[Wesnoth-bugs] [task #8013] Data Export #69 (bugs)

k: Any Effort: 0.00 ___ Details: A new export job has been registered. This task has been created to keep the project informed. However, only Celtic Minstrel, that created the job, can remove the job itself. = Job URL =

[Wesnoth-bugs] [task #8014] Data Export #70 (bugs)

k: Any Effort: 0.00 ___ Details: A new export job has been registered. This task has been created to keep the project informed. However, only Celtic Minstrel, that created the job, can remove the job itself. = Job URL =

[Wesnoth-bugs] [bug #25670] App Crash when canceling or Esc a Lua dialog

Update of bug #25670 (project wesnoth): Status: Confirmed => Fixed ___ Follow-up Comment #1: Fixed in master. ___ Reply to this

[Wesnoth-bugs] [bug #25619] using invalid unit types in debug mode

Update of bug #25619 (project wesnoth): Category: Feature Request => Bug ___ Reply to this item at: ___ Message

[Wesnoth-bugs] [bug #25639] [remove_object] heals units with amla

Follow-up Comment #1, bug #25639 (project wesnoth): I doubt this is a bug with [remove_object]. If the situation is as you suggest, it should also affect objects with an expiry, such as the "stun" weapon special. I don't know of any good way to fix this.

[Wesnoth-bugs] [bug #25625] Translation mark _ in text key breaks event with [animate_unit]

Update of bug #25625 (project wesnoth): Status:None => Ready For Test Assigned to:None => celticminstrel ___ Reply to this item at:

[Wesnoth-bugs] [bug #18679] [ai] blocks are ignored

Update of bug #18679 (project wesnoth): Status: Confirmed => Invalid ___ Follow-up Comment #4: The problem here is that only one of each aspect can be specified in the side's AI - in other

[Wesnoth-bugs] [bug #25390] [for] tag ignores start/end attributes

Update of bug #25390 (project wesnoth): Status:None => Invalid ___ Reply to this item at: ___ Message

[Wesnoth-bugs] [bug #25558] lua error with micro ai

Update of bug #25558 (project wesnoth): Status:None => Fixed ___ Reply to this item at: ___ Message

[Wesnoth-bugs] [bug #25555] Use of [place_shroud] & [remove_shroud] throw errors

Update of bug #2 (project wesnoth): Status: In Progress => Fixed ___ Reply to this item at: ___ Message

[Wesnoth-bugs] [bug #25555] Use of [place_shroud] & [remove_shroud] throw errors

Update of bug #2 (project wesnoth): Status: Fixed => In Progress ___ Reply to this item at: ___ Message

[Wesnoth-bugs] [bug #25555] Use of [place_shroud] & [remove_shroud] throw errors

Follow-up Comment #10, bug #2 (project wesnoth): What exactly goes wrong? Shroud is updated incorrectly? Updated but doesn't take effect until next redraw? Something else? Is it the shroud removal or placement that doesn't work? (Your line numbers seem to be different from master so I can't

[Wesnoth-bugs] [bug #25558] lua error with micro ai

Follow-up Comment #5, bug #25558 (project wesnoth): gfgtdf pointed out something I missed in the previous commit, so try this one: https://github.com/wesnoth/wesnoth/commit/6fe8e65502cc319e61166a3f63b76d5964e947b7 ___ Reply to this item

[Wesnoth-bugs] [bug #25558] lua error with micro ai

Follow-up Comment #4, bug #25558 (project wesnoth): This isn't exactly a fix, but it'll help determine whether I'm on the right track with respect to the cause... Try adding [] to the end of the path strings on lines 99 and 157 of data/ai/micro_ais/micro_ai_helper.lua. (My build isn't working

[Wesnoth-bugs] [bug #25558] lua error with micro ai

Update of bug #25558 (project wesnoth): Assigned to:None => celticminstrel ___ Reply to this item at: ___ Message

[Wesnoth-bugs] [bug #25558] lua error with micro ai

Follow-up Comment #1, bug #25558 (project wesnoth): Is this fixed by the following commit? https://github.com/wesnoth/wesnoth/commit/6de1c9441f25b9c96a524e2c342e9f82483dd1c2 ___ Reply to this item at:

[Wesnoth-bugs] [bug #25561] stderr message about unescaped

Follow-up Comment #4, bug #25561 (project wesnoth): Ah. I'd say it's fine to disallow markup in the author, but I'm sure there would be people who disagree, so I dunno... ___ Reply to this item at:

[Wesnoth-bugs] [bug #25561] stderr message about unescaped

Follow-up Comment #2, bug #25561 (project wesnoth): I disagree on that commit; an exceptional situation should be warning or error, not log. As for the actual issue... it makes sense to allow marked-up text in the add-on or scenario description, so IMO the real problem here is the choice of

[Wesnoth-bugs] [bug #25013] weapon naming: with or without space?

Follow-up Comment #3, bug #25013 (project wesnoth): The "longsword" as a type of weapon in its own right is a construct of games such as Dungeons and Dragons. There is no historical "longsword". Thus, I would say there's no real difference between "longsword" and "long sword" except maybe that

[Wesnoth-bugs] [bug #23841] a [make_tag] tag to create custom wml tags.

Follow-up Comment #8, bug #23841 (project wesnoth): Normally when you overwrite an existing tag, it's to extend it; the original implementation is saved first, and then called from the new implementation. It's not a problem to be able to overwrite an existing tag, but it's not good to be able to

[Wesnoth-bugs] [bug #23752] gui2 listboxes header not working correctly

Follow-up Comment #6, bug #23752 (project wesnoth): When you say no mainline listboxes need horizontal scrollbars, I sure hope that means on 800x600 resolution... ___ Reply to this item at:

[Wesnoth-bugs] [bug #25390] [for] tag ignores start/end attributes

Follow-up Comment #1, bug #25390 (project wesnoth): The array and start/end/step attributes are intended to be mutually exclusive, as documented on the wiki. In effect, the array attribute is a shorthand. If you simply remove array=test from that example, it should work.

[Wesnoth-bugs] [bug #25337] allow custom widget definitions in umc

Follow-up Comment #3, bug #25337 (project wesnoth): You shouldn't be adding definitions before the dialog shows. You should do that early, in a load event or similar. Plus, if you try to add the same definition twice, obviously it won't actually be added twice. So even if you DID add your

[Wesnoth-bugs] [bug #25337] allow custom widget definitions in umc

Follow-up Comment #1, bug #25337 (project wesnoth): I too prefer the second method, but rather than an unregister function, I think it would make more sense to automatically unregister any custom definitions when the Lua kernel is destroyed. (Which means that any definitions registered in the

[Wesnoth-bugs] [bug #25224] add translation friendly string.format function

Follow-up Comment #5, bug #25224 (project wesnoth): If it's in the lua_kernel_base you can simply press the tilde key (more specificall, ` ) at the title screen. If it's in the game_lua_kernel you would need to start a game before pressing that key.

[Wesnoth-bugs] [bug #25224] add translation friendly string.format function

Follow-up Comment #3, bug #25224 (project wesnoth): 1. It makes most sense to me to use the WML variable substitution syntax (as vgettext does), not the % syntax noted by gfgtdf. 2. You might want to think of this as having two parts - first, a "wesnoth.format" or similar function that handles

[Wesnoth-bugs] [bug #25077] memleak in wesnoth help

Follow-up Comment #1, bug #25077 (project wesnoth): This appears to be the same as #25083. ___ Reply to this item at: ___ Message sent via/by Gna!

[Wesnoth-bugs] [bug #25083] leaks in AI vs AI match

Follow-up Comment #1, bug #25083 (project wesnoth): These are all unit attacks being allocated, but in all cases they're stored in a boost::intrusive_ptr, so I don't see how they're being leaked. My guess would be that either there's something wrong with the refcount or the messages are spurious.

[Wesnoth-bugs] [bug #12098] allow [message] to take multiple 'speakers'

Follow-up Comment #9, bug #12098 (project wesnoth): data/gui/dialogs/wml_message.cfg That file contains three separate [window] tags and three separate [window_definition] tags. The issue would be (almost certainly) in the [window] tag that has two portraits.

[Wesnoth-bugs] [bug #12098] allow [message] to take multiple 'speakers'

Follow-up Comment #7, bug #12098 (project wesnoth): I'm not sure exactly of the cause here, but it seems like something in the markup - that is, the GUI2 markup defining the window. I'll try to remember to look into it later (or get Vultraz to do so, as he's better at GUI2 layout), unless you

[Wesnoth-bugs] [bug #25041] Unit types always exist in Lua (segmentation fault)

Update of bug #25041 (project wesnoth): Open/Closed:Open => Closed ___ Follow-up Comment #3: https://github.com/wesnoth/wesnoth/commit/91bd6ab6d09e28048cb12018fea2f6f9c323762b

[Wesnoth-bugs] [bug #25041] Unit types always exist in Lua (segmentation fault)

Follow-up Comment #2, bug #25041 (project wesnoth): Looks like this is simple to fix. I've already done so locally, just need to double-check to make sure it actually works. ___ Reply to this item at:

[Wesnoth-bugs] [bug #24930] image_pos="right" does not display correctly

Follow-up Comment #3, bug #24930 (project wesnoth): I disagree about second_portrait making it irrelevant - it's only irrelevant when second_portrait is actually specified. ___ Reply to this item at:

[Wesnoth-bugs] [bug #12098] allow [message] to take multiple 'speakers'

Follow-up Comment #5, bug #12098 (project wesnoth): I'm not sure. I have no immediate plans to do so, at least. ___ Reply to this item at: ___ Message sent

[Wesnoth-bugs] [bug #22820] GUI2 menus and context menus

Update of bug #22820 (project wesnoth): Release: 1.13.0 => 1.13.6 ___ Follow-up Comment #2: gfgtdf created the menu backend (intended for "comboboxes", which are really not comboboxes),

[Wesnoth-bugs] [bug #11811] Support for gettext plurals in WML

Follow-up Comment #8, bug #11811 (project wesnoth): A general solution to this would be quite hard, I think. The first proposal isn't flexible enough as stated, so that's out (although something similar to it could be implemented more easily). The second proposal requires changes to the WML

[Wesnoth-bugs] [bug #24930] image_pos="right" does not display correctly

Follow-up Comment #1, bug #24930 (project wesnoth): The image_pos=right behaviour should be exactly the same as appending ~RIGHT() to the image path though, so I don't think this is new with the image_pos key. ___ Reply to this item at:

[Wesnoth-bugs] [bug #24931] Allow specifying both a left image and a right image on [message]

Update of bug #24931 (project wesnoth): Status:None => Duplicate ___ Follow-up Comment #2: (Also, this is a duplicate of https://gna.org/bugs/?12098 )

[Wesnoth-bugs] [bug #12098] allow [message] to take multiple 'speakers'

Follow-up Comment #3, bug #12098 (project wesnoth): I have added support for this to the Lua API (the wesnoth.show_message_dialog function), though it has not been wired into the [message] tag at this time. To show a message with two portraits, the portrait and mirror keys in the first argument

[Wesnoth-bugs] [bug #24931] Allow specifying both a left image and a right image on [message]

Follow-up Comment #1, bug #24931 (project wesnoth): I have added support for this to the Lua API (the wesnoth.show_message_dialog function), though it has not been wired into the [message] tag at this time. To show a message with two portraits, the portrait and mirror keys in the first argument

[Wesnoth-bugs] [bug #24933] hotkeys list cannot be sorted by hotkey

Update of bug #24933 (project wesnoth): Status:None => Fixed ___ Reply to this item at: ___ Message

[Wesnoth-bugs] [bug #16107] [modify_side][ai]ai_algorithm= doesn't work

Follow-up Comment #5, bug #16107 (project wesnoth): Also note that [modify_side]switch_ai can already be used to achieve this behaviour, though it needs the path to the AI definition file (under data/ai/ais/ for default AI and data/ai/dev/ for idle AI) rather than the AI ID and is thus not as

[Wesnoth-bugs] [bug #20052] Initiialization for Lua AI new Syntax

Update of bug #20052 (project wesnoth): Status:None => Invalid Open/Closed:Open => Closed ___ Follow-up Comment #1: You can do

[Wesnoth-bugs] [bug #21294] undocumented "backwards compatability" code silently strips out custom ais

Update of bug #21294 (project wesnoth): Status:None => Fixed Open/Closed:Open => Closed ___ Follow-up Comment #4: The backwards

[Wesnoth-bugs] [bug #22777] add a "pixel art friendly" scaling algorithm for zooming in on sprites

Update of bug #22777 (project wesnoth): Status:None => Fixed Open/Closed:Open => Closed ___ Follow-up Comment #2: As of 1.13.5,

[Wesnoth-bugs] [bug #22855] floating point emulation unit test failing, with g++ 4.9

Update of bug #22855 (project wesnoth): Status:None => Wont Fix Open/Closed:Open => Closed ___ Follow-up Comment #1: Since we're not

[Wesnoth-bugs] [bug #24712] Units with select events cannot be selected

Follow-up Comment #4, bug #24712 (project wesnoth): That invites a potential fix, assuming there's some way in Lua to determine whether the speaker is currently selected. Does anyone know of something like that? ___ Reply to this item at:

[Wesnoth-bugs] [bug #24864] Campaign use of random maps causes crash

Follow-up Comment #7, bug #24864 (project wesnoth): Ah, I see, so that's the thing I quickly fixed afterwards? ___ Reply to this item at: ___ Message sent

[Wesnoth-bugs] [bug #24864] Campaign use of random maps causes crash

Update of bug #24864 (project wesnoth): Assigned to:None => celticminstrel ___ Reply to this item at: ___ Message

[Wesnoth-bugs] [bug #24864] Campaign use of random maps causes crash

Update of bug #24864 (project wesnoth): Assigned to: celticminstrel => None ___ Reply to this item at: ___ Message

[Wesnoth-bugs] [bug #24864] Campaign use of random maps causes crash

Follow-up Comment #4, bug #24864 (project wesnoth): gfgtdf: The function in which that fix applied was removed by the PR, so that makes sense. Are you saying the bug resurfaced, or just making a wild observation? ___ Reply to this item at:

[Wesnoth-bugs] [bug #16107] [modify_side][ai]ai_algorithm= doesn't work

Follow-up Comment #4, bug #16107 (project wesnoth): (The preceding comment applies to current master.) ___ Reply to this item at: ___ Message sent via/by Gna!

[Wesnoth-bugs] [bug #16107] [modify_side][ai]ai_algorithm= doesn't work

Follow-up Comment #3, bug #16107 (project wesnoth): There have been significant changes in this area recently, though I believe it has not fixed this bug yet. I'm posting mainly because the test code given below would need to be tweaked a little - to switch back to the default AI, you would need

[Wesnoth-bugs] [bug #24709] Assertion Failure when selecting a UMC random map

Follow-up Comment #2, bug #24709 (project wesnoth): I think I roughly understand the cause of this issue - it looks like a null dereference in the default_map_generator_job::generate_name function. PR 661 deletes this function and no longer holds possibly-null pointers to name generators, so I

[Wesnoth-bugs] [bug #24672] Leadership ability does not work when applied as an object in [modify_unit]

Update of bug #24672 (project wesnoth): Status:None => Fixed Summary: Leadership ability does not work when applied as an object => Leadership ability does not work when applied as an object in [modify_unit]

[Wesnoth-bugs] [bug #24672] Leadership ability does not work when applied as an object

Follow-up Comment #2, bug #24672 (project wesnoth): I don't think this is possible to fix in the leadership definition since it's apparently caused by premature variable substitution. So you should be able to get it to work by adding "delayed_variable_substitution=yes" in your [object]. I guess

[Wesnoth-bugs] [bug #24186] Using a different keyboard layout causes letter based hotkeys not to work.

Follow-up Comment #12, bug #24186 (project wesnoth): ...wait a second. I just marked it as closed, but should it actually be left open until a release happens? ___ Reply to this item at:

[Wesnoth-bugs] [bug #24186] Using a different keyboard layout causes letter based hotkeys not to work.

Update of bug #24186 (project wesnoth): Open/Closed:Open => Closed ___ Reply to this item at: ___ Message

[Wesnoth-bugs] [bug #24550] std::placeholders might clash with boost::bind::placeholders

Update of bug #24550 (project wesnoth): Status:None => Fixed Assigned to:None => celticminstrel Open/Closed:Open => Closed

[Wesnoth-bugs] [bug #24527] AOI S2 broke by a recent change

Update of bug #24527 (project wesnoth): Severity: 3 - Normal => 4 - Important ___ Reply to this item at: ___ Message

[Wesnoth-bugs] [bug #12098] allow [message] to take multiple 'speakers'

Update of bug #12098 (project wesnoth): Assigned to:mordante => celticminstrel ___ Reply to this item at: ___ Message

[Wesnoth-bugs] [bug #24527] AOI S2 broke by a recent change

Update of bug #24527 (project wesnoth): Status:None => Fixed Open/Closed:Open => Closed ___ Reply to this item at:

[Wesnoth-bugs] [bug #24529] Mainline MP Random Maps show no starting postions

Update of bug #24529 (project wesnoth): Item Group: None of the others => Multiplayer ___ Reply to this item at: ___ Message

[Wesnoth-bugs] [bug #24517] Load map dialog should have shortcut to game data

Follow-up Comment #1, bug #24517 (project wesnoth): This is even worse on Mac, where the game data directory is hidden inside the App package, so I definitely agree here. ___ Reply to this item at:

[Wesnoth-bugs] [bug #18679] [ai] blocks are ignored

Update of bug #18679 (project wesnoth): Assigned to:crab => celticminstrel ___ Reply to this item at: ___ Message

[Wesnoth-bugs] [bug #24304] [item]s missing during replay

Follow-up Comment #2, bug #24304 (project wesnoth): Just now I reproduced it on master with the following steps: 1. Start the Son of the Black Eye campaign. This creates a turn 1 auto-save, or you could make an explicit save. Note the pair of big flags near the orc camp. 2. Go to Load Game,

[Wesnoth-bugs] [bug #24304] [item]s missing during replay

URL: Summary: [item]s missing during replay Project: Battle for Wesnoth Submitted by: celticminstrel Submitted on: Tue 12 Jan 2016 05:39:57 AM UTC Category: Bug Severity: 3 -

[Wesnoth-bugs] [bug #24288] WML formulas evaluated too soon

Update of bug #24288 (project wesnoth): Status:None => Fixed Open/Closed:Open => Closed ___ Reply to this item at:

[Wesnoth-bugs] [bug #24288] WML formulas evaluated too soon

Follow-up Comment #2, bug #24288 (project wesnoth): I think commit c2ac138 (in pull request 578) should fix this. ___ Reply to this item at: ___ Message sent

[Wesnoth-bugs] [bug #24200] Dark Forecast maps not aligned

Follow-up Comment #1, bug #24200 (project wesnoth): This might be caused by a change in how [terrain_mask] works (commit 1b22a97). If so, I believe adding "border=no" to the [terrain_mask] should fix the issue. ___ Reply to this item at:

[Wesnoth-bugs] [bug #23917] const [event]s

Follow-up Comment #1, bug #23917 (project wesnoth): I'm a little uncertain about this idea. The intent of saving space in the savefile is good, but somehow the idea of an event that can't be deleted doesn't quite sit right with me. However, should this be implemented, I think "permanent" would

[Wesnoth-bugs] [bug #23918] UI Graphics In OS X El Capitan

Follow-up Comment #3, bug #23918 (project wesnoth): I removed that class because I was getting a runtime warning about multiple definitions of the class, but it looks like for some reason the other definition doesn't exist for you. The class is still defined in SDLmain.m but only enabled for

[Wesnoth-bugs] [bug #23490] Side specific variables

Follow-up Comment #6, bug #23490 (project wesnoth): Oh, right. Somehow I didn't think of that. It's possible without using the save_id directly as a variable name, though. I imagine it'd be a little slower because lookup requires iterating over an array, but it would be possible. As for unit

[Wesnoth-bugs] [bug #23490] Side specific variables

Follow-up Comment #4, bug #23490 (project wesnoth): I can see your point about save_id maybe containing spaces and other special characters, but you can easily ensure that it doesn't (because you control what the save_id is), so I don't see that being a significant issue. The advantage of

[Wesnoth-bugs] [bug #23490] Side specific variables

Follow-up Comment #1, bug #23490 (project wesnoth): My suggestion is to store the inventory in the unit variables of the side leader. Unless you need inventory for sides without a leader? ___ Reply to this item at:

[Wesnoth-bugs] [bug #23490] Side specific variables

Follow-up Comment #2, bug #23490 (project wesnoth): (Another idea: If the issue is that side numbers change, reference something that won't change - the save_id. Probably not possible at the moment, would need save_id added to SSF or something.)

[Wesnoth-bugs] [bug #23896] [show_if] no longer works

Update of bug #23896 (project wesnoth): Assigned to:None => celticminstrel ___ Reply to this item at: ___ Message

[Wesnoth-bugs] [bug #23896] [show_if] no longer works

Update of bug #23896 (project wesnoth): Status:None => Fixed ___ Follow-up Comment #1: Looks like it was a typo in the new [message] implementation (it was checking

[Wesnoth-bugs] [bug #23896] [show_if] no longer works

Follow-up Comment #2, bug #23896 (project wesnoth): (Where "now" means commit d5eea2abbf3d ) ___ Reply to this item at: ___ Message sent via/by Gna!