadata:
Updated on: 2011/05/02 21:03:43 +0000
Start date: 2010/09/29
Relations:
    duplicated #397

Comment by En-Cu-Kou from 2011/05/02 21:03:43 +0000

Eric the Espeon is going through various online dexes and checking the discrepancies. His results (for veekun) are on http://bugs.veekun.com/projects/pokedex/wiki/Corrections.

Rerip Crystal animations

Reported by Zhorken on 2010/05/18 23:28:36 +0000 · Migrated from Redmine issue 272

The current ones are the wrong size, and we lack animations for Unown. And I don’t know where they’re from.

Redmine metadata:
Updated on: 2010/06/29 11:19:19 +0000
Start date: 2010/05/18
Relations:
    blocks #102

Comment by Zhorken from 2010/05/20 00:54:49 +0000

They’re not backwards anymore; I flipped them with gifsicle as a temporary fix.

Comment by Zhorken from 2010/06/29 11:19:20 +0000

Tracker: Bug ⇨ Feature

Make a better type icons

Reported by Zhorken on 2011/01/06 17:58:52 +0000 · Migrated from Redmine issue 524

Make a better Shadow type icon

The current one is modelled after RSE’s type icons, which don’t fit with the DPPt icons we’re using. Also, pick better colours—the current one’s colours are too similar to Dark’s. I was thinking of picking dark purples somewhat like Shadow Lugia’s, significantly purpler than Ghost’s colours, and significantly darker than Poison’s.

Redmine metadata:
Updated on: 2011/03/10 20:18:14 +0000
Start date: 2011/01/06
Relations:
    relates #551

Comment by Eevee from 2011/01/08 23:20:42 +0000

I actually sort of intend to replace all the type icons with those from Gen V; I like the flat style they have and hope they preserve it in the English release.

Comment by Zhorken from 2011/01/08 23:27:35 +0000

Yeah, same, and I guess a ??? icon wouldn’t be hard to whip up.

Comment by Zhorken from 2011/01/28 19:57:25 +0000

I guess we’d also need to make new contest type icons, in that case… Might we want to categorize icons ripped from the game and fanmade extension icons separately in the fs?

Comment by En-Cu-Kou from 2011/03/10 20:18:14 +0000

What about ditching images and serving them as spans with display:inline-block and an appropriate font (#551)?

As the i18n guy I’d like that very much :)

Add latest 3D models

Reported by Eevee on 2010/09/01 20:37:38 +0000 · Migrated from Redmine issue 346

These are pretty sizable so I’m not a huge fan of including every single 3D Pokémon game going back forever.

But it would be cool to have the models and be able to render them in-browser.

This is more an interesting engineering challenge than a serious priority.

Possibly of use:

Redmine metadata:
Updated on: 2010/09/01 20:38:33 +0000
Start date: 2010/09/01
Relations:
duplicated #602

ID column gaps throw random lookup off

Reported by Zhorken on 2011/02/09 22:40:39 +0000 · Migrated from Redmine issue 550

For example, in the moves table, regular moves have IDs 1 to 559, and Shadow moves start at 10001. Random lookup sees that there are 577 moves total, though, and picks an ID from 1 to 577—so once in a while, it tries to look up a move that doesn’t exist.

Other tables with this problem: pokemon (alternate forms), types (??? and Shadow), items (667 does not exist because it was a duplicate live caster), locations (a handful of IDs are skipped for some reason).

Redmine metadata:
Updated on: 2011/02/10 06:49:08 +0000
Start date: 2011/02/09

Comment by Eevee from 2011/02/09 23:37:50 +0000

Location ids are skipped because several locations were merged or removed as junk.

Comment by Zhorken from 2011/02/10 06:49:08 +0000

Project: veekun ⇨ pokedex

Acquire eyecatches

Reported by magical on 2010/09/02 00:08:43 +0000 · Migrated from Redmine issue 348

Eyecatch: Those pictures that pop up upon entering certain locations.
Could probably rip them pretty easily, at least for HG/SS. Or just snag them from Bulbapedia.

Redmine metadata:
Updated on: 2010/09/02 00:08:43 +0000
Start date: 2010/09/01

Pickup tables

Reported by Zhorken on 2010/09/23 14:18:32 +0000 · Migrated from Redmine issue 369

I originally figured I’d put them in the effect proper, but that’s ugly. I’m thinking Pickup items should be stored in the database some other way for spline-pokedex to construct tables out of.

Redmine metadata:
Updated on: 2010/09/23 14:19:21 +0000
Start date: 2010/09/23

Data differences/changes between versions

Reported by Eevee on 2009/08/13 00:08:32 +0000 · Migrated from Redmine issue 78

Have a *_changes table for each table that had something change, I suppose. Schema would be (id, field, version_group_id, old_value). Note the use of field to distinguish from strictly being a column name. I think that should be a (indexed) text field though, rather than a separate table, as the things in it are fairly arbitrary.

Off the top of my head, the following have changed before:

Move PP, power, accuracy, type, effect, effect chance, target?, name (Conversion 2; more?), etc
Pokémon types, base exp, “species” classification
Ability effects (mostly from Emerald)
Type efficacy
Item name, effect, cost, pocket (last two: e.g. fossils)
Location names, areas
I don’t know PLEASE UPDATE ME :( Check dpmoves.html and dppokes.
http://bugs.veekun.com/projects/pokedex/wiki/Changelog

Pokémon types are dual, which poses something of a problem. Type efficacy has a multi-column primary key, which again poses something of a problem.

It would be very, very, very nice if there were some API to reliably ask for a column name and version group and be assured I’m getting back the right data. Perhaps the field should be restricted to a column name. But then I’d have some eight different changes tables, unless I made one big version_changes table. Unnngh.

Smogon has some move differences from RBY: http://www.smogon.com/rb/articles/differences Easier to look over than all of Azure.

Changes from gen4 to gen5 include:

POKÉMON
- Rotom forms: types changed from electric/ghost to electric/(special move type)!
- Castform and Cherrim forms now exist outside of battle

MOVES -- numbers only

- Bind: accuracy 75% -> 85%
- Jump Kick: power 85 -> 100; PP 25 -> 10
- Tackle: power 35 -> 50; accuracy 95% -> 100%
- Wrap: accuracy 85% -> 90%
- Thrash: power 90 -> 120; PP 20 -> 10
- Almost every Pokémon's base exp changed
- Alt. evolution methods for gimmicky crap from gen 4 and some that weren't crap but got screw
- Disable: accuracy 90% -> 100%
- Growth raises both attack stats
- Petal Dance: power 90 -> 120; PP 20 -> 10
- Fire Spin: power 15 -> 35; accuracy 70% -> 85%
- Toxic: accuracy 85% -> 90%
- Meditate: PP 20 -> 40
- Clamp: accuracy 75% -> 85%
- Hi Jump Kick: power 100 -> 130; PP 20 -> 10
- Glare: accuracy 75% -> 90%
- Poison Gas: accuracy 55% -> 90%
- Crabhammer: accuracy 85% -> 90%
- Mind Reader: PP 40 -> 5
- Cotton Spore: accuracy 85% - 100%
- Giga Drain: power 60 -> 75
- Fury Cutter: power 10 -> 20
- Future Sight: power 80 -> 100; PP 15 -> 10; accuracy 90% -> 100%
- Whirlpool: power 15 -> 35; accuracy 70% -> 85%
- Beat Up: power 10 -> 1 (i.e., special; maybe effect changed?)
- Uproar: power 50 -> 90 (!!!!!)
- Sand Tomb: power 15 -> 35; accuracy 70% -> 85%
- Bullet Seed: power 10 -> 25
- Icicle Spear: power 10 -> 25
- Covet: power 40 -> 60
- Rock Blast: accuracy 80% -> 90%
- Doom Desire: power 120 -> 140; accuracy 85% -> 100%
- Me First: power 0 -> 1 (only of interest to nerds)
- Punishment: power 60 -> 1 (??)
- Last Resort: power 130 -> 140
- Drain Punch: power 60 -> 75; PP 5 -> 10
- Magma Storm: accuracy 70% -> 75%

also, every move that never misses is now marked 101% accuracy
        some effect chances MAY have changed, as "effect chance" is now split into parts and is harder to compare

- Wrap, Whirlpool, etc. now last 5–6 turns instead of 2–5
- Sturdy now acts like Focus Sash rather than protecting only against OHKO moves
multiplies 

Snatch will now work on: Conversion, Wish, Recycle, Imprison, Healing Wish, Power Trick, Lucky Chant, Aqua Ring, Magnet Rise, and Lunar DanceBut not Psych Up and Acupressure


Magic  Coat will now work  on: Whirlwind, Roar, Disable, Spite, Spikes, Foresight, Encore, Torment,  Taunt, Odor Sleuth, Miracle Eye, Embargo, Heal Block, Toxic Spikes,  Defog, and Stealth Rock
Protect will now work on: Counter, Mirror Coat, and Metal Burst
But not on Hail

Curse changed from ??? to ghost

    - priority
        - Protect: 3 -> 4
        - Detect: 3 -> 4
        - Endure: 3 -> 4
        - ExtremeSpeed: 1 -> 2
        - Fake Out: 1 -> 3

Some moves may have changed targets; consult data

Black and White also changed base exp for nearly every Pokémon.

Redmine metadata:
Updated on: 2011/09/25 20:21:18 +0000
Start date: 2009/08/13
Done ratio: 50%
Relations:
    relates #190
    duplicated #211
    duplicated #251
    duplicated #530

Comment by Eevee from 2010/04/13 19:28:19 +0000

Target version: 17 ⇨ 16

Comment by Eevee from 2010/04/13 20:11:11 +0000

Target version: 16 ⇨ —

Comment by Zhorken from 2010/05/13 00:09:59 +0000

I’ve been treating changes in ability effects by just writing them into the descriptions. I like the idea of dealing with our written descriptions that way; if we create a model for treating data changes, I think it should be restricted to data changes (and things we can treat like them) rather than prose and such.

Comment by Eevee from 2010/05/13 01:32:29 +0000

Iiiiiiiii don’t know. I like the idea that effect descriptions are always current canon, and not cluttered with bugs and whatever from old versions.

The idea was partly to expose an interface for getting a snapshot of game behavior for any given version, which requires knowing at the db level whether the effect changed.

Comment by Zhorken from 2010/05/14 02:10:59 +0000

Mm, alright. I like that too. Should I go back through and clean those out of the ability effects, then…?

Comment by Eevee from 2010/05/14 03:07:34 +0000

No point until there’s somewhere to put it.

Comment by Eevee from 2010/10/06 04:32:04 +0000

Pasted the list of changes from the old typeme, and made this Q4; it’s sort of getting important now. The db is currently in this awful half~~4 half~~5 state.

Updated subject
priority_id: 4 ⇨ 6
Target version: — ⇨ 2010 Q4

Comment by Eevee from 2010/10/08 08:00:26 +0000

Assignee ⇨ —

Comment by Eevee from 2010/11/28 04:23:32 +0000

Created [[Changelog]] for this; please fill it out because I don’t know anything :(

Comment by Eevee from 2010/12/06 05:21:42 +0000

List of ability changes in Emerald: http://www.psypokes.com/emerald/abilities.php

Comment by surskitty from 2010/12/08 01:36:10 +0000

I DID A BUNCH OF MOVE DIFFERENCES be proud of me.

Comment by Zhorken from 2010/12/30 18:55:08 +0000

Abilities are, as far as I know, entirely done.

Done: 0% ⇨ 50%

Comment by Zhorken from 2010/12/30 18:55:29 +0000

(That 50 isn’t just for abilities, for the record.)

Comment by Zhorken from 2011/01/25 05:49:56 +0000

We neeeed to figure out how to do move effect changes. Look at #530.

Comment by Eevee from 2011/02/06 07:09:20 +0000

What’s missing here?

Comment by Zhorken from 2011/02/28 19:09:39 +0000

Move changes still have a ways to go, item changes aren’t done at all, Pokémon aren’t done at all (though that’s less important), type efficacy also isn’t done though I think that has a separate issue.

Comment by eric the espeon from 2011/05/04 01:47:12 +0000

Quite a few changes can be found in these threads:
http://www.smogon.com/forums/showthread.php?t=65453
http://www.smogon.com/forums/showthread.php?t=85493
http://www.smogon.com/forums/showthread.php?t=85497

Comment by En-Cu-Kou from 2011/08/13 22:34:28 +0000

Adding link to wiki page

Updated description

Comment by En-Cu-Kou from 2011/09/25 20:21:18 +0000

Also, Stadium bug fixes should be covered somehow.

Write more lookup tests

Reported by Eevee on 2010/04/13 21:57:31 +0000 · Migrated from Redmine issue 170

There are a few, but they’re old and forgotten, and probably don’t cover everything lookup does now.

Redmine metadata:
Updated on: 2010/10/08 08:00:27 +0000
Start date: 2010/04/13

Comment by Eevee from 2010/10/08 08:00:27 +0000

Assignee ⇨ —

Effect text cleanup

Reported by Eevee on 2010/09/05 15:48:08 +0000 · Migrated from Redmine issue 353

~~Write a test to hit all the effect text and make sure the markdown is valid~~- no missing closing braces, and all the names of things should be good.-
Change “major status effect” to “major status ailment”. Confusion, attraction, etc. are “minor status ailments”.
~~Change “level” to “stage”.~~
Don’t use types as adjectives; prefer “steel-type move” to “steel move”.
Remove the various synonyms that I figured I’d handwave later, e.g. “paralysis” and “paralyze” both being mechanic links. Pick one and make the others use the new different-title syntax I haven’t implemented yet.
Thunder Wave checks type.
Increase/decrease, raise/lower, heighten/reduce. These are sort of all over the place. Probably want to pick a pair, or at least make it consistent which we use where and why.
Likewise, we’re still a bit inconsistent with “reduce 20%” vs “reduce to 80%” vs “reduce to 0.8×”.
~~Do we have a consistent name for the critical hit chance counter thing?~~
Prefer “friendly Pokémon” and “opposing Pokémon”.
“Leaves the field” vs “leaves battle”.

Maybe we should make a glossary, christ.

Others?

Redmine metadata:
Updated on: 2011/03/29 02:29:06 +0000
Start date: 2010/09/05

Comment by Eevee from 2010/10/07 01:59:16 +0000

Target version: 2010 Q3 ⇨ —

Comment by Eevee from 2010/12/04 06:25:05 +0000

Did a teeny bit of this in the midst of gen 5 stuff.

Convert the "font" images into an actual font

Reported by magical on 2011/02/10 09:39:40 +0000 · Migrated from Redmine issue 551

Surely we can create a bitmap WOFF or something.

This would nicely solve (part of) #365.

Redmine metadata:
Updated on: 2011/03/11 07:11:53 +0000
Start date: 2011/02/10
Relations:
    relates #365
    relates #524

Comment by magical from 2011/03/11 07:11:53 +0000

Looks like UPC has a bunch of Pokémon fonts in TrueType format.

http://upokecenter.com/pkmnfonts.zip

Extend the ability/move-flag thing

Reported by Eevee on 2011/04/28 04:38:02 +0000 · Migrated from Redmine issue 590

A couple abilities are manually associated with particular move flags in spline-pokedex , and show a list of all the matching moves right on the page. (e.g.: Iron Fist)

Extend this to work for items, as well, for e.g. Brightpowder. And move the data into pokedex.

Might be nice to also associate abilities/items that trigger on contact, though inlining a list of every contact move is impractical.

Redmine metadata:
Updated on: 2011/04/28 04:38:02 +0000
Start date: 2011/04/28

Add headbutting trees to the location data

Reported by Anonymous on 2010/11/08 07:51:42 +0000 · Migrated from Redmine issue 443

Redmine metadata:
Updated on: 2011/06/27 21:02:34 +0000
Start date: 2010/11/08
Relations:
    relates #22

Comment by Zhorken from 2010/11/20 17:22:59 +0000

I swear we already had a catchall ticket for funny encounters?

Comment by magical from 2010/12/02 03:50:59 +0000

Project: veekun-incoming ⇨ pokedex

Comment by magical from 2010/12/02 03:51:55 +0000

Need to find the data first.

category_id: None ⇨ 3

Comment by Anonymous from 2011/06/27 21:02:34 +0000

(spam; elided)

Acquire Dream World artwork

Reported by Eevee on 2011/04/22 04:45:58 +0000 · Migrated from Redmine issue 584

Pokémon, items.

I think there are two images for each Pokémon; one used as stock art, and one used within the Dream World itself. Zhorken was looking into this and each of the latter is actually just its own Flash file which nothing can seem to inspect. Super.

Redmine metadata:
Updated on: 2011/04/28 06:51:07 +0000
Start date: 2011/04/22

Comment by Zhorken from 2011/04/27 17:15:34 +0000

Unfortunately it’s also the latter that comes at a decent resolution. :(

Comment by Eevee from 2011/04/28 06:51:07 +0000

For, uh, reference: http://archives.bulbagarden.net/wiki/Category:Pok%C3%A9mon_Dream_World_artwork

Load Black/White locations and encounters

Reported by Eevee on 2010/10/06 04:34:49 +0000 · Migrated from Redmine issue 375

See: http://piratepad.net/bw-encounters

In case that’s vandalized, magical’s data is http://magical.pastebin.com/deZRLmAT for White and http://magical.pastebin.com/m7wUKaCK for Black.

Almost certainly want to reload HGSS first (#297).

Redmine metadata:
Updated on: 2011/04/28 05:41:11 +0000
Start date: 2010/10/06
Done ratio: 90%

Comment by Eevee from 2011/04/28 05:41:11 +0000

So close!

Done: 0% ⇨ 90%

The Terrain Problem

Reported by Eevee on 2011/01/28 04:57:27 +0000 · Migrated from Redmine issue 536

Need new Johto Sugimori art

Reported by Eevee on 2010/08/18 01:42:43 +0000 · Migrated from Redmine issue 336

And old Kanto art, too.

Also I’m still missing Darkrai and Shaymin, I think.

Redmine metadata:
Updated on: 2010/10/08 08:00:29 +0000
Start date: 2010/08/18
Relations:
    relates #75

Comment by Eevee from 2010/10/07 01:59:15 +0000

Target version: 2010 Q3 ⇨ —

Comment by Eevee from 2010/10/08 08:00:29 +0000

Assignee ⇨ —

Obtain spinoff games' imagery

Reported by Eevee on 2009/08/12 23:55:47 +0000 · Migrated from Redmine issue 75

Stadium. Stadium 2. Colosseum. XD.

Pinball. Ranger. PMD. Entire spritesheets, why not.

Old Sugimori images? Anime screenshots, perhaps?

I have a whole page for this; might as well use it.

Redmine metadata:
Updated on: 2012/01/20 21:29:06 +0000
Start date: 2009/08/12
Relations:
    relates #336
    duplicated #199

Comment by Zhorken from 2010/04/21 17:05:34 +0000

category_id: None ⇨ 14

Comment by Eevee from 2010/10/08 08:00:26 +0000

Assignee ⇨ —

Comment by Zhorken from 2012/01/20 20:56:09 +0000

Oh, I have… PMD headshots… still……..

Comment by Zhorken from 2012/01/20 21:29:06 +0000

Oh and I have a TCG sprite ripper that I forgot about apparently.

Smogon parity

Reported by Eevee on 2011/04/28 06:48:40 +0000 · Migrated from Redmine issue 626

Smogon does a whole lot of research on move/ability/item/whatever effects. Alas, it’s sprinkled throughout their forums. It would be nice to collect as much of this insight as possible.

For starters: http://www.smogon.com/forums/showthread.php?t=79340

Perhaps we should approach them about collaborating on this.

Redmine metadata:
Updated on: 2011/06/21 12:20:21 +0000
Start date: 2011/04/28

Comment by Anonymous from 2011/05/02 15:14:01 +0000

Also quite a lot of testing has been done on the PO bugs/fixes forums: http://pokemon-online.eu/forums/forumdisplay.php?9-Bugs-and-Fixes

and if you want to contact people about this, asking on #C&C on irc.synirc.net is a reasonable way to go. It’s where the active contributors tend to hang out.

Comment by Anonymous from 2011/06/21 12:20:21 +0000

(spam; elided)

Combine evolution families logically

Reported by Eevee on 2010/07/18 18:23:21 +0000 · Migrated from Redmine issue 293

Phione and Manaphy should be baseform siblings, with some note that only Phione actually comes from breeding.

Both Nidoran families should be parallel and together, with a note that either Nidoran can result.

Volbeat and Illumise should be siblings, similar to Nidoran.

Unfortunately! Volbeat and Illumise have opposite EXP growth. This requires moving the growth rate into pokemon.

Somewhat related, I believe Shedinja should actually evolve from Ninjask. The game clones after Ninjask evolves, so Shedinja can actually learn moves only Ninjask can.

Redmine metadata:
Updated on: 2011/04/29 15:10:50 +0000
Start date: 2010/07/18
Relations:
    relates #579

Comment by Zhorken from 2010/10/08 20:11:44 +0000

Possibly the Latis, too?

Somewhat related, I believe Shedinja should actually evolve from Ninjask. The game clones after Ninjask evolves, so Shedinja can actually learn moves only Ninjask can.

Only levelup moves, though, and only from one level. It’s more like it evolves into both at once, and anything it learns in the process, it learns both ways.

Comment by En-Cu-Kou from 2011/04/27 15:00:19 +0000

I want to find Ninjask when looking for final evolutions.

Comment by Eevee from 2011/04/28 06:44:54 +0000

Target version: — ⇨ 2011 Q2
Assignee ⇨ Eevee

Comment by En-Cu-Kou from 2011/04/29 15:10:50 +0000

I’m moving exp_growth to Pokemon as part of #579 (http://bugs.veekun.com/issues/579)

Struggle

Reported by eric the espeon on 2010/04/16 13:47:15 +0000 · Migrated from Redmine issue 190

From Gen I to III it causes 50% of damage dealt as recoil, not 25% of max HP. PP would make more sense as n/a imo, you can’t legally use it in a situation that it’s programmed PP value matters. Also, according to Bulba, Rock Head does not prevent Struggle’s recoil damage, so that could be noted.

Redmine metadata:
Updated on: 2010/10/16 11:07:43 +0000
Start date: 2010/04/16
Relations:
    relates #78

Comment by Zhorken from 2010/10/16 11:07:43 +0000

Yeah, Struggle’s recoil seems to be treated separately from regular recoil everywhere—Rock Head, Reckless, Magic Guard, etc. I don’t remember whether I tested any of that myself, but the descriptions I wrote for veekun for those abilities all mention Struggle.

Many tables need readable identifiers

Reported by magical on 2010/04/19 22:07:26 +0000 · Migrated from Redmine issue 207

Any table being used as an Enum should have a readable identifier which does not change. (Unlike to the name column, which can change and is often used as a label by spline-pokedex.)

This includes EncounterCondition, EncounterConditionValue, EvolutionMethod, Generation, MoveDamageClass, MoveTarget, et cetera.

Wiki page: [[Identifiers]].

Redmine metadata:
Updated on: 2011/04/11 18:06:57 +0000
Start date: 2010/04/19
Relations:
    relates #209

Comment by En-Cu-Kou from 2011/04/11 18:06:58 +0000

Also, locations need unique identifiers.

Specify the target schema

Reported by Eevee on 2011/04/28 05:16:25 +0000 · Migrated from Redmine issue 598

PostgreSQL (and any real database heh heh) support multiple schemata per database, which can be used for namespacing without the overhead of multiple databases (or the loss of joining between them).

SQLAlchemy, though, just reads and loads into/from the default. Would be nice if that weren’t the case.

I sense this will be a gigantic headache, as the schema doesn’t appear anywhere within the database URL syntax, and I think you’ll need to specify the schema when running any code that uses pokedex.

This is the best I could find on it: http://stackoverflow.com/questions/2685899/schema-qualified-tables-with-sqlalchemy-sqlite-and-postgresql

Redmine metadata:
Updated on: 2011/04/28 05:16:25 +0000
Start date: 2011/04/28

Load Gen V move metadata

Reported by Eevee on 2010/10/07 02:51:01 +0000 · Migrated from Redmine issue 389

It looks very useful and generally more reliable than my own. Might make all the move categories entirely unnecessary.

Redmine metadata:
Updated on: 2011/09/14 10:46:06 +0000
Start date: 2010/10/07
Relations:
    relates #388

Comment by surskitty from 2010/12/08 13:38:15 +0000

Got everything except changelog and whatever bit 14 is! Well, that’s listed, but I don’t know what it is so the corresponding move flag type is something like ‘mystery’ ‘this does … something’.

Comment by Eevee from 2011/03/06 08:07:04 +0000

Okay, well, loaded the rest into a move_meta table.

Things this needs:

Possibly link to move effects rather than moves. ailment_chance is still different per move, though, so we’d have to split it up.
Ditch effect_chance in favor of ailment_chance and the other chances?
Fix some of the names of enums; I think categories especially are a bit wonky.
Delete “categories”. Update the move search and make the new meta stuff link to it.

Some discrepancies between meta and categories that are probably not worth worrying about:

Acupressure isn’t flagged as raising any stats, because it picks one at random.
Tailwind affects Speed, but isn’t a Speed increase.
Foresight and friends reset evasion rather than altering it.
Memento, Stockpile, Secret Power, and Defog are all just “unique effects” rather than being flagged as stat changers.
Thrash and friends don’t have multi-hit/multi-turn data.

“Categories” that are now missing but I think are interesting enough to deserve a column:

Type of damage calculation: damage formula, damage formula with variable power, fixed amount, etc
Moves with a charge or recovery turn

There are also a few old “categories” that are cool, but not really enough to devote a ton of space to. Maybe good to tack onto /dex/moves, or a mechanics article, or somewhere:

Meta-moves, like Counter or Mirror Move
Weather moves? These are now a subset of field effects
Healing moves; not all are included
Auto-fainting moves

Assignee ⇨ Eevee
Target version: — ⇨ 2011 Q1
Done: 0% ⇨ 60%

Comment by Eevee from 2011/04/11 04:36:11 +0000

All the in-game stuff is live, but I still want to have a damage category and maybe a couple other manual things.

Done: 60% ⇨ 80%

Comment by Eevee from 2011/04/28 04:44:46 +0000

Done: 80% ⇨ 0%
Target version: 2011 Q1 ⇨ —
Assignee ⇨ —

Comment by En-Cu-Kou from 2011/09/14 10:46:06 +0000

Also, what happened to the King’s Rock flag again? It just disappeared, right?

Magnitude

Reported by paperfairy on 2011/02/23 09:22:00 +0000 · Migrated from Redmine issue 555

http://veekun.com/dex/moves/magnitude is displaying incorrect values for percentages.

It should read:

4 - 10 - 5%
5 - 30 - 10%
6 - 50 - 20%
7 - 70 - 30%
8 - 90 - 20%
9 - 110 - 10%
10 - 150 - 5%

Redmine metadata:
Updated on: 2011/09/24 16:33:34 +0000
Start date: 2011/02/23

Comment by Zhorken from 2011/03/05 16:13:47 +0000

Project: veekun-incoming ⇨ pokedex

Comment by En-Cu-Kou from 2011/09/24 16:33:34 +0000

Changed that. Probabilities that don’t add up to 100% adr clearly wrong. (BTW, Smogon still has the wrong list for gens 2 & 3.)

UPC gives slightly different chances for Generation II; is that worth mentioning in the changelog?

Honey tree encounters missing

Reported by eric the espeon on 2010/04/15 19:11:21 +0000 · Migrated from Redmine issue 186

Location data for Pokemon you interact with as an overworld sprite, gift Pokemon (like starters and eevee), and those that “runners” have no encounter data. e.g. http://veekun.com/dex/pokemon/zapdos/locations

Redmine metadata:
Updated on: 2010/04/16 12:45:26 +0000
Start date: 2010/04/15

Comment by eric the espeon from 2010/04/15 19:13:43 +0000

(assigning was accidental, would fix but can’t edit)

Comment by magical from 2010/04/15 19:21:15 +0000

Assignee ⇨ —

Comment by eric the espeon from 2010/04/16 12:45:26 +0000

Add Honey Trees to the list of location data needed. Probably a mechanics page for that too, quite complex things: http://projectpokemon.org/wiki/Honey_Tree_%28Technical%29 http://projectpokemon.org/games/mechanics/munchlax.php

Special encounters and event Pokémon

Reported by Eevee on 2009/05/05 21:58:48 +0000 · Migrated from Redmine issue 22

I would split this up between special encounters and event Pokémon, but I’m not sure how to reasonably do that. That Farfetch’d you traded for in the game is a specific “event” Pokémon, but it’s still in the game!

Special ways to encounter or get a Pokémon. Note that some of these are repeatable and some are not; should a distinction be made here?

(once) Starter Pokémon: Bulbasaur, Mudkip, Chimchar, …
(once) Unique overworld Pokémon: Snorlax, Mewtwo, Lugia, …
(once) In-game trades: Farfetch’d, uh…
(once) In-game giveaways: Riolu egg, Eevee, …
Feebas
Headbutt in GSC
(once) Legendaries that run around
Daily Safari Game Pokémon
Daily Mansion Pokémon
Honey tree Pokémon
(once) Celebi in Crystal in Japan, I guess? I forget how that worked
(once but incidental) Colosseum and XD Snaggables

Special ways to get an INDIVIDUAL (customized somehow) Pokémon:

A zillion Toys R Us, movie, Pokémon Center (http://www.upokecenter.com/nypc.html)), Mystery Gift, etc. giveaways
Crystal’s Extremespeed Dratini
Crystal’s Dizzy Punch Odd Eggs
Stadium 1: Amnesia Psyduck
Stadium 2: Baton Pass Farfetch’d and Earthquake Gligar
“PBR gives you hardy pikachu w/ surf, modest magmortar, and adamant electivire”
Ranch has rewards of some sort
Pokémon Box reward eggs for filling your box: False Swipe Swablu, Extremespeed Zigzagoon, Pay Day Skitty, and Surf Pichu. THERE IS NO WISH RALTS.
Ranger and Ranger 2 have Manaphy egg, Riolu with Aura Sphere, and some sort of Darkrai

Pending questions:

Should Shaymin, Darkrai, and Arceus count as overworld Pokémon? They do appear in the overworld…
Giveaways and ingame trades tend to give you Pokémon with fixed stats, names, moves, OT ID, etc. How do I store/display these without consuming absurd amounts of space?

Some lists of event Pokémon/moves on Smogon:

http://www.smogon.com/forums/showthread.php?t=34197
http://www.smogon.com/forums/showpost.php?p=2418618&postcount=34
http://www.smogon.com/forums/showthread.php?t=63782 (illegal movesets, but may be helpful)

Trying to get more details here on [[SpecialEncounters]]. Please help flesh this out!

Redmine metadata:
Updated on: 2011/09/25 20:04:18 +0000
Start date: 2009/05/05
Relations:
    relates #23
    relates #443
    relates #48
    relates #588

Comment by Eevee from 2009/07/26 21:24:59 +0000

Updated subject

Comment by Eevee from 2009/08/12 21:47:40 +0000

Difficulty: Involved

Comment by Eevee from 2009/08/12 22:41:54 +0000

Target version: — ⇨ 9

Comment by Eevee from 2010/04/13 19:28:44 +0000

Target version: 9 ⇨ 8

Comment by Eevee from 2010/04/13 20:02:40 +0000

category_id: None ⇨ 3

Comment by Eevee from 2010/04/13 20:13:00 +0000

Target version: 8 ⇨ —

Comment by Eevee from 2010/10/08 08:00:25 +0000

Assignee ⇨ —

Comment by En-Cu-Kou from 2011/09/24 18:12:55 +0000

Assignee ⇨ En-Cu-Kou

Comment by En-Cu-Kou from 2011/09/25 20:04:18 +0000

Okay. As an intermediate step between ad-hoc wiki notes and a relational database, I created a googly spreadsheet. It’s read-only to prevent spam, please do request access if you have anything to add.
There I’m collecting the wiki data and whatever else I can find on this. (Generation I is veeeery close to done now.)
And I’m making a script to load it all into the DB.

For the later generations we’ll want to contact Guested of Project Pokémon. Hopefully we can get a hold of his database, and figure out a way to sync as new events are added.

Acquire official anime-style art

Reported by Eevee on 2011/04/28 06:14:46 +0000 · Migrated from Redmine issue 619

I… think this is a thing, right? Different from the old Sugimori art? We should get it.

Redmine metadata:
Updated on: 2011/05/09 03:07:15 +0000
Start date: 2011/04/28

Comment by Zhorken from 2011/05/08 12:29:15 +0000

I’m pretty sure I’ve only seen it for key-character Pokémon like Pikachu?

Comment by Dragonfree from 2011/05/08 19:39:14 +0000

This is on GTS+ (http://gallery.gtsplus.net/) for what looks like all Pokémon, so I’m assuming it exists for all Pokémon.

Comment by Eevee from 2011/05/09 03:07:15 +0000

Original Sugimori: http://archives.bulbagarden.net/wiki/File:001Bulbasaur_RG.png

New Sugimori: http://archives.bulbagarden.net/wiki/File:001Bulbasaur.png

Anime http://archives.bulbagarden.net/wiki/File:001Bulbasaur_AG_anime.png

Missing Crystal Unown sprites

Reported by Eevee on 2009/09/16 02:31:33 +0000 · Migrated from Redmine issue 102

These don’t seem to exist anywhere. I’m not sure what to do about that.

Redmine metadata:
Updated on: 2010/06/29 11:23:50 +0000
Start date: 2009/09/16
Done ratio: 90%
Relations:
    relates #197
    blocked #272

Comment by Eevee from 2009/09/16 02:31:43 +0000

Project: spline-pokedex ⇨ pokedex
Target version: 14 ⇨ —

Comment by Zhorken from 2010/04/21 17:05:00 +0000

category_id: None ⇨ 14

Comment by Zhorken from 2010/05/10 23:54:00 +0000

I’ve ripped these, too, along with most of the other GSC sprites, but I want to figure out Crystal animations and add all the reripped GSC sprites at once. See #197.

Assignee ⇨ Zhorken
Done: 0% ⇨ 90%

Comment by Zhorken from 2010/05/11 00:01:16 +0000

Target version: — ⇨ 2010 Q2
Difficulty: Normal

Comment by Zhorken from 2010/05/18 23:27:25 +0000

I’ve pushed the Unowns I have, but still don’t have Crystal animations, so I’m leaving this open.

Comment by Zhorken from 2010/06/29 11:23:50 +0000

Haha this isn’t happening anytime soon. Reripping Crystal animations isn’t set for any target version, so I’m dropping this issue’s target version, too.

Target version: 2010 Q2 ⇨ —

Parallelism between player's field buff move descriptions

Reported by Vixie on 2010/07/16 01:23:36 +0000 · Migrated from Redmine issue 290

The following moves should have parallel descriptions: Reflect, Light Screen, Safeguard, Mist, Lucky Chant, Tailwind (maybe more I could be forgetting???). Currently, Reflect does not say that it lasts only 5 turns and both “screen” moves do not suggest that the barrier protects the player’s field and not just the user. Lucky Chant and Tailwind are just awkwardly worded and should match Safeguard and Mist for coherency.

Redmine metadata:
Updated on: 2010/10/08 08:00:28 +0000
Start date: 2010/07/16

Comment by Eevee from 2010/08/18 00:35:47 +0000

Assignee ⇨ Eevee
Target version: — ⇨ 2010 Q3
Difficulty: Easy

Comment by Eevee from 2010/08/18 00:35:56 +0000

Project: veekun-incoming ⇨ pokedex

Comment by Eevee from 2010/08/18 00:36:04 +0000

category_id: None ⇨ 6

Comment by Eevee from 2010/10/07 01:59:15 +0000

Target version: 2010 Q3 ⇨ —

Comment by Eevee from 2010/10/08 08:00:28 +0000

Assignee ⇨ —

Fix special accuracy/power values

Reported by Eevee on 2009/11/24 00:15:17 +0000 · Migrated from Redmine issue 124

Power is sometimes 0 or 1. Accuracy is sometimes 0. It’s not entirely clear what these are supposed to mean, especially when there are already multiple other places to specify ‘special damage’ or ‘no accuracy check’ or whatever.

Redmine metadata:
Updated on: 2011/02/26 03:36:21 +0000
Start date: 2009/11/24
Relations:
    relates #317

Comment by Eevee from 2010/04/13 19:28:44 +0000

Target version: 9 ⇨ 8

Comment by Eevee from 2010/04/13 20:03:07 +0000

category_id: None ⇨ 6

Comment by Eevee from 2010/04/13 20:03:53 +0000

Target version: 8 ⇨ 2010 Q3

Comment by eric the espeon from 2010/04/15 17:02:40 +0000

Would putting “variable” (maybe a shorter word would be better) or something like that with a link to exactly how the power is calculated make sense for things like Grass Knot? And for self/field targeting moves — or n/a makes more sense as an acc value imo.

Comment by Eevee from 2010/04/15 17:11:42 +0000

Yeah, the idea was pretty much to just figure out when it should be “n/a” and when it should be “varies”. I would assume 1 means varies and 0 means no damage, and 0% accuracy means no accuracy check, but I haven’t actually checked that. If that’s the case, this is really just a display bug.

The move effect, of course, always explains how the power is calculated.

Comment by En-Cu-Kou from 2010/09/03 13:26:34 +0000

For power, 0 means “no damage”. This can be checked easily:

[m for m in moves if (m.power == 0) ^ (m.damage_class.name == 'None')] == []

1 means varies (just hand-checked); IDK if any other values mean ‘varies’ as well but it’s unlikely.

The following I did not check throroughly:
As far as accuracy goes, the value doesn’t matter at all. It can be 100% (Struggle) or 0% (Pain Split) – the game code probably doesn’t check it at all. But when it’s 0%, it’s a sign that the move effect will ignore accuracy. Rendering 0% as “n/a” would make sense.
The “Rolls Accuracy” flag doesn’t matter either. See for example Swift or Pain Split, where veekun’s flag text and move effect text currently contradict each other ().
What actually matters is the move effect, the other places are most likely leftovers from previous generations (I assume the flag is ripped from the ROM?).

Comment by Eevee from 2010/09/12 23:17:46 +0000

So, without ridiculously complicating the schema, the following seems to be sort of more correct:

NULL accuracy should mean “doesn’t roll accuracy”. 0 accuracy should mean “never hits”.
NULL power should mean “power varies”. 0 power should mean “no damage”.

The uses of NULL are a bit different, but in both cases they mean “the usual system doesn’t apply”.

BTW, we found out the “rolls accuracy” flag is actually about Mirror Move. Durp. So it’s completely irrelevant to this.

Then the Web side can render NULL accuracy as “—” or something, NULL power as “varies” or something, and everything else verbatim. Good?

Comment by En-Cu-Kou from 2010/09/13 04:05:09 +0000

Wouldn’t it be better to keep it how it is in the ROM? I think the pokédex library should stay “close to the metal”.
Advantages:

Ease of checking against any other resource: The 1 for “power varies” is pretty much everywhere.
Keeping the ripping script cleaner: The UI will need special-casing either way; why not just have one place where the code is complicated? (I’m not familiar with the ripping script, I don’t even know if there is one, but there might be some uses of the raw data, and introducing NULLs will complicate them without simplifying anything else.) Related to that:
Poor NULL is abused enough as it is. Why give it two new meanings, which have to be explained somewhere? The current system is good. Used everywhere, pretty standard and non-intrusive (I’m quite surprised you’ve only discovered it now, after making pokédexes since RBY).

The special-case numbers don’t matter, the game doesn’t seem to check them. The reason for the 1 for “power varies” is most likely that in Red and Green there was no Physical/Special/Other category, and the game needed to know what’s damaging and what’s not. Encoding the fact that the game doesn’t check the number into the database sounds iffy.

Besides, how would you find the moves that should have Accuracy set to NULL? Struggle, for example, doesn’t roll accuracy, but is listed as 100%.

Keep it simple.

Comment by Eevee from 2010/09/13 10:32:52 +0000

The database should reflect the data, not whatever whim Game Freak had about how to store it themselves. The inconsistency between 0% and 100% accuracy is exactly one of the reasons why; how on earth am I supposed to fix the UI when the data is bogus?

It’s not just the UI that would be complicated here; it’s anyone else building a UI, too. Crazy data problems should be solved in the data. If it ever needs redumping, well, we can do that pretty easily.

Comment by En-Cu-Kou from 2010/09/14 12:26:48 +0000

If “the database should reflect the data”, there are still problems – mainly because there’s only one NULL in SQL but many special cases in Pokémon. Solving that problem by arbitrarily assigning one to be NULL and the other to be a special-valued integer leaves the problem only half-solved; and what’s worse, there’s no room for future updates to the model.

And there sure are possible updates.
There are variable-power moves that do have a reasonable number in that column: from Facade, Bug Bite, Brine, etc. to Rollout, Spit Up or Punishment. Why don’t those get special treatment?
I wanted to suggest going all the way by having min_power and max_power (that would even allow queries like “power can be ≥ 80”) – but it wouldn’t work either. There are more flavors to “power varies”. It can be that only the base power varies and the usual damage formula is used (Hidden Power, Facade, Magnitude), or that some other formula is used (Sonicboom, Bide, Psywave, Fissure).

That brings us to four special cases now (no damage, variable BP with default, variable BP with no default, and variable without BP, in addition to the “normal” case). Game Freak’s data merges two, you suggest using NULL for one and GF’s representation for all others.

Moving away from move power: For moves that don’t roll accuracy, are you planning to remove the distinction between moves marked 0% and 100%? Are you sure there’s no significance to that distinction? I’m reasonably sure it doesn’t matter in battle mechanics, but that’s it. There’s some data there, and after your efforts of both preserving original flavor texts and making them usable I find it hard to believe you’d want to just erase it.

And, are you planning to set Shedinja’s Base HP to NULL? That one doesn’t use the standard formula either.

…

It’s a question of the scope of the project.

Smogon’s damage formula page has a clear distinction between moves that use base power (in the frst table) and those that use a different formula (listed at the very end). It ignores the “BP=1/BP=reasonable number” distinction altogether. That doesn’t make it a bad page. It has a “human” feel to it – somebody’s prepared the data, looked at all the fine points, categorized and explained everything.
If veekun was like that, I would not trust it. That “human” feel tells me that there will be errors in the numbers somewhere, because it’s just not ripped straight from the game. The reason I use veekun is because it has the raw data that nobody’s played with; it leaves the interpretation to the reader. I go to Smogon beause it has high-quality explanations of the special cases, fine points and general mechanics.
It’s perfectly okay if you want to let veekun have some of that “human” aspect, but please don’t lose the machine precision in the process.
Stick to having data straight from the game in the database, and leave the interpretation to the UI. Or add a flag/column to make it clear that this column contains research-based, “unofficial” data.
Defining the project this way would, at least, help make it more objective than a vague, subjective “the database should reflect the data”.

(Disclaimer: Those are, of course, my opinions. You’re the one steering veekun.)

Comment by En-Cu-Kou from 2010/10/03 07:26:48 +0000

I guess Nintendo solved most of this for us, 101 accuracy should be NULL. Still, be careful about the power.

Anyway, in the same spirit, berry_flavors has a lot of zero rows that should probably just be removed.

Comment by En-Cu-Kou from 2010/10/03 09:10:23 +0000

Also item costs should be NULL when the shops don’t take the item.

Comment by En-Cu-Kou from 2010/10/03 14:24:53 +0000

And the gender rate should be NULL for genderless species.

Comment by Eevee from 2010/10/07 01:59:14 +0000

Target version: 2010 Q3 ⇨ —

Comment by Eevee from 2010/10/08 08:00:27 +0000

Assignee ⇨ —

Comment by Zhorken from 2011/02/26 03:36:21 +0000

Sooo I’ve fixed #317 without actually changing anything in the db, so whenever this side of the issue does get fixed, remember to update spline-pokedex too.

Animated B/W Sprites

Reported by Anonymous on 2010/11/12 23:20:54 +0000 · Migrated from Redmine issue 447

Only the front ones will do. Really like it if you added this - thanks.

Redmine metadata:
Updated on: 2011/09/21 18:34:08 +0000
Start date: 2010/11/12
Relations:
    duplicated #621

Comment by Zhorken from 2010/11/20 17:23:21 +0000

Project: veekun-incoming ⇨ pokedex

Comment by magical from 2010/12/02 02:48:46 +0000

This is definitely something i plan to investigate.

Comment by Eevee from 2010/12/06 05:23:04 +0000

This person has collected some: http://arkeis.com/animatedsprites.html

Comment by magical from 2011/02/06 22:27:47 +0000

Okay, so.

I’m working on this a bit. Code at http://github.com/magical/pokemon-nds-sprites. It can only rip the static sprites at the moment—not animations! (Needs NANR and NMCR support.)

Helpful docs at http://www.romhacking.net/forum/index.php/topic,8407.msg173645.html#msg173645.

Contributions welcome!

Comment by magical from 2011/02/10 06:42:59 +0000

Difficulty: Involved

Comment by En-Cu-Kou from 2011/09/21 18:34:08 +0000

Eevee filed a new issue for this at #621. Copying priority.

priority_id: 4 ⇨ 6

Handle CLI errors more gracefully

Reported by Eevee on 2010/04/24 15:31:40 +0000 · Migrated from Redmine issue 217

Off the top of my head, the following can happen:

Engine URI has a bogus scheme.
Engine URI is just bogus.
Engine is an RDBMS that needs a certain Python module.
Engine is file-based and doesn’t have write permission to the given file.
Engine is not file-based, but the requested database doesn’t exist.
Engine has bad credentials.
Index directory is not readable.
Index directory is not writable.
Index is not populated, and this is a lookup/status/whatever command.
dump tries to dump from an obsolete schema.
load is given an incomplete subtree of dependent tables.
load tries to load into existing tables without -D passed.
Ctrl-C is pressed.

Showing a backtrace is ugly and, for many of these, not very helpful.

Redmine metadata:
Updated on: 2010/10/08 08:00:28 +0000
Start date: 2010/04/24

Comment by Eevee from 2010/10/08 08:00:28 +0000

Assignee ⇨ —

Item locations

Reported by Eevee on 2009/08/12 22:47:34 +0000 · Migrated from Redmine issue 60

Every item (given to you, found in the overworld, bought at a shop) for all the games, ideally.

A start:

Redmine metadata:
Updated on: 2010/10/08 08:00:26 +0000
Start date: 2009/08/12

Comment by Eevee from 2010/05/21 01:20:22 +0000

category_id: None ⇨ 16

Comment by Mystic from 2010/09/14 15:42:21 +0000

Here’s the Poké Mart data from the Generation I games. The game data doesn’t say which shops are where, so I had to work it out by checking with Bulbapedia. There’s one I couldn’t find, but all the towns are covered so I assume it’s just unused.

Adds attachment
Adds attachment

Comment by Eevee from 2010/10/08 08:00:26 +0000

Assignee ⇨ —

Fix issue numbering and migrate from Redmine

Other languages for flavor text

Reported by Anonymous on 2011/03/29 10:41:07 +0000 · Migrated from Redmine issue 573

Add other languages such as french or german flavor texts (At least for the languages ever included in the base)

Redmine metadata:
Updated on: 2011/03/29 16:10:42 +0000
Start date: 2011/03/29

Comment by Zhorken from 2011/03/29 15:53:39 +0000

I can do that, but I’ll need someone to help me distinguish between shy and real hyphens at linebreaks for German/Spanish/Italian RBYGSC text (e.g. “feath~~/ered" vs "bright~~/feathered”.)

Comment by En-Cu-Kou from 2011/03/29 16:10:42 +0000

Get a list, then we can ask for help. uepc.eu or the sites they link to seems like a good place for that.

Investigate B/W abilities, round four or whatever

Reported by Zhorken on 2011/03/10 15:29:54 +0000 · Migrated from Redmine issue 561

Remaining stuff from my original list:

Illusion: Should note that the illusion fades if the ability is lost.
Moody: Do evasion and accuracy count?
Prankster: Note that affected moves don’t get blocked by Quick Guard.

New, messy list of stuff that came up that I didn’t feel like looking into:

Volt Absorb: Cell Battery takes precedence?
Water Absorb: Absorb Bulb takes precedence?
Color Change: Doesn’t react to Sheer Forced moves?
Flash Fire: Still works if lost and regained? If frozen?
Suction Cups et al: Is the “success rate” the chance that something will bite your rod, or does it do something like widen the window in which you can press A to reel the Pokémon in?
Intimidate: Triple battle shenanigans.
Shadow Tag: Triple battle shenanigans.
Static: Overworld shenanigans
Magnet Pull: Overworld shenanigans
Trace: triggering shenanigans
Early Bird: 1-turn sleep shenanigans
Forecast: Soaked Castform
Poison Heal: Mummy???
Lightningrod, Storm Drain: absorbing items
Contrary: how to write
Friend Guard: what? test.
Stench: bypasses Shield Dust?
Skill Link: short effect
Magic Guard: mention Doom Desire/Future Sight
Forewarn: Test new variable/set moves
Slow Start: counter resets?
Honey Gather: effect chance
Flower Gift: new effect totally
Telepathy: test items
Rattled: Normal-type bearers?

Redmine metadata:
Updated on: 2011/09/21 21:45:52 +0000
Start date: 2011/03/10
Done ratio: 80%
Relations:
duplicated #560
duplicated #557
duplicated #622

Comment by En-Cu-Kou from 2011/03/10 15:42:52 +0000

Iron Barbs does’t do anything when the barbed Pokémon touches an opponent.
Tested with my Metal Claw Ferroseed on some wild Joltiks and Klinks.

Comment by Zhorken from 2011/03/10 15:48:25 +0000

That’s what I thought, but the difference between its in-game description and Rough Skin’s made me wonder.

Comment by Dragonfree from 2011/04/13 20:58:17 +0000

I’m pretty darn sure Smogon investigated Prankster pretty thoroughly and found that yes, Growl can go before Quick Attack this way.

Comment by Zhorken from 2011/05/08 12:41:34 +0000

Updated description

Comment by Zhorken from 2011/09/21 21:45:23 +0000

Updated description

Comment by Zhorken from 2011/09/21 21:45:33 +0000

Done: 0% ⇨ 80%

Comment by Zhorken from 2011/09/21 21:45:52 +0000

Updated description

Add Colosseum/XD shadow and purification moves

Reported by Eevee on 2009/07/27 00:30:43 +0000 · Migrated from Redmine issue 48

Don’t bother with Shadow Rush from Colosseum, though; I think that’s rather implied.

Redmine metadata:
Updated on: 2011/09/25 20:12:28 +0000
Start date: 2009/07/27
Relations:
    relates #588
    relates #22

Comment by Eevee from 2009/08/12 22:29:33 +0000

Difficulty: Normal

Comment by Eevee from 2009/08/12 22:41:26 +0000

Target version: — ⇨ 9

Comment by Eevee from 2010/04/13 19:28:44 +0000

Target version: 9 ⇨ 8

Comment by Eevee from 2010/04/13 20:03:07 +0000

category_id: None ⇨ 6

Comment by Eevee from 2010/04/13 20:13:00 +0000

Target version: 8 ⇨ —

Comment by Zhorken from 2010/08/10 00:45:06 +0000

Adding only XD’s shadow moves seems inconsistent, though, even if Colosseum’s are all the same. This feels like saying “don’t bother with machine-learnt moves for Mew.”

Comment by Eevee from 2010/10/08 08:00:25 +0000

Assignee ⇨ —

Comment by En-Cu-Kou from 2011/09/25 20:12:28 +0000

These are similar to overworld encounters (e.g. Zapdos) or gifts (Eevee). Especially if I treat the 3D RPGs as first-class games. I’ll cover this as part of #22 (github #3).

Assignee ⇨ En-Cu-Kou

Lookup by internal id (is also: lookup by number is useless right now)

Reported by magical on 2010/04/19 23:43:45 +0000 · Migrated from Redmine issue 209

Should be able to enter 133 and get both Eevee (whose national dex number is 133) and Octillery (whose Sinnoh dex number is 133). And others, of course.

BUT, you should also get an item from every generation, and a move from every generation, and so forth.

Which means that:

disambiguation needs to indicate which region/generation an id was picked from, and
there needs to be some clever syntax for specifying that beforehand — perhaps pokemon:133/kanto?

Redmine metadata:
Updated on: 2010/12/01 20:05:11 +0000
Start date: 2010/04/19
Relations:
relates #90
relates #207
relates #600

Comment by magical from 2010/04/19 23:44:31 +0000

Assignee ⇨ —

Comment by Eevee from 2010/08/24 22:18:52 +0000

This is a more general problem. Expanding from regional dex numbers to all internal ids.

Updated subject

Comment by Eevee from 2010/12/01 20:05:11 +0000

Updated subject

Sketch sprites from Ruby and Sapphire

Reported by Anonymous on 2010/10/25 11:46:22 +0000 · Migrated from Redmine issue 413

In Ruby and Sapphire contests, the winners get a sketch put on the wall. It does not seem to be derived from the battle sprite to me. A collection of all these would be neato. It might also be the last drop of content wrung from generation 3.

Redmine metadata:
Updated on: 2010/11/20 17:25:13 +0000
Start date: 2010/10/25

Comment by magical from 2010/10/27 06:16:29 +0000

The sketch sprites are generated from the normal sprites; they are passed through a filter of some sort. I believe upokecenter had a script which let you apply the filter to any image.

Comment by Zhorken from 2010/11/20 17:25:13 +0000

Project: veekun-incoming ⇨ pokedex

Obtain Berry growth stage sprites

Reported by Zhorken on 2010/04/29 12:17:33 +0000 · Migrated from Redmine issue 244

They’d be cool! But not so important.

Redmine metadata:
Updated on: 2010/04/29 12:17:33 +0000
Start date: 2010/04/29

Remove fanon "names" for abstract concepts from the db

Reported by Eevee on 2011/04/28 06:13:39 +0000 · Migrated from Redmine issue 618

Move methods, encounter conditions/terrain, etc. We invented the names for these, and it doesn’t seem correct to have them clogging the db.

Of course, the implications for downstream i18n aren’t great either. Open to better suggestions on how to deal with this.

Redmine metadata:
Updated on: 2011/05/08 12:30:05 +0000
Start date: 2011/04/28

Comment by En-Cu-Kou from 2011/04/28 19:25:42 +0000

They should stay where they are.
Every project that presents these to the user will need them. I can’t think of anything that would touch move methods but wouldn’t want to use the names.
The fact that we invented them is not really relevant. Are you considering removing the short effect texts as well?

Comment by Eevee from 2011/04/28 20:56:03 +0000

Short effects are descriptions of a game mechanic. These are practically spline-pokedex UI.

Blurrrhgrhg not sure.

Comment by En-Cu-Kou from 2011/04/28 21:39:27 +0000

Umm… Long effects are the real descriptions. Short ones, with details left out to fit on a line, are spline-pokedex UI.
The thing is that it’s commonly needed UI, which should be in the library, where it can get proper care and translations and stuff. I really don’t want to fork some random piece of spline-pokedex just to get proper names of things.

Include the batteries.

Comment by Zhorken from 2011/05/08 12:27:20 +0000

What? What do I have to do with this?

EDIT: I agree with moving them out of the db, though.

Assignee ⇨ —

Comment by Zhorken from 2011/05/08 12:30:05 +0000

Updated subject

veekun / pokedex Goto Github PK

pokedex's People

Contributors

Stargazers

Watchers

Forkers

pokedex's Issues

Recommend Projects

Recommend Topics

Recommend Org