00:12:54 heh I think I'm done playing fe in tournaments 00:13:03 this time my one game was gnwr 00:17:16 lol I just updated CAO scoring so that the game grids don't show empty rows/cols with obsolete species/class, and somehow managed not to mark MD obsolete 00:44:48 any thoughts on why gdb would be taking up 100% of the cpu on cbro? 00:45:16 a bad crash? 00:45:22 looks like the process has been running for 300 hours 00:45:46 I would kill -9 it 00:45:46 or maybe that's minutes 00:45:56 wasn't sure if there was a valid reason for it to be running for 300 hours 00:46:04 err. minutes. 00:46:10 no, not at 100% cpu 00:46:17 I mean, it could run indefinitely at no cpu 00:46:58 ok. wasn't sure why it would even start up. is that what runs when a game crashes to generate the debug report? 00:47:04 yes 00:47:12 that's what generates the stack trace 00:47:17 !crash cbro 00:47:19 No milestones for cbro (crash). 00:47:33 !crashlog * cbro 00:47:35 2049. Zeldest, XL27 MuWz, T:114247 (milestone): http://crawl.berotato.org/crawl/morgue/Zeldest/crash-Zeldest-20191109-170138.txt 01:24:36 yeah i just kill it when that happens 01:24:36 floraline: You have 2 messages. Use !messages to read them. 01:37:43 Unstable branch on crawl.develz.org updated to: 0.25-a0-87-ga3d287b (34) 02:08:58 Windows builds of master branch on crawl.develz.org updated to: 0.25-a0-87-ga3d287b 02:59:28 Monster database of master branch on crawl.develz.org updated to: 0.24-a0-443-g80245de 03:11:56 Unstable branch on crawl.beRotato.org updated to: 0.25-a0-87-ga3d287b (34) 03:31:22 Fork (bcrawl) on crawl.kelbi.org updated to: 0.23-a0-2708-g79730e94b9 03:53:48 Fork (bcadrencrawl) on crawl.kelbi.org updated to: 0.22.1-237-gd83e9b5da0 06:19:41 -!- amalloy_ is now known as amalloy 06:47:20 03Aidan Holm02 07* 0.25-a0-88-g1777079: Fix redraw issue when game ends (Goratrix, 12124) 10(4 minutes ago, 1 file, 10+ 2-) 13https://github.com/crawl/crawl/commit/177707934945 06:58:46 03Aidan Holm02 07* 0.25-a0-89-g355392f: Partially restore backspace-clears-name behaviour (advil, 12119) 10(4 minutes ago, 1 file, 7+ 0-) 13https://github.com/crawl/crawl/commit/355392fac7f9 07:02:32 Unstable branch on crawl.kelbi.org updated to: 0.25-a0-89-g355392fac7 (34) 07:09:51 03Aidan Holm02 07* 0.25-a0-90-g0e808ab: Fix backspace-clears-name behaviour 10(3 minutes ago, 1 file, 1+ 0-) 13https://github.com/crawl/crawl/commit/0e808ab1dd84 07:17:26 Unstable branch on crawl.kelbi.org updated to: 0.25-a0-90-g0e808ab1dd (34) 08:01:08 OAOBOA 08:29:49 -!- amalloy is now known as amalloy_ 09:34:03 Add support for changing UI image size 13https://crawl.develz.org/mantis/view.php?id=12126 by aidanholm 10:46:27 03Aidan Holm02 07* 0.25-a0-91-g23bb404: Fix overenthusiastic nullptr find-replace 10(47 seconds ago, 2 files, 5+ 5-) 13https://github.com/crawl/crawl/commit/23bb404d27bd 10:56:55 The codeblock that starts at 3848 in tags.cc deletes what the player has seen for the sake of acquirement and good_item generation, right? (I think it does but I'm not sure). 10:58:33 !source tags.cc:3848 10:58:34 https://github.com/crawl/crawl/blob/master/crawl-ref/source/tags.cc#L3848 11:00:03 oof sorry I should have checked that line number with mainline since my fork is still 0.22 based...I'll find the line in mainline 11:00:38 https://github.com/crawl/crawl/blob/master/crawl-ref/source/tags.cc#L3961 11:00:45 there it moved down about 130 lines. 11:02:50 Unstable branch on crawl.kelbi.org updated to: 0.25-a0-91-g23bb404d27 (34) 11:06:19 Bcadren: looks like it fills out you.seen_weapon and you.seen_armour, yep 11:09:17 Thanks. 11:09:31 no worries 11:10:03 advil: I remember you mentioning mantis being painfully slow; I'm guessing that's due to it synchronously sending emails? 11:10:51 separately, it also seems to have only a single worker process, as my GET requests seem to be blocked by issue updates 11:11:48 https://www.mantisbt.org/wiki/doku.php/mantisbt:setting_up_mail_queuing might be worth trying 11:24:34 Stable (0.23) branch on underhound.eu updated to: 0.23.1-91-gf373564dc4 11:31:13 If there's an X next to a commit what does that mean? 11:53:15 aidanh could be, it did seem to me that there was a point where it suddenly became very slow 11:54:06 aidanh unfortunately only nap.kin has access to do anything with mantis on the server and since I think at this point it would be a pretty major project to do any updating, no one has really wanted to bug him about it 11:54:42 not really even sure that upgrading (often a prereq for things like that, though I haven't checked this one) would be feasible, it's extremely far behind current versions 11:56:49 right, that was pretty much my current understanding of the situation, although I thought gammafunk had some access 11:56:55 I don't think that code block *deletes* what the player has seen, it loads it from the save and clears out anything beyond NUM_WEAPONS 11:57:26 NUM_WEAPONS as recorded in the save that is (so the setting to 0 is basically only for weapons that are new since the save) 11:59:26 although that version of mantis is indeed very old, it does support mail queueing 11:59:40 who curates `!nick bot`? 12:06:24 Unstable branch on crawl.akrasiac.org updated to: 0.25-a0-91-g23bb404 (34) 12:26:59 this should be a bit more usable now: http://crawl.akrasiac.org/scoring/fastest-wins-time.html (modulo the repeated games issue of course) 12:32:34 does anyone know why `fastman` is called a bot? it's not extremely obvious from the logs, and e.g. the current rc for that account on cbro looks pretty non-bot-like to me (http://crawl.berotato.org/crawl/rcfiles/crawl-git/FastMan.rc) 12:32:52 (from the name I would have guessed it's just a manman alt?) 12:34:33 the times are also pretty reasonable rts times 12:40:15 Is this the commit that accidentally halved the damage of Ozo's for like six months? fb95405 12:43:07 because if so if I cherrypick that one, I'll ahve to at least make it to the fix before I stop (cherrypicking my fork forward) 13:00:42 DD can't use Heal ability when starving, which is weird, because ability has no hunger cost. 13:07:21 I feel like that issue has been raised before and wasn't fixed then. BUT...it'd be pretty easy to find it in ability.cc and add the abflag::starve_ok flag if people agreed it was appropriate. 13:14:23 What's the cherry-pick decision re the backspace clearing? I think we should unless there's a reason not to I'm not seeing 13:15:44 yes, we should if it is what I think it is 13:16:21 there's a bunch of related things to cherry pick too 13:16:47 I think I've got them all 13:17:25 !time 13:17:26 Time: Nov 10, 2019, 06:17:25 PM, UTC. The 2019 0.24 tournament ends in 1 hour, 42 minutes and 34 seconds. 13:18:23 re abilities when starving, the rule of thumb seems to be either instant things, or things that can directly lead to nutrition 13:20:49 two exceptions are batform and walljump, not sure I can justify batform but walljump is basically for parity with the movement-triggered version 13:23:46 oh, do vampires even get hungry now? 13:24:03 Yea I remembered that it was brought up and turned down before, but not why. >_> 13:34:01 !time 13:34:02 Time: Nov 10, 2019, 06:34:02 PM, UTC. The 2019 0.24 tournament ends in 1 hour, 25 minutes and 57 seconds. 13:35:00 interesting. that cherry-picking from master doesn't seem to reduce the count of "commits behind master". 13:36:41 no, you'd need to rebase from master to do that 13:36:46 cherrypicks don't count 13:40:17 Yea and rebasing would screw with save compatability and the server ability to rebuild or something like that (Floraline told me not to do it anyways) 14:06:24 huh, there was one where I had to combine the changes on one line; a conflict I actually understand why conflicted. neat. 14:07:23 Bcadren: I remind you again not to use this channel as your dev blog 14:08:50 -!- aidanh_ is now known as aidanh 14:38:43 Huh, interesting: abusing multiple Hep wraths to get low XL win. 14:59:16 Out of curiosity could anything in my fork currently be considered for mainline? (I know I straight up have some stuff that's on the WONTDO, but otherwise). 15:07:02 03ebering02 07* 0.25-a0-92-g20dcb12: Remove edge cases from Fedhas protection 10(51 minutes ago, 2 files, 4+ 5-) 13https://github.com/crawl/crawl/commit/20dcb12bf155 15:07:34 cherry-picks incoming 15:08:07 03advil02 {ebering} 07[stone_soup-0.24] * 0.24.0-14-gdd00188: Remove an accidental swastika (Snack, minmay) 10(13 days ago, 1 file, 0+ 16-) 13https://github.com/crawl/crawl/commit/dd00188388c3 15:08:07 03advil02 {ebering} 07[stone_soup-0.24] * 0.24.0-15-g9656144: Fix some odd permastore update conditions (aidanh) 10(13 days ago, 1 file, 5+ 1-) 13https://github.com/crawl/crawl/commit/9656144a79d0 15:08:07 03gammafunk02 {ebering} 07[stone_soup-0.24] * 0.24.0-16-g30ed930: Rework a worm vault (again) 10(9 days ago, 1 file, 50+ 28-) 13https://github.com/crawl/crawl/commit/30ed930bd26d 15:08:07 03ebering02 07[stone_soup-0.24] * 0.24.0-17-g9cefc0a: Check appropriate tracers in monster wand usage (ManMan) 10(5 days ago, 1 file, 14+ 5-) 13https://github.com/crawl/crawl/commit/9cefc0aa56e8 15:08:07 03advil02 {ebering} 07[stone_soup-0.24] * 0.24.0-18-g186447a: Don't crash on a few cases of invalid skill names in lua 10(4 days ago, 1 file, 25+ 7-) 13https://github.com/crawl/crawl/commit/186447a58146 15:08:07 03Aidan Holm02 {ebering} 07[stone_soup-0.24] * 0.24.0-19-g81019e2: Remove weird shift-X binding 10(4 weeks ago, 1 file, 1+ 10-) 13https://github.com/crawl/crawl/commit/81019e2d4dfc 15:08:07 03Aidan Holm02 {ebering} 07[stone_soup-0.24] * 0.24.0-20-g9a60b66: Remove broken back-compat keybindings for hints menu 10(4 weeks ago, 1 file, 3+ 13-) 13https://github.com/crawl/crawl/commit/9a60b66c946e 15:08:07 03advil02 {ebering} 07[stone_soup-0.24] * 0.24.0-21-g26103b2: Improve behavior of `remember_name = false` 10(3 days ago, 5 files, 20+ 8-) 13https://github.com/crawl/crawl/commit/26103b220fcf 15:08:07 03advil02 {ebering} 07[stone_soup-0.24] * 0.24.0-22-g02e3af1: Don't update travel cache on excursion to unvisited levels (odiv) 10(2 days ago, 2 files, 18+ 3-) 13https://github.com/crawl/crawl/commit/02e3af1bf10f 15:08:07 03advil02 {ebering} 07[stone_soup-0.24] * 0.24.0-23-gf00d1e0: Don't crash if there aren't enough spells for an acquired randbook 10(26 hours ago, 1 file, 4+ 1-) 13https://github.com/crawl/crawl/commit/f00d1e050beb 15:08:07 ... and 5 more commits 15:11:28 03ebering02 07* 0.25-a0-93-g1bb4612: Revert "Add a trunk reminder for the 0.24 tournament" 10(23 seconds ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/1bb46122f920 15:18:00 Unstable branch on crawl.kelbi.org updated to: 0.25-a0-93-g1bb46122f9 (34) 16:18:54 Did anything ever happen with multi-floor vaults that aren't encompass vaults; I remember someone talking about wanting to do a sunken ship vault in shoals, but had issues with making sure the stairs would connect with each other like they wanted, was that ever resolved or just abandoned? 16:24:51 -!- amalloy_ is now known as amalloy 16:43:04 it was never completed 16:54:42 Does it still exist as a fork somewhere? (I think it was PleasingFungus that was working on that? but geez it's been so long) 17:07:03 your guess is as good as mine. the best way to dig up ancient crawl stuff is to use search tools 17:09:55 Well team, huzzah on 0.24 17:10:18 I updated the GH wiki with a 0.25 plan, mostly to get my magic ideas down and start the conversation 17:14:07 03Umer Shaikh02 {ebering} 07* 0.25-a0-94-g9e23508: Make the game check the floor when emergency eating 10(3 weeks ago, 1 file, 23+ 8-) 13https://github.com/crawl/crawl/commit/9e23508adf36 17:17:25 Unstable branch on crawl.kelbi.org updated to: 0.25-a0-94-g9e23508adf (34) 17:19:59 That's actually something I wouldn't know how to search for... 17:54:16 !tstats 17:54:39 Stats after 16 days (t): 2880 players, 1046 runers, 540 winners, 1930 wins, 63048 games, winrate 3.06%, total player time 4y+15d+22:29:44. 18:09:20 Bcadren: you could try manually looking through all branches with the most recent commit by PF 18:09:27 Bcadren: there are scripts on stackoverflow to do that 18:13:06 alexjurkiewicz: do you know anything about how curates the bots nick? Iit looked like for your scoreboard that's where you got your bot list? 18:13:11 *bot nick 18:13:22 I think the bot nick is honor system curated? 18:13:34 and/or if a dev notices a bot 18:13:36 I've just never even seen it edited in here 18:13:44 or someone asks a server admin if they can run a bot 18:14:03 crawl botting isn't so exciting these days 18:14:09 maybe it gets edited in the sequell channel 18:14:12 I don't guess it matters, if it had finished and there was a way to have multi-level vaults that would be cool...but if I'm right and PF was working on it and failed, I doubt I could finish it... 18:14:19 yeah it's all qw clones 18:15:46 but I implemented a bot filter for cao scoring, and was trying to make sure I had a reliable list 18:22:38 is qw that much faster than human-time runs? 18:22:50 Unstable branch on underhound.eu updated to: 0.25-a0-94-g9e23508adf (34) 18:23:56 advil: we built the bot list by hand, iirc. Maybe copying from some other source to start with, but we definitely added items by hand 18:24:21 for tournament scoring, you could improve the system by checking for QW-like code in the rc file 18:24:25 oh, hm, ok...it looked a lot like the Sequell one but I didn't look closely so I probably missed some differences 18:25:42 qw is just limited by cpu so yes, it can win in minutes or probably less on a really fast computer; not sure what the lower limit is...it usually runs throttled on servers but even so, the fastest win leaderboard was useless because of bots 18:25:44 of course there could be bots which run client-side 18:26:21 I saw wins in about 3mins on a fast cpu. Most of the time was in message log printing iirc :D 18:27:56 you can see the list I used here: https://github.com/crawl/scoring/commit/ed434600d09c0def6d3a706662c6e8e22dcb6e10 if you care 18:28:32 Do any of the bots have an actually high winrate or just fast time when they do win? I only really remember qw being --Be's and --Fi's but it could be more advanced by now *shrugs* 18:29:52 alexjurkiewicz: is there a tldr on why you dropped that scoreboard project btw? 18:31:29 or any suggestions for things that I should look at that it does that could be worth copying? I've been actively working on incremental improvements to CAO scoring since I got login access 18:31:46 advil: a few: 1) I think we made some bad choices and the system got really complex (eg moving to an external service to provide games rather than relying on logfile parsing) 2) one big goal was to be faster at regenerating data than cao scoring, and it was about equally as fast (with sqlite, to be fair), 3) I realised I didn't really care about scoring 18:32:34 ah I see...yeah one of the interesting things about CAO scoring is just the scale of the full db makes everything hard 18:32:55 a big motivation to start the project was CAO scoring being flaky, and I think your work on that has been fantastic. But the other big one was that CAO scoring is really ugly and is designed for inside baseball nerds, rather than being simpler and more beautiful 18:34:40 yeah, it is a very old-school aesthetic 18:34:46 this suggestion is probably heretical, but I think scoring is better if it doesn't go back to the beginning of time. It's good for the scoreboard to focus on contemporary crawl, even if it means things like the XL11 win or elliptic's streak are left behind 18:35:10 well, I can't say I haven't thought about that, because it would help with the scale problem 18:35:36 yesterday I speculated aloud here as to whether sequell wouldn't be better if it just covered the `recent` timeframe 18:36:04 having more things that are version-specific and get refreshed ever version might be nice too 18:36:05 for one thing !lm would be usable again :D 18:36:15 yeah, that is exactly my main motivation 18:36:45 by "more things" I mean "anything"...but the db is not structured around that 18:37:31 oh! also, giving up on the idea of unifying player names across servers 18:38:08 that one seems more heretical 18:39:04 it's the leading source of griefing in the community 18:39:20 yeah, I realize 18:42:36 but for example that would split someone like Yermak into like 6 accounts 18:43:54 !lg yermak s=src recent 18:43:55 740 games for yermak (recent): 464x cue, 153x cko, 100x cxc, 14x cao, 9x cbro 18:43:59 dang 18:44:09 ok 5 :-P 18:44:19 still not sure what the right solution to that is...I mean in the long run implementing manual account blacklisting for prominent cases isn't too hard, and I actually have that on my list 18:44:25 single sign on 202 18:44:28 2020 18:44:29 heh 18:44:37 I keep having these ideas about federated sign on 18:45:09 there's some kind of backwards compat anger race worry 18:45:10 I would say manual blacklisting is really hard, because you need to keep updating it every year forever. And it's not exactly fun work 18:45:40 we built some heuristics in dcss-scoreboard to detect griefs, it works pretty well but is a losing battle in the long-term 18:46:28 https://github.com/zxc23/dcss-scoreboard/blob/master/scoreboard/scoring.py#L20 18:47:01 interesting 18:47:10 from running that, do you have a sense for how many cases it detected? 18:47:29 I mostly know cases I've heard about directly in this channel, which is like 10 18:48:03 all targeting prominent players or devs 18:49:08 about that number. There were a few cases where griefers seemed to target everyone with >1 active win, but most of those people never complained / weren't high profile 18:50:29 ah 18:54:53 going through stale-ish prs, 18:54:59 !pr 956 18:54:59 https://github.com/crawl/crawl/pull/956 18:55:06 was there an objection to this one, it looks good to me 18:56:51 Bcadren: commenting on old commits sends a github notification to most of the devteam plus people who watch the repo; please save that feature for important content 18:56:53 as long as someone has tested it :D 18:57:33 i'll build an experimental on CPO 18:58:31 should be testable locally I think, if you have gdb installed? (I don't) 18:58:48 with the wizmode crash command 18:58:51 I'm building on an environment w/ gdb at present 19:00:01 i'm on osx so can't test locally 19:00:08 but i'll just test in console on cpo 19:21:40 Stable (0.24) branch on underhound.eu updated to: 0.24.0-28-g17ded55e90 19:24:33 ebering: tested with wizmode ^C crash and worked fine. gdb ran too quick to see the priority though 19:26:59 I had replay running on my terminal and it looks like it ran right 19:27:26 03alexjurkiewicz02 {ebering} 07* 0.25-a0-95-g3c6720d: Run gdb at lower priority 10(10 months ago, 1 file, 3+ 2-) 13https://github.com/crawl/crawl/commit/3c6720d1ae39 19:27:47 A lot of fixedArts have a variable called mondied...what does it mean? 19:32:35 Unstable branch on crawl.kelbi.org updated to: 0.25-a0-95-g3c6720d1ae (34) 19:39:54 ebering: there's another pr of mine which changes the gdb path on osx 🙏 19:41:26 Stigweard (L27 MiBe) ERROR in 'bitary.h' at line 167: bit array range error: 30,-1 / 80,70 (Abyss:3) 19:46:07 !crashlog 19:46:24 21343. Stigweard, XL27 MiBe, T:89971 (milestone): http://crawl.akrasiac.org/rawdata/Stigweard/crash-Stigweard-20191111-004124.txt 19:47:26 hmmmmmmm 19:48:05 alexjurkiewicz: that conflicts with your previous one!!!!! 20:05:13 03alexjurkiewicz02 {ebering} 07* 0.25-a0-96-g4073cf8: Run GDB from /usr/local/bin on OSX 10(9 months ago, 3 files, 9+ 2-) 13https://github.com/crawl/crawl/commit/4073cf80fb89 20:18:08 hm I hope that nice is even present in chroots that have gdb 20:21:08 hm 20:21:28 well 20:21:28 we'll find out on the next crash 20:21:29 but it should be? 20:21:53 it isn't on cao (but cao also doesn't have gdb and I'm not sure is representative) 20:28:27 reverse engineering from the set-up instructions for chroot systems 20:28:57 floraline: if you have a sec (no hurry) could you check if you have `nice` in your chroot, and if so where it is? 20:29:28 really that gdb code should just be made less brittle, and check a bunch of places; if it's doing that it could also check if nice is available 20:29:29 debootstrap should put coreutils into the chroot 20:41:10 Unstable branch on crawl.kelbi.org updated to: 0.25-a0-96-g4073cf80fb (34) 20:50:25 advil: are you still planning on looking at 581 21:27:17 wow, blast from the past 21:27:22 I guess I never did that 21:28:42 I mean, "planning" is a strong word, but I wouldn't want to close it 21:30:53 I mean I'm onboard with the general direction of the change 21:33:27 !pr 581 21:33:28 https://github.com/crawl/crawl/pull/581 22:15:00 advil: looks like my chroot has /usr/bin/nice 22:27:04 https://github.com/crawl/crawl/issues/1128 22:27:12 #tagmajor35in25 22:35:55 SIGNED 22:35:56 never! we'll never bump it! ever! 22:36:03 35 is SACRED 22:36:06 er 22:36:07 34 22:36:45 ebering: i want to defend the extra zany chaos effects: none of them are as weird as cloning a monster, and they happen very rarely, and a broader range of extremely unusual effects is Good Chaos 22:38:00 I think a good goal for chaos would be making it enjoyable for the player to use, and so some of those extremely unusual effects might need to be cut 22:38:08 not sure that cloning is that bad though 22:38:23 cloning might have other issues though 22:39:00 another issue is just that the player needs to be able to reason about their damage when they have chaos in order to have some semblance of strategy 22:39:12 so some effects really might not allow that 22:39:55 if chaos is just some sort of weird eye candy where you just see what happens because it's amusing, it's not a great fit for a game like dcss 22:56:07 New branch created: pull/1198 (1 commit) 13https://github.com/crawl/crawl/pull/1198 22:56:07 03nirrattner02 07https://github.com/crawl/crawl/pull/1198 * 0.25-a0-97-g60f8434: Prevent overtraining skill targets 10(32 minutes ago, 1 file, 47+ 19-) 13https://github.com/crawl/crawl/commit/60f843470b25 22:56:32 what do you mean by "reason about damage"? Do you mean understand the average dps of a chaos weapon as compared to say a vorpal one? 22:57:01 Something like that; at least be semi-informed about what's going to happen 22:57:13 like "probably damage brand" "maybe debuff" "maybe buff" 22:57:46 a lot of chaos effects make it hard to reason about damage but are good for the user. Like petrifying the defender for several turns is great but hard to compare to the higher average damage of flaming 22:59:12 I've always been of the opinion Chaos is good on a ranged weapon or on a spriggan stabber, etc. (distance/speed enough to reset the fight if a really bad effect triggers, but benefits heavily from Petrify/Paralyze etc.) 22:59:25 alexjurkiewicz: sure 22:59:50 that's the "maybe debuff" category 23:00:06 Right now miscasts mean there's a "AAAAAAA" category for effects 23:02:29 Removing that is imo a very good idea; keeping the chaos effect list reasonably pruned is also a good idea, and adding more spell effects isn't a good direction 23:03:13 Also, adding unrand effects to regular chaos that happen very rarely is more eye candy than a meaningful impact 23:05:03 yeah. i think miscasts are good iff cleaned up like you suggested in the wiki writeup 23:05:19 I think even then I'd remove them from chaos brand 23:05:28 The goblin is blasted with wild charm magic and is illuminated! 23:06:24 Is reaping still in chaos? 23:08:03 the other thing that I don't want chaos to do is step on unrand toes, which is why even with a miscast revision those should stay on spellbinder 23:08:58 thematically (uh oh), isn't it alright for chaos to rarely duplicate other effects? 23:10:23 Sure, but it infringes on gameplay 23:54:14 I'm still cherrypicking forward to match my fork to mainline so forgive me if this is an out of date concern but...something should really be done about how annoying alphashops is to maintain...