Minetest logo

IRC log for #minetest-hub, 2018-11-03

| Channels | #minetest-hub index | Today | | Google Search | Plaintext

All times shown according to UTC.

Time Nick Message
00:05 garywhite joined #minetest-hub
00:15 BillyS Ugh
00:15 BillyS I'm trying to understand MT packet code
00:15 BillyS It is ... wierd, to say the least
00:18 BillyS Can anyone here explain it, or give me any pointers?
00:21 sfan5 what exactly are you having trouble with
00:32 BillyS sfan5: I wrote a python script to connect to a minetest server so that I could see what it's doing, and I get messages like 'OEt\x03\x00\x01\x00\x03\xff\xde\x00\x02'.  What do those characters mean?
00:32 BillyS I'm pretty sure the command is the last one
00:33 BillyS Which would mean it's the TOCLIENT_HELLO
00:33 BillyS I know the beginning \x03 means that it's a reliable packet
00:33 BillyS And I think the \x00 after it means that it's a control packet
00:34 sfan5 i'm puzzled how you managed to understand the protocol despite not knowing what "those characters" mean
00:34 BillyS I read https://dev.minetest.net/Network_Protocol
00:34 BillyS And I found an old python script that connected to a server with an old protocal
00:34 BillyS Which I studied quite a bit
00:35 BillyS But I'm not sure what the '\x01\x00\x03\xff\xde\x00' in the middle means
00:35 BillyS I think \xde is the message number, or something like that, because it usually increases by 1 every message
00:36 BillyS Is the \xff just a separator or something?
00:36 sfan5 interpretation doesn't work here, you need to read and understand the documentation
00:38 sfan5 https://github.com/minetest/minetest/blob/master/src/network/connection.h#L136
00:38 sfan5 ^ start there
00:38 BillyS Okay, thx
00:39 BillyS Ooh, so the \x01 is the peer
00:39 BillyS afk
00:40 sfan5 you skipped a few steps again
00:40 sfan5 the peer ID is two bytes, where's the second one?
00:56 BillyS Oh
00:56 BillyS Hmm .... lemme look at this more closely
01:03 BillyS I am confused ... what's the relation between characters and bytes?
01:03 BillyS I think this is utf-8 ... so 1 to 1 in that case
01:06 GreenDimond byte is an amount of data
01:07 GreenDimond standard latin characters are generally a single byte
01:07 BillyS kk
01:07 BillyS I know that byte is an amount of data, but I was pretty sure that the number of bytes per character depends on the encoding of the text
01:08 GreenDimond It does
01:08 BillyS kk
01:08 BillyS Hmmm ...
01:08 T4im utf-8 can vary between 1 to 4 bytes per character
01:08 BillyS Oh, lovely
01:09 BillyS That complicates things for sure
01:09 BillyS I'll just hope that the devs chose 1-byte characters :P
01:09 T4im when i say vary i mean vary with the text, not the implementation
01:10 BillyS I assumed
01:15 BillyS Ah
01:15 BillyS That explains this header ... one of the characters must be two bytes
01:16 BillyS ..wait
01:17 BillyS ooh
01:17 BillyS In the message 'OEt\x03\x00\x01\x00\x03\xff\xdd\x00\x02'
01:17 BillyS \x00\x01\x00\x03\xff\xdd\x00 would be the header, I think
01:17 BillyS Becuase \x03 is the packet type
01:18 BillyS So \x00 is the protocol type (ACK, SET_PEER_ID, PING, or DISCO)
01:19 BillyS .... I think I'm missing something
01:21 T4im exactly what there is supposed to be utf8 again?
01:22 BillyS I think the message is in UTF-8
01:23 T4im you should probably start with the header
01:23 BillyS That's what I'm trying to do
01:23 BillyS This is the complete message I got from the server, header included: \x00\x01\x00\x03\xff\xdd\x00
01:23 BillyS ACK
01:23 BillyS That's wrong
01:23 T4im uhm yes, the 7 bytes header in that comment, theres nothing of utf8 standing anywhere there :o
01:24 BillyS OEt\x03\x00\x01\x00\x03\xff\xdd\x00\x02
01:24 T4im the u8 are unsigned 8 bit, not utf8 :x
01:24 BillyS That ^ is the complete message
01:24 BillyS I'm confusing bytes with strings again ...
01:25 BillyS T4im, sorry, I thought the whole message was sent as a UTF-8 string.  My bad.
01:26 BillyS I still don't get what part of the message is the header though
01:26 T4im the first 7 bytes
01:27 BillyS T4im: So, that is "OEt\x03\x00\x01\x00"?
01:28 T4im 4 bytes protocol id, 2 bytes sender peer id and 1 byte channel, see the comment for details
01:28 T4im maybe; you should probably get a better output there, maybe start up wireshark
01:28 BillyS kk
01:29 T4im and then normalize it to 0prefixed byte everywhere instead of printing strings, that's just hard to read
01:29 T4im hex i mean, 00 00 00 01 etc
01:29 BillyS Good idea; I don't have much experience with this
01:33 T4im s/prefixed/padded/
02:00 BillyS This is wierd
02:01 BillyS in wireshark, it shows traffic between ports 60760 and 30000
02:01 BillyS But the server is on 30000
02:01 BillyS So shouldn't it be between 30000 and 30000?
02:01 BillyS I need to research some of this stuff ...
02:02 BillyS Meh, it's bedtime for me
02:05 garywhite joined #minetest-hub
03:20 Ruslan1 joined #minetest-hub
04:03 sofar BillyS: server port is fixed, but client port is always random
11:25 Gael-de-Sailly joined #minetest-hub
11:26 tenplus1 joined #minetest-hub
11:26 tenplus1 hi folks :)
11:55 rdococ idea: mechanical calculator in Minetest (using no mesecons or digilines, but maybe with some medhanical mod)
11:56 tenplus1 hi rdococ, interesting but how would it work ?!?!
11:56 tenplus1 do you have a schematic
11:56 rdococ no
12:05 aerozoic joined #minetest-hub
12:05 tenplus1 hi aerozoic
12:06 aerozoic wazup tenplus1 !
12:06 tenplus1 :P
12:09 rdococ doubt it's even possible in vanilla minetest_game
12:10 tenplus1 would need some sort of control block
12:11 rubenwardy https://twitter.com/MinetestProject/status/1058690759099121666?s=09
12:13 tenplus1 thx ruben
12:16 aerozoic_ joined #minetest-hub
12:17 aerozoic__ joined #minetest-hub
12:17 sfan5 I'd also like to thank rubenwardy for taking the time and effort to represent Minetest publicly at a conf
12:17 JordachNote8 joined #minetest-hub
12:17 rubenwardy See minetest channel
12:17 rubenwardy Got a donation from the conference
12:17 tenplus1 hi sfan... yeah hope you had fun
12:19 rubenwardy Thanks sfan5
12:19 rubenwardy Also Jordach and Shara
12:19 tenplus1 yay
12:42 T4im joined #minetest-hub
12:42 tenplus1 o/ T4im
12:43 T4im hi :D
12:56 JordachNote81 joined #minetest-hub
12:59 shivajiva for anyone interested https://www.youtube.com/watch?v=dHdU4VRwoxA currently stopped for lunch
12:59 shivajiva Hi tenplus ;)
13:00 tenplus1 hi Shiva :)
13:14 Fixer joined #minetest-hub
13:17 tenplus1 o/ fixer
13:26 Fixer \\o
13:27 rubenwardy So much food
13:28 tenplus1 what ya got ?
13:30 calcul0n joined #minetest-hub
13:30 tenplus1 hi calcul0n
13:30 calcul0n hello
13:36 JordachNote8 joined #minetest-hub
13:41 JordachNote81 joined #minetest-hub
13:42 Krock joined #minetest-hub
13:42 tenplus1 hi Krock
13:42 Krock hi tenplus1
13:43 tenplus1 updated falling pull btw
13:48 Fixer rubenwardy: rip minetest
13:48 tenplus1 ?
13:48 ChimneySwift ??
13:48 rubenwardy ???
13:51 Fixer you need to focus on mt/mtg, not on food
13:52 tenplus1 ffoooooooOOOOoooood!!
13:52 Fixer 00f
13:53 Fixer remember, you are here for minetest, food will be later
13:53 tenplus1 aw
13:53 ChimneySwift you're allowed to eat food, you just can't be really focused on it
13:53 tenplus1 in that case: https://github.com/minetest/minetest/pull/7833
13:58 Krock Somewhen next year Debian will be as good as Windows according to the numbers
13:59 tenplus1 already is in my book
14:04 rubenwardy Fixer: I'm not focusing on food at all
14:04 Fixer i'm focused on your food
14:04 rubenwardy Ohh
14:05 rubenwardy I thought you meant on the stall
14:05 Fixer "The PortSmash paper, titled Port Contention for Fun and Profit, continues to recommend fully disabling SMT, not only to blunt the threat of PortSmash, but also those of TLBleed and two similar attacks known as CacheBleed and MemJam."
14:05 Fixer good to have 4 core without this SMT
14:05 Fixer right on time
14:05 tenplus1 ?
14:06 Fixer "Intel CPUs fall to new hyperthreading exploit"
14:07 tenplus1 oh, maybe that's why newest intel cpu doesnt have HT anymore :D
14:09 calcul0n joined #minetest-hub
14:32 Jordach joined #minetest-hub
14:32 Jordach greetings from #live
14:33 tenplus1 o/////
14:33 Jordach everyone is doing fine :P
14:33 tenplus1 good to hear, how's the noms ?
14:41 Krock o/ Jordach
15:02 rubenwardy We've ditched Jordach for a talk
15:02 rubenwardy Soz
15:08 xerox123 o/
15:08 tenplus1 hi xerox
15:10 xerox123 What's the age range at freenode live?
15:12 rubenwardy Varied
15:12 rubenwardy I've seen a 5 year old and a ~70 year old
15:14 xerox123 Ah
15:14 CWz joined #minetest-hub
15:14 tenplus1 hi CWz
15:14 xerox123 I'll try next year when I'm 18 and don't leave Al my hw to the last weekend™
15:15 CWz Does Linux support SLI
15:22 tenplus1 the last post I could find is dated 2015 CWz and that mentioned SLI being incompatible with OpenGL 3.x cause it doesnt use page flipping
15:23 Ruslan1 joined #minetest-hub
15:25 CWz so no support for two or more graphic cards
15:25 CWz ?
15:27 tenplus1 https://devtalk.nvidia.com/default/topic/617305/?comment=4705640
15:28 rubenwardy It would nice to support one graphics card well, first
15:29 rubenwardy Fuck Nvidia
15:29 tenplus1 lol, ati are doing much more for the open-source driver than nvidia is for theirs
15:33 Jordach rubenwardy: no problem, had a stretch
15:33 Krock is ati still a thing? I though they now sell everything under AMD
15:33 Krock *under the brand
15:33 tenplus1 amd are awesome :)
15:56 CWz nvidia graphic driver seems to work fine on my end
15:56 CWz but the open source driver isn't as good as AMDs
15:57 T4im if i remember correctly it's not nvidia's nor amd's open source drivers, but both are done from kernel developers; just that amd helps out with specs and perhaps even manpower
16:00 tenplus1 yeah, nvidia really need to open up and assist with the open-source driver, it could be soooo much better
16:01 T4im "In addition to providing the necessary documentation, AMD employees contribute code to support their hardware and features." indeed, they are actively contributing :D
16:01 tenplus1 that in my opinion is a good company
16:01 T4im it's just sad that due to the CUDA stuff, you sometimes still depend on nvidia :(
16:02 T4im amd still needs to catch up there
16:17 Gael-de-Sailly joined #minetest-hub
16:52 tenplus1 could I have some mt-game devs look at this please: https://github.com/minetest/minetest/pull/7833
16:52 tenplus1 !title
16:52 MinetestBot tenplus1: falling entity rewrite by tenplus1 · Pull Request #7833 · minetest/minetest · GitHub
16:58 Fixer SLI is dead
16:59 Fixer tenplus1: yet nvidia still works kinda better
16:59 tenplus1 my net-top has nvidia ion and I do like the binary blob, it's faster
17:09 tenplus1 laters folks
17:09 tenplus1 left #minetest-hub
17:42 GreenDimond joined #minetest-hub
18:17 FrostRanger joined #minetest-hub
18:45 Krock !tell tenplus1 Please don't take >the< recent comment of "falling entity rewrite" too serious. I very welcome dev time and PRs, ofc
18:45 MinetestBot Krock: I'll pass that on when tenplus1 is around
18:46 Krock <3 MinetestBot
18:46 MinetestBot <3 Krock
18:46 Krock (because he's not in -dev)
18:47 sfan5 ​!tell will be passed on in whatever channel MinetestBot sees him
18:47 sfan5 MTB is not present in -dev though
18:47 Krock neither ten nor MTB is there
18:53 * VanessaE pokes Krock with some minetest-mods PRs... :P
18:54 Fixer_ joined #minetest-hub
18:55 garywhite joined #minetest-hub
18:56 * Krock opens bookmarks
19:00 Krock framedglass code looks good, non-controversial
19:05 Krock looks like coloring anything other than steel framed obsidian glass wasn't possible?
19:06 VanessaE yeah
19:06 Krock weird that the punch coloring doesn't work for the others even though it has the callbacks registered
19:06 VanessaE I just kept it as it originally was
19:06 VanessaE empty functions.
19:07 VanessaE and anyway only obsidian glass was supposed to be colorable
19:07 Krock yes I see
19:07 VanessaE wait..
19:07 VanessaE what callbacks?
19:08 Krock "on_punch = framedglass.color_on_punch," but I looked at the wrong line. It's alright. Only steel is colorable
19:11 VanessaE :)
19:15 Krock approved. gets somewhen merged.. maybe tomorrow if I don't forget it
19:16 VanessaE \o/
19:16 VanessaE fwiw, about whitespace and code style:  I tend to just stick with whatever's already there.
19:16 VanessaE thanks for the tweaks.
19:17 Krock would've taken longer to write a comment for all those tiny issues
19:18 Krock sticking to old style is consistent in the project, adjusting it is (optimally) consistent with the minetest-mods style
19:19 Krock although I would be happier if more of the mods could follow the MTG/MTE code style rules. It's very easy to notice copypasta in unmaintained mods right now
19:19 VanessaE yepa
19:20 * luk3yx prefers spaces over tabs, they're more consistent for weird whitespacing things.
19:21 * Krock slaps luk3yx and throws them out of the door
19:21 * Krock slams the door
19:22 Krock luk3yx: they're fine as long it's after the first printable character of the line
19:23 Krock i.e. to line up tables or repetitive function arguments
19:23 Krock that's where tabs fail horribly due to their customization
19:24 VanessaE regarding the technic one, technic depends on pipeworks, which depends on basic_materials, if that's of any importance.
19:25 Krock guess who added that dependency
19:25 VanessaE spaces for indenting?  heretic!
19:26 VanessaE Krock: of course I added it :P
19:26 VanessaE but that's been a plan for a years.  just never got around to it.
19:27 Krock https://imgs.xkcd.com/comics/citogenesis.png
19:27 Krock same principle as here ^
19:27 VanessaE granted.
19:30 VanessaE doesn't change the end result:  I simply implemented the basics of a plan RBA and I had percolating before he passed away.
19:31 Krock is it some kind of glue to provide item alternatives if another mod is absent?
19:37 VanessaE it's purpose is to provide those simple items that everyone else seems to keep re-inventing from time to time
19:37 VanessaE though it has aliases so that mods that expected say home decor (plastic) or gloopblocks (cement) will still work (assuming depends.txt is appropriately tweaked)
19:37 VanessaE (and assuming I didn't forget any aliases)
19:37 VanessaE i.e. to reduce repetitiveness and crafting conflicts.
19:37 VanessaE if handled properly, basic_materials shouldn't need to grow too much past where it is now, though I can think of a couple of other items that it should probably have
19:37 VanessaE a/ can/'m sure I could/
19:37 VanessaE you get the idea?
19:38 rubenwardy it's gross, technically
19:38 VanessaE ?
19:38 rubenwardy what's the point of adding items which don't fit in logically?
19:38 VanessaE what doesn't fit into what?
19:38 rubenwardy like, if I have to install the mod just for one part, there's a load of random stuff that isn't used
19:39 rubenwardy it's better to divide things into craft trees
19:39 VanessaE oh here we go again...
19:39 rubenwardy :D
19:39 Krock the get-the-entire-boilerplate-mode or trim-to-necessary-stuff discussion?
19:39 VanessaE can't you just ignore the items that you have no use for?  also, technic, as amended in my PR, uses several items in basic_materials
19:39 Krock s/mode/mod
19:39 VanessaE plus,
19:40 VanessaE basic_materials IS split into parts - metals, plastics, electrical/electronic, misc
19:40 rubenwardy it's a bad user experience
19:41 VanessaE no, a "bad user experience" is showing the user two or more of the same item and having them make the wrong one for the thing they want.
19:41 VanessaE (same as in equivalent in a gaming sense)
19:43 VanessaE if a user is stupid enough to make an item that isn't used anywhere, that's their fault.  if they're shown two options that look like they ought to do the same job and they pick the wrong one, that's the mod's fault.
19:43 VanessaE THAT is the kind of thing basic_materials is meant to address
20:33 Jordach joined #minetest-hub
21:40 behalebabo joined #minetest-hub
22:12 Ruslan1 joined #minetest-hub
22:14 behalebabo joined #minetest-hub
22:14 Jordach joined #minetest-hub
22:25 FrostRanger joined #minetest-hub
23:19 Jordach joined #minetest-hub

| Channels | #minetest-hub index | Today | | Google Search | Plaintext