| Time |
Nick |
Message |
| 00:18 |
Fedora29 |
rubenwardy, u good with leveldb issues? |
| 00:19 |
|
Unarelith joined #minetest |
| 00:19 |
* Fedora29 |
cries for help XD |
| 00:33 |
Fedora29 |
hmm i think i got it |
| 00:43 |
|
est31 joined #minetest |
| 01:27 |
|
riff-IRC joined #minetest |
| 01:32 |
|
puzzlecube joined #minetest |
| 01:52 |
|
piesquared joined #minetest |
| 02:22 |
|
ANAND joined #minetest |
| 04:16 |
|
jas_ joined #minetest |
| 05:02 |
|
Xiong joined #minetest |
| 05:05 |
Xiong |
So since I upgraded to the Dreambuilder technic I noticed the MV Hydro. On a short look it seemed like a good and honest deal and I tore out my LV plant and crafted it up. Ran MV cable and dropped one of the new turbines. |
| 05:08 |
Xiong |
Works like a charm 100%! Only thing is, my plant is deep underwater and I usually have to take special fiddly care to see it's flowing water on the watermills. Didn't think the rules would be different for the MVs. |
| 05:10 |
Xiong |
So I aim for flowing on four sides, level and just below, total of 8 nodes. Actually one side is good for 33% so three sides is enough. |
| 05:11 |
Xiong |
The MV is kicking out full rated power but I'm thinking there might be source, not flowing... I tinker. |
| 05:12 |
Xiong |
End is, I cover all four sides (horizontal) with pie boxes. 0% as expected. I pop one box... 100%!? |
| 05:14 |
Xiong |
Obviously I can rebuild my hydro plant along extremely compact lines if I exploit this. I have just one question: Is this feature or bug? |
| 05:30 |
Xiong |
If feature, fine; I'll figure out the new rules and do my best. But if bug then, I fear, someone may fix it... and I'll be tearing up a big hydro plant to do over for no gain. Rather plan wisely today. |
| 05:35 |
VanessaE |
it doesn't have to be on all 4 sides |
| 05:36 |
VanessaE |
the way I wrote that code, it just adds all four sides' apparent flow and caps it at a max amount. |
| 05:36 |
VanessaE |
but I think at most you're supposed to need at least 2 sides at full blast (vertical flow) |
| 05:37 |
VanessaE |
at best you're* |
| 05:37 |
VanessaE |
just because of the math |
| 05:55 |
Xiong |
The LV watermill code states clearly, three sides is max. This is fair and forgiving. Hard to read the code but the math seems to check. |
| 05:56 |
VanessaE |
it's supposed to be 3 sides' worth of max flow. |
| 05:56 |
Xiong |
That's why one side... and node below... goes 33%. |
| 05:57 |
VanessaE |
yep |
| 05:57 |
VanessaE |
four sides with 25% flow would do also. |
| 05:58 |
Xiong |
What I don't see is detail in how water flow is calculated between nodes. There's an "in pairs" call to what I assume is a core engine function. |
| 05:58 |
|
illwieckz joined #minetest |
| 05:58 |
VanessaE |
the flowing node's param2 is closely mapped to flow rate (even if it's unrealistic) |
| 05:59 |
VanessaE |
so it's just a matter of adding them up and multiplying to get the number of EU. |
| 05:59 |
Xiong |
But... that's a minor point. What happens on the level below may nick that 33%. But there must be water on the same-y side node or 0%. |
| 06:00 |
VanessaE |
a node above or below isn't supposed to matter at all |
| 06:00 |
VanessaE |
only X/Z adjacent |
| 06:00 |
Xiong |
Argh. |
| 06:01 |
VanessaE |
unless someone changed it of course |
| 06:01 |
Xiong |
Not so with LV. The ideal is 8 flowing nodes at y and y-1. |
| 06:01 |
VanessaE |
if it behaves like that, someone changed it. |
| 06:01 |
Xiong |
Been that way for 5 years. |
| 06:01 |
VanessaE |
no |
| 06:01 |
VanessaE |
definitely not. |
| 06:01 |
VanessaE |
I made the flowing water change maybe 2 years ago |
| 06:02 |
Xiong |
Yes. I played then, that was the advice given, that's how I built my plant, it worked. |
| 06:02 |
VanessaE |
and I explicitly made it look for flowing in just the X/Z adjacent nodes, at the same Y level, and not diagonally |
| 06:02 |
Xiong |
Which is different. |
| 06:03 |
Xiong |
With LV. The ideal is 8 flowing nodes at y and y-1. |
| 06:03 |
VanessaE |
*shrug* |
| 06:04 |
Xiong |
MV hydro is uncharted waters for me. |
| 06:04 |
VanessaE |
MV is supposed to just be a redeco of LV + increased EU |
| 06:04 |
crazyR |
Does anyone know if its possible to get player meta data for a player thats offline? |
| 06:04 |
Xiong |
Doesn't seem so. |
| 06:05 |
VanessaE |
I knew I shoulda made the code check for a certain amount of water above the turbine (as in reservoir head :) ) |
| 06:05 |
Xiong |
LV doesn't do that. |
| 06:05 |
VanessaE |
I know. |
| 06:06 |
VanessaE |
it's just something I was considering when I made the source -> flowing changes |
| 06:06 |
Xiong |
So... I have one MV in service... with three y sides covered in pie boxes... pumping 100%. |
| 06:06 |
VanessaE |
pie boxes? |
| 06:07 |
Xiong |
What bozzers me is I have one on a test stand by itself... a minimal network... and it behaves just as the LV does. |
| 06:08 |
Xiong |
They're your pie boxes. Not sure it matters... they're not water. |
| 06:08 |
Xiong |
Want me to test with cobble? |
| 06:09 |
VanessaE |
I don't remember making any pie boxes.... |
| 06:09 |
Xiong |
Nonissue I think. |
| 06:09 |
Xiong |
They're cheap and break fast. Good for fooling around. |
| 06:09 |
* VanessaE |
looks at the MV turbine code. |
| 06:10 |
|
KingBull joined #minetest |
| 06:10 |
VanessaE |
well this looks like my LV code so at least that part's right. |
| 06:10 |
Xiong |
The point is, one unit in service, odd. One unit on test stand, normal. |
| 06:11 |
VanessaE |
if minetest.get_item_group(node.name, "water") == 3 then |
| 06:11 |
VanessaE |
hm. |
| 06:11 |
VanessaE |
odd indeed. |
| 06:11 |
|
KingBull left #minetest |
| 06:12 |
Xiong |
The difficulty with saying definite things about unit in service is, it's deep underwater. Hard to tell certainly what's flowing and what's source. |
| 06:12 |
VanessaE |
I still want to know what these pizza boxes are. |
| 06:13 |
Xiong |
So... let me return to testing. I will put source on all four sides. |
| 06:13 |
VanessaE |
I can only recall a pizza food item, not any boxes. |
| 06:13 |
Xiong |
Pie boxes. |
| 06:13 |
VanessaE |
er pie boxes |
| 06:13 |
VanessaE |
(mental translation, pie == pizza pie :) ) |
| 06:13 |
VanessaE |
OH WAIT |
| 06:13 |
VanessaE |
you mean the baskets |
| 06:14 |
VanessaE |
from the bushes classic mod |
| 06:14 |
Xiong |
Bushes -> berries + sugar etc -> pies. |
| 06:14 |
Xiong |
One can eat an entire full pie box. The empty box has useful handling characteristics. |
| 06:15 |
VanessaE |
yeah now I remember :P |
| 06:15 |
Xiong |
This cannot be germane. I hope. |
| 06:15 |
* VanessaE |
looks at its code. |
| 06:15 |
VanessaE |
groups = { dig_immediate = 3 }, |
| 06:16 |
Xiong |
Cheap, too. |
| 06:16 |
VanessaE |
hm. putting one of those next a turbine should not do anything at all |
| 06:17 |
Xiong |
Let me testmore. |
| 06:24 |
Xiong |
On the test stand with only air at the y+1 level, flooding the y level with source yields 100%. |
| 06:24 |
Xiong |
As I thought, unclear what's going on in the working plant. Hell to tell what's baffled. |
| 06:25 |
VanessaE |
water sources should not make it run at all. |
| 06:25 |
Xiong |
Naturally, setting source on four sides fills in the corners. |
| 06:25 |
VanessaE |
tell me, what's the param2 of one of the neighboring water sources? |
| 06:26 |
Xiong |
What's disturbing is, throwing exactly on pie box... or stone tile... drops the turbine to 0%. |
| 06:27 |
Xiong |
s/on/one |
| 06:27 |
VanessaE |
screenshots? |
| 06:27 |
Xiong |
How would I see an internal parm? |
| 06:28 |
VanessaE |
node inspector |
| 06:28 |
Xiong |
I will shoot a few for you. This will take time. |
| 06:28 |
Xiong |
Sorry... "node inspector" not recognized. Wizard tool? |
| 06:28 |
VanessaE |
i' |
| 06:29 |
VanessaE |
it's one of the mods that comes with dreambuilder, made by sofar |
| 06:29 |
VanessaE |
"inspector" I think |
| 06:30 |
Xiong |
No doubt sth I threw overboard. The phone's running hot as it is. I told you, I cut. |
| 06:30 |
VanessaE |
you DO realize more mods != more CPU, right? |
| 06:31 |
Xiong |
More mods bad. |
| 06:31 |
VanessaE |
most of dreambuilder's mods are idle things, no active code. |
| 06:31 |
Xiong |
Maybe so. But a ton of feathers weighs as much as a ton of lead. |
| 06:32 |
VanessaE |
that's not how mods work. |
| 06:33 |
VanessaE |
ah hah, I can reproduce the MV bug locally |
| 06:33 |
Xiong |
Look, it's foolish to run mt at all on this rig. Takes forever to load. Then I installed technic &co... took forever and a day. Then threw it all out and installed half of dreambuilder. Now I have time to smoke a rasta spliff while it finds room to load. |
| 06:34 |
VanessaE |
water has param2 = 240, this is a unified dyes bug manifesting in the hydro code :) |
| 06:34 |
Xiong |
And I don't even smoke... so it takes a long time to finish. |
| 06:34 |
VanessaE |
actually technically this is an engine bug |
| 06:35 |
Xiong |
Well... you don't need the shots then? |
| 06:36 |
VanessaE |
no |
| 06:36 |
Xiong |
K. |
| 06:37 |
Xiong |
I will rebuild my hydro plant as if 8 nodes if flowing are best. Please don't break that? |
| 06:37 |
VanessaE |
I simply placed some turbines parpendicular to a wall, removed a couple to leave one standing in mid-water, and hooked it up. checked the neighboring water -- that water which the engine filled in when I removed the extra turbines now has param2 = 240, which means unified dyes set that value when I placed the turbines. |
| 06:38 |
VanessaE |
I'm sorry, you're exploiting a bug ;) |
| 06:38 |
Xiong |
Wat? |
| 06:38 |
VanessaE |
now, why the LV turbine isn't affected, I'm not sure. |
| 06:39 |
Xiong |
Maybe it is. |
| 06:39 |
Xiong |
I just proposed to rebuild my plant as if I knew nothing of what I've seen today. What exploit? |
| 06:39 |
VanessaE |
either way, the turbine is only working because of an invalid param2 value in one of its adjacent water nodes. |
| 06:39 |
VanessaE |
that^ is the exploti |
| 06:40 |
VanessaE |
I'll submit a fix later to make technic ignore non-flowing water, and I'll figure out and fix why unified dyes is setting that value to begin with (see, there's a global on-place, and 240 is a special value) |
| 06:41 |
|
Markow joined #minetest |
| 06:41 |
Xiong |
If I set flowing water on four sides, y and y-1, for a total of 8... it will not work? |
| 06:41 |
VanessaE |
stop with the y-1 |
| 06:41 |
VanessaE |
the turbine does. not. look. there. |
| 06:41 |
Xiong |
Why? |
| 06:41 |
Xiong |
Maybe it will someday. |
| 06:42 |
VanessaE |
it ONLY looks at the same Y level as the turbine itself, and no it won't. right, nore? ;) |
| 06:42 |
|
illwieckz joined #minetest |
| 06:42 |
VanessaE |
if you have flowing water adjacent, it'll work. that's the correct behavior |
| 06:43 |
Xiong |
That's how the LV works. I can't be sure someone, somewhere, will not make the MV device conform. |
| 06:43 |
VanessaE |
I am just about to. |
| 06:43 |
Xiong |
? |
| 06:44 |
Xiong |
Sorry but I have a relentlessly logical mind. Either you just contradicted yourself or I don't understand you. |
| 06:44 |
VanessaE |
I am about to fix the bug so that both MV and LV require flowing water only. |
| 06:45 |
VanessaE |
then the param2 value will behave as intended |
| 06:46 |
Xiong |
That's not my point. You said to ignore the four y-1 nodes. That's not okay for LV watermill. |
| 06:46 |
VanessaE |
yes, it is. LV turbine does not care what's at Y-1 |
| 06:47 |
VanessaE |
https://github.com/minetest-mods/technic/blob/master/technic/machines/LV/water_mill.lua#L36 |
| 06:47 |
Xiong |
Okay. But... actual practice differs. |
| 06:48 |
VanessaE |
not for long. :P |
| 06:50 |
Xiong |
Well... in what version? Look, I'm afraid we're talking in circles. This is how it was, how it is, version after version, world after world. If you can prevail to change that, okay. But you will have to hobnob with your steamroll your fellow wizards. |
| 06:51 |
VanessaE |
I'm a technic contributor. I can fix it. :) |
| 06:52 |
Xiong |
That's drastic. Go ahead. Always tough to set up a hydro plant. If only 3 flowing nodes required at y level, significantly easier. |
| 06:52 |
Xiong |
... much easier. |
| 06:53 |
Xiong |
You might also poke around and see if the LV OP bug is related |
| 06:53 |
Xiong |
... does not surface in dreambuilder. |
| 06:54 |
VanessaE |
https://github.com/minetest-mods/technic/pull/471 |
| 06:54 |
VanessaE |
nolsen: ^ |
| 06:54 |
VanessaE |
er |
| 06:54 |
VanessaE |
nore: ^ |
| 06:54 |
VanessaE |
they behave as intended now. |
| 06:55 |
VanessaE |
wait a minute |
| 06:55 |
Cornelia |
And suddenly I want to play Minetest Technic |
| 06:55 |
VanessaE |
derp mode engaged |
| 06:55 |
* Cornelia |
shall wait for 5.0.0 :p |
| 06:55 |
Xiong |
Before loading dbr, I was getting 1575 W out of a single LV mill. |
| 06:55 |
VanessaE |
there. derp mode disengaged :) |
| 06:56 |
VanessaE |
a single string.find() fixed the MV turbine |
| 06:56 |
Xiong |
After dbr, output power dropped to advertised 180. |
| 06:56 |
VanessaE |
LV already worked as intended, I'm not sure how I managed to mix up the code, sorry about that |
| 06:57 |
nore |
VanessaE: lgtm |
| 06:57 |
VanessaE |
the LV had an explicit check for flowing water, the MV didn't. |
| 06:57 |
* nore |
goes back to sleep |
| 06:57 |
VanessaE |
mergy merge merge ;) |
| 06:58 |
nore |
On my phone right now, merging is hard :p |
| 06:58 |
Xiong |
Okay... I'm going to rebuild my plant. The lights are turning brown. |
| 06:58 |
VanessaE |
Xiong: you may not have to, just let the water floooooow |
| 06:58 |
VanessaE |
nore: oh ok |
| 06:58 |
nore |
Will do later |
| 06:58 |
VanessaE |
I just used to click the merge button and let the site handle it :) |
| 06:59 |
nore |
Yep, but I need to be logged in for that... |
| 06:59 |
Xiong |
VE I do not have a hydro plant now. I cut it down to build MV. Waiting avails nothing. |
| 06:59 |
VanessaE |
I'll look into the "240" glitch tomorrow, I'm pretty sure I know where that lies. |
| 06:59 |
VanessaE |
Xiong: oh ok, misunderstood. |
| 07:00 |
Xiong |
Glad we made progress. Gotta go dig and place. I think I will not do this underwater. |
| 07:01 |
VanessaE |
ah hah |
| 07:01 |
VanessaE |
it is not UD doing the 240. just a coincidence. |
| 07:10 |
VanessaE |
can reproduce without UD. that's a relief. |
| 07:11 |
VanessaE |
nore: thanks :) |
| 07:11 |
nore |
you're welcome :) |
| 07:11 |
* nore |
*really* needs to get out of bed now |
| 07:14 |
VanessaE |
heh |
| 07:19 |
VanessaE |
https://github.com/minetest/minetest/blob/stable-0.4/src/map.cpp#L900 |
| 07:19 |
VanessaE |
there's the cause. |
| 07:19 |
VanessaE |
so the 240 isn't really, precisely a bug |
| 07:19 |
|
SanskritFritz joined #minetest |
| 07:20 |
VanessaE |
Xiong: I'll issue a new dreambuilder build. |
| 07:23 |
VanessaE |
done. |
| 07:29 |
|
illwieckz joined #minetest |
| 07:29 |
|
Helwyr joined #minetest |
| 07:33 |
|
CWz joined #minetest |
| 07:36 |
* VanessaE |
pokes Xiong with a build ID :P |
| 08:01 |
|
illwieckz joined #minetest |
| 08:20 |
|
marisag joined #minetest |
| 08:41 |
|
illwieckz joined #minetest |
| 08:53 |
|
nowhere_man joined #minetest |
| 09:05 |
|
illwieckz joined #minetest |
| 09:17 |
|
Beton joined #minetest |
| 09:31 |
|
troller joined #minetest |
| 09:49 |
|
illwieckz joined #minetest |
| 09:55 |
|
troller joined #minetest |
| 10:05 |
|
troller joined #minetest |
| 10:05 |
|
LaCosa joined #minetest |
| 10:05 |
|
LaCosa joined #minetest |
| 10:09 |
|
aheinecke joined #minetest |
| 10:10 |
|
Fixer joined #minetest |
| 10:15 |
|
illwieckz joined #minetest |
| 10:19 |
|
jluc joined #minetest |
| 10:21 |
Xiong |
VanessaE Installed hydro_turbine.lua from ec25fd83 in place of _ from 856b8685 and experimented. |
| 10:24 |
Xiong |
Hydros are arranged in a staircase with cables under-behind. Top mill reports 100% (with source directly above). Next mill down reports 64% in the following environment: |
| 10:27 |
Xiong |
y, rear face cable, otherwise flowing; y-1, cable directly under, to rear, and to one side. |
| 10:27 |
Xiong |
Rm the side cable improves output. |
| 10:28 |
|
troller joined #minetest |
| 10:29 |
Xiong |
That's fine, can work with that. Banks must be connected somehow but I can do that well below the entire staircase. |
| 10:30 |
Xiong |
But... not as expected. Output still depends on flowing water at y-1 neighborhood. |
| 10:31 |
Xiong |
I assume the rear cable costs sth too... but the cap math means three is enough... as before. |
| 10:33 |
Xiong |
Um... this is exactly as it was in the previous version. I only replaced the indicated file. Please say if the hydro fix spans other files. |
| 10:39 |
Xiong |
On an unrelated note... gonna need a better forceload anchor strat. Currently one anchor cranked up to theoretically cover entire base. But when I reentered world at new hydro plant all mills reported no network. When I checked the switch it reported plenty of power. When I returned to plant, sure enough, mills on. |
| 10:40 |
Xiong |
I detest heisenbugs |
| 10:41 |
Xiong |
Night all! |
| 10:48 |
dogetest |
What did the Ethereum said to Minetest? |
| 10:49 |
|
troller joined #minetest |
| 10:51 |
dogetest |
Let's go to MetaMask party? |
| 10:57 |
|
jspiros joined #minetest |
| 11:02 |
|
ensonic joined #minetest |
| 11:05 |
|
illwieckz joined #minetest |
| 11:16 |
|
LaCosa joined #minetest |
| 11:16 |
|
LaCosa joined #minetest |
| 11:31 |
|
troller joined #minetest |
| 11:31 |
|
illwieckz joined #minetest |
| 11:53 |
|
rtjure joined #minetest |
| 11:55 |
|
troller joined #minetest |
| 12:03 |
|
Beton joined #minetest |
| 12:11 |
|
riff-IRC joined #minetest |
| 12:19 |
|
illwieckz joined #minetest |
| 12:36 |
|
ANAND joined #minetest |
| 12:43 |
|
Gael-de-Sailly joined #minetest |
| 12:54 |
|
illwieckz joined #minetest |
| 12:56 |
|
piesquared joined #minetest |
| 13:20 |
|
illwieckz joined #minetest |
| 13:47 |
|
illwieckz joined #minetest |
| 14:07 |
|
scr267 joined #minetest |
| 14:30 |
|
illwieckz joined #minetest |
| 14:31 |
|
Wuzzy joined #minetest |
| 14:34 |
|
troller joined #minetest |
| 14:47 |
|
troller joined #minetest |
| 14:53 |
|
Foz joined #minetest |
| 15:03 |
|
scr267 joined #minetest |
| 15:03 |
|
illwieckz joined #minetest |
| 15:10 |
|
ANAND joined #minetest |
| 15:19 |
|
MindUser0_0 joined #minetest |
| 15:37 |
|
troller joined #minetest |
| 15:41 |
|
Rafi59 joined #minetest |
| 15:42 |
|
Unarelith joined #minetest |
| 15:45 |
|
illwieckz joined #minetest |
| 15:54 |
|
nowhere_man joined #minetest |
| 15:55 |
|
piesquared joined #minetest |
| 15:59 |
|
Markow joined #minetest |
| 16:03 |
|
Unarelith joined #minetest |
| 16:04 |
|
Markow joined #minetest |
| 16:17 |
|
Akberid joined #minetest |
| 16:39 |
|
illwieckz joined #minetest |
| 16:52 |
|
troller joined #minetest |
| 16:52 |
|
jas_ joined #minetest |
| 16:55 |
|
proller__ joined #minetest |
| 16:57 |
|
sec^nd joined #minetest |
| 17:10 |
|
piesquared joined #minetest |
| 17:11 |
|
Cornelia joined #minetest |
| 17:14 |
|
illwieckz joined #minetest |
| 17:16 |
|
piesquared joined #minetest |
| 17:27 |
|
Gael-de-Sailly joined #minetest |
| 17:35 |
|
illwieckz joined #minetest |
| 17:40 |
|
GeHa joined #minetest |
| 17:40 |
|
nowhere_man joined #minetest |
| 17:50 |
|
illwieckz joined #minetest |
| 17:51 |
|
piesquared joined #minetest |
| 17:53 |
|
LMD joined #minetest |
| 18:38 |
|
Krock joined #minetest |
| 19:00 |
|
ensonic joined #minetest |
| 19:04 |
|
Unarelith joined #minetest |
| 19:14 |
|
piesquared joined #minetest |
| 19:17 |
|
Xiong joined #minetest |
| 19:17 |
|
puzzlecube joined #minetest |
| 19:21 |
|
piesquared joined #minetest |
| 19:25 |
|
red-001 joined #minetest |
| 19:38 |
|
illwieckz joined #minetest |
| 19:53 |
VanessaE |
Xiong: only the one file is replaced, and the behavior is exactly as intende |
| 19:53 |
VanessaE |
intended |
| 19:53 |
VanessaE |
s/replaced/affected/ |
| 20:16 |
|
proller joined #minetest |
| 20:59 |
|
Cornelia joined #minetest |
| 21:05 |
|
jimvideo joined #minetest |
| 21:24 |
|
GreenDimond joined #minetest |
| 21:24 |
|
jluc joined #minetest |
| 21:30 |
|
GreenDimond left #minetest |
| 22:06 |
|
Fixer_ joined #minetest |
| 22:10 |
|
nowhere_man joined #minetest |
| 22:14 |
|
proller joined #minetest |
| 22:22 |
|
FreeFull joined #minetest |
| 22:35 |
|
Fixer joined #minetest |
| 23:27 |
|
piesquared joined #minetest |
| 23:44 |
|
riff-IRC joined #minetest |
| 23:59 |
|
LaCosa joined #minetest |
| 23:59 |
|
LaCosa joined #minetest |
