ScummVM IRC Log - 10 May 2026

[Back to Index]

[01:54] <-- Dominus left irc: Ping timeout: 255 seconds
[01:54] --> DominusExult joined #scummvm.
[01:55] Nick change: DominusExult -> Dominus
[02:20] <Scummette> [scummvm] tunnelsociety opened pull request #7503: NANCY: Sync JournalData Entry sceneID (master...nancy-journalentryhotspot) https://is.gd/vHyoZK
[02:32] <grogbot> [discord] <npjg> Yes, most of Lion King should be playable, but the two minigames (stalking zazu and the constellations) don't work yet becuase they have hardcoded logic instead of scripts. In dalmatians, the last minigame (the fetch maze) doesn't work for similar reasons. I am working on those now actually. Puzzle Castle starts but doesn't get very far because the maze navigation is hardcoded in there too. I don't believe the Puzzle Castle CD audio track
[02:36] <grogbot> [discord] <_plunderscore_> Honestly listening to them out of context after three decades, I think the games are probably better off without them. They sound like they took helium before singing.
[03:36] <Scummette> [scummvm] dreammaster pushed 1 new commits to master: https://is.gd/PHhBf7
[03:36] <Scummette> scummvm/master 27334ea dreammaster: MADS: DRAGONSPHERE: Added room 105
[04:45] <Scummette> [scummvm] dreammaster pushed 2 new commits to master: https://is.gd/YhHJwm
[04:45] <Scummette> scummvm/master 411be8b dreammaster: MADS: DRAGONSPHERE: Added room 106
[04:45] <Scummette> scummvm/master bcecd8a dreammaster: MADS: DRAGONSPHERE: Update remaining room skeleton files
[04:51] <ScummBot> Build [#17676](https://buildbot.scummvm.org/#/builders/54/builds/17676) of `master-android-arm` failed: failed compile (failure)
[04:51] <ScummBot> Build [#17679](https://buildbot.scummvm.org/#/builders/23/builds/17679) of `master-android-arm64` failed: compile (failure)
[04:51] <ScummBot> Build [#10298](https://buildbot.scummvm.org/#/builders/181/builds/10298) of `master-appletv` failed: failed compile (failure)
[04:51] <ScummBot> Build [#17088](https://buildbot.scummvm.org/#/builders/108/builds/17088) of `master-ios7-arm64` failed: compile (failure)
[04:51] <ScummBot> Build [#17675](https://buildbot.scummvm.org/#/builders/80/builds/17675) of `master-android-x86` failed: compile (failure)
[04:52] <ScummBot> Build [#17665](https://buildbot.scummvm.org/#/builders/24/builds/17665) of `master-android-x86-64` failed: failed compile (failure)
[04:53] <ScummBot> Build [#17686](https://buildbot.scummvm.org/#/builders/39/builds/17686) of `master-debian-x86-64-clang` failed: compile (failure)
[04:56] <ScummBot> Build [#17522](https://buildbot.scummvm.org/#/builders/93/builds/17522) of `master-macosx-arm64` failed: compile (failure)
[04:56] <ScummBot> Build [#17660](https://buildbot.scummvm.org/#/builders/78/builds/17660) of `master-macosx-i386` failed: compile (failure)
[04:56] <ScummBot> Build [#17518](https://buildbot.scummvm.org/#/builders/89/builds/17518) of `master-macosx-x86_64` failed: failed compile (failure)
[05:01] <ScummBot> Build [#15757](https://buildbot.scummvm.org/#/builders/34/builds/15757) of `fetch-master` failed: failed triggered build-scheduler-master, 29 successes, 10 failures
[05:44] <Scummette> [scummvm] dreammaster pushed 1 new commits to master: https://is.gd/8pe9Eh
[05:44] <Scummette> scummvm/master 14e600e dreammaster: MADS: DRAGONSPHERE: Added newline at the end of room header files
[05:49] <-- actioninja left irc: Remote host closed the connection
[05:50] --> actioninja joined #scummvm.
[05:51] <ScummBot> Build [#10299](https://buildbot.scummvm.org/#/builders/181/builds/10299) of `master-appletv` completed successfully.
[05:51] <ScummBot> Build [#17680](https://buildbot.scummvm.org/#/builders/23/builds/17680) of `master-android-arm64` completed successfully.
[05:53] <ScummBot> Build [#17677](https://buildbot.scummvm.org/#/builders/54/builds/17677) of `master-android-arm` completed successfully.
[05:53] <ScummBot> Build [#17687](https://buildbot.scummvm.org/#/builders/39/builds/17687) of `master-debian-x86-64-clang` completed successfully.
[05:54] <ScummBot> Build [#17676](https://buildbot.scummvm.org/#/builders/80/builds/17676) of `master-android-x86` completed successfully.
[05:56] <ScummBot> Build [#17089](https://buildbot.scummvm.org/#/builders/108/builds/17089) of `master-ios7-arm64` completed successfully.
[05:56] <ScummBot> Build [#17523](https://buildbot.scummvm.org/#/builders/93/builds/17523) of `master-macosx-arm64` completed successfully.
[05:56] <ScummBot> Build [#17661](https://buildbot.scummvm.org/#/builders/78/builds/17661) of `master-macosx-i386` completed successfully.
[05:56] <ScummBot> Build [#17666](https://buildbot.scummvm.org/#/builders/24/builds/17666) of `master-android-x86-64` completed successfully.
[05:57] <ScummBot> Build [#17519](https://buildbot.scummvm.org/#/builders/89/builds/17519) of `master-macosx-x86_64` completed successfully.
[06:02] <ScummBot> Build [#15758](https://buildbot.scummvm.org/#/builders/34/builds/15758) of `fetch-master` completed successfully.
[06:11] <Scummette> [scummvm] dreammaster pushed 1 new commits to master: https://is.gd/IWhUKt
[06:11] <Scummette> scummvm/master d2676bb dreammaster: MADS: DRAGONSPHERE: Added room 107
[06:22] --> DominusExult joined #scummvm.
[06:23] <Scummette> [scummvm] SegMash opened pull request #7504: SCI: Adding Hebrew Translation For KQ7 (master...kq7-hebrew-translation) https://is.gd/x6dxgW
[06:25] <-- Dominus left irc: Ping timeout: 244 seconds
[06:25] Nick change: DominusExult -> Dominus
[06:52] <Scummette> [scummvm] dreammaster pushed 2 new commits to master: https://is.gd/R4lxm2
[06:52] <Scummette> scummvm/master 6a188af dreammaster: MADS: DRAGONSPHERE: Added room 108
[06:52] <Scummette> scummvm/master 555fe3d dreammaster: MADS: DRAGONSPHERE: Added room 109
[08:02] --> borosky joined #scummvm.
[08:48] <Scummette> [scummvm] dreammaster pushed 4 new commits to master: https://is.gd/zKo71K
[08:48] <Scummette> scummvm/master e09d07f dreammaster: MADS: PHANTOM: Fix potential write to invalid memory in quote_load
[08:48] <Scummette> scummvm/master 6f1ef5c dreammaster: MADS: DRAGONSPHERE: Added room 110
[08:48] <Scummette> scummvm/master 7b4550f dreammaster: MADS: DRAGONSPHERE: Add commonly needed includes to remaining room files
[09:36] <-- ced117 left irc: Ping timeout: 255 seconds
[09:45] <Scummette> [scummvm] dreammaster pushed 4 new commits to master: https://is.gd/Cvl4Rt
[09:45] <Scummette> scummvm/master f125a12 dreammaster: MADS: DRAGONSPHERE: Added room 112
[09:45] <Scummette> scummvm/master 9c90161 dreammaster: MADS: DRAGONSPHERE: Added room 113
[09:45] <Scummette> scummvm/master 9706500 dreammaster: MADS: DRAGONSPHERE: Added room 114
[09:57] --> ced117 joined #scummvm.
[09:59] <Scummette> [scummvm] sev- closed pull request #7497: CONFIGURE: MACOS: Fix TTS & SDK feature checks for Tiger/Leopard (master...fix/build-osx-improve-objc-portability-checks) https://is.gd/Hwokw3
[09:59] <Scummette> [scummvm] sev- pushed 2 new commits to master: https://is.gd/MLBEoo
[09:59] <Scummette> scummvm/master d9a9f83 dwatteau: CONFIGURE: MACOS: Replace `-ObjC++` compiler flag with `-x objective-c++`
[09:59] <Scummette> scummvm/master b73f6e3 dwatteau: CONFIGURE: MACOS: Improve NSSpeechSynthesizerDelegate test for Tiger/Leopard
[10:01] <Scummette> [scummvm] bluegr closed pull request #7504: SCI: Adding Hebrew Translation For KQ7 (master...kq7-hebrew-translation) https://is.gd/x6dxgW
[10:01] <Scummette> [scummvm] bluegr pushed 1 new commits to master: https://is.gd/Axsdf2
[10:01] <Scummette> scummvm/master 8bf7656 SegMash: SCI: Adding Hebrew Translation For KQ7
[10:01] <Scummette> [scummvm] ScummVM-Translations pushed 1 new commits to master: https://is.gd/zcECWK
[10:01] <Scummette> scummvm/master 22000cb weblate: I18N: Update translations templates
[10:02] <Scummette> [scummvm] ScummVM-Translations pushed 1 new commits to master: https://is.gd/x619LQ
[10:02] <Scummette> scummvm/master a04992e weblate: I18N: Update translation files
[10:03] <Scummette> [scummvm] sev- pushed 1 new commits to master: https://is.gd/ZRMufx
[10:03] <Scummette> scummvm/master 145e5db sev-: MADS: DRAGONSPHERE: Remove extraneous parenthesis
[10:42] <Scummette> [scummvm] sev- pushed 2 new commits to master: https://is.gd/ci9LBV
[10:42] <Scummette> scummvm/master 760e57a sev-: DIRECTOR: Added more debug output on movie unloading
[10:42] <Scummette> scummvm/master f240c04 sev-: DIRECTOR: Do not reload movie if next movie is set to the same
[11:14] <Scummette> [scummvm] sev- pushed 1 new commits to master: https://is.gd/j3PhYq
[11:14] <Scummette> scummvm/master 56e697f sev-: DIRECTOR: Print out shared cast full path in debugger
[12:29] --> Guest4 joined #scummvm.
[12:29] <-- Guest4 left irc: Quit: Client closed
[13:13] <Scummette> [scummvm] bluegr pushed 1 new commits to master: https://is.gd/Hxivmb
[13:13] <Scummette> scummvm/master 6725db8 bluegr: NANCY: Fix broken web search links after saving/loading in Nancy9
[13:14] <Scummette> [scummvm] bluegr closed pull request #7503: NANCY: Sync JournalData Entry sceneID (master...nancy-journalentryhotspot) https://is.gd/vHyoZK
[15:39] --> ny00123 joined #scummvm.
[16:09] <Scummette> [scummvm] bluegr closed pull request #7494: NANCY: Add detection for Wii version of nancy16 (master...patch-11) https://is.gd/DbSGSN
[16:09] <Scummette> [scummvm] bluegr pushed 1 new commits to master: https://is.gd/wauAFX
[16:09] <Scummette> scummvm/master 824d7d4 einstein95: NANCY: Add detection for Wii version of nancy16
[17:19] <-- ced117 left irc: Ping timeout: 244 seconds
[17:36] --> ced117 joined #scummvm.
[18:02] <-- APic left irc: Quit: fffffffffffffffffffffffffffffffffuuuuuuuuuuuuuuu-
[18:07] --> APic joined #scummvm.
[21:00] <-- ny00123 left irc: Quit: Leaving
[21:40] <grogbot> [discord] <tunnelsociety> finished a test play of Nancy Drew: Danger on Deception Island 🐋 🚲! 34 more tickets in the tracker 😅 (some already closed!)
[21:40] <grogbot> [discord] <tunnelsociety> say, this game has ADGF_TESTING, but does there really continue to be such a thing as a testing period for games as of version 2026.x? I am reading the Google Doc "proposed release procedure" and I guess not anymore. But as for this game, the wiki says supported as of "2026.3.0" which of course doesn't exist yet, and it's not in any non-DEV compatibility table yet, but the news post which has been distributed doesn't really imply
[21:40] <grogbot> [discord] <tunnelsociety> It led to what I consider a very annoying situation trying to do a fix for an issue in this Nancy game, where we are *already* worried about breaking saves for a game that is "supported" in a future release, that has not come yet. Maybe just a growing pain. was new implication (or loss of meaning) of ADGF_TESTING foreseen? Note that flag only presents a warning to user in release build, not otherwise...
[21:44] <grogbot> [discord] <tunnelsociety> ...35! i forgot one, to file right now 😆
[21:53] <grogbot> [discord] <sev____> you first come with a wrong assumption, and then make conclusion based on it, which is logically also wrong
[21:54] <grogbot> [discord] <sev____> ADGF_TESTING is required and is used. There is a time gap between game announced as supported (ADGF_TESTING) and release, when that flag is cleared
[21:54] <grogbot> [discord] <tunnelsociety> great, that's what i want to hear! would you mind chiming in at github PR 7503?
[21:58] <grogbot> [discord] <sev____> I already discussed it with @.bluegr
[21:58] <grogbot> [discord] <tunnelsociety> ok. what is the outcome of that?
[21:58] <grogbot> [discord] <sev____> like 7 hours ago
[22:00] <grogbot> [discord] <tunnelsociety> it is not really my assumption, it is what i learned from 7503. i am waiting to be corrected on that...
[22:02] <grogbot> [discord] <sev____> I suggested to introduce save versioning like we have in other engines, and store this value in save version 2
[22:03] <grogbot> [discord] <tunnelsociety> ok, so that implies that we should protect existing saves, right? i am wondering why that is. what is valuable about existing saves for this game which is in testing?
[22:04] <grogbot> [discord] <sev____> it is not about a game, but the engine
[22:09] <grogbot> [discord] <sev____> once a game has been announced public, we require all the saves to be backward-compatible
[22:10] <grogbot> [discord] <sev____> we break compatibility only for in-development engines
[22:12] <grogbot> [discord] <tunnelsociety> but ADGF_TESTING is per-game not per-engine right? so overall, is Nancy engine in development or not? consider this conversation with nancy engine's author a couple years ago for a similar situation:
[22:12] <grogbot> [discord] <tunnelsociety> ```
[22:12] <grogbot> [discord] <tunnelsociety> tunnelsociety: oh? i had a patch for it, uhh
[22:12] <grogbot> [discord] <tunnelsociety> fracturehill: hmmm and does that patch handle both existing saves and new ones? ;)
[22:12] <grogbot> [discord] <tunnelsociety> tunnelsociety: no but DOG isn't supported in current right?
[22:12] <grogbot> [discord] <tunnelsociety> tunnelsociety: do we have freedom to break DOG save files for the moment?
[22:12] <grogbot> [discord] <tunnelsociety> fracturehill: yes
[22:12] <grogbot> [discord] <tunnelsociety> fracturehill: oh wait is this dog-specific?
[22:12] <grogbot> [discord] <tunnelsociety> ```
[22:12] <grogbot> [discord] <tunnelsociety> (at that time, incidentally, it turned out NOT to be DOG-specific but also affecting a couple earlier games 😄)
[22:13] <grogbot> [discord] <tunnelsociety> when is the right time to test a game and discover save-breaking issues that are convenient to fix, before we need to protect existing saves? or maybe that phase does not come
[22:14] <grogbot> [discord] <sev____> I do not see that the nancy engine is using game-specific save formats
[22:15] <grogbot> [discord] <sev____> e.g. it is just the same routines to store. Since it is using marshalling, I understand, that certain objects will be present only in specifc games
[22:15] <grogbot> [discord] <sev____> and the save system is still same.
[22:16] <grogbot> [discord] <sev____> thus, it would better have save versions, and for the cases like this, the save version bumped
[22:16] <grogbot> [discord] <sev____> the dialog which I read with @fracturehill does not sound right to me
[22:17] <grogbot> [discord] <tunnelsociety> i guess my understanding of saves might be inside-out since Nancy engine is where i spend almost all my time in ScummVM
[22:17] <grogbot> [discord] <sev____> I was not aware of this happening, and I can conclude based on the info I have, that Kaloyan did not know about it either
[22:18] <grogbot> [discord] <tunnelsociety> for sure, that is fine. i think my question still stands, about when is the appropriate time to test while having the freedom to break saves. do i need to be a ScummVM team member to keep my finger on the pulse and get to testing before it's complex to make some changes? like, i was actually waiting for an announcement of nancy8 and nancy9 **to begin** testing; i could have started earlier if i believed it was the right time.
[22:20] <grogbot> [discord] <sev____> I think I've answered that question: when an engine is in-development. The moment we announce public testing, saves must have pertain the compatibility
[22:21] <grogbot> [discord] <sev____> and in this particular case, the engine is in public since 2023041
[22:21] <grogbot> [discord] <sev____> and in this particular case, the engine is in public since 20230411 (edited)
[22:23] <grogbot> [discord] <sev____> even if you'd test these games before the announcement, which happened on April 14th, the save compatibility must be pertained
[22:23] <grogbot> [discord] <sev____> these games = 8 & 9
[22:24] <grogbot> [discord] <sev____> we cannot break saves for 0-7
[22:24] <grogbot> [discord] <tunnelsociety> ok, you have said 0-7 there. what about 8 and 9, right now?
[22:24] <grogbot> [discord] <sev____> I do not know how to explain it differently. Let me try, though
[22:25] <grogbot> [discord] <sev____> it is one engine
[22:25] <grogbot> [discord] <sev____> it supports now 9 games
[22:25] <grogbot> [discord] <sev____> the save code is the same
[22:25] <grogbot> [discord] <bluegr> Oh my...
[22:25] <grogbot> [discord] <sev____> it must be able to read any saves made after april 11, 2023
[22:25] <grogbot> [discord] <sev____> the date when we announced support for the first games
[22:26] <grogbot> [discord] <bluegr> I thought I explained the situation about broken saved games in the PR
[22:26] <grogbot> [discord] <bluegr> I totally agree with @sev____ here
[22:26] <grogbot> [discord] <bluegr> Let me repeat what was discussed
[22:26] <grogbot> [discord] <bluegr> We have saved games for a game
[22:29] <grogbot> [discord] <bluegr> And it was proposed to alter them in a way that BREAKS them without any warning, or way to detect that they are broken
[22:29] <grogbot> [discord] <bluegr> This is frustrating, prone to errors, and will bring additional bug reports for no reason
[22:29] <grogbot> [discord] <bluegr> I explained that, we CAN break existing saved games, but we have two options:
[22:29] <grogbot> [discord] <bluegr> 1) TELL the user they are broken, and refuse to load them
[22:29] <grogbot> [discord] <bluegr> 2) TELL the user they are broken, and try to fix them
[22:29] <grogbot> [discord] <bluegr> In that PR, none of these two options was done
[22:29] <grogbot> [discord] <bluegr> The CORRECT way to do such changes is by introducing versioning, and avoid broken saved games that can't be detected
[22:29] <grogbot> [discord] <bluegr> So, we'll introduce versioning to nancy saves, and save the needed data. That is correct design
[22:29] <grogbot> [discord] <bluegr> Suddenly breaking saved games without warning, or a way to determine they're broken, is bad design
[22:30] <grogbot> [discord] <tunnelsociety> the proposed code of that PR *only* affected nancy9. another way to say that is, nany0 to nancy8 were *totally unaffected* by that PR. nancy0 to nancy7 are considered supported, so of course those should not break. the question is specifically about nancy8 and nancy9 which have AGDF_TESTING.
[22:31] <grogbot> [discord] <tunnelsociety> with all due respect, there are unanswered questions from me in that PR too, so no, the explanation in that PR is not complete
[22:32] <grogbot> [discord] <bluegr> This is a global approach
[22:32] <grogbot> [discord] <tunnelsociety> what i am hearing is that saves cannot be broken for games in testing, which contradicts for example wiki article "Release Testing" where it is written: `It may happen that at some point of our testing we break savegame compatibility, or do something which affects a specific engine deeply. In that case, you could be asked to restart the game you're currently playing from the beginning. You will be explicitly told to do so, so don
[22:33] <grogbot> [discord] <sev____> that statement is about horribly broken saves, which we had in the past
[22:33] <grogbot> [discord] <bluegr> SCI has been in development for years, for example, and we changed its saved game format about 8 times or so
[22:33] <grogbot> [discord] <bluegr> Each change was versioned, and all changes were either backwards compatible, or told the user that his games were broken
[22:33] <grogbot> [discord] <bluegr> Same with SCUMM
[22:33] <grogbot> [discord] <sev____> horribly broken = unfixable
[22:33] <grogbot> [discord] <tunnelsociety> i agree very strongly that save versioning is a good thing to do ASAP, for the record
[22:34] <grogbot> [discord] <bluegr> Exactly. "In that case, you could be asked to restart the game you're currently playing from the beginning. You will be explicitly told to do so, so don't worry if we say nothing in this aspect."
[22:34] <grogbot> [discord] <bluegr> BUT we can't do that without DETECTING that the saved games are broken! We need versioning
[22:34] <grogbot> [discord] <sev____> > Same with SCUMM
[22:34] <grogbot> [discord] <sev____> LOL, @.bluegr
[22:34] <grogbot> [discord] <bluegr> Well, at least we try 🙂
[22:34] <grogbot> [discord] <sev____> scumm/aaveload.cpp
[22:34] <grogbot> [discord] <sev____> ```C++
[22:34] <grogbot> [discord] <sev____> #define CURRENT_VER 124
[22:34] <grogbot> [discord] <sev____> ```
[22:34] <grogbot> [discord] <bluegr> Yeah, ok, true 😛
[22:35] <grogbot> [discord] <bluegr> So, imagine breaking saved games 124 times
[22:35] <grogbot> [discord] <bluegr> And not telling users about it
[22:35] <grogbot> [discord] <bluegr> How would they react?
[22:35] <-- borosky left irc: Ping timeout: 245 seconds
[22:35] <grogbot> [discord] <tunnelsociety> please, i understand all of this about breaking saved games *for supported games*!
[22:36] <grogbot> [discord] <ccawley2011> Are you able to 100% guarantee that the changes will _only_ affect nancy9 and won't have unforeseen side effects on the other titles?
[22:36] <grogbot> [discord] <sev____> let's try wrapping it up.
[22:36] <grogbot> [discord] <sev____> @tunnelsociety is the picture about the save compatibility clear to you now?
[22:36] <grogbot> [discord] <bluegr> In this case, the broken saves would be only for nancy9
[22:36] <grogbot> [discord] <tunnelsociety> what the heck does `if (>= kNancy9)` mean? honestly, it was such a simple proposed change. it could not have been simpler, actually. the save versioning system will need the same logic as that, right?
[22:37] <grogbot> [discord] <bluegr> OK, let's clear this up, in all capitals: we DON'T WANT TO BREAK ANY SAVED GAME, UNLESS WE CAN TELL THE USER ABOUT IT
[22:37] <grogbot> [discord] <bluegr> Better?
[22:37] <grogbot> [discord] <tunnelsociety> the thing i seem to have missed was the window of opportunity to test games *before* we were worried about this save compatibility. it seems that this window *does not exist* which is bizarre to me
[22:38] <grogbot> [discord] <sev____> there is no need to attack us, please
[22:38] <grogbot> [discord] <sev____> such window of opportunity exists before a game is announced as supported
[22:38] <grogbot> [discord] <sev____> which already happened for nancy8 and 9
[22:38] <grogbot> [discord] <tunnelsociety> no, i am sorry for the tone, it is certainly not an attack. i love ScummVM's Nancy engine and want it to be good. i am sad that i am missing out entirely on how i can contribute in a useful way to it.
[22:39] <grogbot> [discord] <bluegr> But... honestly, I'll just wire up the versioning check and let's wrap this up
[22:39] <grogbot> [discord] <sev____> so, technically, before that announcement last month, during the dev, the code within (if >= Nancy9) could still change unconditionally
[22:39] <grogbot> [discord] <sev____> not anymore
[22:39] <grogbot> [discord] <tunnelsociety> wiki says 2026.3.0, please explain that version number...
[22:39] <grogbot> [discord] <tunnelsociety> i guess that is effectively every build following 2026.2.0?
[22:40] <grogbot> [discord] <sev____> no
[22:40] <grogbot> [discord] <sev____> you somehow missing parts of this discussion. maybe you're a bit overwhelmed
[22:40] <grogbot> [discord] <sev____> this is the announcement about the game being supported: https://www.scummvm.org/news/20260414/
[22:41] <grogbot> [discord] <sev____> this is the date after which the users may have saves, we want to be able to load them
[22:42] <grogbot> [discord] <tunnelsociety> ok, i will just overlook the wiki article for the game. then, looking to the future, when is a very suitable time for me to try nancy10+? (this is why i mentioned being a team member above -- it is not clear when is the transition from "of course it is broken, most stuff is not even implemented yet" to "this game *SHOULD* work but we need to discover issues")
[22:43] <grogbot> [discord] <sev____> here is an example of a bad save handling, cine/saveload.cpp
[22:43] <grogbot> [discord] <sev____> ```C++
[22:43] <grogbot> [discord] <sev____> } else if (saveGameFormat == TEMP_OS_FORMAT) {
[22:43] <grogbot> [discord] <sev____> GUI::MessageDialog alert(_("WARNING: The savegame you are loading is using "
[22:43] <grogbot> [discord] <sev____> "a temporary broken format. Things will be broken. Please consider starting "
[22:43] <grogbot> [discord] <sev____> "Operation Stealth from beginning using new savegames."),
[22:43] <grogbot> [discord] <sev____> _("Load anyway"), _("Cancel"));
[22:44] <grogbot> [discord] <sev____> load = (alert.runModal() == GUI::kMessageOK);
[22:44] <grogbot> [discord] <sev____> }
[22:44] <grogbot> [discord] <sev____> ```
[22:44] <grogbot> [discord] <bluegr> The important thing here is: Even though that's very broken, at least the user knows about it
[22:44] <grogbot> [discord] <bluegr> The important thing here is: even though that's very broken, at least the user knows about it (edited)
[22:45] <grogbot> [discord] <bluegr> With the proposed change, the engine would read garbage from existing saved games, without telling anyone.
[22:45] <grogbot> [discord] <bluegr> That's the definition of bad design
[22:45] <grogbot> [discord] <sev____> the engine devs know. this time @.bluegr may have an idea when
[22:46] <grogbot> [discord] <bluegr> You are free to start any game isn't announced. Just don't expect anything to work 🙂
[22:46] <grogbot> [discord] <bluegr> You are free to start any game that isn't announced. Just don't expect anything to work 🙂 (edited)
[22:47] <grogbot> [discord] <sev____> and for the nancy8&9. My understanding that @.bluegr has plans to add versioning and store that field in the newer saves, while still loading the previous saves, or asking the user to restart
[22:48] <grogbot> [discord] <bluegr> Exactly
[22:48] <grogbot> [discord] <sev____> and that is thanks to your research
[22:48] <grogbot> [discord] <tunnelsociety> actually, I think i did not explain my own understanding well enough. i genuinely believed that nancy8 and nancy9 saves were *unimportant* because the games were AGDF_TESTING. you can see how the logic protects games 1-7 very well.
[22:48] <grogbot> [discord] <tunnelsociety> but, unfortunately, this kind of takes me back to the beginning--at this point, i am not sure what is the usefulness of ADGF_TESTING
[22:48] <grogbot> [discord] <bluegr> What does a testing flag have to do with this?
[22:48] <grogbot> [discord] <sev____> the usefulness is to show warning to a user
[22:49] <grogbot> [discord] <tunnelsociety> the warning is not shown!
[22:49] <grogbot> [discord] <bluegr> Why are we even discussing about breaking existing saves without telling about this to the users?
[22:49] <grogbot> [discord] <tunnelsociety> i can launch CAR and DDI in ScummVM right now and i see nothing, the game simply starts
[22:49] <grogbot> [discord] <sev____> then it is broken
[22:50] <grogbot> [discord] <tunnelsociety> hmm, i read about its intent earlier, i think it was searching this Discord. one moment, i will try to find a reference
[22:50] <grogbot> [discord] <ccawley2011> `AGDF_TESTING` is distinct from `ADGF_UNSTABLE` - the former is for when public testing has been announced (and the save format should be preserved, the latter is for games that are incomplete (and existing saves may break at any time).
[22:50] <grogbot> [discord] <ccawley2011> `AGDF_TESTING` is distinct from `ADGF_UNSTABLE` - the former is for when public testing has been announced (and the save format should be preserved), the latter is for games that are incomplete (and existing saves may break at any time). (edited)
[22:51] <grogbot> [discord] <bluegr> OK, I honestly don't understand the whole situation. There is literally 0 (zero) reason to break something without knowing it's broken. For me, there's nothing more to add to that reasoning
[22:52] <grogbot> [discord] <sev____> we are far past this point
[22:52] <grogbot> [discord] <sev____> discussing nancy10
[22:52] <grogbot> [discord] <tunnelsociety> i agree, there is nothing more to say on breaking saves
[22:52] <grogbot> [discord] <bluegr> Good
[22:53] <grogbot> [discord] <tunnelsociety> thanks, that makes sense. I remember reading somewhere that --enable-release does the warning, but the default (debug, or not-release) does not. maybe it was that Google Doc, which i now find myself unable to read without an account login i do not have :/
[22:53] <grogbot> [discord] <ccawley2011> Also regarding `ADGF_UNSTABLE`, while we _can_ break saves for unstable games, it doesn't necessarily mean we _should_ if there's a better alternative.
[22:54] <grogbot> [discord] <tunnelsociety> ah, here we go! https://discord.com/channels/581224060529148060/581224061091446795/590661662630084630
[22:54] <grogbot> [discord] <sev____> @tunnelsociety by chance, maybe you have `enable_unsupported_game_warning=no` in your scummvm.ini?
[22:55] <grogbot> [discord] <tunnelsociety> that is possible, i will check now. but i am noticing the word `unsupported` in there 😉
[22:56] <grogbot> [discord] <sev____> ah, I see the problem...
[22:56] <grogbot> [discord] <tunnelsociety> no, it is not there. and of course it is not, i just remembered that i see a warning as i launch Nancy10 :p
[22:56] <grogbot> [discord] <sev____> ```C++
[22:56] <grogbot> [discord] <sev____> bool showTestingWarning = false;
[22:56] <grogbot> [discord] <sev____> #ifdef RELEASE_BUILD
[22:56] <grogbot> [discord] <sev____> showTestingWarning = true;
[22:56] <grogbot> [discord] <sev____> #endif
[22:56] <grogbot> [discord] <sev____> ```
[22:56] <grogbot> [discord] <tunnelsociety> yes, thank you
[22:56] <grogbot> [discord] <sev____> will change that tomorrow
[22:58] <grogbot> [discord] <tunnelsociety> now i had better get to Amerzone testing which i put off for so long 🥲 thanks for talking this through with me!
[23:06] <grogbot> [discord] <bluegr> Thanks for all your help with testing the Nancy games! 🙂
[23:07] <grogbot> [discord] <bluegr> Also, turns out we do have versioning in Nancy saves, so I'll just go and implement correct synchronization
[00:00] --- Mon May 11 2026