DDraceNetwork - developer

DDraceNetwork

Development / developer

Development discussion. Logged to https://ddnet.tw/irclogs/ Connected with DDNet's IRC channel, Matrix room and GitHub repositories — IRC: #ddnet on Quakenet | Matrix: #ddnet-developer:matrix.org GitHub: https://github.com/ddnet

Between 2018-02-13 00:00:00Z and 2018-02-14 00:00:00Z

[quakenet] <Learath2> aah the good old days when I didn't know how to even code and still tried my best

[ddnet/ddnet] New branch created: staging.tmp

02:02

[ddnet:staging.tmp] 1 new commit

aaec4e0 [ci skip] - bors[bot]

02:02

[ddnet:staging.tmp] 5 new commits

9e9b57e Move generated files to the CMake build directory - heinrich5991 7975815 Remove spurious engine/server include in game - heinrich5991 22164d5 Squelch warnings in bundled ZLIB dependency - heinrich5991 94818ec Try to find external GLEW, pnglite and Wavpack - heinrich5991 d1bf484 [ci skip] -bors-staging-tmp-1034 - bors[bot]

02:02

[ddnet:staging] 6 new commits

5587e6d Remove unused inp_grab - def- 9e9b57e Move generated files to the CMake build directory - heinrich5991 7975815 Remove spurious engine/server include in game - heinrich5991 22164d5 Squelch warnings in bundled ZLIB dependency - heinrich5991 94818ec Try to find external GLEW, pnglite and Wavpack - heinrich5991

02:05

[ddnet/ddnet] Pull request closed: #1027 Move to new-style CMake, allow external GLEW, pnglite, wavpack

02:12

[ddnet:master] 5 new commits

9e9b57e Move generated files to the CMake build directory - heinrich5991 7975815 Remove spurious engine/server include in game - heinrich5991 22164d5 Squelch warnings in bundled ZLIB dependency - heinrich5991 94818ec Try to find external GLEW, pnglite and Wavpack - heinrich5991 c0fa444 Merge #1034 - bors[bot]

02:12

[ddnet/ddnet] Issue closed: #1023 Update wavpack

02:12

[ddnet/ddnet] Issue closed: #1016 Glew should be able to be loaded externally

02:12

[ddnet/ddnet] Pull request closed: #1034 Try to find external GLEW, pnglite and Wavpack

[ddnet/ddnet] Pull request opened: #1035 Fix #1031

Not sure if this is the best way to do this though :D

[quakenet] <heinrich5991> deen: I altered tables on ddnet.tw but it hasn't replicated to ger

12:09

[quakenet] <heinrich5991> I don't find logs or antything, maybe I should have checked first

12:09

[quakenet] <Learath2> there should be binlogs somewhere

12:10

[quakenet] <heinrich5991> nothing in systemd logs if you mean that

12:10

[quakenet] <Learath2> no actual relay ogs in mysql bin format

12:10

[quakenet] <heinrich5991> Learath2: how to check that?

12:12

[quakenet] <Learath2> they should be in /var/lib/mysql

12:13

[quakenet] <heinrich5991> Learath2: do you know the replication strategy?

12:13

[quakenet] <heinrich5991> i.e. where do I have to look? ddnet.tw or ger.ddnet.tw?

12:13

[quakenet] <heinrich5991> or is it on both?

12:13

[quakenet] <Learath2> not really

12:17

[quakenet] <Learath2> the ALTER TABLE shows up on neither

12:18

[quakenet] <heinrich5991> Learath2: not even on ddnet.tw?

12:18

[quakenet] <Learath2> not on the relay logs n

12:18

[quakenet] <heinrich5991> so what do I do?

12:19

[quakenet] <heinrich5991> and how do I access the relay logs?

12:19

[quakenet] <heinrich5991> and do you know more credentials?

12:19

[quakenet] <heinrich5991> > SHOW SLAVE STATUS;

12:19

[quakenet] <heinrich5991> ERROR 1227 (42000): Access denied; you need (at least one of) the SUPER, REPLICATION CLIENT privilege(s) for this operation

12:20

[quakenet] <Learath2> mysqlbinlog --no-defaults file

12:20

[quakenet] <Learath2> use root

12:22

[quakenet] <heinrich5991> Learath2: shall I just issue the ALTER TABLE on all machines?

12:22

[quakenet] <heinrich5991> would that fix stuff?

12:22

[quakenet] <heinrich5991> or would it make things worse :P

12:24

[quakenet] <Learath2> I don't get our replication strategy

12:24

[quakenet] <Learath2> it looks more like a chain then a tree :P

12:24

[quakenet] <Learath2> Last_Error: Error 'Got error 64 'Temp file write failure' from InnoDB' on query. Default database: 'teeworlds'. Query: 'ALTER TABLE record_race ADD COLUMN GameID VARCHAR(64) AFTER cp25

12:25

[quakenet] <Learath2> that's on CAN

hm

[quakenet] <Learath2> actually on USA

I'm applying it manually to GER now, let's see if I can replicate a record from GER to ddnet.tw then

[quakenet] <Learath2> inb4 it adds 2 columns

Does somebody know who created the web map rendering feature (e.g. https://ddnet.tw/maps/?map=Quick), where to download the sourcecode and what license it is under?

[quakenet] <Learath2> timakro something tells me it was eastbyte but not sure

do you know how i can contact him?

[quakenet] <Learath2> heinrich5991: it's a chain yeah, @ > usa > can > ger > @

12:38

[quakenet] <Learath2> it propagated fine to usa

12:38

[quakenet] <Learath2> can gave an error so it couldn't go further

12:38

[quakenet] <Learath2> how to fix it I have no idea

12:38

[quakenet] <Learath2> :D

12:39

[quakenet] <timakro> oh EastByte is here in irc -.-

12:40

[quakenet] <Learath2> heinrich5991: might need to restart slave on can and see if it fixes itself, it said couldn't write to temp maybe it was full on diskspace

@timakro https://github.com/east/tw-webgl ?

tw-webgl - teeworlds webgl

@Sonix danke!

Gerne

monkaS

1

Probably disk full on CAN

@deen can disk size be increased on CAN? what's the hosting provider?

[quakenet] <Learath2> heinrich5991: did you manage to get it to propagate all the way?

14:20

[quakenet] <heinrich5991> Learath2: dunno. can you retry propagation to CAN?

14:20

[quakenet] <heinrich5991> restart didn't seem to trigger it (?)

14:21

[quakenet] <Learath2> it errored again

14:21

[quakenet] <Learath2> we don't have much diskspace available at CAN at all :/

14:22

[quakenet] <heinrich5991> yea, I tried optimizing it a bit

14:23

[quakenet] <heinrich5991> but it only freed around 500MB

14:25

[quakenet] <Learath2> heinrich5991: can we maybe take a dump from ddnet.tw and just drop table and reimport on CAN?

14:25

[quakenet] <heinrich5991> that sounds super easy to get wrong

14:25

[quakenet] <heinrich5991> :D

14:26

[quakenet] <Learath2> it's better then replication being stuck right here

14:26

[quakenet] <Learath2> god knows how big that dump is going to be :P

14:27

[quakenet] <heinrich5991> probably bigger than the DB

14:27

[quakenet] <heinrich5991> which is 5GB in itself

14:34

[quakenet] <Learath2> oh you did add the column on GER didn't you? :D

14:35

[quakenet] <Learath2> it propagated back to ddnet.tw and caused a duplicate column error

@heinrich5991 it's OVH. Let's just remove the DB there instead

[quakenet] <heinrich5991> sometimes re-ordering a server can increase its specs

14:37

[quakenet] <Learath2> skipped the bad queries on ddnet.tw

throw out CAN db from all servers, connect the other ones directly

14:37

and shut down CAN db

14:37

enough space

[quakenet] <Learath2> deen: CAN should dump records to USA, stop slave on CAN, make GER master USA instead, done?

something like that

14:42

why dump to USA?

14:42

why not just connect GER <-> USA directly

[quakenet] <Learath2> well CAN needs somewhere to send it's records, USA sounds closer then GER

yes

14:43

they always have multiple

14:43

so you can add USA and GER as fallback

[quakenet] <Learath2> and don't we want to keep the loop GER->USA->ddnet.tw->GER?

yes, keep loop

14:43

I shut down GER2 in a week?

[quakenet] <Learath2> do you happen to know how to remove CAN from the loop? idk what master_log_pos and master_log_file should be tbh

I can do that when I'm home

14:45

I don't know either, but I guess I'll figure it out along the way

14:45

you can get the master log values from CAN

[quakenet] <Learath2> I'll replace GER2 with a kimisufi node, see if that holds up btw

on CAN: stop slave; on USA: stop slave; CHANGE MASTER TO MASTER_HOST='ger.ddnet.tw', MASTER_USER='repl', MASTER_PASSWORD='SECRETREPL', MASTER_LOG_FILE='mysql-bin.000010', MASTER_LOG_POS=37941939; start slave;

14:46

something like that

[quakenet] <Learath2> I'll do it as soon as I figure out how master_log_file and master_log_pos are supposed to work

14:47

[quakenet] <heinrich5991> thanks

14:47

[quakenet] <heinrich5991> sorry for destroying everything :(

14:48

[quakenet] <Learath2> my shower is dripping... I'm about to go insane

Replace the rubber circle thingy inside, they get brittle with time

14:49

heinrich5991: np ,i remember having the same problem with CAN^^

[quakenet] <Learath2> heinrich5991: you didn't break anything, just CAN

[quakenet] <Learath2> should be fine now \o/

15:06

[quakenet] <heinrich5991> you fixed everything™?

15:08

[quakenet] <Learath2> just need to let CAN connect to USA instead

15:08

[quakenet] <heinrich5991> Learath2: shall I disable the mysqld service on CAN?

15:08

[quakenet] <heinrich5991> so it doesn't start on next boot?

15:08

[quakenet] <Learath2> yeah

15:09

[quakenet] <Learath2> and could you check the log for why CAN cannot connect to usa mysql?

15:09

[quakenet] <heinrich5991> which log?

15:09

[quakenet] <heinrich5991> CAN's ddnet server logs?

15:10

[quakenet] <Learath2> nvm, old pw on CAN

15:10

[quakenet] <Learath2> yeah all should be fine (tm)

15:12

[quakenet] <Learath2> should try to propagate sth around to see if it actually works fine

15:13

[quakenet] <Learath2> removing one point from myself

15:14

[quakenet] <heinrich5991> and re-adding 10 points afterwards?

15:14

[quakenet] <heinrich5991> classic

15:14

[quakenet] <Learath2> :D

15:14

[quakenet] <heinrich5991> Learath2: the DB is still on CAN btw

15:15

[quakenet] <Learath2> yeah just stopped it while working on it

15:15

[quakenet] <heinrich5991> ok

15:17

[quakenet] <Learath2> yay it works(tm)

15:18

[quakenet] <Learath2> CAN still messes up my terminal...

15:20

[quakenet] <Learath2> DROP DATABASE teeworlds; inbound :P

15:23

[quakenet] <heinrich5991> sounds safe

15:23

[quakenet] <Learath2> very

15:23

[quakenet] <heinrich5991> de*en has backups ;)

15:30

[quakenet] <deen> what's the state?

15:30

[quakenet] <deen> tell me if you need help

15:31

[quakenet] <Learath2> can is no longer in the loop

15:31

[quakenet] <Learath2> deen: the blocker pw in the db doesn't match the one in mysqlblock.cfg do we no longer use mysql for block?

15:31

[quakenet] <deen> dunno

15:31

[quakenet] <deen> no one cares about ranks on block

15:32

[quakenet] <deen> but fix them please

15:32

[quakenet] <Learath2> it must have been broken for quite some time on all servers

15:32

[quakenet] <heinrich5991> playing on the blocker servers, I thought that was by design :D

15:33

[quakenet] <heinrich5991> that there are no permanent scores ^^

15:33

[quakenet] <Learath2> and whoever changed the pw scrubbed the .mysql_history :D

15:33

[quakenet] <Learath2> guess we need a new pw for them

15:34

[quakenet] <deen> or just keep them without ranks

15:34

[quakenet] <deen> i don't mind

15:34

[quakenet] <deen> and no one complained

15:38

[quakenet] <heinrich5991> I don't care either, was just my experience :D

15:40

[quakenet] <Learath2> do any of you know whether zcatch sql implementation would accept a domain instead of an ip?

15:44

[quakenet] <Learath2> well I went with ip just incase

15:44

[quakenet] <Learath2> anyways done with all the mysql stuff, everything I can think of is fixed now

15:44

[quakenet] <deen> i think so, but I prefer ip too

15:45

[quakenet] <deen> since dns sometimes breaks

15:45

[quakenet] <deen> especially with dns amplification attacks hitting us

15:46

[quakenet] <Learath2> deen: do you happen to remember what fixes the weird TERM issue I kept having with zaf?

15:46

[quakenet] <Learath2> it happens with can now

15:48

[quakenet] <Learath2> heinrich5991: guess you can merge #1025 now if you don't see any issues with it

15:48

[quakenet] <deen> what's the issue?

15:48

[quakenet] <Learath2> deen: backspace doesn't seem to work, colors don't work

15:49

[quakenet] <Learath2> if I do TERM="xterm-256color" it does work tho

15:49

[quakenet] <deen> what's your normal TERM?

15:50

[quakenet] <Learath2> rxvt-unicode-256color

15:50

[quakenet] <deen> weird

15:51

[quakenet] <deen> i also use rxvt-unicode with 256 colors and for me TERM=xterm-256colo

15:51

[quakenet] <deen> r

15:52

[quakenet] <deen> either terminfo or ~/.zkbd

15:53

[quakenet] <heinrich5991> y'all should just fish with default options

15:54

[quakenet] <heinrich5991> :P

15:55

[quakenet] <heinrich5991> Learath2: CAN's hdd is now at 42%. good work. thanks ;)

15:55

[quakenet] <Learath2> can store so much teehistorian data

15:55

[quakenet] <heinrich5991> ^^

[ddnet/ddnet] New branch created: staging.tmp

16:01

[ddnet:staging.tmp] 1 new commit

10252bb [ci skip] - bors[bot]

16:01

[ddnet:staging.tmp] 6 new commits

4748d5f Add client version to the record - Learath2 4a52bca Include the teehistorian uuid instead - Learath2 0aeb912 Escape the UUID just in case - Learath2 94d98e8 Fix mistakes. Fix CREATE TABLE - Learath2 bc11fc2 Add GameUuid to teamscore aswell - Learath2

16:01

[ddnet:staging] 6 new commits

4748d5f Add client version to the record - Learath2 4a52bca Include the teehistorian uuid instead - Learath2 0aeb912 Escape the UUID just in case - Learath2 94d98e8 Fix mistakes. Fix CREATE TABLE - Learath2 bc11fc2 Add GameUuid to teamscore aswell - Learath2

[ddnet:master] 6 new commits

4748d5f Add client version to the record - Learath2 4a52bca Include the teehistorian uuid instead - Learath2 0aeb912 Escape the UUID just in case - Learath2 94d98e8 Fix mistakes. Fix CREATE TABLE - Learath2 bc11fc2 Add GameUuid to teamscore aswell - Learath2

16:13

[ddnet/ddnet] Pull request closed: #1025 Add Teehistorian UUIDs to every record

[ddnet-web:master] 1 new commit

af7fce1 Forgot this one - Learath2

+----------+--------------+------+---------------------+--------------------------------------+
| map      | name         | time | timestamp           | gameid                               |
+----------+--------------+------+---------------------+--------------------------------------+
| run_blue | heinrich5991 | 8.36 | 2017-01-28 20:28:37 | NULL                                 |
| run_blue | heinrich5991 | 8.36 | 2017-01-28 20:30:19 | NULL                                 |
| run_blue | heinrich5991 | 7.74 | 2017-01-28 20:31:05 | NULL                                 |
| run_blue | heinrich5991 |  7.5 | 2017-01-28 20:31:31 | NULL                                 |
| run_blue | heinrich5991 | 16.7 | 2018-02-13 17:26:01 | 84663ce1-12cf-4871-9651-643120960dc1 |
+----------+--------------+------+---------------------+--------------------------------------+

[quakenet] <heinrich5991> Learath2: it worx :)

16:34

[quakenet] <Learath2> yay

16:35

[quakenet] <Learath2> we should probably populate the null ones in reverse btw

16:35

[quakenet] <Learath2> after we have better tools of analyzing teehistorian files that is

16:35

[quakenet] <heinrich5991> :)

gameid? an uuid for runs?

[quakenet] <heinrich5991> nah, for server runs

16:40

[quakenet] <heinrich5991> it's the file name of the teehistorian file that corresponds to the run

16:40

[quakenet] <heinrich5991> i.e. we can replay the race with that id

I see thanks

what is the gameid for?

[quakenet] <Learath2> 17:40 <@heinrich5991> it's the file name of the teehistorian file that corresponds to the run

16:51

[quakenet] <Learath2> heinrich5991: uhm there seems to be a cmake problem on circleci

16:51

[quakenet] <Learath2> circle didn't like the wavpack update maybe?

Image attachment

17:05

.git/info/exclude doesn't work anymore?

17:05

is this due to the new .gitignore changes?

17:05

or i am blind?

[quakenet] <Learath2> well it works just fine for autoexec_server.cfg no?

it isnt edited

17:07

thats why it isnt shown

[quakenet] <Learath2> hmm, doubt gitignore overrides .git/info/exclude anyways, idk

this is weird

17:12

it might have relation to

*
!**/
!*.*

(edited)

any thoughts on this?

check the doc of gitignore. it should tell you the preferences

level of precedence is command line > .gitignore > .git/info/exclude > core.excludesFile

19:40

so yeah, gitignore is killing my exclude file :c

19:40

https://git-scm.com/docs/gitignore

19:41

oh

19:41

"the last pattern decides the outcome" (edited)

Creating part , creating new idea thinkind yes cool , making , but not possible , fixing part , fixed nice , start part not possible , fixing ( maked harder a bit ) , becamed more better part then i intendet to make !

hard to understand

@Deleted User this is #developer not #off-topic. I am interested in your map tho.

Exported 213 message(s)