| Time |
Nick |
Message |
| 02:33 |
|
YuGiOhJCJ joined #minetest-dev |
| 03:38 |
|
olliy joined #minetest-dev |
| 04:00 |
|
MTDiscord joined #minetest-dev |
| 04:37 |
|
calcul0n joined #minetest-dev |
| 06:38 |
|
calcul0n_ joined #minetest-dev |
| 08:25 |
|
Fixer joined #minetest-dev |
| 09:51 |
sfan5 |
~tell Desour I added a CI run with sdl, can you rebase all of your PRs so that runs? |
| 09:51 |
ShadowBot |
sfan5: OK. |
| 10:09 |
sfan5 |
pushing http://sprunge.us/0Erjbx?diff in 10m |
| 10:19 |
|
proller joined #minetest-dev |
| 10:29 |
|
_Zaizen_[m] joined #minetest-dev |
| 10:48 |
|
_Zaizen_[m]1 joined #minetest-dev |
| 10:48 |
|
Goobax[m] joined #minetest-dev |
| 10:48 |
|
MisterE[m] joined #minetest-dev |
| 10:48 |
|
programmerjake joined #minetest-dev |
| 10:48 |
|
freshreplicant[m joined #minetest-dev |
| 10:48 |
|
fgaz joined #minetest-dev |
| 10:48 |
|
k327[m] joined #minetest-dev |
| 10:48 |
|
Zughy[m] joined #minetest-dev |
| 10:48 |
|
Whanake[m] joined #minetest-dev |
| 10:48 |
|
schwarzwald[m] joined #minetest-dev |
| 10:48 |
|
lebruhgamer[m] joined #minetest-dev |
| 11:03 |
sfan5 |
can I get any opinions on #12790? |
| 11:03 |
ShadowBot |
https://github.com/minetest/minetest/issues/12790 -- Take geographic distance into account for server list ordering by sfan5 |
| 11:40 |
|
Desour joined #minetest-dev |
| 11:53 |
Krock |
is the sole purpose of get_once() to escape the async environment? |
| 11:54 |
Krock |
oh wait. couldn't you use the Settings object for that? |
| 11:55 |
Krock |
concept-wise it does look good |
| 12:09 |
proller |
in first network stack should be replaced, and with good network this have sense.. now it works terrible even with 1ms ping 1G local network... |
| 12:15 |
Krock |
that's somewhat unrelated to that PR |
| 13:17 |
sfan5 |
Krock: I can but I didn't want to clutter the settings |
| 13:51 |
Krock |
well, now you're cluttering the C++ side. I don't know what's better though |
| 13:55 |
Krock |
hmm I think that using the settings would be a saner approach (does not introduce helper code) |
| 13:56 |
|
Noisytoot joined #minetest-dev |
| 13:57 |
MTDiscord |
<MisterE> TBH, I do not like that the cryptomine server is very high on the serverlist. If it is a scam, that looks bad. If it is not a scam, then its fine I guess. |
| 14:17 |
|
fluxionary joined #minetest-dev |
| 14:37 |
|
behalebabo joined #minetest-dev |
| 14:40 |
MTDiscord |
<Warr1024> How do we as a community actually define a "scam"? If they give nothing and take nothing but your time, is that a scam or is that just the same premise as any normal game? How liable can servers be expected to be for the existence or non-existence of external markets for in-game resources? |
| 14:41 |
rubenwardy |
paying players to play incentivises AFKers, and also is a form of manipulating the server list |
| 14:44 |
MTDiscord |
<Warr1024> Sure, but what do we consider "paying"? Does minegeld or other ostensibly non-external-market resources count? Do game mechanics that encourage automated farming or manufacturing count? It seems like servers would always rather incentivize players to be active if possible, and only to AFK if they can't make the incentive targeted enough. |
| 14:52 |
rubenwardy |
another thing it incentivises is bots |
| 14:55 |
MTDiscord |
<Warr1024> From the perspective of the server list and of other players, there isn't actually all that much difference between a "bot" and any other AFK player. |
| 14:56 |
MTDiscord |
<Warr1024> I'd actually really like to have a way to hide specific accounts from the serverlist, so that bots actually become usable without messing with serverlist statistics. |
| 14:56 |
rubenwardy |
if a package author spammed a package with fake downloads and reviews on ContentDB, I'd penalise or ban them |
| 14:57 |
MTDiscord |
<Warr1024> Right, it just requires you to have a good idea of what's "fake", and a way to tell whether it's done with the author's collusion, or whether it was done in order to trigger penalties for that author. |
| 14:58 |
rubenwardy |
ContentDB also has rules against paying for reviews |
| 15:01 |
rubenwardy |
Even if it is legit, having crypto as top servers is not a good luck |
| 15:01 |
rubenwardy |
*look |
| 15:04 |
MTDiscord |
<Warr1024> Yeah, it's definitely ick. It's just hard to do something about it without compromising some other principles we want to have. |
| 15:04 |
MTDiscord |
<Warr1024> The geographical thing might help, though, a little or a lot, depending on how strongly it's done. |
| 15:06 |
rubenwardy |
I don't see any conflict with principles |
| 15:06 |
sfan5 |
I still don't know which principles we as a project are apparently supposed to have |
| 15:07 |
rubenwardy |
we're not banning them from using minetest. It would just be reducing the spread of their shit and making the serverlist more relevant to actual players |
| 15:13 |
MTDiscord |
<Warr1024> I thought we were big on user choice. The decision about whether or not to try out a crypto server should be left to the user, and we should do the minimum amount possible to impose our will on that decision. |
| 15:13 |
MTDiscord |
<Warr1024> It's probably acceptable to have at least some sorting in the server list, rather than just dumping it in random order every time, but we have to consider the consequences of our choice to sort it in a particular way. |
| 15:14 |
sfan5 |
so far "payment incentivises afk" is the closest we have got to an actual arguments, "crypto is bad", "their server is bad", "it's not a good look" are not usable arguments |
| 15:14 |
MTDiscord |
<Warr1024> And we can't just make arbitrary decisions to push one server's score down, or to block selected content just because we don't like it. |
| 15:16 |
MTDiscord |
<Warr1024> Right, paying people to inflate your server list score is about the only thing we can use to justify penalizing this particular server, and if we start finding ourselves doing it more often, we are going to probably have to be prepared to answer questions about what would actually count. |
| 15:16 |
MTDiscord |
<Warr1024> As much as I'd love to just have an "I knows it when I sees it" standard and rarely ever have to apply it, that's not always an option. Sometimes an enforcement action makes people feel nervous about which side of the line they're on, and they want to know where that line is. |
| 15:18 |
MTDiscord |
<Warr1024> tbh, though, it'd be nice if we could actually justifiably kick this server at least off the top of the listing already, just so we can stop having to keep having this discussion 😑 |
| 15:19 |
sfan5 |
as I mentioned last time with the afk time tracking Warr mentioned there are way too many factors to decided whether a particular online player is worth it or relevant. you quickly leave the "preventing cheating" territory and enter into "curation" territory |
| 15:19 |
sfan5 |
IMO we can stop having this discussion at any moment, personalyl the server doesn't bother me all that much |
| 15:20 |
rubenwardy |
If we wanted to maximise user choice, we'd randomise the server list every 5 minutes |
| 15:22 |
Pexin |
is the server fun? what experience are players having? if it's just MT base game with fake money thrown at you, will players just go find a fun unique server anyway? |
| 15:23 |
MTDiscord |
<Warr1024> I'm in favor of mixing in a fairly significant amount of randomness into the server list. I don't think I'd go quite so far as to ignore scoring factors altogether and go pure random, though. I'm not entirely sure how far I'd go, quantitatively, though; I guess I'd have to fiddle with the knobs to see at what extremes things start to look nonsensical. |
| 15:26 |
MTDiscord |
<Warr1024> Heh, as far as I know, nobody who is proposing to penalize the server is likely to have extensively reviewed the actual gameplay there. Some people have poked around there but I haven't heard much about anybody actually trying to engage with it honestly. There is a lot of "crypto -> scam" knee-jerk reaction; if they had titled the server differently then we may never have had this conversation. |
| 15:28 |
sfan5 |
unrelated but I just noticed that the claimed age of 203 days is very high, I don't think this server has existed for more than a month |
| 15:28 |
sfan5 |
(not that this would change much about the ranking) |
| 15:28 |
sfan5 |
and actually the age nubmer is totally broken anyway because it counts the number of in-game days |
| 15:29 |
rubenwardy |
it counts in-game seconds and converts to real life days |
| 15:29 |
rubenwardy |
to clarify |
| 15:30 |
rubenwardy |
although that's a crappy clarification |
| 15:30 |
sfan5 |
ah well it can take into account the current time_speed but eh |
| 15:30 |
rubenwardy |
I thought time_Speed just effected time of day, and not game time |
| 15:31 |
sfan5 |
oops yeah nevermind |
| 15:31 |
sfan5 |
game_time strictly increases by real-time while the game is active |
| 15:31 |
sfan5 |
I was thinking of day_count |
| 15:33 |
rubenwardy |
I guess the solution to a lot of these issues would be to make the serverlist harder to game |
| 15:34 |
rubenwardy |
track age server-side |
| 15:34 |
rubenwardy |
user reviews could be used as a metric, but also argh user reviews |
| 15:35 |
sfan5 |
or stop attempting to sort the list |
| 15:36 |
Pexin |
topic seems to go in circles. it was already noted a while ago "ideal" solution is for many more factors to be objectively quantifiable, and have a robust means for players to choose their own priorities |
| 15:37 |
Pexin |
look at http://vgmusic.com "new-files" sort page |
| 17:17 |
Zughy[m] |
another argument could be "we're against crypto, deal with it". It's our software, we choose, we don't need to find excuses |
| 17:18 |
Zughy[m] |
that server bothers me for two reasons: 1. as a server admin that shit is gonna always surpass my server 2. people thinking MT is dead and it's just some place for some financial speculation instead of actual content |
| 17:19 |
Zughy[m] |
we already have a lot of debatable content, that won't help at all |
| 17:24 |
|
Fleckenstein joined #minetest-dev |
| 17:25 |
schwarzwald[m] |
Another solution is to not have a public serverlist. |
| 17:39 |
MTDiscord |
<Warr1024> There are some problems with that solution of course. If anybody volunteers to run one they may end up becoming the de facto default anyway. At least with sfan5 in charge, there is much less risk of over-curation.... |
| 17:39 |
MTDiscord |
<Warr1024> If we could have a public serverlistlist then we could punt the problem further down the chain and let list operators curate as they want, but then how do we sort THAT list? 😆 |
| 18:27 |
MTDiscord |
<GoodClover> From a technical viewpoint, I don't think Thee serverlist should censor any server out unless it's illegal or attempting to abuse the list. If I wanted to make my own list, I have the choice of convincing all relevant admins to change their serverlist_url, or copying from Thee serverlist. If Thee list starts curating, then people can't make their own lists including things the master list decided to clonk. |
| 18:28 |
MTDiscord |
<GoodClover> (note that at present serverlist_url can only be one URL, but that's not the main issue - if a new server comes along it wouldn't have your list's URL to report to) |
| 18:28 |
MTDiscord |
<GoodClover> From my own thoughts, heck yes banish the crypto to hell |
| 18:30 |
MTDiscord |
<GoodClover> another way to put it is I think Thee serverlist should act like an unopinionated relay |
| 18:30 |
MTDiscord |
<luatic> we need a way for players to mark servers as "ignored" in their clients |
| 18:31 |
MTDiscord |
<luatic> I agree with olive |
| 18:31 |
MTDiscord |
<Warr1024> We really want the project to be as decentralized as possible, I think, but we are sort of forced into some level of centralization by usability needs. Asking each user to enter their own serverlist url would just not really fly. |
| 18:31 |
MTDiscord |
<Warr1024> This topic feels like it's getting very main-channel and not so -dev... |
| 18:32 |
MTDiscord |
<GoodClover> heh, serverlistlist, a list of serverlists to choose from |
| 18:41 |
MTDiscord |
<savilli> You can change your serverlist already if you really want to |
| 19:31 |
|
vampirefrog joined #minetest-dev |
| 20:38 |
|
behalebabo joined #minetest-dev |
| 20:39 |
|
calcul0n_ joined #minetest-dev |
| 22:35 |
|
panwolfram joined #minetest-dev |
