11:23 <+sctt_> hi
11:23 <@heinrich5991> hi
11:29 < ddnet-commits> [ddnet] heinrich5991 opened pull request #845: Add Windows CMake compilation instructions (master...pr_ddnet_cmake_readme) https://git.io/v5ZAX
15:06 <@heinrich5991> sctt_: is the pull request more or less done now?
15:27 <@heinrich5991> is there someone capable of making good screenshot guides for windows? https://github.com/ddnet/ddnet/pull/845#discussion_r135739812
15:27 <@heinrich5991> Ryozuki: are you, or do you maybe know someone who is? :)
15:38 <+Ryozuki> i am able
15:41 <+Ryozuki> heinrich5991, you could extend https://wiki.ddnet.tw/Compiling_DDRaceNetwork 
15:41 <+Ryozuki> we got a nice wiki e.e why not use it
15:41 <+Ryozuki> (that page is outdated/incomplete btw)
15:44 <@heinrich5991> I updated the README, but I can also update that page
15:46 <@deen> Please only keep information in one place
15:46 <@deen> when people go to compile ddnet, they will be on github already, so keep the info there. Merge all the additional info from the wiki into that and make the wiki a link to github readme
15:47 <@heinrich5991> it doesn't look like the wiki page has extra info right now
15:47 <@heinrich5991> looks like a copy of the old readme
15:47 <+Ryozuki> yes...
15:47 <@deen> Don't make copies of information...
15:47 < ddnet-commits> [ddnet] def- closed pull request #845: Add CMake compilation instructions (master...pr_ddnet_cmake_readme) https://git.io/v5ZAX
15:47 <@deen> No one wants to maintain multiple versions
15:47 <+Ryozuki> 07:21, 18 June 2016‎ 
15:47 <+Ryozuki> anyways its rly old
15:48 <+Ryozuki> (last edit)
16:00 <+Ryozuki> http://i.imgur.com/NGMMxIp.png
16:00 <+Ryozuki> http://i.imgur.com/8HoUF90.png
16:00 <+Ryozuki> good?
16:00 <+Ryozuki> or to confusing
16:02 <+Ryozuki> (visual studio adds *lots* of files over the place, idk if there is some kind of clean cmd, also the end build ends up at Debug/ and the dlls needed arent there
16:03 <@heinrich5991> but if you execute it with visual studio it runs, right?
16:03 <@heinrich5991> I vote for "too confusing" btw
16:03 <@heinrich5991> soo many arrows
16:03 <+Ryozuki> so
16:03 <+Ryozuki> you want lot of images instead?
16:03 <@heinrich5991> I think that would actually be better
16:04 <@heinrich5991> interesting points also are the install C++ extensions for MSVS and the install python for *all users*
16:05 <+Ryozuki> visual studio fills the ddnet folder with shit
16:05 <+Ryozuki> and it is run from ddnet/build
16:05 <+Ryozuki> i dont see why it does that
16:05 <@heinrich5991> does it? which files for example?
16:08 <+Ryozuki> http://i.imgur.com/01wm6Oq.png
16:08 <+Ryozuki> http://i.imgur.com/F2TND1o.png
16:09 <@heinrich5991> :D that's not good
16:09 <@heinrich5991> let me boot my windows computer
16:10 <+sctt_> heinrich5991: yes, this morning i've tested a lot map_replace_image tool and it seems to work well in many situations, so i guess im not going to add any other commits
16:11 <@heinrich5991> could you squash all these commits into one?
16:11 <@heinrich5991> wait a second, I can probably tell you the right command
16:13 <@heinrich5991> sctt_: you with me? :)
16:13 <+sctt_> oh ok
16:13 <+sctt_> yes :D
16:13 <+sctt_> i thought you could do something like "Squash and merge"
16:13 <+sctt_> i saw i somewhere
16:13 <@heinrich5991> are you in the git command line with inside the repository?
16:13 <+sctt_> yes
16:14 <@heinrich5991> what is the upstream git repository named, do you already have a name for that?
16:14 <@heinrich5991> git remote -v can show you that
16:15 <+sctt_> yes wait 1 sec
16:16 <@heinrich5991> Ryozuki: it hasn't created these folders for me, they're probably artifacts from an earlier build
16:16 <+Ryozuki> mmm
16:16 <+Ryozuki> let me test again
16:18 <+sctt_> heinrich5991: origin	https://github.com/sctt/ddnet.git (fetch) origin	https://github.com/sctt/ddnet.git (push)
16:18 <@heinrich5991> sctt_: do "git remote add remote https://github.com/ddnet/ddnet"
16:18 <+sctt_> done
16:19 <@heinrich5991> git fetch upstream
16:19 <@heinrich5991> and then you can do git rebase upstream/master
16:19 <+sctt_> fatal: 'upstream' does not appear to be a git repository
16:19 <+sctt_> :(
16:19 <@heinrich5991> duh
16:19 <@heinrich5991> git remote rename remote upstream
16:19 <@heinrich5991> a typo from me
16:20 <+Ryozuki> https://help.github.com/articles/syncing-a-fork/
16:20 <+Ryozuki> oh
16:20 <+Ryozuki> its not that :P
16:20 <+sctt_> heinrich5991: ok it worked now^^
16:20 <+sctt_> Ryozuki: thanks i will read that :D
16:20 <@heinrich5991> then do git rebase upstream/master
16:21 <+sctt_> ok it applied
16:21 <+Ryozuki> i recloned ddnet
16:21 <+Ryozuki> im getting a error with cmake now
16:21 <+Ryozuki> :D
16:21 <@heinrich5991> :(
16:21 <+Ryozuki> it doesnt use bundled libs?
16:21 <@heinrich5991> Ryozuki: btw, you can clean your repository with git clean
16:21 <+Ryozuki> CMake Error at CMakeLists.txt:126 (message):
16:21 <+Ryozuki>   You must install Curl to compile the DDNet client
16:21 <+Ryozuki> + with all libs
16:21 <@heinrich5991> Ryozuki: did you clone with --recursive?
16:22 <@heinrich5991> otherwise do a git submodule update --init
16:22 <+ddnet-discord> <Ryozuki> https://cdn.discordapp.com/attachments/293493549758939136/352095607076159489/Screenshot_6.png
16:22 <+Ryozuki> hein
16:22 <+Ryozuki> oh
16:22 <+Ryozuki> im dumb
16:22 <+Ryozuki> :D
16:22 <+Ryozuki> always forget that..
16:22 <@heinrich5991> sctt_: do a git rebase upstream/master again and mark all but the first commit as "squash"
16:22 <@heinrich5991> and don't reclone ^^
16:23 <+sctt_> heinrich5991:git rebase upstream/master Current branch master is up to date.
16:23 <@heinrich5991> ah right
16:23 <@heinrich5991> I ment git rebase -i upstream/master
16:23 <@heinrich5991> -i for interactve
16:23 <+sctt_> oh i see
16:24 <+sctt_> my editor opened, how to mark the commits as squash? xD
16:24 <+sctt_> i see pick 46d51e2 Update map_resave.cpp
16:24 <@heinrich5991> replace pick by squash
16:24 <+sctt_> should i rename pick into squash>
16:24 <@heinrich5991> yes
16:24 <+sctt_> oh ok
16:24 <+Ryozuki> heinrich5991, using the cmake gui and building with vs2017 fills the ddnet folder also
16:25 <@heinrich5991> I have vs2015. mh
16:25 <@heinrich5991> have you selected a different build folder?
16:25 <+sctt_> can i rename the first commit name during this process?
16:25 <+Ryozuki> Desktop/ddnet/build
16:25 <@heinrich5991> you'll be prompted afterwards, sctt_
16:25 <+sctt_> ok :D
16:26 <+Ryozuki> heinrich5991, looks like vs is generating the object files at ddnet folder instead of build
16:26 <+Ryozuki> idk why
16:26 <@heinrich5991> so I should install msvs 2017? :(
16:27 <+Ryozuki> it currently builds generating the object files at ddnet and then moves those to the build folder
16:27 <+Ryozuki> but the folders remain there
16:27 <@heinrich5991> hmm
16:27 <+sctt_> heinrich5991: ok it finished
16:27 <+sctt_> but i still see all the commits
16:28 <@heinrich5991> then do git push origin +master
16:29 <+sctt_> heinrich5991: ok it worked :D i see only one commit now, thank you
16:30 <+Ryozuki> git clean -f doesnt remove the folders :/
16:30 <@heinrich5991> -fd
16:30 <@heinrich5991> but do -nd first
16:30 <@heinrich5991> to see which it would remove
16:31 <+Ryozuki> to late
16:31 <+Ryozuki> λ git clean -fd
16:31 <+Ryozuki> Removing build/data/
16:31 <+Ryozuki> doesnt remove folders
16:31 <@heinrich5991> git clean -nxd
16:31 <@heinrich5991> that also removes files ignored by git
16:31 <@heinrich5991> (the -x)
16:31 <+Ryozuki> that looks good
16:32 <+Ryozuki> and worked
16:32 <+Ryozuki> -xdf
16:33 <+Ryozuki> mmm what about writing a bat script to build it
16:33 <+Ryozuki> and cleaning the mess
16:33 <+ddnet-discord> <deen> what mess?
16:33 <+ddnet-discord> <deen> build should be simple cmake && make
16:33 <+Ryozuki> the mess that vs2015 produces
16:33 <+Ryozuki> 2017*
16:33 <+ddnet-discord> <deen> use a separate build directory and tell microsoft not to make a mess?
16:34 <+Ryozuki> i cant, the program does what it wants
16:34 <+Ryozuki> i built in a separate folder
16:34 <+Ryozuki> yet it generates a bunch of folders on ddnet root folder
16:34 <@heinrich5991> I'm currently installing msvs 2017, I'll check what went wrong
16:34 <+ddnet-discord> <deen> that sounds like a bug
16:34 <@heinrich5991> I don't even know how it knows about the existence of that folder
16:35 <@heinrich5991> other than the fact that it is the shared root of all source files
16:35 <+Ryozuki> im gonna try from command line and using msbuild
16:36 <+Ryozuki> oh
16:36 <+Ryozuki> mmm
16:36 <+Ryozuki> forget what i say
16:36 <+Ryozuki> said*
16:37 <+Ryozuki> (tried again, still makes folders on ddnet)
16:37 <@heinrich5991> is the problem resolved?
16:37 <+Ryozuki> no
16:37 <@heinrich5991> I just aborted my installation of msvs 2017 ^^
16:37 <@heinrich5991> meh
16:37 <+Ryozuki> D:
16:38 <+Ryozuki> im sry for confusing you
16:38 <+Ryozuki> https://github.com/opencv/opencv/issues/7796
16:38 <+Ryozuki> ^
16:39 <+Ryozuki> another project, with same issue
16:40 <+Ryozuki> and this is the cmake bug issue https://gitlab.kitware.com/cmake/cmake/issues/16458
16:40 <+Ryozuki> (or looks like)
16:41 <+Ryozuki> and yes, looks like its a vs bug
16:41 <+Ryozuki> can't do nothing then
16:42 <@heinrich5991> "This issue has been fixed and is now available in our latest update. You can download the update via the in-product notification or from here: https://www.visualstudio.com/vs/ Thanks to everyone who provided feedback, votes and solutions for this problem."
16:42 <@heinrich5991> Ryozuki: could you try updating your vs? this post is from august the 14th
16:43 <+Ryozuki> hmm maybe wait
16:43 <@heinrich5991> "Closing this, as I can confirm this has been fixed in Visual Studio 15.3.0."
16:43 <+Ryozuki> (i got 15.0 new is 15.3, hope this fixes it)
16:43 <@heinrich5991> in the cmake issue
16:43 <+Ryozuki> oh
16:43 <+Ryozuki> nice
16:43 <+Ryozuki> will update
16:45 <@heinrich5991> sctt_: unfortunately there are quite some minor style issues remaining, but that shouldn't be hard to fix
16:45 <@heinrich5991> I could either annotate the appropriate locations in the pull request
16:45 <@heinrich5991> or, since I'm in an awesome mood today, fix it myself :)
16:46 <+sctt_> xD
16:46 <@heinrich5991> the former is probably better for you, the latter is easier for you :)
16:46 <+Ryozuki> haha 
16:46 <+sctt_> if you tell me what's the problem i can fix it, np
16:46 <@heinrich5991> kk, then let me begin the review :)
16:46 <+sctt_> ok :D
16:46 <+sctt_> i have to go out in 20 mins so i'll fix it tonight
16:47 <@heinrich5991> sure, no pressure
16:47 <+Ryozuki> i should have thought about updating before, my bad :/
16:47 <@heinrich5991> np
17:22 <@heinrich5991> deen: have you tried this function already? https://msdn.microsoft.com/en-us/library/windows/desktop/ms633543(v=vs.85).aspx
17:27 <@deen> yeah, didn't work according to user reports
17:28 <@deen> i think SDL2 has functions for dpi stuff, but they actually do nothing
17:29 <@heinrich5991> this one? SDL_WINDOW_ALLOW_HIGHDPI
17:29 <@deen> yes
17:29 <@heinrich5991> https://wiki.libsdl.org/SDL_WindowFlags
17:29 <@deen> it might be implemented on macosx, but apparently not yet on windows
17:33 <+Ryozuki> updated vs and the bug is gone
17:33 <@heinrich5991> good :)
17:34 <@heinrich5991> I added a comment to the opencv issue: https://github.com/opencv/opencv/issues/7796#issuecomment-325700775 :)
17:34 <+Ryozuki> nice
17:35 <+Ryozuki> tho the dlls arent on the output folder
17:35 <+Ryozuki> but on the build folder
17:35 <+Ryozuki> maybe its possible to add a postbuild command for vs to move these?
17:36 <@heinrich5991> but can you start the programs from the msvs UI?
17:36 <+Ryozuki> yep
17:37 <@heinrich5991> then it should be good, right?
17:38 <+Ryozuki> not if i run outside vs
17:38 <+Ryozuki> cause dlls arent copied
17:38 <+Ryozuki> to the output folder
17:39 <+Ryozuki> https://stackoverflow.com/questions/9994045/copy-target-file-to-another-location-in-a-post-build-step-in-cmake
17:39 <+Ryozuki> smth like that would fix it
18:25 <@Nimda> Zadrotos_final by Opana just released on Oldschool at 2017-08-29 18:21
20:34 <+Toer> hey, it seems that my isp latley has problems with the master servers of tw, does anyone has a script to fetch the addresses i could put on my server, or already set up something like this?
21:04 <+sctt_> heinrich5991: i've read your revision, that's great! but i have some doubts and observations about some point... can we brefly discuss?
21:04 <@heinrich5991> sure
21:04 <+sctt_> ok so, a general thing
21:05 <+sctt_> many things you pointed out are about pieces of code i just copied from tw source
21:05 <+sctt_> like LoadPNG and ExtractName
21:05 <+sctt_> or general structure of main, it is from map_resave
21:06 <+sctt_> does it mean i should also change also the methods i copied
21:06 <+sctt_> or just change map_replace_image?
21:06 <@heinrich5991> if you want, you can also change map_resave
21:07 <@heinrich5991> but it would suffice if the new code is better :)
21:07 <@heinrich5991> ddnet source code does have some dark corners that border on unreadability
21:07 <@heinrich5991> but we won't fix it all at once, no big refactor or anything
21:08 <+sctt_> ok so i will focus on the tool
21:08 <@heinrich5991> sure :)
21:08 <+sctt_> consider that the things you pointed out on LoadPNG and ExtractName
21:08 <+sctt_> are like this also inside the main code
21:10 <@heinrich5991> hm. maybe I'll do a follow-up fix of the main code
21:10 <+sctt_> ok :D
21:10 <@heinrich5991> extractname e.g. is from 2010 :)
21:10 <+sctt_> than i have a couple of more questions
21:10 <@heinrich5991> go ahead :)
21:10 <+sctt_> where you say
21:10 <+sctt_> "Is this line still needed?"
21:10 <+sctt_> you are referring to an include list
21:11 <@heinrich5991> yes
21:11 <+sctt_> which include were you referring exactly?
21:11 <@heinrich5991> to the graphics
21:11 <+sctt_> oh
21:11 <@heinrich5991> engine/graphics.cpp
21:11 <+sctt_> thats for image info structure
21:11 <@heinrich5991> .h
21:11 <@heinrich5991> ah ok
21:11 <@heinrich5991> so yes, it's still needed
21:11 <+sctt_> ok^^
21:11 <+sctt_> than
21:11 <+sctt_> then*
21:12 <+sctt_> This allocates and leaks the CMapItemImage. Either make it global (meh) or pass in a pointer. Ask me about this one if you don't know what to do. :)
21:12 <+sctt_> why leakead? it is needed for datafile.addData
21:12 <+sctt_> addItem sorry
21:12 <@heinrich5991> leaked means: it is never deleted
21:12 <+sctt_> yes but i dont want to delete it
21:12 <+sctt_> cause it is needed
21:12 <+sctt_> where datafile.Finished
21:13 <@heinrich5991> after the call to additem, you don't need it anymore
21:13 <@heinrich5991> additem makes a copy
21:13 <+sctt_> oh
21:13 <+sctt_> i didnt know that
21:13 <+sctt_> i thought he kept the reference
21:13 <@heinrich5991> yea, the joys of c/c++ :)
21:13 <+sctt_> but i noticed, when i do getItem
21:13 <+sctt_> i don't have a copy
21:13 <+sctt_> i have his reference
21:14 <+sctt_> that's why i thought it was the same with addItem
21:16 <@heinrich5991> but even if it were the case, you'd still leak it after finish
21:16 <@heinrich5991> – for every new you must have a delete somewhere
21:16 <@heinrich5991> but in this case, it should be doable without new/delete
21:17 <+sctt_> if that were the case, i would just have to delete the datawriter
21:17 <+sctt_> his destroyer should deallocate the items i guess
21:18 <@heinrich5991> no
21:18 <+sctt_> so it doesnt :/
21:18 <+sctt_> so i have to save all the items into an array and deallocate them after the finish
21:18 <@heinrich5991> or jsut don't allocate using new at all
21:19 <+sctt_> but considering that after the .finish the program quits and all its memory willl be released, i dont think its worth it
21:19 <+sctt_> i can't not allocating them
21:19 <+sctt_> the function needs the return a void pointer
21:19 <+sctt_> and i can't use the same Item i'm reading, i have to clone it into another memory area
21:19 <+ddnet-discord> <Savander> Did you ever think about some kind of Code Reformatter? To standardize whole code, at least in structure? Without changing naming convention?
21:20 <+sctt_> i know that because before i was using the same Item
21:20 <+sctt_> and the result map was messy, cause the reader probably was doing his calculation on the item i modified while i shouldnt have done that
21:20 <+sctt_> so i decided to clone it instead of directly modify it
21:20 <+sctt_> and that solved the problem
21:21 <@heinrich5991> you could passi n a pointer to a CMapItemImage
21:21 <@heinrich5991> given by the caller
21:21 <@heinrich5991> and just write in there
21:21 <@heinrich5991> the caller just puts the CMapItemImage on the stack
21:21 <@heinrich5991> like CMapItemImage Buffer;
21:22 <+sctt_> yes but it is the main who does addItem
21:22 <+sctt_> so the new Item can't be destroyed by the ReplaceItemFunction
21:22 <@heinrich5991> ReplaceImageItem(.., &Buffer);
21:23 <+sctt_> oh
21:23 <+sctt_> so add a new parameter on ReplaceImageItem
21:23 <+sctt_> referring to the new Item
21:24 <+sctt_> that will oblige me to do more checks inside the main
21:24 <+sctt_> to see if it is allocated
21:24 <@heinrich5991> (feel free not to fix ExtractName if you mean)
21:24 <@heinrich5991> *if you want
21:25 <@heinrich5991> not really
21:25 <+sctt_> why? you want to fix it? xD
21:25 <@heinrich5991> I can fix it later, yes
21:25 <@heinrich5991> see, you can do it like this:
21:26 <@heinrich5991> ReplaceImage item works like before, but gets passed the pointer to CMapItemImage *pBuffer
21:26 <@heinrich5991> if you change the image stuff, just return pBuffer
21:26 <@heinrich5991> if not, do the stuff like before
21:27 <@heinrich5991> this way, you either get back a pointer to the Buffer (which is a variable in main) or to the DatafileReader item
21:27 <@heinrich5991> you don't have to deallocate either of these, so you don't need any extra code in main
21:27 <+sctt_> so you mean
21:27 <+sctt_> the main will create the new CMapItemImage
21:28 <+sctt_> as an object
21:28 <@heinrich5991> without any "new" keywords involved
21:28 <@heinrich5991> yes
21:28 <+sctt_> and pass it as a pointer
21:28 <@heinrich5991> yes
21:28 <+sctt_> and the function will return that pointer
21:28 <+sctt_> if it was actually the wanted item
21:28 <+sctt_> yeah that could work :D
21:29 <+sctt_> heinrich5991: ok that's clear now
21:30 <@heinrich5991> cool
21:30 <+sctt_> heinrich5991: last question, you said to print messages like usage
21:30 <+sctt_> what do you use in ddnet for that?
21:30 <+sctt_> dgb_msg?
21:30 <@heinrich5991> just use dbg_msg
21:30 <@heinrich5991> doesn't look nice, but it's something
21:30 <+sctt_> ok^^
21:30 <+sctt_> so i will have to compile in debug mode to see them, correct?
21:31 <@heinrich5991> no, it always displays them
21:31 <+sctt_> oh good
21:31 <+sctt_> in stderr?
21:31 <+sctt_> or stdout?
21:31 <@heinrich5991> no, stdout
21:31 <@heinrich5991> IIRC
21:31 <+sctt_> okok
21:31 <+sctt_> very last question xD
21:31 <+sctt_> you said you want to fix ExtractName
21:31 <+sctt_> so should i still fix LoadPNG?
21:32 <+sctt_> it's a tw pasted code that too
21:32 <@heinrich5991> yes, that seems very easy to fix
21:32 <+sctt_> ok cool^^
21:33 <+sctt_> ok so i will start to rework the code in a while and let you know :D
21:34 <@heinrich5991> cool
21:34 <@heinrich5991> just ping me on github when you're done
22:46 <+sctt_> heinrich5991: curiosity, if you are using hungarian notation, why you don't put 'n' before integers? xD
22:47 <@heinrich5991> it's only used to distinguish where a variable comes from in TW
22:47 <@heinrich5991> g_ for globals, m_ for members, s_ for statics
22:47 <@heinrich5991> and then a for arrays and p for pointers, mh.
22:47 <+sctt_> yeah that's hungarian notation
22:48 <+sctt_> you use it everywhere but not for integers
22:48 <+sctt_> xD
23:51 <+sctt_> @heinrich5991 you still here?
23:52 <+sctt_> heinrich5991:
23:52 <@heinrich5991> yes
23:52 <@heinrich5991> :)
23:52 <+sctt_> :)
23:52 <+sctt_> i've finished the fixes
23:52 <+sctt_> but...
23:52 <+sctt_> i've noticed i don't get any output from dbg_mesg
23:52 <+sctt_> :/
23:53 <@heinrich5991> put dbg_logger_stdout() into your main function
23:54 <@heinrich5991> before any dbg_msg statements
23:55 <+sctt_> ok :D
23:56 <+sctt_> heinrich5991: oh god
23:56 <+sctt_> i've got like 10 lines of garbage
23:56 <+sctt_> like [2017-08-29 23:50:47][storage]: added path '$USERDIR' ('/root/.teeworlds')
23:56 <@heinrich5991> that's ok
23:56 <@heinrich5991> WAIT
23:56 <@heinrich5991> why do you execute stuff as root?
23:56 <+sctt_> well if he spams, maybe i should spam too xD
23:56 <+sctt_> i'm on a VM
23:56 <@heinrich5991> ok
23:57 <@heinrich5991> (still doesn't hurt to use a user account, but ok)
23:58 <+sctt_> it's a live
23:58 <+sctt_> he has only root by default