00:48 -!- Menche has quit [Quit: Leaving] 00:54 -!- Menche has joined #hardfought 00:55 -!- K2 has quit [Quit: Rico, smoke! *poof*] 00:56 -!- mode/Beholder_ [+Zi] by Beholder_ 00:56 -!- tolkien.freenode.net changed the topic on #hardfought to: Welcome to Hardfought - https://www.hardfought.org/ - public irc logs available on website ¤ Public NetHack server - ssh nethack@hardfought.org ¤ /dev/null/nethack tribute tournament hosted here starting midnight PST on Nov 1st - https://redd.it/7184xn 00:57 -!- K2 has joined #hardfought 00:57 -!- mode/#hardfought [+o K2] by ChanServ 00:57 -!- raisse has joined #hardfought 00:57 -!- mode/#hardfought [+v raisse] by ChanServ 00:58 -!- K2 has quit [Client Quit] 01:05 -!- mode/Beholder [+Zi] by Beholder 01:05 -!- tolkien.freenode.net changed the topic on #hardfought to: Welcome to Hardfought - https://www.hardfought.org/ - public irc logs available on website ¤ Public NetHack server - ssh nethack@hardfought.org ¤ /dev/null/nethack tribute tournament hosted here starting midnight PST on Nov 1st - https://redd.it/7184xn 01:05 -!- mode/#hardfought [+v Beholder] by ChanServ 01:05 -!- K2 has joined #hardfought 01:05 -!- mode/#hardfought [+o K2] by ChanServ 01:24 Tangles: new game 01:24 people, at least me, like to mess with it 01:26 jonadab: zapm ,like nethack, has a somewhat unique price category for the ID floppy disk 01:26 but if you don't find a computer (software engineers start with one), you can't use it regardless 01:28 !who 01:28 K2: Yuring [slex] 01:29 in general I almost always have more things I want to ID than I can, if I can at all 01:30 the one best off here is the software engineer role since 1, they start with floppies (which can be ID, but will at least ID some floppy disks for you), a computer (so they can actually use it) and can sense bugginess of items innately, meaning you don't have to ID armor at all except in edge cases 01:34 that said, the game is fairly short, if you can deal with BOFH (which isn't nearly as bad as Rodney, the NetHac equavilent) and the radiation in gamma caves you've probably won 01:34 especially if you can perform controlled teleporting and mapping, allowing you to skip most of the endgame 01:35 *NetHack 01:39 [slex] Yuring (Pok Dro Fem Neu), 2598 points, T:1924, killed by an icy touch 01:51 -!- LarienTelrunya has joined #hardfought 01:52 LarienTelrunya: mail daemon viruses doesn't contaminate, but rather, gives the same sick condition as the sewers, allbeit temporarily 01:52 morning FIQ! what does that condition do? 01:52 which stops natural HP and Cha recovery (Cha in zapm works somewhat similar to pw) for some time 01:53 but wasn't there some side effect of the sewers that kills you if you don't have an air filter or something? 01:53 yes.. the sickness condition 01:53 it doesn't kill directly 01:53 but makes it much more dangerous 01:53 since HP regeneration is frozen 01:53 I did it regardless in my winning game yesterday because I hadn't found a means of air filtering at that point 01:54 I had a tissue regenerator implant though so I still recovered some HP 01:54 ah, so it's not impossible 01:55 sewers was probably the hardest part for me that game 01:55 some monsters there can hurt you badly 01:56 you don't want that with bad HP regen :P 01:56 and I guess it's even worse with no HP regen at all? 01:57 hi 01:57 Hello bug_sniper, Welcome to #hardfought 01:57 so can you watch people play zapm? 01:58 yes 01:58 via dgl or hterm 01:58 can play via hterm also 02:34 I dunno how to say this... but I'm playing nethack from hospital again 02:36 what did you do 02:37 -!- elenmirie has quit [Ping timeout: 248 seconds] 02:37 my ankle wound hasn't healed properly and got infected 02:39 -!- raisse has quit [Ping timeout: 252 seconds] 02:39 how badly? 02:40 so I have had the pleasure of my wound being scraped clean and being hooked up to drip-feed antibiotics 02:41 lovely 02:41 not too bad, but infections of wounds like this have to be taken seriously (think gangrene, amputations) 02:43 -!- bug_sniper has left #hardfought 02:43 -!- bug_sniper has joined #hardfought 02:43 -!- bug_sniper has left #hardfought 02:43 -!- bug_sniper has joined #hardfought 02:57 actually, the scraping out the would bit made me feel for Jamie Lannister in the Game of Thrones (when the Meister cleans infection out of his arm-stump). Like my pain was maybe 0.5% of what he would have suffered 02:57 [slex] Yuring (Pok Dro Fem Neu) killed the Kisuke, on T:1095 03:00 could have used maggots 03:00 painless 03:08 [slex] Yuring (Pok Dro Fem Neu), 1688 points, T:1152, quit 03:10 Is Kisuke something need mention？ 03:11 it's a boss monster; the bot mentions every time a boss is killed 03:15 -!- raisse has joined #hardfought 03:15 -!- mode/#hardfought [+v raisse] by ChanServ 03:25 -!- raisse has quit [Ping timeout: 248 seconds] 03:27 -!- raisse has joined #hardfought 03:27 -!- mode/#hardfought [+v raisse] by ChanServ 03:34 wear a green spell helmet and it autocurse 03:35 yeah, they do that; they also make the entire dungeon colored in a beautiful fleecy green, and casting a spell while having green spells will cause map amnesia 03:35 -!- firemonkey has quit [Read error: Connection reset by peer] 03:35 :( 03:40 -!- raisse has quit [Ping timeout: 248 seconds] 03:45 K2: I hope you never become my doctor 03:48 ;) 04:10 LarienTelrunya: I saw a grass cloak in last game. It's described making green monster become pet. 04:10 Any co 04:11 yeah, it does that; the specifics are that any green or bright green monster that is spawned while you wear the cloak has 2% chance of being a pet. 04:13 But green spell not make them real green, right? 04:13 no, green spell cloak only changes what they look like; for the grass cloak, foxes are e.g. still red and thus ineligible 04:45 Saw a tool shop that tools embedded in Wall 04:45 Have to kick them down 04:47 yeah, you can kick items in walls to move them to your feet; just make sure you don't accidentally kick the items out of the shop without paying or you'll anger the shopkeeper :) 04:51 -!- elenmirie has joined #hardfought 04:51 -!- mode/#hardfought [+v elenmirie] by ChanServ 06:06 back 06:06 "surgery" done 06:06 simple process, was just off for a few hours 06:07 huh, Grasshopper was in hospital too 06:07 allthough his thing seems far more critical 06:07 Grasshopper: hope everything gets osrted out 06:08 LarienTelrunya: in caves of qud, night vision makes everything green 06:08 it's horrible 06:09 well, slex's green spell effect looks fleecy IMHO, I don't know about Qud 06:11 well in slex' case it is an interface screw 06:11 how would you like the entire screen being green 06:11 as a normal "beneifical" effect 06:11 it's horrible :P 06:11 *beneficial 06:13 do you have a screenshot of that? 06:16 of caves of qud?`not really 06:26 FIQ, thanks 06:26 Looks like I'll be in hospital another 2 nights at least 06:29 There's a TV but everything is in Thai (yes I speak Thai but there's only so much Thai TV I can take). However, the view over the city through my window is A-maaaaaaaaa-zing 06:31 Luckily I have my Kindle so can read books on that 06:48 ok 06:49 heh 06:49 I remember as a kid before the benefit of handheld devices and such 06:49 if there was a tV in first place 06:50 it would be near the ceiling a very small thing impossible to see (vision impairment) 06:50 it was 06:50 booooooooooooring 06:50 during these times 06:54 i was just reading books 07:06 the idea was probably that if the TV is at the ceiling, it makes us of otherwise unused space 07:17 FIQ, I rarely watch TV much at all, even when it is in English 07:18 I'd be happy if I could use my laptop more (play NH!), but I can only manage sitting up for fairly short periods 07:18 and am meant to be horz in order to help my foot/ankle heal 07:19 -!- noty has quit [Quit: WeeChat 1.9] 07:19 what I need to invent is a bed-tray that can be angled to suit someone lying down 07:19 with something to stop your laptop or whatever falling off 08:26 -!- LarienTelrunya has quit [Quit: Page closed] 09:18 [4k] Zappi (philbo) (Wiz Elf Fem Cha) heard a rumor from The Oracle, on T:3507 09:22 [4k] Zappi (philbo) (Wiz Elf Fem Cha), 12993 points, T:3681, killed by a gray unicorn 09:27 -!- stennowork has quit [Quit: Leaving] 09:34 -!- rikersan has joined #hardfought 09:35 [dnh] rikerw (Pri Dro Mal Neu), 6796 points, T:3146, quit 09:38 [dnh] rikerw (Pri Dro Mal Neu), 32 points, T:378, killed by a water demon 09:49 [dnh] rikerw (Pri Dro Mal Neu), 3414 points, T:772, killed by an arrow 09:52 [nd] sluggoman (Wiz Hum Mal Neu), 372 points, T:933, killed by a gnome zombie 09:56 -!- rikersan has quit [Quit: Lost terminal] 10:06 [nd] sluggoman (Wiz Hum Mal Cha), 1025 points, T:1724, killed by a killer bee 10:16 -!- ais523 has joined #hardfought 10:25 [dnh] rikerw (Hdr-Pri Dro Mal Neu) entered Minetown for the first time, on T:1190 10:27 -!- rikersan has joined #hardfought 10:30 -!- ais523 has quit [Remote host closed the connection] 10:30 -!- ais523 has joined #hardfought 10:37 -!- rikersan has quit [Quit: Lost terminal] 10:55 -!- rikersan has joined #hardfought 11:01 -!- rikersan has quit [Quit: Lost terminal] 11:05 so, I just found out that killer_xname in 3.6 actually included a concession to devnull and other tournaments with unique deaths 11:06 it will hardcode your death to "choked on a deadly slime mold" regardless of fruitname 11:08 quite a few variants have unique deaths fixes like that 11:13 is it just assumed that tournaments are expected to filter out monster names for unique deaths? 11:14 aosdict: Tournaments don't necessarily _want_ to filter out monster names from unique deaths; although if they have slex in them, it becomes a bit of a pressing issue. 11:15 (Though slex itself has made changes to make that easier.) 11:15 Not monster names like "jackal", but stuff like "jackal called escaped (with the Amulet)". 11:16 ...actually, I don't think you can escape with the Amulet except in disgrace, but whatever. 11:16 I assume junethack at least will treat "killed by a little dog" and "killed by a little dog called Hachi" the same for unique deaths purposes. 11:17 -!- ais523 has quit [Quit: It seems most convenient to apologise for my connection in the quit message, given how often it comes up… If I immediately reconnect, it's probably because I could send but not receive.] 11:17 -!- ais523 has joined #hardfought 11:19 aosdict: Oh, that. 11:19 Yeah, those should be marked someway; NH4 puts them in quotation marks, IIRC. 11:22 I'm doing some analysis on NAO data, and the many deaths due to named items, named fruit, and named pets are quite annoying. 11:22 Hmm, there are two ascensions in less than 10 turns. Interesting 11:22 one is paxedtest 11:23 ais523: what do you think is a feasible number of turns greater than 2000 for which a legitimate ascension is possible? 11:23 since you're trying to find the smallest such number 11:23 aosdict: the other one overflowed the turn counter, I believe 11:23 -!- ais523 has quit [Remote host closed the connection] 11:23 -!- ais523 has joined #hardfought 11:23 it was really awkward as the game kept getting killed by paxed/dtype because they thought it was a stuck process 11:24 given that it was running at 100% CPU for a long period of time 11:24 yes, it's a negative turncount 11:24 ais523: there's a process that automatically does that 11:24 Was that on 32-bit, I hope? 11:25 jonadab: NAO is 32-bit, I think 11:25 brb 11:28 ais523: _still_? Interesting. 11:28 Hmm, maybe for save compatibility, I guess. 11:29 it might be a 32-bit VM? 11:29 Actually, yeah, as soon as I say it, that makes a lot of sense. 11:29 Because going to 64-bit would likely break save compat. 11:29 ais523: The VM moved not long ago, from Amazon US west-coast to Amazon US east-coast. 11:29 A year or so ago, I guess. 11:29 32-bit NetHack runs on 64-bit systems, though 11:30 you need to install 32-bit versions of its dependencies but apart from that it just works 11:30 And anyway, I doubt if Amazon _provides_ 32-bit only VMs. But the OS could be 32-bit, potentially. 11:30 I have a 32-bit NetHack build for TAS testing 11:30 Ah. 11:30 I thought for TAS you used the old DOS build of 343. Or are you going to do a 3.6 TAS? 11:31 OS is 64 bit, chroot is 32 bit libraries for nethack 11:31 Ah. 11:32 jonadab: the actual playback is on DOS, but DOS is a terrible platform for doing research on 11:32 so we have a Linux version of NetHack set up to act identically to the DOS version (except for some added cheat codes) 11:32 ais523: Wait, does the DOS version use the same RNG as the *nix version? 11:32 e.g. we replaced the RNG with the algorithm used on DOS 11:33 Ah, I see. 11:33 jonadab: and 343 is 32 bit, 360+ is 64 bit 11:33 on nao 11:33 Custom build. 11:33 dtype: Ah, that _does_ make sense. 11:33 and yes, for old compat 11:33 strangely enough, the RNG algorithm used by DJGPP also happens to be provided in the NetHack source as an alternative RNG algo for use on platforms with bad-quality RNGs 11:33 so all we needed to do to use it was to flip a #define 11:33 Interesting. 11:33 I think I've been told that before and forgot, because now that you say it, it sounds familiar. 11:33 jonadab: I think nao has had 343 save compat for a decade or some such 11:34 dtype: Right, since the move to 343, I believe. 11:34 dtype: this is why no #overview. 11:35 I just hadn't previously thought about the fact that that would necessitate still using a 32-bit build of it. 11:35 But yes, with the 3.x save system, it would. 11:46 I always thought of the 32 bit counters to be kind of part of the whole 343 experience. :) 11:51 NetHack is more internally consistent with 32-bit variables 11:51 mostly because it was set up as long=32 because it used to run on some int=16 platforms 11:51 in retrospect it should probably have used defines or typedefs for variable sizes 11:52 Hi 11:52 Hello FIQ, Welcome to #hardfought 11:53 I am increasingly of the opinion that C's char/short/int/long stuff is very badly designed; that int should exist because sometimes you just want whatever takes up one machine word on the platform you're on, but the others should not exist or should have specific numbers of bits in the spec. 11:54 Because yes, a lot of old software is using a particular one of those _in order to get a certain number of bits_. Which is bad and wrong. 11:54 talking about NH3's nonportable save compat reminds me of the hack bhaak had to go through 11:54 to keep save compat after a major server upgrade 11:54 So there should've been int, which is a machine word, and int8, int16, int32, int64. 11:54 and that was just for *changing the user id* 11:55 This was with Un? 11:55 allthough that wasn't quite the same thing 11:55 jonadab: bhaak made a server upgrade on un's server, the server has a ton of un versions kept around for save compat reasons 11:55 Sure. 11:55 and nethack has a check that you are running as the proper user id before it allows you to run the save 11:55 or it will delete it 11:55 Ah. 11:56 So he probably, for server purposes, wanted to just #ifdef that out. 11:56 and removing this check was not feasible with the amount of versions that was around, I'm not even sure if bhaak knew the exact commits 11:56 Ah. 11:56 so he resorted to hacking the check out of the binaries 11:56 The plot thickens. 11:56 Yeep. 11:57 on nao nethack is always just user #5. 11:57 games. :) 12:00 Heh. 12:00 I know for a fact that NH4 does not have this check, because people send me save files all the time so I can fix bugs. 12:01 So there should've been int, which is a machine word, and int8, int16, int32, int64. ← Rust works like that: isize, i8, i16, i32, i64 12:01 although isize is the size of one /pointer/ because "machine word" isn't always uniquely defined 12:01 Well, "all the time" meaning "a few times a year. Mostly in june. 12:01 (whereas the pointer size lets you know what, say, the maximum array index is) 12:01 ais523: Ok, that's not unreasonable. 12:02 I was also thinking register size, but it's conceivable that some platforms have more than one register size. i686, for instance, has I think three register sizes. 12:02 ais523: btw, one thing I don't understand 12:02 *why* does my save revisioning work? 12:02 jonadab: modern x86_64 has a huge number of registers of varying sizes 12:02 afaik, something I was only made aware of later, the game will sanity check a save at first by reloading, saving again, and reloading 12:03 this would obviously break if you make an "userspace" revisioning system 12:03 so why does it work? 12:03 In the case of x86, the largest available register size would obviously be the one you'd use, because the smaller ones are sub-registers that exist for backward compatibility. But a platform could exist with ... yeah, with that. 12:03 the largest general-purpose ones are 256 bits, I think? 12:03 FIQ: there's a maybe_old_version flag used in various parts of the save system 12:04 that's designed to allow for saves becoming slightly different from version to version (including different commits) 12:04 ahh 12:04 so this is flag used on initial load? 12:04 *so this flag is used on initial load 12:05 I can't remember the details and it's even possible they don't work 12:05 well I know at least one bug 12:05 when I tripped the internal version check, instead of erroring, the game segfaulted because it attempted to save a null gamestate 12:06 I mentioned this a while ago afaik 12:07 -!- stennowork has joined #hardfought 12:08 jonadab: do you think it would be reasonable 12:08 instead of books having "charges" 12:08 you can only read them once and they then disappear 12:08 but you get 3x the memory? 12:08 so you don't have to do spellbook micromanagement 12:08 I don't see how this micromanagement makes the game interesting, just annoying 12:09 FIQ: I would not, as a player, complain about that, on the _condition_ that spellbooks are common enough you don't need read-identified books as polyfodder. 12:09 well, both fh and 4k addresses this, allthough differently 12:09 4k spawns statues when trice are involved 12:09 Which you might be able to achieve by having monsters that get spells get the book slightly more often. 12:09 this makes books significantly more common 12:10 fiqhack spawns books on spellcasters 12:10 It only happens at level gen time, but yes. 12:10 from the spells they get generated with 12:10 I am not sure if Fourk's changes here are enough. 12:10 (meaning the later into the game you are, the more highlevel spells you'll get this way) 12:10 But FIQHack's should be, and if not, you can probably just adjust the monster-with-spell-also-gets-the-book probability slightly upward. 12:10 FIQ: the main positive effect of the whole spellbook charge thing is to cause there to be a tradeoff about whether you stash spellbooks or whether you polypile them 12:11 ais523: polypiling is a bug 12:11 er 12:11 this isn't a hugely large effect, so I can see wanting to change it 12:11 the way book polypile work, taht is 12:11 ais523: In practice, you stash ones you care about, like magic missile and identify, and polypile the other 95% which are worthless spells. 12:11 *that 12:11 there is a bug in book polypiling but this would apply even if the bug didn't exist, I think 12:11 if it takes 20k turns to un-learn a spellbook, one could require the player to just obtain a new spellbook after 20k turns 12:11 ais523: you read a spell book once 12:11 1 charge exhausted 12:11 now you poly it 12:12 1 charge exhausted 12:12 i like the: read once, book vanishes approach 12:12 now you read the new spell 12:12 and the book is now used up 12:12 and it can be poly'd as often as you like 12:12 would simplify things greatly 12:12 this isn't that major of a strategy I feel 12:12 stennowork: That makes read-identifying spellbooks extremely suboptimal. 12:12 (not @ stennowork but @ what I was going on) 12:13 jonadab, why? 12:13 stennowork: that's interesting, but I feel it would hurt people who play very long games 12:13 ais523: that is the thing 12:13 stennowork: Because most of them aren't the spells you want, so you'd want to poly them instead. 12:13 I don't want to do this 12:13 it could probably be fixed just by increasing spellbook drop chance, though 12:13 lets increase it to 30k turns? 12:13 so I wanted to just give them the 60k duration 12:13 (either the way FIQhack does it or simply by changing the deathdrop formula) 12:13 read-ID unknown spellbooks and having them vanish sounds fine for m 12:13 e 12:14 note that you'd have to increase the rate spellbooks drop over time, rather than the rate at which they generate with levels, as long games have more drops-over-time 12:14 Ultimately, I think I like FIQHack's solution to spellbook availability. It even manages to scale spell level with danger level. 12:14 Mostl. 12:14 the way fh does spellbooks with casters is one of my more successful balance changes I think 12:14 *Mostly. 12:14 I'm quite fond of it myself 12:14 but yeah 12:14 assuming it works like fiqhack 12:14 then I think ^ is a reasonable change 12:14 and removes tedium 12:15 i just rushed into the discussion here i have to admit v_v 12:15 i will read the backlog first 12:15 OK, my verdict on the change you mention is that it's a good change but only if combined with FIQhack's existing change to spellbook gen, or some other change with a similar effect 12:16 FIQ: I agree that making spellbooks read-once reduces tedium, if and only if enough spellbooks are available that the player doesn't feel compelled to polypile all the undesirable-spell ones. 12:16 (Before reading them.) 12:16 but balancing around read-ID only really makes sense for illiterate, and there reading spellbooks isn't an issue anyway 12:16 i think 12:17 um 12:17 illiterate users can't read books anyway :P 12:17 are you referring to some ID-less conduct rather 12:17 illiterate users don't have any _use_ for books, AFAIK. 12:17 yes. thats why i said that reading spellbooks isn't an issue anyway 12:18 hmm 12:18 i think the risk is fine for reading unIDed spellbooks 12:18 the risk that it would vanish 12:18 errr 12:18 in this case, the risk with reading an unIDed spellbook would be the risk that you could have better used the book for something else 12:18 atm I use 30000 as a magic number for "permanent knowledge" 12:18 the risk that you learn a useless spell and lose your book 12:19 this was clearly very poor forethought of me 12:19 this is similar to the risk with quaffing an unIDed potion that you've already dip-tested 12:19 (in fiqhack, if you are crowned, you get your role's bonus spell permanently) 12:19 stennowork: If you have a wand of polymorph, the risk of losing your chance at a much better spell would be significant, at least to me. 12:19 Especially if the books are rare. 12:19 Like in 3.x 12:19 you have to identify first before poly 12:19 if you have wands of poly 12:19 you can probably make ID scrolls 12:19 use them 12:19 would you poly price-IDed potions? 12:19 stennowork: yes, and that's tedious to do. 12:20 Unless you can cast ID. 12:20 hmm 12:20 Because you end up stashing all your books until you have 30 of them to hold at once while you read a blessed scroll of ID. 12:20 And price-identifying them to see which ones are _likely_ to be the ones you want, etc. 12:20 jonadab: the tradeoff there is not getting the spells until later 12:20 "I know I can poly all these 500+ zorkmid books because I don't care about those spells as a Valkyrie." 12:21 ais523: Yes, I know. 12:21 Which is why it's only relevant if the books are rare like in 3.x 12:21 ais523: btw, not sure if I pointed this out 12:21 If books are fairly common, the motivation to do that goes away. 12:21 but I fixed a bug a long time ago regarding menu page handling 12:21 you mentioned it, I think? 12:21 jonadab apparently thinks it's a *feature* 12:21 oh, you didn't mention that 12:21 do you consider it a bug or feature? 12:22 yes, because I suggested jonadab to port it yesterday because the issue was annoying me in 4k :P 12:22 I can't remember what it is now 12:22 basically 12:22 this 12:22 http://home.fiq.se/4k.png 12:22 FIQ: it wasn't me, it was octe, un.nethack.nu's admin. he wrote it up here: https://unnethack.wordpress.com/2014/09/15/saving-save-games/ 12:22 in fiqhack, instead of that ^ happening 12:22 the scroll will "snap" back up to show a full page 12:23 TL;DR: Patched 47 binaries to achieve backwards compatibility with old saves 12:23 there's also a real bug going on but thats a seperate issue that should be fixed regardless, off-by-1 error: http://home.fiq.se/4k2.png 12:23 (the former is what jonadab argues to be a feature) 12:23 what do you think? 12:23 bhaak: ah, I thought you made that post 12:23 it wasn't clear 12:23 sorry :) 12:23 FIQ: OK, so the main issue with snapping up is that you ideally want pgup and pgdn to be opposites 12:23 but yeah I was referring to that 12:24 and that's not the case after the fiqhack fix 12:24 you can press pgdn, then pgup, and end up somewhere else 12:24 the alternative would be a bit weird I feel 12:24 making pageup go back to the previous state, suddenly not scrolling a full page 12:24 also it means that if you're near the bottom and pgdn, you then find it hard to work out what text is new and what was already shown on the page 12:24 that annoys me when using file viewers, often 12:24 this is the exact thing jonadab brought up actually 12:25 those are the only two drawbacks I can think of, though 12:25 getting lost 12:25 in the menus 12:25 and the alternative also has issues 12:25 ais523: my issue with the non-fiqhack behaviour is that if I know something is near the bottom 12:25 I scroll down to the bottom and work from there 12:25 with the old behaviour I can't do this 12:26 can't you just press end? 12:26 that'd scroll to exactly the place you want in one keystroke 12:26 hmm 12:26 I guess I didn't know 12:26 but does it do the right thing outside fiqhack 12:26 let's see 12:27 yep 12:27 it does 12:27 hmmm 12:27 that was my main issue wit hthe thing :P 12:27 *with the thing 12:28 so maybe it is not a good change? 12:28 also, as for file viewers, most I've used in open/save dialogs and such scroll sideways 12:29 so it works akin to non-fiqhack 12:29 I mean things that view files (less, evince, etc.), not things that view lists of files 12:29 ah 12:30 ais523: ok so if we go under the premise that it is a bad fix 12:30 that leaves the bug I linked where nothing shows 12:30 one issue with fixing this 12:30 I quite like the fact that when you are at the end 12:30 there is an extra row at the bottom 12:30 to signify "end of list" 12:31 you'd have to remove this, or come up with a good way for how pageup/pagedn should behave on this corner case, to fix the above 12:32 right, so the problem here is that if you want an end-of-list indication, but the list is a whole number of pages long, where do you put it? 12:32 I suppose one fix is to show the last non-empty line and nothing else, similar to if you scrolled down from 1 row above, but that also "desyncs" pagedn+up 12:32 about the only thing I can think of is to make the end-of-list indication be nonempty 12:39 hmm 12:40 so dgl uses COLS/LINES in the watchmode menu to figure out how large your terminal is 12:40 I suppose this is why it is the only screen in the entire dgl to actually notice if your terminal size changed 12:41 allowing ingame resizing would probably be awkward with ttyrec handling 12:41 but at least it should be able to notice if size changed before starting a game 12:42 and update stuff appropriately 12:42 is this dgl we're talking about? NH4 derivatives handle resizing just fine 12:42 and there's no reason why vanilla couldn't 12:42 yes, that is why I am mentioning dgl multiple times :P 12:43 right, that's what made me consider it 12:43 I didn't think dgl cared about the codes sent in-game at all, though 12:43 is this just to update the size shown on the watchmode menu if the game is resized in-game? 12:44 it will do weird things if you attempt to start a game, NH4 or otherwise, with a terminal size that changed from that you first connecte 12:44 d 12:45 http://home.fiq.se/fail.png 12:45 this happens if you fire up 4k (and any other game) while you changed terminal size before you started the game, but while remaining connected to dgl 12:45 the only thing that will correctly know your term size after a resize is the watch menu 12:46 which will highlight stuff in red or not appropriately 12:46 I have no idea why 12:46 oh that channel still exists, who would've thought 12:46 what channel? 12:47 a social one in that list, i think i am even +F 12:47 what list 12:47 sorry, in the channel list of the screenshot you just sent 12:47