Merge pull request #1 from sanderland/master

Merge upstream changes

.github/FUNDING.yml

@@ -1,7 +1,7 @@
 # These are supported funding model platforms
 
 github: sanderland # Replace with up to 4 GitHub Sponsors-enabled usernames e.g., [user1, user2]
-patreon: katrainbaduk # Replace with a single Patreon username
+patreon: # Replace with a single Patreon username
 open_collective: # Replace with a single Open Collective username
 ko_fi: # Replace with a single Ko-fi username
 tidelift: # Replace with a single Tidelift platform-name/package-name e.g., npm/babel

.gitignore

@@ -16,6 +16,8 @@ tmp.pickle
 my
 logs
 callgrind.*
+cpp/out
+.vs
 
 # debug
 outdated_log.txt

CONTRIBUTIONS.md

@@ -52,6 +52,7 @@ Many thanks to these additional authors:
 * "fohristiwhirl" for the Gibo and NGF formats parsing code.
 * "kaorahi" for bug fixes and SGF parser improvements.
 * "ajkenny84" for the red-green colourblind theme.
+* Lukasz Wierzbowski for the ability to paste urls for sgfs and helping fix alt-gr issues.
 
 ## Translators
 
@@ -62,7 +63,7 @@ Many thanks to the following contributors for translations.
 * German: "nimets123" with contributions from "trohde"
 * Spanish: Sergio Villegas ("serpiente") with contributions from the Spanish OGS community
 * Russian: Dmitry Ivankov and Alexander Kiselev
-* Simplified Chinese: Qing Mu with contributions from "Medwin"
+* Simplified Chinese: Qing Mu with contributions from "Medwin" and Viktor Lin
 * Japanese: "kaorahi"
 
 ## Additional thanks to

INSTALL.md

@@ -53,7 +53,7 @@ Simply download and run, everything is included.
 
 ### <a name="LinuxQuick"></a>Quick install guide
 
-If you have a working Python 3.6/3.7 available, you should be able to simply:
+If you have a working Python 3.6/3.7/3.8 available, you should be able to simply:
 
 * Run `pip3 install -U katrain`
 * Run the program by executing `katrain` in a terminal.

LICENSE

@@ -1,10 +1,11 @@
 This repository includes:
 
-1. Binaries for 'KataGo', which is Copyright 2019 David J Wu et al.
+1. Binaries for 'KataGo', which is Copyright David J Wu et al.
 For on related licenses for these binaries and libraries see https://github.com/lightvector/KataGo
 
 2. Icons from www.flaticon.com, used with permission with the following attributions:
 - Equalize icon and Thrash Icon: derived from work by bqlqn from www.flaticon.com
+- Local search icon: derived from work by Pixel Perfect from www.flaticon.com
 - Other Menu icons and Flag icons: derived from work by Freepik from www.flaticon.com
 
 3. The True Type Font DIGITAL-7 version 1.02 by Sizenko Alexander, which is free for non-commercial use.

README.md

@@ -120,6 +120,7 @@ This estimate should be reasonably accurate as long as you have not changed the
     * **Calibrated Rank Bot** was calibrated on various bots (e.g. GnuGo and Pachi at different strength settings) to play a balanced
      game from the opening to the endgame without making serious (DDK) blunders. Further discussion can be found
       [here](https://github.com/sanderland/katrain/issues/44) and [here](https://github.com/sanderland/katrain/issues/74).
+    * **Simple Style** Prefers moves that solidify both player's territory, leading to relatively simpler moves.
     * **ScoreLoss** is KataGo analyzing as usual, but
       choosing from potential moves depending on the expected score loss, leading to a varied style with mostly small mistakes.
     * **Policy** uses the top move from the policy network (it's 'shape sense' without reading).
@@ -166,6 +167,7 @@ Keyboard shortcuts are shown with **[key]**.
     * **[s]**: Equalize visits: Re-evaluate all currently shown next moves with the same visits as the current top move. Useful to increase confidence in the suggestions with high uncertainty.
     * **[d]**: Analyze all moves: Evaluate all possible next moves. This can take a bit of time even though 'fast_visits' is used, but can be useful to see how many reasonable next moves are available.
     * **[f]**: Find alternatives: Increases analysis of current candidate moves to at least the 'fast visits' level, and request a new query that excludes all current candidate moves.    
+    * **[g]**: Find local responses: Searches within 5 spaces of the last move only.
     * **[spacebar]**: Turn continuous analysis on/off. This will continuously improve analysis of the current position, similar to Lizzie's 'pondering', but only when there are no other queries going on.
     * **[enter]** AI move. Makes the AI move for the current player regardless of current player selection.
 
@@ -180,6 +182,8 @@ In addition to shortcuts mentioned above and those shown in the main menu:
 * **[arrow left]** or **[z]**: Undo move. Hold alt for 10 moves at a time, or ctrl to skip to the start.
 * **[arrow right]** or **[x]**: Redo move. Hold alt for 10 moves at a time, or ctrl to skip to the start.
 * **[arrow up/down]** Switch branch, as would be expected from the move tree.
+* **[n]** As in clicking the red arrow, go to one move before the next mistake (orange or worse) by a human player.
+* **[home/end]** Go to the beginning/end of the game.
 * **[scroll up]**: Undo move. Only works when hovering the cursor over the board.
 * **[scroll down]**: Redo move. Only works when hovering the cursor over the board.
 * **[click on a move]**: See detailed statistics for a previous move, along with expected variation that was best instead of this move.

i18n.py

@@ -26,7 +26,7 @@ for lang in locales:
     pofile[lang] = os.path.join(localedir, lang, "LC_MESSAGES", "katrain.po")
     po[lang] = polib.pofile(pofile[lang])
     for entry in po[lang].translated_entries():
-        if "TODO" in entry.comment:
+        if "TODO" in entry.comment and not "DEPRECATED" in entry.comment:
             todos[lang].append(entry)
         strings_to_langs[entry.msgid][lang] = entry
         strings_to_keys[entry.msgid][lang] = set(re.findall("{.*?}", entry.msgstr))

katrain/KataGo/analysis_config.cfg

@@ -109,6 +109,7 @@ nnMaxBatchSize = 96
 
 # Other General GPU Settings-------------------------------------------------------------------------------
 
+
 # Cache up to 2 ** this many neural net evaluations in case of transpositions in the tree.
 nnCacheSizePowerOfTwo = 20
 # Size of mutex pool for nnCache is 2 ** this
@@ -116,7 +117,6 @@ nnMutexPoolSizePowerOfTwo = 16
 # Randomize board orientation when running neural net evals?
 nnRandomize = true
 
-
 # TO USE MULTIPLE GPUS:
 # Set this to the number of GPUs you have and/or would like to use...
 # AND if it is more than 1, uncomment the appropriate CUDA or OpenCL section below.

katrain/__main__.py

@@ -30,11 +30,13 @@ Config.set("graphics", "width", max(400, int(WINDOW_X * WINDOW_SCALE_FAC)))
 Config.set("graphics", "height", max(400, int(WINDOW_Y * WINDOW_SCALE_FAC)))
 Config.set("input", "mouse", "mouse,multitouch_on_demand")
 
+import re
 import signal
 import sys
 import threading
 import traceback
 from queue import Queue
+import urllib3
 import webbrowser
 
 from kivy.base import ExceptionHandler, ExceptionManager
@@ -130,9 +132,9 @@ class KaTrainGui(Screen, KaTrainBase):
         return self.play_mode.mode
 
     def toggle_continuous_analysis(self):
-        self.idle_analysis = not self.idle_analysis
-        if not self.idle_analysis:
+        if self.idle_analysis:
             self.controls.set_status("", STATUS_INFO)
+        self.idle_analysis = not self.idle_analysis
         self.update_state()
 
     def start(self):
@@ -157,7 +159,6 @@ class KaTrainGui(Screen, KaTrainBase):
         self.board_controls.mid_circles_container.clear_widgets()
         self.board_controls.mid_circles_container.add_widget(bot)
         self.board_controls.mid_circles_container.add_widget(top)
-        self.board_controls.branch.disabled = not cn.parent or len(cn.parent.children) <= 1
         self.controls.players["W"].captures = prisoners["W"]
         self.controls.players["B"].captures = prisoners["B"]
 
@@ -168,6 +169,7 @@ class KaTrainGui(Screen, KaTrainBase):
             self.board_controls.engine_status_col = ENGINE_READY_COL
         else:
             self.board_controls.engine_status_col = ENGINE_BUSY_COL
+        self.board_controls.queries_remaining = len(self.engine.queries)
 
         # redraw board/stones
         if redraw_board:
@@ -193,21 +195,21 @@ class KaTrainGui(Screen, KaTrainBase):
             teaching_undo = cn.player and last_player.being_taught and cn.parent
             if (
                 teaching_undo
-                and cn.analysis_ready
-                and cn.parent.analysis_ready
+                and cn.analysis_complete
+                and cn.parent.analysis_complete
                 and not cn.children
                 and not self.game.end_result
             ):
                 self.game.analyze_undo(cn)  # not via message loop
             if (
-                cn.analysis_ready
+                cn.analysis_complete
                 and next_player.ai
                 and not cn.children
                 and not self.game.end_result
                 and not (teaching_undo and cn.auto_undo is None)
             ):  # cn mismatch stops this if undo fired. avoid message loop here or fires repeatedly.
                 self._do_ai_move(cn)
-                Clock.schedule_once(self.board_gui.play_stone_sound, 0)
+                Clock.schedule_once(self.board_gui.play_stone_sound, 0.25)
         if len(self.engine.queries) == 0 and self.idle_analysis:
             self("analyze-extra", "extra", continuous=True)
         Clock.schedule_once(lambda _dt: self.update_gui(cn, redraw_board=redraw_board), -1)  # trigger?
@@ -218,7 +220,7 @@ class KaTrainGui(Screen, KaTrainBase):
             self.controls.update_players()
             self.update_state()
         for player_setup_block in PlayerSetupBlock.INSTANCES:
-            player_setup_block.update_players(bw, self.players_info[bw])
+            player_setup_block.update_player_info(bw, self.players_info[bw])
 
     def set_note(self, note):
         self.game.current_node.note = note
@@ -293,6 +295,10 @@ class KaTrainGui(Screen, KaTrainBase):
         self.board_gui.animating_pv = None
         self.game.redo(n_times)
 
+    def _do_next_mistake(self):
+        self.board_gui.animating_pv = None
+        self.game.redo(999, stop_on_mistake=self.config("trainer/eval_thresholds")[-4])
+
     def _do_cycle_children(self, *args):
         self.board_gui.animating_pv = None
         self.game.cycle_children(*args)
@@ -324,7 +330,7 @@ class KaTrainGui(Screen, KaTrainBase):
         self.controls.timer.paused = True
         if not self.timer_settings_popup:
             self.timer_settings_popup = I18NPopup(
-                title_key="timer settings", size=[dp(350), dp(400)], content=ConfigTimerPopup(self)
+                title_key="timer settings", size=[dp(600), dp(500)], content=ConfigTimerPopup(self)
             ).__self__
             self.timer_settings_popup.content.popup = self.timer_settings_popup
         self.timer_settings_popup.open()
@@ -351,12 +357,12 @@ class KaTrainGui(Screen, KaTrainBase):
         self.controls.timer.paused = True
         if not self.ai_settings_popup:
             self.ai_settings_popup = I18NPopup(
-                title_key="ai settings", size=[dp(600), dp(650)], content=ConfigAIPopup(self)
+                title_key="ai settings", size=[dp(750), dp(750)], content=ConfigAIPopup(self)
             ).__self__
             self.ai_settings_popup.content.popup = self.ai_settings_popup
         self.ai_settings_popup.open()
 
-    def load_sgf_file(self, file, fast=False, rewind=False):
+    def load_sgf_file(self, file, fast=False, rewind=True):
         try:
             move_tree = KaTrainSGF.parse_file(file)
         except ParseError as e:
@@ -388,6 +394,7 @@ class KaTrainGui(Screen, KaTrainBase):
             popup_contents.filesel.on_success = readfile
             popup_contents.filesel.on_submit = readfile
         self.fileselect_popup.open()
+        self.fileselect_popup.content.filesel.ids.list_view._trigger_update()
 
     def _do_output_sgf(self):
         msg = self.game.write_sgf(self.config("general/sgf_save"))
@@ -399,6 +406,14 @@ class KaTrainGui(Screen, KaTrainBase):
         if not clipboard:
             self.controls.set_status(f"Ctrl-V pressed but clipboard is empty.", STATUS_INFO)
             return
+
+        url_match = re.match(r"(?P<url>https?://[^\s]+)", clipboard)
+        if url_match:
+            self.log("Recognized url: " + url_match.group(), OUTPUT_INFO)
+            http = urllib3.PoolManager()
+            response = http.request("GET", url_match.group())
+            clipboard = response.data.decode("utf-8")
+
         try:
             move_tree = KaTrainSGF.parse_sgf(clipboard)
         except Exception as exc:
@@ -440,7 +455,9 @@ class KaTrainGui(Screen, KaTrainBase):
             "s": ("analyze-extra", "equalize"),
             "d": ("analyze-extra", "sweep"),
             "f": ("analyze-extra", "alternative"),
+            "g": ("analyze-extra", "local"),
             "p": ("play", None),
+            "n": ("next-mistake",),
             "down": ("switch-branch", 1),
             "up": ("switch-branch", -1),
             "f5": ("timer-popup",),
@@ -466,7 +483,9 @@ class KaTrainGui(Screen, KaTrainBase):
                 return
             else:
                 return
-
+        ctrl_pressed = "ctrl" in modifiers
+        alt_pressed = "alt" in modifiers
+        shift_pressed = "shift" in modifiers
         shortcuts = self.shortcuts
         if keycode[1] == "tab":
             self.play_mode.switch_ui_mode()
@@ -474,26 +493,30 @@ class KaTrainGui(Screen, KaTrainBase):
             self.nav_drawer.set_state("toggle")
         elif keycode[1] == "spacebar":
             self.toggle_continuous_analysis()
-        elif keycode[1] == "b" and "ctrl" not in modifiers:
+        elif keycode[1] == "b" and ctrl_pressed:
             self.controls.timer.paused = not self.controls.timer.paused
-        elif keycode[1] in ["`", "~", "m"] and "ctrl" not in modifiers:
+        elif keycode[1] in ["`", "~", "m"] and ctrl_pressed:
             self.zen = (self.zen + 1) % 3
         elif keycode[1] in ["left", "z"]:
-            self("undo", 1 + ("alt" in modifiers) * 9 + ("ctrl" in modifiers) * 999)
+            self("undo", 1 + (alt_pressed or shift_pressed) * 9 + (ctrl_pressed and not alt_pressed) * 999)
         elif keycode[1] in ["right", "x"]:
-            self("redo", 1 + ("alt" in modifiers) * 9 + ("ctrl" in modifiers) * 999)
-        elif keycode[1] == "n" and "ctrl" in modifiers:
+            self("redo", 1 + (alt_pressed or shift_pressed) * 9 + (ctrl_pressed and not alt_pressed) * 999)
+        elif keycode[1] == "home":
+            self("undo", 999)
+        elif keycode[1] == "end":
+            self("redo", 999)
+        elif keycode[1] == "n" and ctrl_pressed:
             self("new-game-popup")
-        elif keycode[1] == "l" and "ctrl" in modifiers:
+        elif keycode[1] == "l" and ctrl_pressed:
             self("analyze-sgf-popup")
-        elif keycode[1] == "s" and "ctrl" in modifiers:
+        elif keycode[1] == "s" and ctrl_pressed:
             self("output-sgf")
-        elif keycode[1] == "c" and "ctrl" in modifiers:
+        elif keycode[1] == "c" and ctrl_pressed:
             Clipboard.copy(self.game.root.sgf())
             self.controls.set_status(i18n._("Copied SGF to clipboard."), STATUS_INFO)
-        elif keycode[1] == "v" and "ctrl" in modifiers:
+        elif keycode[1] == "v" and ctrl_pressed:
             self.load_sgf_from_clipboard()
-        elif keycode[1] in shortcuts.keys() and "ctrl" not in modifiers:
+        elif keycode[1] in shortcuts.keys() and not ctrl_pressed:
             shortcut = shortcuts[keycode[1]]
             if isinstance(shortcut, Widget):
                 shortcut.trigger_action(duration=0)

katrain/config.json

@@ -17,11 +17,13 @@
         "anim_pv_time": 0.5,
         "debug_level": 0,
         "lang": "en",
-        "version": "1.5.0"
+        "version": "1.6.1"
     },
     "timer": {
         "byo_length": 30,
         "byo_periods": 5,
+        "minimal_use": 0,
+        "main_time": 0,
         "sound": true
     },
     "game": {
@@ -86,6 +88,22 @@
         "ai:policy": {
             "opening_moves": 22.0
         },
+        "ai:simple": {
+            "max_points_lost": 1.75,
+            "settled_weight": 1.0,
+            "opponent_fac": 0.5,
+            "min_visits": 3,
+            "attach_penalty": 1,
+            "tenuki_penalty": 0.5
+        },
+        "ai:settle": {
+            "max_points_lost": 1.75,
+            "settled_weight": 1.0,
+            "opponent_fac": 0.5,
+            "min_visits": 3,
+            "attach_penalty": 1,
+            "tenuki_penalty": 0.5
+        },
         "ai:p:weighted": {
             "weaken_fac": 1.25,
             "pick_override": 1.0,

katrain/core/ai.py

@@ -8,29 +8,31 @@ from katrain.core.constants import (
     AI_DEFAULT,
     AI_HANDICAP,
     AI_INFLUENCE,
+    AI_INFLUENCE_ELO_GRID,
     AI_JIGO,
     AI_LOCAL,
+    AI_LOCAL_ELO_GRID,
     AI_PICK,
+    AI_PICK_ELO_GRID,
     AI_POLICY,
     AI_RANK,
     AI_SCORELOSS,
+    AI_SCORELOSS_ELO,
+    AI_SIMPLE_OWNERSHIP,
     AI_STRATEGIES_PICK,
     AI_STRATEGIES_POLICY,
     AI_STRENGTH,
     AI_TENUKI,
+    AI_TENUKI_ELO_GRID,
     AI_TERRITORY,
+    AI_TERRITORY_ELO_GRID,
     AI_WEIGHTED,
+    AI_WEIGHTED_ELO,
+    CALIBRATED_RANK_ELO,
     OUTPUT_DEBUG,
     OUTPUT_ERROR,
     OUTPUT_INFO,
-    AI_WEIGHTED_ELO,
-    AI_SCORELOSS_ELO,
-    CALIBRATED_RANK_ELO,
-    AI_LOCAL_ELO_GRID,
-    AI_TENUKI_ELO_GRID,
-    AI_TERRITORY_ELO_GRID,
-    AI_INFLUENCE_ELO_GRID,
-    AI_PICK_ELO_GRID,
+    AI_SETTLE_STONES,
 )
 from katrain.core.game import Game, GameNode, Move
 from katrain.core.utils import var_to_grid
@@ -162,13 +164,14 @@ def request_ai_analysis(game: Game, cn: GameNode, extra_settings: Dict) -> Optio
     error = False
     analysis = None
 
-    def set_analysis(a):
+    def set_analysis(a, partial_result):
         nonlocal analysis
+        if not partial_result:
             analysis = a
 
     def set_error(a):
         nonlocal error
-        game.katrain.log("Error in PDA-based analysis", a)
+        game.katrain.log(f"Error in additional analysis query: {a}")
         error = True
 
     engine = game.engines[cn.player]
@@ -203,7 +206,7 @@ def generate_ai_move(game: Game, ai_mode: str, ai_settings: Dict) -> Tuple[Move,
             game.katrain.log(f"Error getting handicap-based move", OUTPUT_ERROR)
             ai_mode = AI_DEFAULT
 
-    while not cn.analysis_ready:
+    while not cn.analysis_complete:
         time.sleep(0.01)
         game.engines[cn.next_player].check_alive(exception_if_dead=True)
 
@@ -328,7 +331,10 @@ def generate_ai_move(game: Game, ai_mode: str, ai_settings: Dict) -> Tuple[Move,
             candidate_ai_moves = handicap_analysis["moveInfos"]
 
         top_cand = Move.from_gtp(candidate_ai_moves[0]["move"], player=cn.next_player)
-        if top_cand.is_pass and ai_mode not in [AI_DEFAULT, AI_HANDICAP]:  # don't play suicidal to balance score
+        if top_cand.is_pass and ai_mode not in [
+            AI_DEFAULT,
+            AI_HANDICAP,
+        ]:  # don't play suicidal to balance score
             aimove = top_cand
             ai_thoughts += f"Top move is pass, so passing regardless of strategy. "
         else:
@@ -352,6 +358,80 @@ def generate_ai_move(game: Game, ai_mode: str, ai_settings: Dict) -> Tuple[Move,
                 topmove = weighted_selection_without_replacement(moves, 1)[0]
                 aimove = topmove[2]
                 ai_thoughts += f"ScoreLoss strategy found {len(candidate_ai_moves)} candidate moves (best {top_cand.gtp()}) and chose {aimove.gtp()} (weight {topmove[1]:.3f}, point loss {topmove[0]:.1f}) based on score weights."
+            elif ai_mode in [AI_SIMPLE_OWNERSHIP, AI_SETTLE_STONES]:
+                stones_with_player = {(*s.coords, s.player) for s in game.stones}
+                next_player_sign = cn.player_sign(cn.next_player)
+                if ai_mode == AI_SIMPLE_OWNERSHIP:
+
+                    def settledness(d, player_sign, player):
+                        return sum([abs(o) for o in d["ownership"] if player_sign * o > 0])
+
+                else:
+                    board_size_x, board_size_y = game.board_size
+
+                    def settledness(d, player_sign, player):
+                        ownership_grid = var_to_grid(d["ownership"], (board_size_x, board_size_y))
+                        return sum(
+                            [abs(ownership_grid[s.coords[0]][s.coords[1]]) for s in game.stones if s.player == player]
+                        )
+
+                def is_attachment(move):
+                    if move.is_pass:
+                        return False
+                    attach_opponent_stones = sum(
+                        (move.coords[0] + dx, move.coords[1] + dy, cn.player) in stones_with_player
+                        for dx in [-1, 0, 1]
+                        for dy in [-1, 0, 1]
+                        if abs(dx) + abs(dy) == 1
+                    )
+                    nearby_own_stones = sum(
+                        (move.coords[0] + dx, move.coords[1] + dy, cn.next_player) in stones_with_player
+                        for dx in [-2, 0, 1, 2]
+                        for dy in [-2 - 1, 0, 1, 2]
+                        if abs(dx) + abs(dy) <= 2  # allows clamps/jumps
+                    )
+                    return attach_opponent_stones >= 1 and nearby_own_stones == 0
+
+                def is_tenuki(d):
+                    return not d.is_pass and not any(
+                        not node
+                        or not node.move
+                        or node.move.is_pass
+                        or max(abs(last_c - cand_c) for last_c, cand_c in zip(node.move.coords, d.coords)) < 5
+                        for node in [cn, cn.parent]
+                    )
+
+                moves_with_settledness = sorted(
+                    [
+                        (
+                            move,
+                            settledness(d, next_player_sign, cn.next_player),
+                            settledness(d, -next_player_sign, cn.player),
+                            is_attachment(move),
+                            is_tenuki(move),
+                            d,
+                        )
+                        for d in candidate_ai_moves
+                        if d["pointsLost"] < ai_settings["max_points_lost"]
+                        and "ownership" in d
+                        and (d["order"] <= 1 or d["visits"] >= ai_settings.get("min_visits", 1))
+                        for move in [Move.from_gtp(d["move"], player=cn.next_player)]
+                        if not (move.is_pass and d["pointsLost"] > 0.75)
+                    ],
+                    key=lambda t: t[5]["pointsLost"]
+                    + ai_settings["attach_penalty"] * t[3]
+                    + ai_settings["tenuki_penalty"] * t[4]
+                    - ai_settings["settled_weight"] * (t[1] + ai_settings["opponent_fac"] * t[2]),
+                )
+                if moves_with_settledness:
+                    cands = [
+                        f"{move.gtp()} ({d['pointsLost']:.1f} pt lost, {d['visits']} visits, {settled:.1f} settledness, {oppsettled:.1f} opponent settledness{', attachment' if isattach else ''}{', tenuki' if istenuki else ''})"
+                        for move, settled, oppsettled, isattach, istenuki, d in moves_with_settledness[:5]
+                    ]
+                    ai_thoughts += f"{ai_mode} strategy. Top 5 Candidates {', '.join(cands)} "
+                    aimove = moves_with_settledness[0][0]
+                else:
+                    raise (Exception("No moves found - are you using an older KataGo with no per-move ownership info?"))
             else:
                 if ai_mode not in [AI_DEFAULT, AI_HANDICAP]:
                     game.katrain.log(f"Unknown AI mode {ai_mode} or policy missing, using default.", OUTPUT_INFO)

katrain/core/base_katrain.py

@@ -49,13 +49,13 @@ class KaTrainBase:
 
     """Settings, logging, and players functionality, so other classes like bots who need a katrain instance can be used without a GUI"""
 
-    def __init__(self, force_package_config=False, debug_level=0, **kwargs):
-        self.debug_level = debug_level
+    def __init__(self, force_package_config=False, debug_level=None, **kwargs):
+        self.debug_level = debug_level or 0
         self.game = None
 
         self.logger = lambda message, level=OUTPUT_INFO: self.log(message, level)
         self.config_file = self._load_config(force_package_config=force_package_config)
-        self.debug_level = debug_level or self.config("general/debug_level", OUTPUT_INFO)
+        self.debug_level = self.config("general/debug_level", OUTPUT_INFO) if debug_level is None else debug_level
 
         Config.set("kivy", "log_level", "warning")
         if self.debug_level >= OUTPUT_DEBUG:

katrain/core/constants.py

@@ -1,6 +1,6 @@
-VERSION = "1.5.0"
+VERSION = "1.6.1"
 HOMEPAGE = "https://github.com/sanderland/katrain"
-CONFIG_MIN_VERSION = "1.4.0"  # keep config files from this version
+CONFIG_MIN_VERSION = "1.6.0"  # keep config files from this version
 
 OUTPUT_ERROR = -1
 OUTPUT_KATAGO_STDERR = -0.5
@@ -33,10 +33,12 @@ AI_TENUKI = "ai:p:tenuki"
 AI_INFLUENCE = "ai:p:influence"
 AI_TERRITORY = "ai:p:territory"
 AI_RANK = "ai:p:rank"
+AI_SIMPLE_OWNERSHIP = "ai:simple"
+AI_SETTLE_STONES = "ai:settle"
 
 AI_CONFIG_DEFAULT = AI_RANK
 
-AI_STRATEGIES_ENGINE = [AI_DEFAULT, AI_HANDICAP, AI_SCORELOSS, AI_JIGO]
+AI_STRATEGIES_ENGINE = [AI_DEFAULT, AI_HANDICAP, AI_SCORELOSS, AI_SIMPLE_OWNERSHIP, AI_JIGO]  # AI_SETTLE_STONES
 AI_STRATEGIES_PICK = [AI_PICK, AI_LOCAL, AI_TENUKI, AI_INFLUENCE, AI_TERRITORY, AI_RANK]
 AI_STRATEGIES_POLICY = [AI_WEIGHTED, AI_POLICY] + AI_STRATEGIES_PICK
 AI_STRATEGIES = AI_STRATEGIES_ENGINE + AI_STRATEGIES_POLICY
@@ -44,6 +46,8 @@ AI_STRATEGIES_RECOMMENDED_ORDER = [
     AI_DEFAULT,
     AI_RANK,
     AI_HANDICAP,
+    AI_SIMPLE_OWNERSHIP,
+    #    AI_SETTLE_STONES,
     AI_SCORELOSS,
     AI_POLICY,
     AI_WEIGHTED,
@@ -67,6 +71,8 @@ AI_STRENGTH = {  # dan ranks, backup if model is missing. TODO: remove some?
     AI_INFLUENCE: -7,
     AI_TERRITORY: -7,
     AI_RANK: float("nan"),
+    AI_SIMPLE_OWNERSHIP: 2,
+    AI_SETTLE_STONES: 2,
 }
 
 AI_OPTION_VALUES = {
@@ -85,8 +91,23 @@ AI_OPTION_VALUES = {
     "threshold": [2, 2.5, 3, 3.5, 4, 4.5],
     "automatic": "bool",
     "pda": [(x / 10, f"{'W' if x<0 else 'B'}+{abs(x/10):.1f}") for x in range(-30, 31)],
+    "max_points_lost": [x / 10 for x in range(51)],
+    "settled_weight": [x / 4 for x in range(0, 17)],
+    "opponent_fac": [x / 10 for x in range(-20, 11)],
+    "min_visits": range(1, 10),
+    "attach_penalty": [x / 10 for x in range(-10, 51)],
+    "tenuki_penalty": [x / 10 for x in range(-10, 51)],
+}
+AI_KEY_PROPERTIES = {
+    "kyu_rank",
+    "strength",
+    "weaken_fac",
+    "pick_frac",
+    "pick_n",
+    "automatic",
+    "max_points_lost",
+    "min_visits",
 }
-AI_KEY_PROPERTIES = {"kyu_rank", "strength", "weaken_fac", "pick_frac", "pick_n", "automatic"}
 
 
 CALIBRATED_RANK_ELO = [

katrain/core/engine.py

@@ -13,6 +13,7 @@ from kivy.utils import platform
 from katrain.core.constants import OUTPUT_DEBUG, OUTPUT_ERROR, OUTPUT_EXTRA_DEBUG, OUTPUT_KATAGO_STDERR
 from katrain.core.game_node import GameNode
 from katrain.core.lang import i18n
+from katrain.core.sgf_parser import Move
 from katrain.core.utils import find_package_resource
 
 
@@ -85,10 +86,10 @@ class KataGoEngine:
     def start(self):
         try:
             self.katrain.log(f"Starting KataGo with {self.command}", OUTPUT_DEBUG)
-            startupinfo = None # stop command box popups on windows/pyinstaller
+            startupinfo = None
             if hasattr(subprocess, "STARTUPINFO"):
                 startupinfo = subprocess.STARTUPINFO()
-                startupinfo.dwFlags |= subprocess.STARTF_USESHOWWINDOW
+                startupinfo.dwFlags |= subprocess.STARTF_USESHOWWINDOW  # stop command box popups on win/pyinstaller
             self.katago_process = subprocess.Popen(
                 self.command,
                 startupinfo=startupinfo,
@@ -107,6 +108,8 @@ class KataGoEngine:
 
     def on_new_game(self):
         self.base_priority += 1
+        for query_id in list(self.queries.keys()):
+            self.terminate_query(query_id)
         self.queries = {}
 
     def restart(self):
@@ -131,11 +134,14 @@ class KataGoEngine:
             raise EngineDiedException(died_msg)
         return ok
 
+    def wait_to_finish(self):
+        while self.queries and self.katago_process and self.katago_process.poll() is None:
+            time.sleep(0.1)
+
     def shutdown(self, finish=False):
         process = self.katago_process
         if finish and process:
-            while self.queries and process.poll() is None:
-                time.sleep(0.1)
+            self.wait_to_finish()
         if process:
             self.katago_process = None
             process.terminate()
@@ -196,14 +202,18 @@ class KataGoEngine:
                 elif "warning" in analysis:
                     self.katrain.log(f"{analysis} received from KataGo", OUTPUT_DEBUG)
                 else:
+                    partial_result = analysis.get("isDuringSearch", False)
+                    if not partial_result:
                         del self.queries[query_id]
                     time_taken = time.time() - start_time
                     self.katrain.log(
-                        f"[{time_taken:.1f}][{query_id}] KataGo Analysis Received: {analysis.keys()}", OUTPUT_DEBUG,
+                        f"[{time_taken:.1f}][{query_id}][{'....' if partial_result else 'done'}] KataGo Analysis Received: {analysis.keys()}",
+                        OUTPUT_DEBUG,
                     )
                     self.katrain.log(line, OUTPUT_EXTRA_DEBUG)
                     try:
-                        callback(analysis)
+                        if callback and not analysis.get("noResults", False):
+                            callback(analysis, partial_result)
                     except Exception as e:
                         self.katrain.log(f"Error in engine callback for query {query_id}: {e}", OUTPUT_ERROR)
                 if getattr(self.katrain, "update_state", None):  # easier mocking etc
@@ -225,7 +235,10 @@ class KataGoEngine:
                 self.katago_process.stdin.flush()
             except OSError as e:
                 self.check_alive(os_error=str(e), exception_if_dead=True)
-                return  # do not raise, since there's nothing to catch it
+
+    def terminate_query(self, query_id):
+        if query_id is not None:
+            self.send_query({"action": "terminate", "terminateId": query_id}, None, None)
 
     def request_analysis(
         self,
@@ -236,10 +249,12 @@ class KataGoEngine:
         analyze_fast: bool = False,
         time_limit=True,
         find_alternatives: bool = False,
+        find_local: bool = False,
         priority: int = 0,
         ownership: Optional[bool] = None,
         next_move: Optional[GameNode] = None,
         extra_settings: Optional[Dict] = None,
+        report_every: Optional[float] = None,
     ):
         nodes = analysis_node.nodes_from_root
         moves = [m for node in nodes for m in node.moves]
@@ -253,6 +268,8 @@ class KataGoEngine:
             if analyze_fast and self.config.get("fast_visits"):
                 visits = self.config["fast_visits"]
 
+        size_x, size_y = analysis_node.board_size
+
         if find_alternatives:
             avoid = [
                 {
@@ -261,10 +278,26 @@ class KataGoEngine:
                     "untilDepth": 1,
                 }
             ]
+        elif find_local:
+            distance = 5
+            last_move = analysis_node.move
+            if last_move is None or last_move.is_pass:
+                return
+            avoid = [
+                {
+                    "moves": [
+                        Move((x, y)).gtp()
+                        for x in range(0, size_x)
+                        for y in range(0, size_y)
+                        if max(abs(x - last_move.coords[0]), abs(y - last_move.coords[1])) > distance
+                    ],
+                    "player": analysis_node.next_player,
+                    "untilDepth": 1,
+                }
+            ]
         else:
             avoid = []
 
-        size_x, size_y = analysis_node.board_size
         settings = copy.copy(self.override_settings)
         if time_limit:
             settings["maxTime"] = self.config["max_time"]
@@ -277,15 +310,19 @@ class KataGoEngine:
             "analyzeTurns": [len(moves)],
             "maxVisits": visits,
             "komi": analysis_node.komi,
-            "avoidMoves": avoid,
             "boardXSize": size_x,
             "boardYSize": size_y,
             "includeOwnership": ownership and not next_move,
+            "includeMovesOwnership": ownership and not next_move,
             "includePolicy": not next_move,
             "initialStones": [[m.player, m.gtp()] for m in initial_stones],
             "initialPlayer": analysis_node.root.next_player,
             "moves": [[m.player, m.gtp()] for m in moves],
             "overrideSettings": {**settings, **(extra_settings or {})},
         }
+        if report_every is not None:
+            query["reportDuringSearchEvery"] = report_every
+        if avoid:
+            query["avoidMoves"] = avoid
         self.send_query(query, callback, error_callback, next_move)
         analysis_node.analysis_visits_requested = max(analysis_node.analysis_visits_requested, visits)

katrain/core/game.py

@@ -79,12 +79,14 @@ class Game:
             self.root.set_property("RU", katrain.config("game/rules"))
 
         self.set_current_node(self.root)
+        self.main_time_used = 0
         threading.Thread(
             target=lambda: self.analyze_all_nodes(-1_000_000, analyze_fast=analyze_fast), daemon=True
         ).start()  # return faster, but bypass Kivy Clock
 
     def analyze_all_nodes(self, priority=0, analyze_fast=False):
         for node in self.root.nodes_in_tree:
+            node.clear_analysis()
             node.analyze(self.engines[node.next_player], priority=priority, analyze_fast=analyze_fast)
 
     # -- move tree functions --
@@ -183,11 +185,21 @@ class Game:
                 cn = cn.parent
         self.set_current_node(cn)
 
-    def redo(self, n_times=1):
+    def redo(self, n_times=1, stop_on_mistake=None):
         cn = self.current_node  # avoid race conditions
-        for _ in range(n_times):
+        for move in range(n_times):
             if cn.children:
                 cn = cn.ordered_children[0]
+            if (
+                move > 0
+                and stop_on_mistake is not None
+                and cn.points_lost is not None
+                and cn.points_lost >= stop_on_mistake
+                and self.katrain.players_info[cn.player].player_type != PLAYER_AI
+            ):
+                self.set_current_node(cn.parent)
+                return
+        if stop_on_mistake is None:
             self.set_current_node(cn)
 
     def cycle_children(self, direction):
@@ -328,6 +340,7 @@ class Game:
             else:
                 visits = cn.analysis_visits_requested + engine.config["max_visits"]
                 self.katrain.controls.set_status(i18n._("extra analysis").format(visits=visits), STATUS_ANALYSIS)
+            self.katrain.controls.set_status(i18n._("extra analysis").format(visits=visits), STATUS_ANALYSIS)
             cn.analyze(engine, visits=visits, priority=-1_000, time_limit=False)
             return
         if mode == "game":
@@ -338,13 +351,13 @@ class Game:
                 min_visits = min(node.analysis_visits_requested for node in nodes)
                 visits = min_visits + engine.config["max_visits"]
             for node in nodes:
-                node.analyze(engine, visits=visits, priority=-1_000_000, time_limit=False)
+                node.analyze(engine, visits=visits, priority=-1_000_000, time_limit=False, report_every=None)
             self.katrain.controls.set_status(i18n._("game re-analysis").format(visits=visits), STATUS_ANALYSIS)
             return
 
         elif mode == "sweep":
             board_size_x, board_size_y = self.board_size
-            if cn.analysis_ready:
+            if cn.analysis_exists:
                 policy_grid = (
                     var_to_grid(self.current_node.policy, size=(board_size_x, board_size_y))
                     if self.current_node.policy
@@ -369,32 +382,35 @@ class Game:
             visits = engine.config["fast_visits"]
             self.katrain.controls.set_status(i18n._("sweep analysis").format(visits=visits), STATUS_ANALYSIS)
             priority = -1_000_000_000
-        elif mode in ["equalize", "alternative"]:
-            if not cn.analysis_ready:
+        elif mode == "local":  # also do a quick update on current candidates so it doesn't look too weird
+            self.katrain.controls.set_status(i18n._("local analysis"), STATUS_ANALYSIS)
+            cn.analyze(engine, priority=-500, time_limit=False, find_local="local")
+            return
+        elif mode in ["equalize", "alternative", "local"]:
+            if not cn.analysis_complete and mode != "local":
                 self.katrain.controls.set_status(i18n._("wait-before-equalize"), STATUS_INFO, self.current_node)
                 return
-
-            analyze_moves = [Move.from_gtp(gtp, player=cn.next_player) for gtp, _ in cn.analysis["moves"].items()]
-            priority = -1_000
             if mode == "alternative":  # also do a quick update on current candidates so it doesn't look too weird
                 self.katrain.controls.set_status(i18n._("alternative analysis"), STATUS_ANALYSIS)
-                cn.analyze(engine, priority=-500, time_limit=False, find_alternatives=True)
+                cn.analyze(engine, priority=-500, time_limit=False, find_alternatives="alternative")
                 visits = engine.config["fast_visits"]
-            else:
+            else:  # equalize
                 visits = max(d["visits"] for d in cn.analysis["moves"].values())
                 self.katrain.controls.set_status(i18n._("equalizing analysis").format(visits=visits), STATUS_ANALYSIS)
+            priority = -1_000
+            analyze_moves = [Move.from_gtp(gtp, player=cn.next_player) for gtp, _ in cn.analysis["moves"].items()]
         else:
             raise ValueError("Invalid analysis mode")
         for move in analyze_moves:
             if cn.analysis["moves"].get(move.gtp(), {"visits": 0})["visits"] < visits:
                 cn.analyze(
-                    engine, priority, visits=visits, refine_move=move, time_limit=False
+                    engine, priority=priority, visits=visits, refine_move=move, time_limit=False, report_every=None
                 )  # explicitly requested so take as long as you need
 
     def analyze_undo(self, node):
         train_config = self.katrain.config("trainer")
         move = node.move
-        if node != self.current_node or node.auto_undo is not None or not node.analysis_ready or not move:
+        if node != self.current_node or node.auto_undo is not None or not node.analysis_complete or not move:
             return
         points_lost = node.points_lost
         thresholds = train_config["eval_thresholds"]

katrain/core/game_node.py

@@ -13,7 +13,6 @@ class GameNode(SGFNode):
 
     def __init__(self, parent=None, properties=None, move=None):
         super().__init__(parent=parent, properties=properties, move=move)
-        self.analysis = {"moves": {}, "root": None}
         self.ownership = None
         self.policy = None
         self.auto_undo = None  # None = not analyzed. False: not undone (good move). True: undone (bad move)
@@ -21,9 +20,13 @@ class GameNode(SGFNode):
         self.note = ""
         self.move_number = 0
         self.time_used = 0
-        self.analysis_visits_requested = 0
         self.undo_threshold = random.random()  # for fractional undos
         self.end_state = None
+        self.clear_analysis()
+
+    def clear_analysis(self):
+        self.analysis_visits_requested = 0
+        self.analysis = {"moves": {}, "root": None, "completed": False}
 
     def sgf_properties(self, save_comments_player=None, save_comments_class=None, eval_thresholds=None):
         properties = copy.copy(super().sgf_properties())
@@ -34,8 +37,8 @@ class GameNode(SGFNode):
             show_class = False
         if (
             self.parent
-            and self.parent.analysis_ready
-            and self.analysis_ready
+            and self.parent.analysis_exists
+            and self.analysis_exists
             and (note or ((save_comments_player or {}).get(self.player, False) and show_class))
         ):
             candidate_moves = self.parent.candidate_moves
@@ -77,16 +80,22 @@ class GameNode(SGFNode):
         refine_move=None,
         analyze_fast=False,
         find_alternatives=False,
+        find_local=False,
+        report_every=0.25,
     ):
         engine.request_analysis(
             self,
-            lambda result: self.set_analysis(result, refine_move, find_alternatives),
+            callback=lambda result, partial_result: self.set_analysis(
+                result, refine_move, find_alternatives or find_local, partial_result
+            ),
             priority=priority,
             visits=visits,
             analyze_fast=analyze_fast,
             time_limit=time_limit,
             next_move=refine_move,
             find_alternatives=find_alternatives,
+            find_local=find_local,
+            report_every=report_every,
         )
 
     def update_move_analysis(self, move_analysis, move_gtp):
@@ -102,24 +111,26 @@ class GameNode(SGFNode):
             if cur["visits"] < move_analysis["visits"]:
                 cur.update(move_analysis)
 
-    def set_analysis(self, analysis_json: Dict, refine_move: Optional[Move], alternatives_mode: bool):
+    def set_analysis(
+        self, analysis_json: Dict, refine_move: Optional[Move], additional_moves: bool, partial_result: bool = False
+    ):
         if refine_move:
             pvtail = analysis_json["moveInfos"][0]["pv"] if analysis_json["moveInfos"] else []
             self.update_move_analysis(
                 {"pv": [refine_move.gtp()] + pvtail, **analysis_json["rootInfo"]}, refine_move.gtp()
             )
         else:
-            if alternatives_mode:
+            if additional_moves:
                 for m in analysis_json["moveInfos"]:
-                    m["order"] += 10  # offset for not making this top
-            if refine_move is None and not alternatives_mode:
+                    del m["order"]  # avoid changing order
+            if refine_move is None and not additional_moves:
                 for move_dict in self.analysis["moves"].values():
                     move_dict["order"] = 999  # old moves to end
             for move_analysis in analysis_json["moveInfos"]:
                 self.update_move_analysis(move_analysis, move_analysis["move"])
             self.ownership = analysis_json.get("ownership")
             self.policy = analysis_json.get("policy")
-            if not alternatives_mode:
+            if not additional_moves:
                 self.analysis["root"] = analysis_json["rootInfo"]
             if self.parent and self.move:
                 analysis_json["rootInfo"]["pv"] = [self.move.gtp()] + (
@@ -128,14 +139,20 @@ class GameNode(SGFNode):
                 self.parent.update_move_analysis(
                     analysis_json["rootInfo"], self.move.gtp()
                 )  # update analysis in parent for consistency
+            is_normal_query = refine_move is None and not additional_moves
+            self.analysis["completed"] = self.analysis["completed"] or (is_normal_query and not partial_result)
 
     @property
-    def analysis_ready(self):
+    def analysis_exists(self):
         return self.analysis["root"] is not None
 
+    @property
+    def analysis_complete(self):
+        return self.analysis["completed"] and self.analysis["root"] is not None
+
     @property
     def score(self) -> Optional[float]:
-        if self.analysis_ready:
+        if self.analysis_exists:
             return self.analysis["root"].get("scoreLead")
 
     def format_score(self, score=None):
@@ -145,7 +162,7 @@ class GameNode(SGFNode):
 
     @property
     def winrate(self) -> Optional[float]:
-        if self.analysis_ready:
+        if self.analysis_exists:
             return self.analysis["root"].get("winrate")
 
     def format_winrate(self, win_rate=None):
@@ -176,12 +193,12 @@ class GameNode(SGFNode):
             return ""
 
         text = i18n._("move").format(number=self.depth) + f": {single_move.player} {single_move.gtp()}\n"
-        if self.analysis_ready:
+        if self.analysis_exists:
             score = self.score
             if sgf:
                 text += i18n._("Info:score").format(score=self.format_score(score)) + "\n"
                 text += i18n._("Info:winrate").format(winrate=self.format_winrate()) + "\n"
-            if self.parent and self.parent.analysis_ready:
+            if self.parent and self.parent.analysis_exists:
                 previous_top_move = self.parent.candidate_moves[0]
                 if sgf or details:
                     if previous_top_move["move"] != single_move.gtp():
@@ -202,20 +219,21 @@ class GameNode(SGFNode):
                     if currmove_pol_rank is not None:
                         policy_rank_msg = i18n._("Info:policy rank")
                         text += policy_rank_msg.format(rank=currmove_pol_rank, probability=currmove_pol_prob) + "\n"
-                    if currmove_pol_rank is None or currmove_pol_rank != 1 and (sgf or details):
+                    if currmove_pol_rank != 1 and policy_ranking and (sgf or details):
                         policy_best_msg = i18n._("Info:policy best")
                         pol_move, pol_prob = policy_ranking[0][1].gtp(), policy_ranking[0][0]
                         text += policy_best_msg.format(move=pol_move, probability=pol_prob) + "\n"
             if self.auto_undo and sgf:
                 text += i18n._("Info:teaching undo") + "\n"
-                top_pv = self.analysis_ready and self.candidate_moves[0].get("pv")
+                top_pv = self.analysis_exists and self.candidate_moves[0].get("pv")
                 if top_pv:
                     text += i18n._("Info:undo predicted PV").format(pv=f"{self.next_player}{' '.join(top_pv)}") + "\n"
-            if self.ai_thoughts and (sgf or details):
-                text += "\n" + i18n._("Info:AI thoughts").format(thoughts=self.ai_thoughts)
         else:
             text = i18n._("No analysis available") if sgf else i18n._("Analyzing move...")
 
+        if self.ai_thoughts and (sgf or details):
+            text += "\n" + i18n._("Info:AI thoughts").format(thoughts=self.ai_thoughts)
+
         if "C" in self.properties:
             text += "\n[u]SGF Comments:[/u]\n" + "\n".join(self.properties["C"])
 
@@ -224,7 +242,7 @@ class GameNode(SGFNode):
     @property
     def points_lost(self) -> Optional[float]:
         single_move = self.move
-        if single_move and self.parent and self.analysis_ready and self.parent.analysis_ready:
+        if single_move and self.parent and self.analysis_exists and self.parent.analysis_exists:
             parent_score = self.parent.score
             score = self.score
             return self.player_sign(single_move.player) * (parent_score - score)
@@ -236,8 +254,8 @@ class GameNode(SGFNode):
             single_move
             and self.parent
             and self.parent.parent
-            and self.analysis_ready
-            and self.parent.parent.analysis_ready
+            and self.analysis_exists
+            and self.parent.parent.analysis_exists
         ):
             parent_parent_score = self.parent.parent.score
             score = self.score
@@ -249,7 +267,7 @@ class GameNode(SGFNode):
 
     @property
     def candidate_moves(self) -> List[Dict]:
-        if not self.analysis_ready:
+        if not self.analysis_exists:
             return []
         if not self.analysis["moves"]:
             polmoves = self.policy_ranking

katrain/gui.kv

@@ -32,7 +32,7 @@
 #:set CP_SPACING dp(6)
 #:set CP_SMALL_SPACING dp(3)
 #:set CP_PADDING dp(6)
-#:set NOTES_FONT_SIZE dp(18)
+#:set NOTES_FONT_SIZE sp(18)
 
 #:set EPSILON 1e-9
 
@@ -126,7 +126,7 @@
         pos: [root.pos[i] + (root.size[i] - root.icon_size[i])/2 for i in [0,1]]
         source: root.icon
         mipmap: True
-        color: [1,1,1,0.4] if root.disabled else [1,1,1,1]
+        color: [1,1,1,0.4] if root.disabled else root.color
 
 
 <I18NSpinnerOption@SpinnerOption>:
@@ -171,8 +171,6 @@
         size: image.size
         font_size: root.min_size * 1.2 / (1 + 1 * len(self.text))
 
-
-
 # UI components
 
 <BadukPanWidget>:
@@ -187,7 +185,6 @@
 
 <BadukPanControls>:
     katrain: app.gui
-    branch: branch
     circles: [black_circle,white_circle]
     mid_circles_container: mid_circles_container
     pass_btn: pass_btn
@@ -209,7 +206,7 @@
             on_press: root.katrain('undo',999)
         NavIconButton:
             icon: 'img/Previous-5.png'
-            on_press: root.katrain('undo',5)
+            on_press: root.katrain('undo',10)
         NavIconButton:
             icon: 'img/Previous.png'
             on_press: root.katrain('undo')
@@ -233,20 +230,19 @@
             on_press: root.katrain('redo')
         NavIconButton:
             icon: 'img/Next-5.png'
-            on_press: root.katrain('redo',5)
+            on_press: root.katrain('redo',10)
         NavIconButton:
             icon: 'img/Next-End.png'
             on_press: root.katrain('redo',999)
         NavIconButton:
-            id: branch
-            icon: 'img/Branch.png'
-            on_press: root.katrain('cycle-children',1)
-            disabled: True
+            icon: 'img/Next-Mistake.png'
+            color: [0.65,0.05,0.05,1]
+            on_press: root.katrain('next-mistake')
     ClickableLabel:
         pos_hint: {'center_x': 0.85,'center_y':0.5}
         size_hint: None, 1
         width: self.height
-        canvas:
+        canvas.before:
             Color:
                 rgba: YELLOW
             Ellipse:
@@ -264,6 +260,9 @@
             Line:
                 circle: self.pos[0] + self.width/2, self.pos[1] + self.height/2, self.width/6
                 width: 1
+        text: str(root.queries_remaining) if root.queries_remaining >= 3 else ''
+        color: BLACK
+        font_size: 0.5 * self.width / max(2,len(self.text))
         on_press: app.gui.controls.set_status(i18n._('ai-ponder'),STATUS_INFO)
 
 
@@ -357,14 +356,14 @@
     spacing: CP_SPACING
     padding: CP_PADDING * 4, CP_PADDING * 2
     paused: True
-    timeout: self.state[1] <= 0
+    timeout: root.state[1] is not None and root.state[1] <= 0
     MDBoxLayout:
         size_hint: 0.7,1
         MDFloatLayout:
             size_hint: 0.4,1
             TimerLabel:
                 color: BACKGROUND_COLOR
-                text: '8:88'
+                text: '{}:88'.format('188' if root.state[0] >= 6000 else '88')
                 font_size: 0.85*self.height
                 pos_hint: {'right':1,'center_y':0.5}
                 halign: 'right'
@@ -372,27 +371,27 @@
             TimerLabel:
                 id: timer
                 timeout: root.timeout
-                text: "{:01d}:{:02d}".format(int(root.state[0]+0.99)//60,int((root.state[0]+0.99) % 60) )
+                text: "{:2d}:{:02d}".format(int(root.state[0]+0.99)//60,int((root.state[0]+0.99) % 60) )
                 font_size: 0.85*self.height
                 pos_hint: {'right':1,'center_y':0.5}
                 halign: 'right'
                 valign: 'middle'
         TimerLabel:
             size_hint: 0.05,1
-            text: "x"
+            text: "x" if  root.state[1] is not None else ""
             font_size: self.height * 0.3
             text_size: self.width, self.height * 0.95
             halign: 'right'
             valign: 'top'
             timeout: root.timeout
         TimerLabel:
-            size_hint: 0.1,1
+            size_hint: 0.075,1
             font_size: self.height * 0.35
             text_size: self.size
             halign: 'left'
             valign: 'top'
             id: periods
-            text: str(root.state[1])
+            text: str(root.state[1]) if root.state[1] is not None else ""
             max_lines: 1
             timeout: root.timeout
     MDBoxLayout:
@@ -507,8 +506,6 @@
     katrain: app.gui
     orientation: 'vertical'
     players: {'B':B_player,'W':W_player}
-    engine_status_col: [1,1,1,1]
-    engine_status_pondering: -1
     timer:timer
     graph: graph
     stats: stats
@@ -685,6 +682,7 @@
         color: LIGHTGREY
 
 
+
 <AnalysisToggle>:
     adaptive_width: True
     spacing: CP_SPACING
@@ -700,15 +698,18 @@
         color: checkbox.disabled_color if root.disabled else (checkbox.selected_color if checkbox.active else checkbox.unselected_color)
         size_hint_x: None
         width: self.texture_size[0] + 4
-        on_left_press: checkbox.active = not checkbox.active
-    MDCheckbox:
+        on_left_press: checkbox.trigger_action()
+    TriStateMDCheckbox:
         id: checkbox
         disabled: root.disabled
+        tri_state: root.tri_state
         size_hint: None,0.5
         pos_hint: {'center_y': 0.5}
         width: self.height
         active: root.default_active
         on_active: if app.gui: app.gui.update_state()
+        on_slashed: if app.gui: app.gui.update_state()
+
 
 <AnalysisMenuItem@MenuItem>:
     on_close: app.gui.analysis_controls.close_dropdown()
@@ -749,6 +750,11 @@
         icon: 'img/Alternative.png'
         shortcut: 'f'
         on_action: root.katrain("analyze-extra", "alternative")
+    AnalysisMenuItem:
+        text: i18n._("analysis:local")
+        icon: 'img/local.png'
+        shortcut: 'g'
+        on_action: root.katrain("analyze-extra", "local")
     AnalysisMenuItem:
         text: i18n._("analysis:game")
         icon: 'img/analysis.png'
@@ -792,6 +798,7 @@
             AnalysisToggle:
                 id: eval
                 text: i18n._("analysis:dots")
+                tri_state: True
             AnalysisToggle:
                 id: hints
                 text: i18n._("analysis:topmoves")
@@ -853,11 +860,12 @@
     player_subtype_ai: player_subtype_ai
     MDBoxLayout:
         size_hint: 0.25, 1
-        CircleWithText:
+        ClickableCircle:
             size_hint: 0.8, 0.8
             pos_hint: {'center_x':0.5,'center_y':0.5}
             text: root.player[0]
             player: root.player
+            on_left_press: root.parent.swap_players()
     MDBoxLayout:
         size_hint: 0.5, 1
         orientation: 'vertical'
@@ -1059,14 +1067,14 @@
                         AnalysisControls:
                             id: analysis_controls
                             size_hint_y: None
-                            height: 1 if root.zen > 0 else max(dp(50),min(dp(75),self.width / 13.5))
+                            height: 1 if root.zen > 0 else max(sp(50),min(sp(75),self.width / 13.5))
                             opacity: 0 if root.zen > 0 else 1
                         BadukPanWidget:
                             id: board_gui
                         BadukPanControls:
                             id: board_controls
                             size_hint_y: None
-                            height: 1 if root.zen > 0 else max(dp(50),min(dp(75),self.width / 13.5))
+                            height: 1 if root.zen > 0 else max(sp(50),min(sp(75),self.width / 13.5))
                             opacity: 0 if root.zen > 0 else 1
                     MDBoxLayout:
                         size_hint_x: None

katrain/gui/badukpan.py

@@ -42,7 +42,7 @@ class BadukPanWidget(Widget):
         Window.bind(mouse_pos=self.on_mouse_pos)
         self.redraw_board_contents_trigger = Clock.create_trigger(self.draw_board_contents)
         self.redraw_trigger = Clock.create_trigger(self.redraw)
-        self.bind(size=self.redraw_trigger)
+        self.bind(size=self.redraw_trigger, pos=self.redraw_trigger)
         Clock.schedule_interval(self.animate_pv, 0.1)
 
     # stone placement functions
@@ -106,6 +106,21 @@ class BadukPanWidget(Widget):
             return
         katrain = self.katrain
         if self.ghost_stone and ("button" not in touch.profile or touch.button == "left"):
+            game = self.katrain and self.katrain.game
+            current_node = game and self.katrain.game.current_node
+            if (
+                current_node
+                and not current_node.children
+                and not self.katrain.next_player_info.ai
+                and not self.katrain.controls.timer.paused
+                and self.katrain.play_analyze_mode == MODE_PLAY
+                and self.katrain.config("timer/main_time", 0) * 60 - game.main_time_used <= 0
+                and current_node.time_used < self.katrain.config("timer/minimal_use", 0)
+            ):
+                self.katrain.controls.set_status(
+                    i18n._("move too fast").format(num=self.katrain.config("timer/minimal_use", 0)), STATUS_TEACHING
+                )
+            else:
                 katrain("play", self.ghost_stone)
                 self.play_stone_sound()
         elif not self.ghost_stone:
@@ -125,7 +140,7 @@ class BadukPanWidget(Widget):
                     katrain.log(f"\nRoot Stats:\n{nodes_here[-1].analysis['root']}", OUTPUT_DEBUG)
                     katrain.controls.info.text = nodes_here[-1].comment(sgf=True)
                     katrain.controls.active_comment_node = nodes_here[-1]
-                    if nodes_here[-1].parent.analysis_ready:
+                    if nodes_here[-1].parent.analysis_exists:
                         self.set_animating_pv(nodes_here[-1].parent.candidate_moves[0]["pv"], nodes_here[-1].parent)
 
         self.ghost_stone = None
@@ -257,6 +272,7 @@ class BadukPanWidget(Widget):
             current_node = katrain.game.current_node
             game_ended = katrain.game.end_result
             full_eval_on = katrain.analysis_controls.eval.active
+            all_dots_off = katrain.analysis_controls.eval.checkbox.slashed
             has_stone = {}
             drawn_stone = {}
             for m in katrain.game.stones:
@@ -280,7 +296,9 @@ class BadukPanWidget(Widget):
                 placements = node.placements
                 for m in node.moves + placements:
                     if has_stone.get(m.coords) and not drawn_stone.get(m.coords):  # skip captures, last only for
-                        move_eval_on = show_dots_for.get(m.player) and (i < show_n_eval or full_eval_on)
+                        move_eval_on = (
+                            not all_dots_off and show_dots_for.get(m.player) and (i < show_n_eval or full_eval_on)
+                        )
                         if move_eval_on and points_lost is not None:
                             evalcol = self.eval_color(points_lost, show_dots_for_class)
                         else:
@@ -454,7 +472,7 @@ class BadukPanWidget(Widget):
                             source="img/topmove.png",
                         )
                         if self.trainer_config["text_point_loss"] and text_on:
-                            if move_dict["pointsLost"] < 0.05:
+                            if -0.05 < move_dict["pointsLost"] < 0.05:
                                 ptloss_text = "0.0"
                             else:
                                 ptloss_text = f"{-move_dict['pointsLost']:+.1f}"
@@ -484,7 +502,7 @@ class BadukPanWidget(Widget):
                 for child_node in current_node.children:
                     move = child_node.move
                     if move and move.coords is not None:
-                        if child_node.analysis_ready:
+                        if child_node.analysis_exists:
                             self.active_pv_moves.append(
                                 (move.coords, [move.gtp()] + child_node.candidate_moves[0]["pv"], current_node)
                             )
@@ -619,3 +637,4 @@ class AnalysisControls(MDBoxLayout):
 class BadukPanControls(MDFloatLayout):
     engine_status_col = ListProperty(ENGINE_DOWN_COL)
     engine_status_pondering = NumericProperty(-1)
+    queries_remaining = NumericProperty(0)

katrain/gui/controlspanel.py

@@ -71,6 +71,9 @@ class ControlsPanel(BoxLayout):
         self.active_comment_node = None
         self.last_timer_update = (None, 0, False)
         self.beep = SoundLoader.load("sounds/countdownbeep.wav")
+        self.boing = SoundLoader.load("sounds/boing.wav")
+        if self.boing:
+            self.boing.volume = 0.1
         self.beep_start = 5.2
         self.timer_interval = 0.07
 
@@ -121,7 +124,7 @@ class ControlsPanel(BoxLayout):
                 self.active_comment_node = current_node.children[-1]
             elif current_node.parent:
                 self.active_comment_node = current_node.parent
-        elif both_players_are_robots and not current_node.analysis_ready and current_node.parent:
+        elif both_players_are_robots and not current_node.analysis_exists and current_node.parent:
             self.active_comment_node = current_node.parent
 
         lock_ai = katrain.config("trainer/lock_ai") and katrain.play_analyze_mode == MODE_PLAY
@@ -132,7 +135,7 @@ class ControlsPanel(BoxLayout):
                 teach=katrain.players_info[self.active_comment_node.player].being_taught, details=details
             )
 
-        if self.active_comment_node.analysis_ready:
+        if self.active_comment_node.analysis_exists:
             self.stats.score = self.active_comment_node.format_score() or ""
             self.stats.winrate = self.active_comment_node.format_winrate() or ""
             self.stats.points_lost = self.active_comment_node.points_lost
@@ -149,11 +152,13 @@ class ControlsPanel(BoxLayout):
         self.info.text = info
 
     def update_timer(self, _dt):
-        current_node = self.katrain and self.katrain.game and self.katrain.game.current_node
+        game = self.katrain and self.katrain.game
+        current_node = game and self.katrain.game.current_node
         if current_node:
             last_update_node, last_update_time, beeping = self.last_timer_update
             new_beeping = beeping
             now = time.time()
+            main_time = self.katrain.config("timer/main_time", 0) * 60
             byo_len = max(1, self.katrain.config("timer/byo_length"))
             byo_num = max(1, self.katrain.config("timer/byo_periods"))
             sounds_on = self.katrain.config("timer/sound")
@@ -161,8 +166,15 @@ class ControlsPanel(BoxLayout):
             ai = player.ai
             used_period = False
 
+            min_use = self.katrain.config("timer/minimal_use", 0)
+            boing_at_remaining = byo_len - min_use
+            main_time_remaining = main_time - game.main_time_used
+
             if not self.timer.paused and not ai and self.katrain.play_analyze_mode == MODE_PLAY:
                 if last_update_node == current_node and not current_node.children:
+                    if main_time_remaining > 0:
+                        game.main_time_used += now - last_update_time
+                    else:
                         current_node.time_used += now - last_update_time
                 else:
                     current_node.time_used = 0
@@ -173,6 +185,7 @@ class ControlsPanel(BoxLayout):
                     time_remaining += byo_len
                     player.periods_used += 1
                     used_period = True
+
                 if (
                     self.beep_start - 2 * self.timer_interval < time_remaining < self.beep_start
                     and player.periods_used < byo_num
@@ -180,6 +193,18 @@ class ControlsPanel(BoxLayout):
                     new_beeping = True
                 elif time_remaining > self.beep_start:
                     new_beeping = False
+
+                if (
+                    min_use
+                    and not new_beeping
+                    and self.boing
+                    and boing_at_remaining - self.timer_interval
+                    < time_remaining
+                    < boing_at_remaining + self.timer_interval
+                    and player.periods_used < byo_num
+                ):
+                    self.boing.play()
+
             else:
                 new_beeping = False
 
@@ -198,4 +223,7 @@ class ControlsPanel(BoxLayout):
 
             self.last_timer_update = (current_node, now, new_beeping)
 
+            if main_time_remaining > 0:
+                self.timer.state = (main_time_remaining, None, ai)
+            else:
                 self.timer.state = (max(0, time_remaining), max(0, periods_rem), ai)

katrain/gui/kivyutils.py

@@ -23,6 +23,7 @@ from kivymd.uix.behaviors import CircularRippleBehavior, RectangularRippleBehavi
 from kivymd.uix.boxlayout import MDBoxLayout
 from kivymd.uix.button import BaseFlatButton, BasePressedButton
 from kivymd.uix.navigationdrawer import MDNavigationDrawer
+from kivymd.uix.selectioncontrol import MDCheckbox
 from kivymd.uix.textfield import MDTextField
 
 from katrain.core.constants import AI_STRATEGIES_RECOMMENDED_ORDER, GAME_TYPES, MODE_PLAY, PLAYER_AI
@@ -112,6 +113,7 @@ class AutoSizedRectangleToggleButton(ToggleButtonMixin, AutoSizedRectangleButton
 
 
 class TransparentIconButton(CircularRippleBehavior, Button):
+    color = ListProperty([1, 1, 1, 1])
     icon_size = ListProperty([25, 25])
     icon = StringProperty("")
     disabled = BooleanProperty(False)
@@ -342,9 +344,12 @@ class PlayerSetupBlock(MDBoxLayout):
     def swap_players(self):
         player_dump = {bw: p.player_type_dump for bw, p in self.players.items()}
         for bw in "BW":
-            self.update_players(bw, player_dump["B" if bw == "W" else "W"])
+            self.update_player_params(bw, player_dump["B" if bw == "W" else "W"])
 
-    def update_players(self, bw, player_info):  # update sub widget based on gui state change
+    def update_player_params(self, bw, params):
+        self.players[bw].update_widget(**params)
+
+    def update_player_info(self, bw, player_info):  # update sub widget based on gui state change
         self.players[bw].update_widget(player_type=player_info.player_type, player_subtype=player_info.player_subtype)
 
 
@@ -368,14 +373,36 @@ class Timer(BGBoxLayout):
     timeout = BooleanProperty(False)
 
 
+class TriStateMDCheckbox(MDCheckbox):
+    tri_state = BooleanProperty(False)
+    slashed = BooleanProperty(False)
+
+    def _do_press(self):
+        if not self.tri_state:
+            return super()._do_press()
+        if self.slashed:
+            self.state = "normal"
+            self.slashed = False
+            self.icon = "checkbox-blank-outline"
+        elif self.state == "down":
+            self.state = "normal"
+            self.slashed = True
+            self.icon = "checkbox-blank-off-outline"
+        else:
+            self.state = "down"
+            self.slashed = False
+            self.icon = "checkbox-marked-outline"
+
+
 class AnalysisToggle(MDBoxLayout):
     text = StringProperty("")
     default_active = BooleanProperty(False)
     font_name = StringProperty(DEFAULT_FONT)
     disabled = BooleanProperty(False)
+    tri_state = BooleanProperty(False)
 
     def trigger_action(self, *args, **kwargs):
-        return self.checkbox.trigger_action(*args, **kwargs)
+        return self.checkbox._do_press()
 
     def activate(self, *_args):
         self.checkbox.active = True
@@ -559,6 +586,10 @@ class ClickableLabel(LeftButtonBehavior, Label):
     pass
 
 
+class ClickableCircle(LeftButtonBehavior, CircleWithText):
+    pass
+
+
 class ScrollableLabel(ScrollView, BackgroundMixin):
     __events__ = ["on_ref_press"]
     outline_color = ListProperty([0, 0, 0, 0])  # mixin not working for some reason

katrain/gui/popups.py

@@ -238,6 +238,7 @@ class ConfigTimerPopup(QuickConfigGui):
             p.periods_used = 0
         self.katrain.controls.timer.paused = True
         self.katrain.game.current_node.time_used = 0
+        self.katrain.game.main_time_used = 0
         self.katrain.update_state()
 
 
@@ -245,7 +246,7 @@ class NewGamePopup(QuickConfigGui):
     def __init__(self, katrain):
         super().__init__(katrain)
         for bw, info in katrain.players_info.items():
-            self.player_setup.update_players(bw, info)
+            self.player_setup.update_player_info(bw, info)
 
         self.rules_spinner.value_refs = [name for abbr, name in katrain.engine.RULESETS_ABBR]
 
@@ -259,6 +260,20 @@ class NewGamePopup(QuickConfigGui):
             self.katrain.update_player(bw, **player_setup.player_type_dump)
         self.katrain("new-game")
 
+    def update_game(self):
+        props = self.collect_properties(self)
+        root = self.katrain.game.root
+        changed = False
+        for k, v in [("RU", props["game/rules"]), ("KM", props["game/komi"])]:
+            current = root.get_property(k)
+            if current != v:
+                changed = True
+                self.katrain.game.root.set_property(k, v)
+        if changed:
+            self.katrain.engine.on_new_game()
+            self.katrain.game.analyze_all_nodes()
+        self.popup.dismiss()
+
 
 def wrap_anchor(widget):
     anchor = AnchorLayout()
@@ -341,7 +356,7 @@ class ConfigAIPopup(QuickConfigGui):
         self.options_grid.clear_widgets()
         self.help_label.text = i18n._(strategy.replace("ai:", "aihelp:"))
         for k, v in sorted(mode_settings.items(), key=lambda kv: (kv[0] not in AI_KEY_PROPERTIES, kv[0])):
-            self.options_grid.add_widget(DescriptionLabel(text=k, size_hint_x=0.25))
+            self.options_grid.add_widget(DescriptionLabel(text=k, size_hint_x=0.275))
             if k in AI_OPTION_VALUES:
                 values = AI_OPTION_VALUES[k]
                 if values == "bool":
@@ -482,20 +497,20 @@ class ConfigPopup(QuickConfigGui):
 
     KATAGOS = {
         "win": {
-            "OpenCL v1.6.1": "https://github.com/lightvector/KataGo/releases/download/v1.6.1/katago-v1.6.1-gpu-opencl-windows-x64.zip",
-            "Eigen AVX2 (Modern CPUs) v1.6.1": "https://github.com/lightvector/KataGo/releases/download/v1.6.1/katago-v1.6.1-cpu-eigen-avx2-windows-x64.zip",
-            "Eigen (CPU, Non-optimized) v1.6.1": "https://github.com/lightvector/KataGo/releases/download/v1.6.1/katago-v1.6.1-cpu-eigen-windows-x64.zip",
+            "OpenCL v1.7.0": "https://github.com/lightvector/KataGo/releases/download/v1.7.0/katago-v1.7.0-gpu-opencl-windows-x64.zip",
+            "Eigen AVX2 (Modern CPUs) v1.7.0": "https://github.com/lightvector/KataGo/releases/download/v1.7.0/katago-v1.7.0-cpu-eigen-avx2-windows-x64.zip",
+            "Eigen (CPU, Non-optimized) v1.7.0": "https://github.com/lightvector/KataGo/releases/download/v1.7.0/katago-v1.7.0-cpu-eigen-windows-x64.zip",
             "OpenCL v1.6.1 (bigger boards)": "https://github.com/lightvector/KataGo/releases/download/v1.6.1%2Bbs29/katago-v1.6.1+bs29-gpu-opencl-windows-x64.zip",
         },
         "linux": {
-            "OpenCL v1.6.1": "https://github.com/lightvector/KataGo/releases/download/v1.6.1/katago-v1.6.1-gpu-opencl-linux-x64.zip",
-            "Eigen AVX2 (Modern CPUs) v1.6.1": "https://github.com/lightvector/KataGo/releases/download/v1.6.1/katago-v1.6.1-cpu-eigen-avx2-linux-x64.zip",
-            "Eigen (CPU, Non-optimized) v1.6.1": "https://github.com/lightvector/KataGo/releases/download/v1.6.1/katago-v1.6.1-cpu-eigen-linux-x64.zip",
+            "OpenCL v1.7.0": "https://github.com/lightvector/KataGo/releases/download/v1.7.0/katago-v1.7.0-gpu-opencl-linux-x64.zip",
+            "Eigen AVX2 (Modern CPUs) v1.7.0": "https://github.com/lightvector/KataGo/releases/download/v1.7.0/katago-v1.7.0-cpu-eigen-avx2-linux-x64.zip",
+            "Eigen (CPU, Non-optimized) v1.7.0": "https://github.com/lightvector/KataGo/releases/download/v1.7.0/katago-v1.7.0-cpu-eigen-linux-x64.zip",
             "OpenCL v1.6.1 (bigger boards)": "https://github.com/lightvector/KataGo/releases/download/v1.6.1%2Bbs29/katago-v1.6.1+bs29-gpu-opencl-linux-x64.zip",
         },
         "just-descriptions": {
-            "CUDA v1.6.1 (Windows)": "https://github.com/lightvector/KataGo/releases/download/v1.6.1/katago-v1.6.1-gpu-cuda10.2-windows-x64.zip",
-            "CUDA v1.6.1 (Linux)": "https://github.com/lightvector/KataGo/releases/download/v1.6.1/katago-v1.6.1-gpu-cuda10.2-linux-x64.zip",
+            "CUDA v1.7.0 (Windows)": "https://github.com/lightvector/KataGo/releases/download/v1.7.0/katago-v1.7.0-gpu-cuda10.2-windows-x64.zip",
+            "CUDA v1.7.0 (Linux)": "https://github.com/lightvector/KataGo/releases/download/v1.7.0/katago-v1.7.0-gpu-cuda10.2-linux-x64.zip",
         },
     }
 

katrain/gui/widgets/movetree.py

@@ -3,7 +3,7 @@ from collections import defaultdict
 from kivy.graphics.context_instructions import Color
 from kivy.graphics.vertex_instructions import Line, Rectangle
 from kivy.lang import Builder
-from kivy.properties import ObjectProperty, Clock, NumericProperty, BooleanProperty
+from kivy.properties import BooleanProperty, Clock, NumericProperty, ObjectProperty
 from kivy.uix.dropdown import DropDown
 from kivy.uix.scrollview import ScrollView
 from kivy.uix.widget import Widget
@@ -11,15 +11,15 @@ from kivymd.app import MDApp
 
 from katrain.gui.kivyutils import BackgroundMixin, draw_circle, draw_text
 from katrain.gui.style import (
-    WHITE,
-    STONE_COLORS,
-    LIGHTGREY,
     BACKGROUND_COLOR,
-    OUTLINE_COLORS,
     LIGHTER_BACKGROUND_COLOR,
-    YELLOW,
-    STONE_TEXT_COLORS,
+    LIGHTGREY,
+    OUTLINE_COLORS,
     RED,
+    STONE_COLORS,
+    STONE_TEXT_COLORS,
+    WHITE,
+    YELLOW,
 )
 
 
@@ -85,6 +85,17 @@ class MoveTreeCanvas(Widget):
             self.set_game_node(parent)
         self.is_open = False
 
+    def make_selected_node_main_branch(self):
+
+        if self.menu_selected_node and self.menu_selected_node.parent:
+            node = self.menu_selected_node
+            while node.parent is not None:
+                node.parent.children.remove(node)
+                node.parent.children.insert(0, node)
+                node = node.parent
+            self.set_game_node(self.menu_selected_node)
+        self.is_open = False
+
     def switch_branch(self, direction=1):
         pos = self.move_pos.get(self.scroll_view_widget.current_node)
         if not self.scroll_view_widget:
@@ -96,7 +107,7 @@ class MoveTreeCanvas(Widget):
         self.set_game_node(same_x_moves[new_index][1])
 
     def draw_move_tree(self, current_node):
-        if not self.scroll_view_widget:
+        if not self.scroll_view_widget or not current_node:
             return
 
         spacing = 5
@@ -109,6 +120,7 @@ class MoveTreeCanvas(Widget):
         stack = root.ordered_children[::-1]
         next_y_pos = defaultdict(int)  # x pos -> max y pos
         children = defaultdict(list)  # since AI self-play etc may modify the tree between layout and draw!
+        children[root] = root.ordered_children
         while stack:
             move = stack.pop()
             x = move.depth
@@ -176,6 +188,9 @@ class MoveTree(ScrollView, BackgroundMixin):
     def delete_selected_node(self):
         self.move_tree_canvas.delete_selected_node()
 
+    def make_selected_node_main_branch(self):
+        self.move_tree_canvas.make_selected_node_main_branch()
+
     def scroll_to_pixel(self, x, y):
         if not self._viewport:
             return
@@ -219,5 +234,11 @@ Builder.load_string(
         shortcut: ''
         on_action: root.katrain.controls.move_tree.delete_selected_node()
         -background_color: LIGHTER_BACKGROUND_COLOR
+    MoveTreeDropdownItem:
+        text: i18n._("Make Main Branch")
+        icon: 'img/Branch.png'
+        shortcut: ''
+        on_action: root.katrain.controls.move_tree.make_selected_node_main_branch()
+        -background_color: LIGHTER_BACKGROUND_COLOR
     """
 )

katrain/gui/widgets/selection_slider.py

@@ -127,7 +127,7 @@ KV = """
         Label:
             id: label
             text: slider.values[slider.index][1]
-            font_size: dp(12)
+            font_size: sp(12)
             lang_change_tracking: i18n._('') # for font
             halign: "center"
             color: root.thumb_color if root.active else [0, 0, 0, 0]

katrain/i18n/locales/cn/LC_MESSAGES/katrain.po

@@ -521,7 +521,6 @@ msgstr "读秒声音开启"
 msgid "tab:info-details"
 msgstr "详细"
 
-#. TODO: updated
 msgid "teaching undo message"
 msgstr "撤销第 {move} 手棋因为损失超过 {points_lost:.1f} 目. 悬停在该手棋 上方查看信息. 请再次尝试."
 
@@ -640,3 +639,54 @@ msgstr "寻找其他走法"
 #. status message
 msgid "alternative analysis"
 msgstr "正在搜索其他走法"
+
+msgid "minimal time use"
+msgstr "最少思考时间 (秒)"
+
+msgid "move too fast"
+msgstr "请思考 {num} 秒再下棋"
+
+msgid "main time"
+msgstr "保留时间 (分钟)"
+
+msgid "aihelp:simple"
+msgstr ""
+"简化局面的风格。 "
+"所有损失最高'max_points_lost'目和至少'min_visits'计算深度的着法当中，偏向于损失较小，乙方空定型（偏向值settled_weight），对方空定型（偏向值settled_weight"
+" * opponent "
+"fac），避免脱先（5路以上，偏向值tenuki_penalty），避免碰（偏向值attachment_penalty）的手法。偏向值可以设为复数捣乱局面，宽根噪声（通用设置）可以棋力代价更加符合风格。"
+" "
+
+msgid "ai:simple"
+msgstr "简明风格"
+
+msgid "analysis:local"
+msgstr "寻找局部走法"
+
+msgid "local analysis"
+msgstr "正在寻找局部走法"
+
+msgid "change current game"
+msgstr ""
+"更新贴目\n"
+"规则在当前对局"
+
+msgid "Make Main Branch"
+msgstr "设为实战变化"
+
+#. TODO - DEPRECATED
+msgid "aihelp:settle"
+msgstr ""
+"Plays moves that lead to clarity on stones' status, favouring solid moves "
+"rather than complex sabaki. Of moves that lose at most 'max_points_lost' "
+"points and with at least 'min_visits', prefer moves that lose fewer points, "
+"settle own stones more (with importance settled_weight), settle opponent's "
+"stones more (importance settled_weight * opponent_fac), avoid tenuki (more "
+"than 5 spaces, importance tenuki_penalty), and avoid attachments (importance"
+" attachment_penalty). Weights can be negative to make it complicate the game"
+" instead, and wide root noise (in general settings) can be used to further "
+"play according to the style at the cost of strength. "
+
+#. TODO - DEPRECATED
+msgid "ai:settle"
+msgstr "Settle Style"

katrain/i18n/locales/de/LC_MESSAGES/katrain.po

@@ -671,12 +671,12 @@ msgstr ""
 msgid "show point loss as text"
 msgstr "Zeig den Punkteverlust auf Beste Züge"
 
-#. TODO - updated
 msgid "teaching undo message"
 msgstr ""
-"Zug {move} zurückgenommen, da er {points_lost:.1f} Punkte verliert. Bewege "
-"Mauszeiger über den Zug,um die erwartete Widerlegung zu sehen. Bitte "
-"versuche es erneut."
+"Zug {move} zurückgenommen, da er {points_lost:.1f} Punkte verliert. Bitte "
+"versuche es erneut. Bewege den Mauszeiger über den Zug, um die erwartete "
+"Widerlegung zu sehen, oder setze ein Häkchen bei “Erwartetes Gebiet” (t), um"
+" die Bereiche des Bretts zu sehen, wo Punktverlust zu erwarten ist."
 
 #. resign the game
 msgid "resign"
@@ -689,3 +689,49 @@ msgstr "Finde alternative Züge"
 #.  - status message
 msgid "alternative analysis"
 msgstr "Suche nach alternativen Zügen"
+
+#. TODO - ignore moves if timer is on and less than this is used
+msgid "minimal time use"
+msgstr "Minimal time use in byo-yomi (s)"
+
+msgid "move too fast"
+msgstr "Denke mindestens {num} Sekunden nach, bevor Du setzt."
+
+msgid "main time"
+msgstr "Hauptbedenkzeit  (Min.)"
+
+msgid "aihelp:simple"
+msgstr ""
+"Spielt Züge, die dazu zu einer Vereinfachung des Brettzustandes führen. Von "
+"Zügen,die höchstens 'max_points_lost' Punkte verlieren und mit wenigstens "
+"'min_visits' Visits, bevorzuge die Züge, die weniger Punkte verlieren und"
+"eigenes Gebiet mehr sichern."
+
+msgid "ai:simple"
+msgstr "Einfacher Stil"
+
+#. TODO - DEPRECATED
+msgid "aihelp:settle"
+msgstr ""
+"Spielt Züge, die dazu zu einer Vereinfachung des Brettzustandes führen. Von "
+"Zügen,die höchstens 'max_points_lost' Punkte verlieren und mit wenigstens "
+"'min_visits' Visits, bevorzuge die Züge, die weniger Punkte verlieren, "
+"eigenes Gebiet mehr sichern"
+
+#. TODO - DEPRECATED
+msgid "ai:settle"
+msgstr "Settle Style"
+
+msgid "analysis:local"
+msgstr "Finde lokale Antworten"
+
+msgid "local analysis"
+msgstr "Suche nach lokalen Antworten"
+
+msgid "change current game"
+msgstr ""
+"Aktualisiere Komi/Regeln\n"
+" im aktuellen Spiel"
+
+msgid "Make Main Branch"
+msgstr "Zum Hauptzweig machen"

katrain/i18n/locales/en/LC_MESSAGES/katrain.po

@@ -106,6 +106,10 @@ msgstr "Deeper full game analysis"
 msgid "analysis:alternatives"
 msgstr "Find alternative moves"
 
+#. Analyze again but allow only nearby moves
+msgid "analysis:local"
+msgstr "Find local responses"
+
 #. Keep improving current position analysis when there is nothing else to do
 msgid "analysis:continuous"
 msgstr "Toggle continuous analysis"
@@ -125,6 +129,10 @@ msgstr "Analyzing entire board using {visits} visits/move"
 msgid "alternative analysis"
 msgstr "Searching for alternative moves"
 
+#. status message
+msgid "local analysis"
+msgstr "Searching for local responses"
+
 #. status message
 msgid "equalizing analysis"
 msgstr "Equalizing analysis of candidate moves to {visits} visits"
@@ -293,14 +301,25 @@ msgid "timer settings"
 msgstr "Adjust Timer Settings"
 
 msgid "byoyomi length"
-msgstr "Time per Byo-yomi Period"
+msgstr "Time per byo-yomi period (s)"
 
 msgid "byoyomi periods"
-msgstr "Number of Periods"
+msgstr "Number of periods"
 
 msgid "count down sound"
 msgstr "Sound on"
 
+#. in minutes
+msgid "main time"
+msgstr "Main time (min.)"
+
+#. ignore moves if timer is on and less than this is used
+msgid "minimal time use"
+msgstr "Minimal time use in byo-yomi (s)"
+
+msgid "move too fast"
+msgstr "Think for at least {num} seconds before playing."
+
 #. submit button
 msgid "update timer"
 msgstr "Update Timer"
@@ -351,6 +370,12 @@ msgstr "Use x:y (e.g. 19:9) for non-square size"
 msgid "new game"
 msgstr "Start New Game"
 
+#. change komi etc in current game
+msgid "change current game"
+msgstr ""
+"Update komi/rules\n"
+"in active game"
+
 # teacher settings from here
 msgid "teacher settings"
 msgstr "Edit Teacher Settings"
@@ -560,6 +585,38 @@ msgstr ""
 "perspective. When `automatic` is set, KaTrain will find a suitable value "
 "automatically."
 
+msgid "ai:simple"
+msgstr "Simple Style"
+
+msgid "aihelp:simple"
+msgstr ""
+"Plays moves that lead to simplifying the board state. Of moves that lose at "
+"most 'max_points_lost' points and with at least 'min_visits', prefer moves "
+"that lose fewer points, settle own territory more (with importance "
+"settled_weight), settle opponent's territory more (importance settled_weight"
+" * opponent_fac), avoid tenuki (more than 5 spaces, importance "
+"tenuki_penalty), and avoid attachments (importance attachment_penalty). "
+"Weights can be negative to make it complicate the game instead, and wide "
+"root noise (in general settings) can be used to increase play according to "
+"the style at the cost of strength. "
+
+#. DEPRECATED
+msgid "ai:settle"
+msgstr "Settle Style"
+
+#. DEPRECATED
+msgid "aihelp:settle"
+msgstr ""
+"Plays moves that lead to clarity on stones' status, favouring solid moves "
+"rather than complex sabaki. Of moves that lose at most 'max_points_lost' "
+"points and with at least 'min_visits', prefer moves that lose fewer points, "
+"settle own stones more (with importance settled_weight), settle opponent's "
+"stones more (importance settled_weight * opponent_fac), avoid tenuki (more "
+"than 5 spaces, importance tenuki_penalty), and avoid attachments (importance"
+" attachment_penalty). Weights can be negative to make it complicate the game"
+" instead, and wide root noise (in general settings) can be used to further "
+"play according to the style at the cost of strength. "
+
 msgid "ai:jigo"
 msgstr "KataJigo"
 
@@ -691,6 +748,10 @@ msgstr "Number of visits per move"
 msgid "Delete Node"
 msgstr "Delete Node"
 
+#. on move tree editing
+msgid "Make Main Branch"
+msgstr "Make Main Branch"
+
 #. theme label
 msgid "theme"
 msgstr "Colour Theme"

katrain/i18n/locales/es/LC_MESSAGES/katrain.po

@@ -697,3 +697,58 @@ msgstr "Find alternative moves"
 #. TODO - status message
 msgid "alternative analysis"
 msgstr "Searching for alternative moves"
+
+#. TODO - ignore moves if timer is on and less than this is used
+msgid "minimal time use"
+msgstr "Minimal time use in byo-yomi (s)"
+
+#. TODO
+msgid "move too fast"
+msgstr "Think for at least {num} seconds before playing."
+
+#. TODO - in minutes
+msgid "main time"
+msgstr "Main time (min.)"
+
+#. TODO
+msgid "aihelp:simple"
+msgstr "Plays moves that lead to simplifying the board state."
+
+#. TODO
+msgid "ai:simple"
+msgstr "Simple Style"
+
+#. TODO - DEPRECATED
+msgid "aihelp:settle"
+msgstr ""
+"Plays moves that lead to clarity on stones' status, favouring solid moves "
+"rather than complex sabaki. Of moves that lose at most 'max_points_lost' "
+"points and with at least 'min_visits', prefer moves that lose fewer points, "
+"settle own stones more (with importance settled_weight), settle opponent's "
+"stones more (importance settled_weight * opponent_fac), avoid tenuki (more "
+"than 5 spaces, importance tenuki_penalty), and avoid attachments (importance"
+" attachment_penalty). Weights can be negative to make it complicate the game"
+" instead, and wide root noise (in general settings) can be used to further "
+"play according to the style at the cost of strength. "
+
+#. TODO - DEPRECATED
+msgid "ai:settle"
+msgstr "Settle Style"
+
+#. TODO - Analyze again but allow only nearby moves
+msgid "analysis:local"
+msgstr "Find local responses"
+
+#. TODO - status message
+msgid "local analysis"
+msgstr "Searching for local responses"
+
+#. TODO - change komi etc in current game
+msgid "change current game"
+msgstr ""
+"Update komi/rules\n"
+"in active game"
+
+#. TODO - on move tree editing
+msgid "Make Main Branch"
+msgstr "Make Main Branch"

katrain/i18n/locales/fr/LC_MESSAGES/katrain.po

@@ -93,7 +93,7 @@ msgid "analysis:extra"
 msgstr "Analyser davantage"
 
 msgid "analysis:equalize"
-msgstr "Égaliser l’analyse des coups"
+msgstr "Égaliser l'analyse des coups"
 
 msgid "analysis:sweep"
 msgstr "Analyser tous les possibles"
@@ -105,7 +105,7 @@ msgid "analysis:continuous"
 msgstr "Lancer/Arrêter l'analyse continue"
 
 msgid "analysis:aimove"
-msgstr "Jouer le coup de l’IA"
+msgstr "Jouer le coup de l'IA"
 
 #. label in little popup for analyzing entire game
 msgid "reanalyze max visits"
@@ -526,8 +526,9 @@ msgstr ""
 msgid "teaching undo message"
 msgstr ""
 "Le coup {move} a été annulé car il perdait {points_lost:.1f} points.Vous "
-"pouvez survoler le coup pour voir la réfutation. Veuillez jouer un nouveau "
-"coup."
+"pouvez consulter les Prévisions territoriales (t) pour identifier les "
+"régions concernées, et survoler le coup pour connaître sa réfutation.Pour "
+"continuer la partie, veuillez jouer un nouveau coup."
 
 msgid "sweep analysis"
 msgstr "Analyse de tous les possibles ({visits} visites/coup)."
@@ -721,3 +722,57 @@ msgstr "Proposer des coups différents"
 
 msgid "alternative analysis"
 msgstr "Recherche de coups différents."
+
+msgid "minimal time use"
+msgstr ""
+"Temps de réflexion \n"
+"minimum (s/période)"
+
+msgid "move too fast"
+msgstr "Réfléchissez au moins {num} secondes avant de jouer."
+
+msgid "main time"
+msgstr "Temps principal (min)"
+
+msgid "aihelp:simple"
+msgstr ""
+"Par défaut, le moteur de jeu cherche à simplifier la position. \n"
+"Il joue un coup parmi ceux considérés au moins 'min_visits' fois et perdant au plus 'max_points_lost'. \n"
+"'settled_weight' le pousse à consolider ses territoires, et ceux adverses une fois multiplié par 'opponent_fac'. \n"
+"'tenuki_penalty' le pousse à jouer localement ; 'attach_penalty' à éviter les coups de contact direct. \n"
+"Quand elles sont possibles, les valeurs négatives favorisent au contraire un style de jeu compliqué. \n"
+"Élargir le Champ des possibles (cf. Paramètres généraux) peaufine le style au détriment du niveau de jeu."
+
+msgid "ai:simple"
+msgstr "Modéré"
+
+#. TODO - DEPRECATED
+msgid "aihelp:settle"
+msgstr ""
+"Plays moves that lead to clarity on stones' status, favouring solid moves "
+"rather than complex sabaki. Of moves that lose at most 'max_points_lost' "
+"points and with at least 'min_visits', prefer moves that lose fewer points, "
+"settle own stones more (with importance settled_weight), settle opponent's "
+"stones more (importance settled_weight * opponent_fac), avoid tenuki (more "
+"than 5 spaces, importance tenuki_penalty), and avoid attachments (importance"
+" attachment_penalty). Weights can be negative to make it complicate the game"
+" instead, and wide root noise (in general settings) can be used to further "
+"play according to the style at the cost of strength. "
+
+#. TODO - DEPRECATED
+msgid "ai:settle"
+msgstr "Settle Stones Style"
+
+msgid "analysis:local"
+msgstr "Proposer des coups locaux"
+
+msgid "local analysis"
+msgstr "Recherche de coups locaux."
+
+msgid "change current game"
+msgstr ""
+"Mettre à jour\n"
+"la partie en cours"
+
+msgid "Make Main Branch"
+msgstr "Mettre en branche principale"

katrain/i18n/locales/jp/LC_MESSAGES/katrain.po

@@ -679,3 +679,55 @@ msgstr ""
 #. status message
 msgid "alternative analysis"
 msgstr "他の候補手を分析中"
+
+#. ignore moves if timer is on and less than this is used
+msgid "minimal time use"
+msgstr "早打ち禁止秒数"
+
+msgid "move too fast"
+msgstr "早打ちすぎです。{num}秒は考えてください。"
+
+#. in minutes
+msgid "main time"
+msgstr "持ち時間（分）"
+
+msgid "aihelp:simple"
+msgstr ""
+"盤面の単純化をめざして, 次の基準で手を選びます (括弧内は重視度): 損する目数が少ない, 自分の確定地が多い (settled_weight), "
+"相手も確定地が多い (settled_weight × opponent fac), 相手が打った手の近辺…縦横とも3間以内 "
+"(tenuki_penalty), 相手の石に接触しない (attachment_penalty). ただし, 損が大きすぎたり "
+"(max_points_lost目以上) 探索手数が少なすぎたり (min_visits手未満) する候補手は除きます. 重視度を負にすれば, "
+"逆に盤面の複雑化をめざします. また, 一般設定のwide root noiseを大きくすれば, 強さは犠牲になりますが, "
+"より前述の基準に忠実な打ち方をします."
+
+msgid "ai:simple"
+msgstr "簡明派"
+
+msgid "aihelp:settle"
+msgstr ""
+"複雑なサバキよりも, 石の生死をはっきりさせる堅実な手を好みます. 具体的には, 次の基準で手を選びます (括弧内は重視度): 損する目数が少ない, "
+"自分の生死が明確 (settled_weight), 相手の生死も明確 (settled_weight × opponent fac), "
+"相手が打った手の近辺…縦横とも3間以内 (tenuki_penalty), 相手の石に接触しない (attachment_penalty). ただし, "
+"損が大きすぎたり (max_points_lost目以上) 探索手数が少なすぎたり (min_visits手未満) する候補手は除きます. "
+"重視度を負にすれば, 逆に盤面の複雑化をめざします. また, 一般設定のwide root noiseを大きくすれば, 強さは犠牲になりますが, "
+"より前述の基準に忠実な打ち方をします."
+
+msgid "ai:settle"
+msgstr "堅実派"
+
+#. Analyze again but allow only nearby moves
+msgid "analysis:local"
+msgstr "近辺の応手に絞って分析"
+
+#. status message
+msgid "local analysis"
+msgstr "近辺の応手を分析中"
+
+msgid "change current game"
+msgstr ""
+"現対局のコミ・\n"
+"ルールを変更"
+
+#. on move tree editing
+msgid "Make Main Branch"
+msgstr "こちらを主分岐に"

katrain/i18n/locales/ko/LC_MESSAGES/katrain.po

@@ -546,11 +546,6 @@ msgstr ""
 "`threshold`선 밑의 수를 더 많이 뽑은 후 그 중 최선의 수를 둡니다. `line_weight`을 증가시키면 바둑판 중앙에 있는"
 " 수를 더욱 꺼리게 됩니다. 판이 `endgame` 비율만큼 차면 평범하게 둡니다."
 
-#. TODO: updated
-msgid "teaching undo message"
-msgstr ""
-"{points_lost:.1f}집 손해이므로 {move}을/를 물렀습니다. 커서를 올려 상대의 대응책을 보십시오. 다시 두어 주세요."
-
 msgid "aihelp:p:rank"
 msgstr ""
 "제한적인 착수 가능 목록에서 무작위로 다음 착수를 선택한 후 그 중 최선의 수를 둡니다. 더 강한 설정은 최선의 수를 더 큰 목록에서 "
@@ -654,3 +649,58 @@ msgstr "다른 수 찾기"
 
 msgid "alternative analysis"
 msgstr "다른 수를 찾는 중입니다"
+
+msgid "teaching undo message"
+msgstr ""
+"{points_lost:.1f}집 손해이므로 {move}을/를 물렀습니다. 다시 두어 주세요. 힌트를 보려면, 커서를 올려 상대의 "
+"대응책을 보거나, 어디서 집 손해를 보는지 집 모양을 통해 (t) 알아보십시오. "
+
+msgid "minimal time use"
+msgstr "이보다 빨리 착수한 수 무시"
+
+msgid "move too fast"
+msgstr "최소한 {num}초만큼 생각하고 착수해 주십시오"
+
+msgid "main time"
+msgstr "제한 시간(분)"
+
+msgid "aihelp:simple"
+msgstr ""
+"국면을 단순화하는 수를 둡니다. 최대 'max_point_lost'집 손해를 보고 최소 'min_visits'번 탐색한 수 중에, 집 "
+"손해가 적고, 자신의 집을 굳히며(settled_weight에 따라 조절), 상대의 집도 굳히고(settled_weight * "
+"opponent fac에 따라 조절), 붙임수를 피하는(attachment_penalty에 따라 조절) 수를 둡니다. 가중치가 음수일 "
+"때는 반대로 국면을 복잡하게 만듭니다."
+
+msgid "ai:simple"
+msgstr "간명한 기풍"
+
+msgid "analysis:local"
+msgstr "국지적인 응수 찾기"
+
+msgid "local analysis"
+msgstr "국지적인 응수를 찾고 있습니다"
+
+msgid "change current game"
+msgstr ""
+"이 판에 대해\n"
+"덤과 규칙 등 조절"
+
+#. TODO - DEPRECATED
+msgid "ai:settle"
+msgstr "Settle Style"
+
+#. TODO - DEPRECATED
+msgid "aihelp:settle"
+msgstr ""
+"Plays moves that lead to clarity on stones' status, favouring solid moves "
+"rather than complex sabaki. Of moves that lose at most 'max_points_lost' "
+"points and with at least 'min_visits', prefer moves that lose fewer points, "
+"settle own stones more (with importance settled_weight), settle opponent's "
+"stones more (importance settled_weight * opponent_fac), avoid tenuki (more "
+"than 5 spaces, importance tenuki_penalty), and avoid attachments (importance"
+" attachment_penalty). Weights can be negative to make it complicate the game"
+" instead, and wide root noise (in general settings) can be used to further "
+"play according to the style at the cost of strength. "
+
+msgid "Make Main Branch"
+msgstr "주 변화로 만들기"

katrain/i18n/locales/ru/LC_MESSAGES/katrain.po

@@ -407,7 +407,7 @@ msgid "engine:max_time"
 msgstr "Максимальное время анализа"
 
 msgid "engine:wide_root_noise"
-msgstr "Высокий уровень шума (увеличивает разнообразие рассматриваемых ходов)"
+msgstr "Уровень шума (увеличивает разнообразие рассматриваемых ходов)"
 
 msgid "engine:wide_root_noise:hint"
 msgstr ""
@@ -684,3 +684,57 @@ msgstr "Поискать альтернативные ходы"
 
 msgid "alternative analysis"
 msgstr "Поиск альтернативных ходов"
+
+msgid "minimal time use"
+msgstr "Минимально потраченное время в бёёми (сек.)"
+
+msgid "move too fast"
+msgstr "Подумайте как минимум {num} секунд."
+
+msgid "main time"
+msgstr "Основное время (мин.)"
+
+msgid "aihelp:simple"
+msgstr ""
+"Делает ходы, упрощающие позицию на доске. Из ходов, теряющих не более чем "
+"'max_points_lost' очков и имеющих как минимум 'min_visits', предпочитает "
+"ходы, укрепляющие собственную территорию (коэффициент settled_weight), "
+"укрепляющие территорию оппонента (коэффициент settled_weight * "
+"opponent_fac), избегает тэнуки (более 5 пунктов, коэффициент "
+"tenuki_penalty), и избегает прилипаний (коэффициент attachment_penalty). "
+"Коэффициенты могут быть отрицательными, чтобы напротив сделать игру сложнее."
+" 'Уровень шума' (в 'Настройках системы') может быть использован для "
+"увеличения вариативности игры за счёт небольшой потери силы. "
+
+msgid "ai:simple"
+msgstr "Простой стиль"
+
+msgid "aihelp:settle"
+msgstr ""
+"Делает ходы, ведущие к ясности в статусе камней, предпочитая надёжные ходы, "
+"нежели сложное сабаки. Из ходов, теряющих не более чем 'max_points_lost' "
+"очков и имеющих как минимум 'min_visits', предпочитает ходы, укрепляющие "
+"собственную территорию (коэффициент settled_weight), укрепляющие территорию "
+"оппонента (коэффициент settled_weight * opponent_fac), избегает тэнуки "
+"(более 5 пунктов, коэффициент tenuki_penalty), и избегает прилипаний "
+"(коэффициент attachment_penalty). Коэффициенты могут быть отрицательными, "
+"чтобы напротив сделать игру сложнее. 'Уровень шума' (в 'Настройках системы')"
+" может быть использован для увеличения вариативности игры за счёт небольшой "
+"потери силы. "
+
+msgid "ai:settle"
+msgstr "Надёжный стиль"
+
+msgid "analysis:local"
+msgstr "Найти локальные ответы"
+
+msgid "local analysis"
+msgstr "Поиск локальных ответов"
+
+msgid "change current game"
+msgstr ""
+"Изменить коми/правила\n"
+"в текущей игре"
+
+msgid "Make Main Branch"
+msgstr "Сделать главной веткой"

katrain/popups.kv

@@ -8,8 +8,8 @@
 #:set CP_SMALL_SPACING 3
 #:set CP_PADDING 6
 
-#:set INPUT_FONT_SIZE dp(25)
-#:set DESC_FONT_SIZE  dp(20)
+#:set INPUT_FONT_SIZE sp(25)
+#:set DESC_FONT_SIZE  sp(20)
 
 #:set LIGHTGREY [0.7,0.7,0.7,1]
 #:import BACKGROUND_COLOR katrain.gui.style.BACKGROUND_COLOR
@@ -32,7 +32,7 @@
     line_color_focus: TEXT_COLOR
 
 <LabelledPathInput>:
-    font_size: dp(16)
+    font_size: sp(16)
     multiline: False
     helper_text_mode: "on_error"
     helper_text: i18n._("file not found")
@@ -179,13 +179,18 @@
         orientation: 'horizontal'
         GridLayout:
             cols: 2
-            rows: 4
+            rows: 5
             spacing: CP_SPACING
             DescriptionLabel:
                 text: i18n._("general:sgf_save")
             AnchorLayout:
                 LabelledPathInput:
                     input_property: "general/sgf_save"
+            DescriptionLabel:
+                text: i18n._('count down sound')
+            AnchorLayout:
+                LabelledCheckBox:
+                    input_property: "timer/sound"
             DescriptionLabel:
                 text: i18n._("general:anim_pv_time")
             AnchorLayout:
@@ -250,7 +255,7 @@
             AutoSizedRoundedRectangleButton:
                 padding_x: 15
                 size_hint: None,None
-                height: dp(40)
+                height: sp(40)
                 text: i18n._("update settings")
                 on_press: root.update_config(True)
 
@@ -319,9 +324,15 @@
 <ConfigTimerPopup>:
     MDGridLayout:
         cols: 2
-        rows: 3
+        rows: 4
         spacing: 1.5 * CP_SPACING
         size_hint: 1,2.5
+        DescriptionLabel:
+            text: i18n._('main time')
+        AnchorLayout:
+            LabelledIntInput:
+                text: "10"
+                input_property: "timer/main_time"
         DescriptionLabel:
             text: i18n._('byoyomi length')
         AnchorLayout:
@@ -335,10 +346,11 @@
                 text: "5"
                 input_property: "timer/byo_periods"
         DescriptionLabel:
-            text: i18n._('count down sound')
+            text: i18n._('minimal time use')
         AnchorLayout:
-            LabelledCheckBox:
-                input_property: "timer/sound"
+            LabelledIntInput:
+                text: "0"
+                input_property: "timer/minimal_use"
     AnchorLayout:
         size_hint: 1,1
         AutoSizedRoundedRectangleButton:
@@ -443,11 +455,20 @@
                 input_property: 'game/clear_cache'
         SmallDescriptionLabel:
             text: i18n._('avoids replaying')
+    BoxLayout:
+        size_hint: 1,0.75
         AnchorLayout:
-        size_hint: 1,1
+            size_hint: 0.5, 1
             AutoSizedRoundedRectangleButton:
                 padding_x: 15
                 size_hint: None,0.5
+                _font_size: self.height*0.35
+                text: i18n._("change current game")
+                on_press: root.update_game()
+        AnchorLayout:
+            AutoSizedRoundedRectangleButton:
+                padding_x: 15
+                size_hint: None,0.66
                 text: i18n._("new game")
                 on_press: root.update_config(True)
 
@@ -457,7 +478,7 @@
     textbox: textbox
     SelectionSlider:
         id: slider
-        size_hint: 2,1
+        size_hint: 2.25,1
         values: root.values
         on_change: textbox.text = str(slider.value)
         track_color: LIGHTGREY

setup.py

@@ -41,16 +41,19 @@ setup(
     install_requires=[
         "wheel",
         "setuptools",
+        "urllib3",
         "importlib_resources ;python_version<'3.7'",
-        "pygame;platform_system=='Darwin'",  # some mac versions need this for kivy
         "cython>=0.24,<=0.29.14,!=0.27,!=0.27.2",  # kivy wants this
+        "kivy==2.0.0rc2;platform_system=='Darwin'",  # rc3 failing on mac
+        "kivy>=2.0.0rc2;platform_system!='Darwin'",  # just use the latest on linux?
+        "pygame;platform_system=='Darwin'",  # some mac versions need this for kivy
         "kivy_deps.glew;platform_system=='Windows'",
         "kivy_deps.sdl2;platform_system=='Windows'",
         "kivy_deps.gstreamer;platform_system=='Windows'",
-        "kivy==2.0.0rc2",  # rc3 failing on mac
         "kivymd>=0.104.1",
         "screeninfo;platform_system!='Darwin'",  # for screen resolution, has problems on macos
     ],
+    dependency_links=["https://kivy.org/downloads/simple/"],
     python_requires=">=3.6, <4",
     entry_points={"console_scripts": ["katrain=katrain.__main__:run_app"]},
     classifiers=[