Present: jK, hoijui, Tobi, abma, Kloot
Agenda
* Welcome
* Release plan
* communication enhancement with mod-devs
* what exactly is the scope of the answer site?
* include FLOZi's changes?
* Next meeting
* Anything else? (WVTTK)
=== Welcome ===
<hoijui>who will do the minutes?
<hoijui>should i?
<abma>hm..
me!<hoijui>.. mm ok :D
<hoijui>thanks!
=== Release plan ===
<hoijui>Kloot.. i guess you have done the stuff you wanted to by now, right?
<hoijui>for a .7 release, i mean
<jK>I have to do ~2 commits for .7
<hoijui>ok
<hoijui>any estimate on that?
<Kloot>yeah, nothing else that cannot wait (based on lack of feedback)
<hoijui>ook
<hoijui>for the record: what feedback would you want?
<hoijui>just general testing of latest changes?
<Kloot>something like "I have tested build 0xab12cd and issue 123 that I complained about is now fixed"
<hoijui>ahh ok
<hoijui>i did tests with S44, and to me it seems to work well now.
<hoijui>in respect to tanks getting stuck
<hoijui>jk?
<hoijui>cant give any estimate?
<hoijui>a week, two?
<hoijui>just asking casue.. people usually ask me this a lot
<Kloot>it'd be nice if some of the more vocal BA crowd could be bothered to confirm that too :)
<hoijui>when next release will be
<jK>just need to merge master back in
<hoijui>mmm
<hoijui>ah ok
<hoijui>ah yeah kloot, have you seen my KAIK crash report?
<Kloot>yeah but the trace didn't seem to make sense
<hoijui>mm
<hoijui>hmm.. it seems to look ok to me
<hoijui>except.. that i can not see what the problem shoudl be at the line it crashes
<hoijui>also.. all the crashess had hte same stack trace
<hoijui>the stack trace itsself seems coherent (often one line off though, as so often)
<Kloot>[21:19:52] <Kloot> yeah but the trace didn't seem to make sense
<Kloot>[21:20:00] <[ARP]hoijui_g5> mm
<hoijui>well.. maybe we can discuss this later...
<Kloot>trace itself is coherent yes, but the crash location does not look
<Kloot>viable
<hoijui>mm
<hoijui>ok
<hoijui>i may try if reverting that same commit that caused the hangs will fix it
<hoijui>so we coudl use that as a workaround for .7, if we can nto track it down otherwise
<Kloot>oh, is it reproducable consistently?
<Kloot>even with fd13290a01271904b0da81f3bb814b767ad25249 applied?
<hoijui>here yes,, i had it in every game
<hoijui>3 out of 3
<hoijui>aehm..
<hoijui>ahh no, did not try with that yet
<hoijui>good call, will do that after meeting
<Kloot>k
<hoijui>so theres the two jk commits nad that KAIK problem.
<hoijui>if that is solved, i will ask some guys to test and .. then we make .7, if it looks good
<Kloot> http://github.com/spring/spring/commit/ ... 3b8e94a506 should also go into .7 btw (already promised someone
)<hoijui>ok, sure
=== communication enhancement with mod-devs ===
<hoijui>look at etherpad maybe, for sub topics
<hoijui>shoudl we make each one its own topic?
<Kloot>hmm, the one about scope maybe
<hoijui>ok
<hoijui>by koshi: take the answer site out of bootstrap mode? (awarding more karma then in normal mode)
<hoijui>he said we shoudl decide this
<hoijui>bootsrap mode is for getting the dynamics running
<hoijui>so we get some guys with lots of karma fast, which then act as moderators
<hoijui>if we have enough of them, we can turn that off
<Kloot>can't karma be handed out manually by admins?
<hoijui>i dont know...
<hoijui>mayb this, or you can manually make someone moderator.. yeah is kind of the same
<hoijui>to me it is ok too if we turn it off already
<hoijui>it seems to be used enough
<hoijui>and we have at least 3 admins
<Kloot>agreed
<Tobi>we could post that we will turn it off in X days, prodding people into quickly participating a bit more to get some more rep :p
<hoijui>:D
<abma>with an big counter on top?
<hoijui>also ok wiht me
<abma>yep, turn it off
<hoijui>maybe we should have a look how many have how much karma, and how much is needed
<Tobi>doubt there is much theory on how much is needed
<hoijui>:D :D
<hoijui>i mena.. shoudl be some setting somewhere
<Tobi>oh you mean for particular features
<hoijui>yes
<Tobi>I have that written somewhere in a nice list
<Tobi>or maybe not.. only badges and how many rep certain actions gives
<hoijui>Minimum reputation to edit others posts: 400
<hoijui>close others posts: 600
<hoijui>delete comments: 400
<Tobi>people can still flag much earlier though
<hoijui>yeah..
<hoijui>this is the highest stuff
<Tobi>and iirc if a post is flagged 5 times or so it is hidden
<hoijui>this is stuff that like.. makes you moderator, kind of..
<Tobi>yeah
<hoijui>Minimum reputation to close own question: 60
<Tobi>well I also think it is ok to turn off bootstrap sometime this week
<abma>then wait until two people have >400?
<abma>that should be sometime this week :)
<Tobi>nah I prefer just setting a date
<hoijui>only one guy is close to 400 (excluding admins)
<Tobi>simply change at next meeting?
<Tobi>then I will post later this week that on monday we change it
<Tobi>hopefully attracting some more people
<hoijui>+1
<Kloot>hoping people care enough about some number on a website :)
<hoijui>:D
<hoijui>it is magic man! it's karma!!
=== what exactly is the scope of the answer site? ===
<Tobi>post count on spring forum was hidden because of that IIRC
<hoijui>:D
<Kloot>heh
<hoijui>so yeah.. the initial idea was, that it is only for mod-dev -> engine-dev
<hoijui>and mod-dev <-> mod-dev
<Tobi>well, that was the focus, but at least I personally didn't really mean to 100% exclude other things
<hoijui>thats also what was communicated in Tobis thread for example
<hoijui>ok
<Tobi>on the forum it might have been a bit different though
<hoijui>so you thing we should endorse all spring related stuff there?
<Tobi>but atm my opinion is that we should not try to kill otherwise good Spring related questions solely because they don't have to do with modding
<hoijui>think*
<jK>imo keep it clean
<hoijui>yeah
<hoijui>i want a clear rule
<jK>non modding related questions should be removed, and forum should be used for it
<hoijui>either mod stuff only, or everything (orwhatever)
<Tobi>so mapping and AI questions should be removed?
<hoijui>i do not care, but i want a clear rule
<jK>mapping & AI are modding related
<hoijui>:D :D
<Tobi>and building installers?
<abma>IMHO, too
<jK>it would be nice if the engine installer would have a plugin for mod installations
<abma>the quality matters i think...
<hoijui>welll.. we defined it as mod-dev communication enhancement
<jK>so all games/mods can use the engine installer for their ones
<abma>(quality of questions + answers)
<hoijui>i personalyy do not include AI devs and map makers in that category
<hoijui>what do you mena wiht plugin?
<jK>simple 1-2 steps to generate a custom installer for a specific game
<Tobi>[LCC]jK: look at koshi's installer creator thing
<jK>e.g. place files in folder xyz and then start make_nsis.sh
<Tobi>I don't know exactly what is needed for it or how many steps it is but it might be pretty close to what you want
<hoijui>i also remember that koshis thing allwos skinnign the instalelr or soemthing
<hoijui>and i guess it will be easier to use then what i woudl come up with
<hoijui>when trying to allow what you propose, jk
<hoijui>aehm.. we kind of got off-topic :D
<hoijui>so ...
<hoijui>it is for all sorts of questions between devs related to spring?
<hoijui>may it be engine, mod, map or AI?
<jK>yeah, just not user related things
<Tobi>hmm true yeah
<hoijui>"I get only white units (ATI), what can i do?"
<hoijui>these not, yeah?
<abma>thats for the forum...yep
<Tobi>so lets put the line at no end-user questions | only any spring related development (whether it be AI/map/mod/custom installer building/widget making/etc.) questions
<hoijui>ok!
<Tobi>we may need to take care of the tags a little though, soonish
<hoijui>yeah :D
<Tobi>maybe I'll re-tag some questions on wednesday so tags are more consistent
<hoijui>good
<hoijui>i might also do some, if i get bored
<hoijui>next?
<hoijui>ouh..
<hoijui>abma mentioned quality
<hoijui>have you seen any you think are not good enough?
<hoijui>we said we either down rate them or delete?
<hoijui>i saw two so far
<hoijui>one of them, i was able to convince the author that he shoudl instead put it on forum
<hoijui>so he deleted it
<hoijui>and the other is still there
<hoijui>http://answers.springlobby.info/questio ... chitecture
<jK>no design discussions
<hoijui>which is also off-topic, in my eyes
<hoijui>yeah
<jK>no discussions at all
<hoijui>yeah.. i propose to enforce deleting there, instead of downvoting
<hoijui>next?
<Tobi>faq updated - http://answers.springlobby.info/faq/
<hoijui>
<hoijui>nice
<hoijui>"AI, games, maps, gadgets or widgets"
<hoijui>ouh ... :D doh
<hoijui>i though mods is missing, but there is games
<hoijui>perfect for me then
=== include FLOZi's changes ===
<hoijui> http://github.com/spring/spring/pull/3
<hoijui>i think, jk made one of them obsolete already
<hoijui>the other one is this: http://github.com/FLOZi/spring/commit/a ... 763fc91003
<hoijui>should i merge it in?
<hoijui>looks good to me
<hoijui>he siad the smae is done for features and units already
<hoijui>(defs)
<Kloot>that's fine
<hoijui>ok
<hoijui>will include it then
<jK>why the revert of PlaySoundFile?
<Kloot>he re-applied it
<hoijui>on top, there is a link/button: Files Changed
=== Next meeting ===
<hoijui>next meeting, smae day, same time?
<hoijui>for me ok
<Tobi>same
<Kloot>+1
<jK>fine
=== Anything else? (WVTTK) ===
<Tobi>thx for the website fixes abma, I will apply them on wednesday probably, still a bit busy with paper writing and other stuff
<hoijui>jk, the last two of the commits you pushed some minutes ago..
<hoijui>are the ones you meant to be in the release?
<abma>hm.. what does WVTTK mean? :-D
<Tobi>lol
<hoijui>:D i remember.. some dutch thing
<abma>"Wat Verder Ter Tafel Komt" ?
<Tobi>it means: Wat Verder Ten Tafel Komt, i.e. whatever else you want to bring in
<Tobi>yeah
<abma>aah, ok
<Tobi>it's funny it got stuck in the title
<jK><[ARP]hoijui_g5> jk, the last two of the commits you pushed some minutes ago.. > all except the loadingscreen one
<hoijui>ok
<jK>also commiting now some other stuff, but that isn't needed to be in the release
<hoijui>ok
<abma>[LCC]jK: rts/Game/Game.cpp:2352: error: ÔÇÿconst struct CMapInfo::map_tÔÇÖ has no member named ÔÇÿhumanNameÔÇÖ
<abma>(could be, that i was to quick in pulling/compiling :)
<jK>happens when you have 10000 uncommited changes and do a textsearch
<abma>:)
<hoijui>hehe :D
<hoijui>ouh man!
<hoijui>now that you have the new machine
<hoijui>and surely plenty of disc space
<hoijui>you should maybe have some separate local repos of spring
<hoijui>and keep the one you are workign on mostly clean
<jK>didn't got a new hdd
<hoijui>ah :D
<jK>hdd + PSU are the only old things left in the case ^^
<hoijui>mmm
<hoijui>and no much free space on hd?
<abma>hmm... a bit off-topic: why do we have nearly no people from asia here? do they play in general no RTS games, or ...?
<Kloot>no, they all play starcraft
<hoijui>they have star craft 1
<abma>hmm, ok... impressive-double-answer :D
<hoijui>:D
<Kloot>there was some chinese TA/spring community at one point
<hoijui>they had a mod, right?
<Kloot>yes iirc
<hoijui>for my checking-AI-Interface-ABI-compatibility problem...
<hoijui>woudl anyone have an idea?
<hoijui>one other solution woudl be, to usa a manual version
<hoijui>so whenever you change the callback, you have to change the version
<hoijui>i just cant imagine this to work in practise
<hoijui>what do you think?
<hoijui>well.. its mainly Kloot and me doing changes there
<hoijui>but.. it is tedious enough wihtout that already
<hoijui>but the only automatic tools that seem well fit for the job that i foudn so far, are written in perl
<hoijui>that woudl have to be run by CMake so... perl woudl be a build dep
<hoijui>and i am not even sure if it woudl work for windows builds.. (i see changes as high for MinGW, low for other compilers)
<Kloot>the only way I see is to compare hashes over the parsed binary representation of a header
<hoijui>yeah.. that could work, though i am not sure if we would get the smae results there for different compilers
<hoijui>and. i am also not sure how exactly to do it
<hoijui>would need somethign like .. a separate project in cmake for each header, cause only one precompiled header is allowed per project
<hoijui>or an empty main.cpp that includes all hte headers
<Kloot>well, if that method does not generate the same hash for different compilers then there is no way to decide ABI compatibility other than writing a custom pseudocompiler for C
<hoijui>yet an other option, is to stripp the headers with .. eg an AWK script (get rid of comments mainly) and hash the results
<hoijui>mm :/
<hoijui>i dont know how this perl tool does it
<hoijui>it takes headers and compield lib, and generates some intermediate file(s)
<Kloot>can you read perl? ;p
<hoijui>:D
<hoijui>no
<hoijui>yeah i know..
<hoijui>i also dont want that
<Kloot>hmm
<hoijui>using an AWK script is relatively easy, as we already require AWK anyway
<hoijui>it might indicate incompatibility when there is none, but not the other way around
<Kloot>but that can not do anything more than verify equal textual representations
<hoijui>yeah
<hoijui>for me, this would be enough already
<hoijui>this version shoudl be used to comapre if the smae version is used (header wise)
<hoijui>if it is incmpatible due to other reasons.. liek different callign convention, or 64/32 bitness or so..
<hoijui>it is basically a bug in code or packaging or something
<Kloot>or possibly different build flags
<Kloot>it just depends on how sure you want to be :)
<hoijui>yeah true
<hoijui>with this thing, i really just want to comapre different API versions, ABI wise
<hoijui>if we get problems with different compielrs or flags or so, we can still solve that
<hoijui>i cant remember we had that so far
<hoijui>as AFs Shard compiles from VS seem ot work fine on MinGW spring
