[00:05] sev (~sev@scummvm/sev) left irc: Quit: This computer has gone to sleep [00:42] speachy (~speachy@rockbox/developer/speachy) left irc: Quit: WeeChat 4.9.0 [01:02] borosky (~borosky@213-134-187-71.dynamic.play.pl) left irc: Ping timeout: 245 seconds [06:45] [scummvm] dreammaster pushed 1 new commits to master: Error, database insert failed [06:45] scummvm/master f700845 dreammaster: MADS: FOREST: Fix crash in intro in room 104 [07:12] ldevulder (~ldevulder@82.66.145.28) joined #scummvm. [07:25] sev (~sev@81-205-47-249.fixed.kpn.net) joined #scummvm. [07:25] sev (~sev@81-205-47-249.fixed.kpn.net) left irc: Changing host [07:25] sev (~sev@scummvm/sev) joined #scummvm. [07:25] #scummvm: mode change '+o sev' by ChanServ!ChanServ@services.libera.chat [07:33] sev (~sev@scummvm/sev) left irc: Quit: Leaving [07:37] [discord] Music playback in Once Upon a Forest \o/ [07:45] borosky (~borosky@213-134-187-71.dynamic.play.pl) joined #scummvm. [07:50] ldevulder (~ldevulder@82.66.145.28) left irc: Quit: Leaving [07:55] [scummvm] mgerhardy pushed 1 new commits to master: Error, database insert failed [07:55] scummvm/master 7be75cb mgerhardy: MACS2: removed own imgui debug channel - use the global registered one [07:56] [scummvm] dreammaster pushed 2 new commits to master: Error, database insert failed [07:56] scummvm/master 4182473 dreammaster: MADS: FOREST: Cleanup of some room 101 switches [07:56] scummvm/master 616403e dreammaster: MADS: FOREST: Implementing original cheat mode [08:25] [scummvm] sev- pushed 1 new commits to master: Error, database insert failed [08:25] scummvm/master 44d8a71 sev-: DIRECTOR: LINGO: Improve 'the clickOn' behavior [08:32] TAS_2012v (~2012@h-98-128-246-108.A498.priv.bahnhof.se) left irc: Remote host closed the connection [08:40] [scummvm] dreammaster pushed 1 new commits to master: Error, database insert failed [08:40] scummvm/master f1359cd dreammaster: MADS: DRAGONSPHERE: Fix Dragonsphere cheat password [09:09] [discord] It's a bit sad to see how everything is shifting towards AI. I guess in a few years anyone could be a reverse-engineer just feeding up Codex/Claude with a proper prompts with zero understanding how to write code/inner knowledge... I guess even now you can. I do not like the way industry goes too, and if AI will do the code and the reverse parts - what is your work? Stare into text window with Claude, steering him into right direction? [09:11] [discord] Good for you 🙂 I have less and less time to dedicate to this, I'm lucky to have a busy life outside the internet, so if I want to keep doing this and give back to this community I have to speed it up. [09:11] [discord] Otherwise I might just abandon this hobby entirely, which would be a shame given how much I learned in 6 years and how much time I previously poured into this [09:16] [discord] Sure, I am not actually speaking about people, who were in this for years, as I guess by now most of them already have a lot of knowledge they could get from this hobby. But as AI slop extends further into all fields, it will soon transform into "feed it to AI, man", which is truly sad. AI takes jobs, takes hobby, takes knowledge. We get a sloppy result, thats right. No idea, what I wanted to say :) Just do not like where all that is he [12:04] speachy (~speachy@rockbox/developer/speachy) joined #scummvm. [12:08] [scummvm] dreammaster pushed 1 new commits to master: Error, database insert failed [12:08] scummvm/master 133a862 dreammaster: MADS: PHANTOM: Adding support for demo [12:08] gsi (~gsi@user/gsi) left irc: Ping timeout: 252 seconds [12:25] [discord] hi guys, i was looking on the gh but i couldnt find any chores or any problems a newcomer could help out with? scummvm is really important to me, and it would mean the world for me to be able to contribute, could anyone link any issues that still need attention? i'd need to learn the codebase and whats going on, but if someone from the dev team could offer a suggestion to an issue, i would really appreciate it [12:26] [discord] hi guys, i was looking on the gh but i couldnt find any chores or any problems a newcomer could help out with? scummvm is really important to me, and it would mean the world for me to be able to contribute, could anyone link any issues that still need attention? i'd need to learn the codebase and whats going on, but if someone from the dev team could offer a suggestion to an issue, i would really appreciate it (btw i dont u [12:40] [scummvm] whoozle closed pull request #7616: PHOENIXVR: Simplify blitting routines (master...phoenixvr-blit) Error, database insert failed [12:40] [scummvm] whoozle pushed 4 new commits to master: Error, database insert failed [12:40] scummvm/master 91cd7e7 ccawley2011: VIDEO: Remove extra conversion step from FourXMRawVideoTrack [12:40] scummvm/master 58dc974 ccawley2011: PHOENIXVR: Make more use of ManagedSurface::simpleBlitFrom [12:40] scummvm/master e8513e2 ccawley2011: PHOENIXVR: Combine scaling with vertical flip where possible [13:28] [discord] Is support for Return of the Phantom now in nightly builds then? [13:42] [scummvm] mgerhardy pushed 2 new commits to master: Error, database insert failed [13:42] scummvm/master b9b75e7 mgerhardy: MACS2: pass by ref for renderString [13:42] scummvm/master ddb9646 mgerhardy: MACS2: fixed center of mouse cursor sprite [14:01] [discord] I don't think it's that simple tbh. AI is doing a lot of bad stuff, but I'm also seeing people go from prompting LLMs to learning more about the underlying tech and genuinely becoming more knowledgeable about programming. I've heard of people using LLMs and getting all kinds of medical misinformation, but I've also heard about people using them to search medical literature and to become better informed about complex medical conditions [14:01] [discord] I don't think I consider LLMs a net positive for the world but since we can't undiscover them (I think "discover" is the right word. Individual LLMs are created but as a class they're kind of an emergent property of language at scale, which means that every poetic metaphor writers have ever used about the power of language is truer than they could have possibly imagined, but let's not get into that), the best we can do is figure out h [14:19] [discord] I agree with all that. Except people becoming knowledgeable with AI - I think people who were already into something - who will know it old way too, just slower, by mining databases/code/search - will get real good search engine (alas not 100% trustworthy). So in general benefits are not that evident - means they will not lead to more knowledgeable people than before, but will lead to more fakes, who think knowledge is making a proper p [14:20] [scummvm] 11-andy-11 opened pull request #7628: SDL: Fix game cursor position at scale factors above 1x (master...surfacesdl-cursor-hotspot) Error, database insert failed [14:21] [discord] I agree with all that. Except people becoming knowledgeable with AI - I think people who were already into something - who will know it old way too, just slower, by mining databases/code/search - will get real good search engine (alas not 100% trustworthy). So in general benefits are not that evident - means they will not lead to more knowledgeable people than before, but will lead to more fakes, who think knowledge is making a proper p [14:23] [discord] I do not agree with all those absolutes... I learned a lot which I could apply on my own e.g. on m68000 and 16bit x86 reverse engineering for which, I swear, I couldn't find literally any written resources or guides on how to work with them... [14:24] [discord] It always depends on a case by case basis, but if anything, things which were either underdocumented or completely undocumented are more easily accessible now [14:26] [discord] I guess we will see soon enough 😫 All popular repos already plagued with AI PRs no one will maintain after merge with hundreds lines of code. Anyway, I will stick to old way, learning m68k myself and atari restarts and special code right now... [14:46] [discord] I'm not entirely done with it yet; I'm still looking into the track to device mapping and want to make the AdLib driver more accurate to the original. But the current code provides decent playback. [14:59] [discord] Well, what happens is that people who *wouldn't* have gotten into those things before, because the barrier to getting anywhere using the "old ways" was too high now have a starting point to work their way up. [14:59] [discord] You have to remember that an astonishing number of people never learned how to craft a good search query. Hell, crafting a good search query can be a knowledge check in and of itself even if you know how, because often you won't get good results unless you know the terms of art for the thing you're looking forwhich, frequently, newcomers to a field *do not*. LLMs are actually really useful for search problems (RAG, aka duct-taping [14:59] [discord] LLMs will probably increase the number of frauds more than it increases the number of actually knowledgeable people, mind, and that is a troubling problem (clearly LLM-written junk is already making its way into respectable publications). But they are *actually useful* for people seeking knowledge. You could say the same thing about Google and if you go back about 25 years people totally did! [15:04] [scummvm] 11-andy-11 opened pull request #7629: CHAMBER: Support the Amiga US (Chamber) release in the Amiga renderer (master...week6) Error, database insert failed [15:39] [discord] So we replace programming with prompt engineering.and still name it programming? 👍 I do not clearly see your point here, I am good if you only search with LLM, but MCP decompilation? LLM conversion of said decompiled code to C code and after telling "yes, i am a reverse engineer and prpgrammer"? I guess mileage may vary, but I will do it old fashioned way. Spending years, maybe. But at least havinf fun in process, not in the getting [15:41] [discord] So we replace programming with prompt engineering.and still name it programming? 👍 I do not clearly see your point here, I am good if you only search with LLM, but MCP decompilation? LLM conversion of said decompiled code to C code and after telling "yes, i am a reverse engineer and prpgrammer"? I guess mileage may vary, but I will do it old fashioned way. Spending years, maybe. But at least having fun in process, not in the getting [16:01] [discord] Definitely, anyone's mileage IS going to vary 🙂 [16:10] [scummvm] sev- closed pull request #7628: SDL: Fix game cursor position at scale factors above 1x (master...surfacesdl-cursor-hotspot) Error, database insert failed [16:10] [scummvm] sev- pushed 1 new commits to master: Error, database insert failed [16:10] scummvm/master 641e3d9 11-andy-11: SDL: Fix game cursor position at scale factors above 1x [16:11] [scummvm] ScummVM-Translations pushed 1 new commits to master: Error, database insert failed [16:11] scummvm/master 8024443 weblate: I18N: Update translations templates [16:11] [scummvm] ScummVM-Translations pushed 1 new commits to master: Error, database insert failed [16:11] scummvm/master 9ed7ad0 weblate: I18N: Update translation files [16:21] [discord] I'm trying to use compress_queen but I'm getting this error : [16:21] [discord] ``Fatal Error : You are using an incorrect (outdated?) version of the queen.tbl file`` [16:23] [discord] Apparently there are (at least) three versions of the file, but in any case the one [linked to in the wiki]() doesn't work with the uncompressed FOTAQ in the Downloads section of the website. [16:23] [discord] Any idea ? [16:31] [discord] Nevermind, [this one]() works. [16:32] [discord] -# There's got to be a rule in the universe that says that you're bound to find a solution right after you've asked for help. [16:55] [discord] Why do you want that even 😂 [16:56] [discord] A retro handheld with limited storage space. [17:09] [discord] You save like 30 megabytes or less [17:09] [discord] Nothing in the modern storage sizes [17:11] [discord] Fair point but then why allow compression at all ? [17:12] [discord] Doesn't matter much anyway. I got my answer, and if nobody cares, well& cheers ? 😄 [17:13] borosky (~borosky@213-134-187-71.dynamic.play.pl) left irc: Ping timeout: 276 seconds [17:14] [discord] We allowed it 25 years ago [17:14] [discord] No kidding [17:15] [discord] Most of our compression tools are from an era when larger drives were expensive. It's still relevant for a few platforms (the RiscPC for example is notoriously fussy about IDE compatibility, and I'm currently using a 512 MB disc in mine), but for everything else, larger SD cards are quite cheap these days. [17:16] [discord] &or they were six months ago. 😂 [17:17] [discord] https://xkcd.com/979/ [17:17] [discord] (Seriously, I wanted to buy a 256GB card the other day and they're like 60 now.) [17:18] [discord] [17:18] [discord] https://en.wikipedia.org/wiki/Rubber_duck_debugging (edited) [17:19] Last message repeated 1 time(s). [17:21] [discord] I talk to my pens. They're in a pot beneath the monitor, so they can spy on what happens behind the scenes. [17:22] [discord] -# As long as they don't start answering, I'm fine. [17:23] [discord] Well, you are the talking yogurt after all