Time  Nick         Message
05:36 hmmmm        hmmm
05:37 hmmmm        how do i use treegen's copy of make_tree() when now it takes a specific seed, screwing up backwards compatibility
05:40 hmmmm        i need to make a way to get myrand's next and pseudorandom's next.
05:50 hmmmm        i need a way to return the seed of PseudoRandom used in make_tree()... this is tough
05:52 hmmmm        going to need to modify make_tree to take a PseudoRandom * and just use that.  no other way to do this with using disgusting hacks.  sorry RBA, your modification had unforseen consequences.
06:20 RealBadAngel that was not my idea to add seeds to trees
06:20 RealBadAngel i was PilzAdam's
06:21 RealBadAngel *it
06:22 hmmmm        nevermind, i'm just spazzing out.  it's fine.  i added the seed parameter back.
06:22 hmmmm        it's random enough.
06:23 hmmmm        i just can't help but feel like our randomness seriously sucks, both because our LCG sucks and the way previous values are re-used by breaking them off into PseudoRandom objects and throwing away the resulting enthropy
16:07 PrEAroD21H   Hi everyone
17:03 rubenwardy   changelog completed: http://dev.minetest.net/User:Rubenwardy/changelog
17:03 rubenwardy   can I put it into http://dev.minetest.net/Changelog?
17:03 rubenwardy   ping: PilzAdam, Calinou, hmmmm
17:10 celeron55    the credits seem a bit random at some points
17:12 celeron55    "Map generation is now slightly faster and can be tweaked in minetest.conf (celeron55 & RealBadAngel & PilzAdam)" <- i don't think i have had anything to do with this, and most of that is the work of kwolekr
17:13 celeron55    "Player physics are now tweakable by server admin (celeron55 && kwolekr)" <- nope; the authors are completely wrong, the real author is Taoki
17:14 RealBadAngel who made those texts?
17:14 celeron55    i guess most are originally by Calinou, a few are by hmmmm and the rest is by rubenwardy
17:15 RealBadAngel i shouldnt be mentioned in mapgen stuff
17:15 rubenwardy   100% of the text came from previous changlog
17:15 rubenwardy   I just reformated it to be easier to read
17:15 rubenwardy   feel free to change it
17:16 RealBadAngel put there just hmmm(kwolekr)
17:17 rubenwardy   adjusted
17:17 RealBadAngel i dont want to be blamed for mapgen being slow ;)
17:17 celeron55    changed those two that i mentioned
17:39 hmmmm        people still consider mapgen slow
17:39 hmmmm        even though it's like 5000x faster than it was two versions ago
17:41 Exio         everything is slow
17:41 Exio         (for the people)
17:41 Exio         "the world takes more than 2 seconds to load, it is TOO SLOW!!!!!"
17:42 hmmmm        i do agree that it's too slow, but people shouldn't be exactly whining about it
17:44 Jordach      hmmmm, yeah, it's faster than the 0.2.x mapgen
19:16 thexyz       https://github.com/minetest/minetest/pull/536
19:16 thexyz       hmmmm, darkrose, PilzAdam, celeron55 ^
19:17 hmmmm        oh god @_@
19:17 hmmmm        i really hope everything there works 100%
19:18 darkrose     this looks like a "merge and pray" commit
19:18 hmmmm        we need to make certain it's not going ot screw things up, but at the same time, we have to merge that asap or else it'll become problematic
19:19 Exio         for something is the git
19:19 Exio         it doesn't mean "super tested and everything working at the 100%"!
19:22 thexyz       one more suggestion: remove all #if 0 blocks
19:22 hmmmm        those are mostly in the mapgen
19:22 hmmmm        which is under heavy development at the moment
19:23 thexyz       of course, this commit breaks nothing ^^
19:23 thexyz       (because it's not merged yet)
19:24 hmmmm        you're absolutely sure you're not inadvertently inserting something that wasn't there by using []?
19:24 hmmmm        that's the #1 mistake really
19:25 thexyz       hm?
19:26 thexyz       any example?
19:27 hmmmm        well i don't mean your code in particular, i'm just saying that modifying a value like somemap[pos] = 5; would insert (pos, 5) if pos wasn't already there without you knowing that it was inserted
19:28 thexyz       doesn't irrlicht's .insert(key, value) work the same?
19:28 hmmmm        *shrug*
19:28 hmmmm        ah i like that you use std::set where appropriate
19:28 thexyz       oh, i see the difference
19:28 thexyz       it doesn't replace node when it already exists
19:29 thexyz       http://irrlicht.sourceforge.net/docu/classirr_1_1core_1_1map.html#af9f8f34cab620e3bdc1ae72715ab9d15
19:29 hmmmm        did you make that mistake at all?
19:29 hmmmm        :p
19:31 thexyz       I don't c++ that much
19:31 PilzAdam     thexyz, throw it in and wait for bug report
19:32 hmmmm        oh dear
19:32 PilzAdam     we cant find the bugs when simply not use the code
19:33 thexyz       PilzAdam: can you make a windows build and check if it's working right?
19:33 PilzAdam     sure
19:40 PilzAdam     http://ompldr.org/vaHBxZw/minetest-0.4.5-stl2-win32.zip
19:41 PilzAdam     no problems while compiling and works in wine
19:41 thexyz       great, let's find some windows users now
19:41 hmmmm        i am really excited about the code cleanup that's been going on lately
19:42 hmmmm        things are looking neat, consistent, professional
19:43 Jordach      ^^ i doubt it
19:55 thexyz       is everybody fine with me updating menulogo.png (Minetest-c55 → Minetest)?
19:55 PilzAdam     yes
19:55 hmmmm        yeah
19:55 hmmmm        the whole menulogo stuff is being updated along with the cloud thing anyway
19:56 hmmmm        i wonder if they got the two issues with it ironed out yet or not