| Time |
Nick |
Message |
| 00:09 |
|
TempestMancer joined #minetest |
| 00:27 |
|
AliasAlreadyTake joined #minetest |
| 00:37 |
|
Wikiwide joined #minetest |
| 00:52 |
|
Verticen joined #minetest |
| 00:53 |
|
specing_ joined #minetest |
| 01:12 |
|
garywhite joined #minetest |
| 01:13 |
|
MTDiscord1 joined #minetest |
| 01:40 |
MTDiscord |
<MisterE> how do I get the stack from an '__builtin:item' objectreference? |
| 01:41 |
|
SuperMiner joined #minetest |
| 01:42 |
MTDiscord |
<Jonathon> ItemStack(obj:get_luaentity().itemstring) |
| 01:43 |
MTDiscord |
<MisterE> great, thx |
| 01:52 |
|
absurb joined #minetest |
| 01:57 |
|
evlc joined #minetest |
| 02:04 |
|
troller joined #minetest |
| 02:22 |
|
turtleman joined #minetest |
| 02:33 |
|
Hawk777 joined #minetest |
| 02:48 |
|
fluxionary joined #minetest |
| 03:20 |
|
fluxionary joined #minetest |
| 03:28 |
|
queria^clone joined #minetest |
| 03:34 |
|
queria^clone joined #minetest |
| 04:11 |
|
v-rob joined #minetest |
| 04:28 |
|
Verticen joined #minetest |
| 04:35 |
|
fluxionary joined #minetest |
| 04:49 |
|
delta23 joined #minetest |
| 05:00 |
|
MTDiscord joined #minetest |
| 05:36 |
BoostedMenchers |
seems that minetest still gets the DPI for fonts the same way it always has |
| 05:36 |
BoostedMenchers |
when the X server is configured with the screen's actual dimensions minetest renders the font too large |
| 05:36 |
BoostedMenchers |
at least on displays that have a real DPI higher than 96 |
| 05:36 |
BoostedMenchers |
which is … most modern displays |
| 05:39 |
|
leo_rockway joined #minetest |
| 05:55 |
erlehmann |
BoostedMenchers yeah it should use font DPI |
| 05:55 |
erlehmann |
not DPI |
| 06:01 |
|
riff-IRC joined #minetest |
| 06:04 |
|
v-rob joined #minetest |
| 06:18 |
|
epoch joined #minetest |
| 06:22 |
erlehmann |
BoostedMenchers you tested https://github.com/minetest/minetest/pull/11848 or what? |
| 06:29 |
BoostedMenchers |
looking at it |
| 06:36 |
erlehmann |
BoostedMenchers i remain a bit sceptical of the 72 dpi claim, but if it works … |
| 06:36 |
erlehmann |
v-rob do you have a mac? |
| 06:36 |
v-rob |
I do not. |
| 06:37 |
v-rob |
However, the definition of a point is 1/72 of an inch, i.e. 72 DPI |
| 06:37 |
v-rob |
Or PPI, more accurately |
| 06:37 |
v-rob |
But it amounts to the same thing |
| 06:37 |
erlehmann |
well it is good that you looked into it i guess |
| 06:37 |
erlehmann |
so how about font dpi? |
| 06:38 |
erlehmann |
every other application i have tried uses font dpi, not dpi |
| 06:38 |
erlehmann |
those are two different settings, at least on X11 |
| 06:39 |
BoostedMenchers |
the two things I've seen that display oversized fonts with the real DPI configured are GTK2 and Irrlicht |
| 06:39 |
v-rob |
Not sure about what that is. All I know is that fonts are generally measured in points (unless you use CSS px or something), and Minetest has been measuring them in some non-existent measurement unit that is kind of like a point but not really. In any case, using points is an improvement, and Unifont works correctly at a size of 12 like it should. |
| 06:41 |
erlehmann |
v-rob look at this https://linuxreviews.org/HOWTO_set_DPI_in_Xorg#How_DPI_Values_Are_Used |
| 06:41 |
erlehmann |
v-rob, you have linux? |
| 06:42 |
v-rob |
I use Windows. I used to have a Linux VM sitting around, but not ATM. |
| 06:42 |
erlehmann |
v-rob wait, i give you a screenshot |
| 06:44 |
BoostedMenchers |
“This also sets the value reported as physical size of the X screen as a whole (union of all configured monitors). In configurations with multiple monitors with different DPIs, the value has no physical meaning, but it may be used by some legacy clients which do not support RandR version 1.2 to compute a reference font scaling.” |
| 06:44 |
BoostedMenchers |
from the xrandr manpage, regarding "--dpi from-output" |
| 06:44 |
erlehmann |
v-rob the bottom setting changes font dpi https://mister-muffin.de/p/7mRE.png |
| 06:48 |
v-rob |
OK, so from what I can see, Minetest just doesn't check font DPI -- it just uses the system DPI to scale everything uniformly. Hence, my code uses the system DPI and won't reflect the font DPI. However, if you provided some code to get font DPI for X11, it probably wouldn't be too difficult to make it use font DPI |
| 06:49 |
BoostedMenchers |
what exactly is the font_require_multiple setting? |
| 06:49 |
BoostedMenchers |
the description seems to imply it is a boolean but it takes an integer |
| 06:49 |
erlehmann |
the naming could use work lol |
| 06:50 |
BoostedMenchers |
an integer >1 |
| 06:50 |
v-rob |
Unifont is a pixel font, so it only looks good at size 12, 24, 36, 48, etc. If you set that setting to 12, it'll only allow the font to be one of those values |
| 06:50 |
erlehmann |
12 what |
| 06:50 |
v-rob |
points |
| 06:50 |
v-rob |
The name definitely could use work |
| 06:50 |
BoostedMenchers |
pixel font == bitmap font? |
| 06:51 |
BoostedMenchers |
ok |
| 06:51 |
erlehmann |
the problem is also that unifont is a 16 pixel high bitmap font |
| 06:51 |
erlehmann |
so where do i set “16 pixel” as the size? |
| 06:52 |
v-rob |
They're measured in points, like said before. If you look at the Unifont page (http://unifoundry.com/unifont/index.html) you can find that the default size is 12 points, which, for this font in particular, will turn out to be 16 pixels tall. |
| 06:53 |
v-rob |
It's unfortunate that fonts are measured in points, but points really doesn't correspond well to their actual height. That's why things like CSS em are a good idea. |
| 06:53 |
erlehmann |
i have contributed to unifont. i am more thinking about how to actually configure a pixel font base size. |
| 06:54 |
BoostedMenchers |
it has font_path, font_path_bold, font_path_italic, font_path_bold_italic… |
| 06:54 |
BoostedMenchers |
are you reimplementing fontconfig in minetest or irrlicht :P |
| 06:54 |
v-rob |
Also, pixel font != bitmap font. A pixel font is a design, but a bitmap font is a format for storing a font, as opposed to TTF |
| 06:54 |
erlehmann |
v-rob, i'll research font dpi. btw, i think it is a thing in windows too. it is called “custom sizing” https://www.top-password.com/blog/wp-content/uploads/2016/11/custom-sizing-options.png |
| 06:55 |
erlehmann |
i might be wrong about this |
| 06:55 |
BoostedMenchers |
ah |
| 06:55 |
v-rob |
Pretty sure that Windows doesn't have font DPI. It has scaling factors that scale stuff universally though. |
| 06:56 |
erlehmann |
ok |
| 06:57 |
erlehmann |
ok so the standard is standard is 1 pt = 1/72" |
| 06:58 |
BoostedMenchers |
I'm still not entirely sure what's happening |
| 06:58 |
BoostedMenchers |
the X server by default reports one of it's DPI values as 96 no matter what |
| 06:58 |
BoostedMenchers |
they changed it, briefly, and the result was some clients showed oversized fonts (GTK2 and Irrlicht) |
| 06:59 |
BoostedMenchers |
it's been reverted |
| 06:59 |
erlehmann |
so 12 pt should be 0.16" tall on a 96dpi display |
| 06:59 |
erlehmann |
BoostedMenchers that was the *default* |
| 06:59 |
erlehmann |
you can still change it |
| 06:59 |
BoostedMenchers |
yeah |
| 06:59 |
BoostedMenchers |
I have it reporting the real DPI now |
| 07:00 |
erlehmann |
and do you get crisp or blurry unifont? |
| 07:00 |
BoostedMenchers |
haven't tried |
| 07:00 |
erlehmann |
pls try |
| 07:00 |
BoostedMenchers |
lets see I have it installed |
| 07:01 |
BoostedMenchers |
found the path |
| 07:01 |
erlehmann |
i do not agree with v-rob that it is ok to remove pixel font support in a minor release (als note that the current state is broken so that you can end up in a situation where you have no working font), but i do agree with him that there is a significant benefit to having non-blurry fonts work with a single code path and pixel fonts can probably be deprecated if *all* of that works. |
| 07:02 |
erlehmann |
als → also |
| 07:02 |
erlehmann |
i meant “bitmap fonts can be deprecated” |
| 07:03 |
erlehmann |
v-rob btw, how can i prevent the rainbow spam with ttf font, is there any option? |
| 07:03 |
v-rob |
Rainbow? |
| 07:03 |
erlehmann |
some ppl make every letter a different color using CSMs |
| 07:03 |
erlehmann |
it is very hard to read |
| 07:03 |
erlehmann |
so i use non-colored pixel font |
| 07:05 |
v-rob |
Oh, disabling colorization of fonts. I'm not sure where I stand on that TBH, but it wouldn't be difficult. If the other core devs think that's a good idea, then I can add it to the PR. |
| 07:06 |
BoostedMenchers |
I don't see a difference with font_require multiple being 1 or 12 |
| 07:06 |
BoostedMenchers |
http://menche.us/files/font2.png |
| 07:06 |
erlehmann |
BoostedMenchers try 16 |
| 07:06 |
erlehmann |
or 13 |
| 07:07 |
v-rob |
Try setting the font to a different size than 12 |
| 07:07 |
erlehmann |
and turn off antialiasing |
| 07:07 |
v-rob |
If multiple is 12 and size is 13, the size should be overridden to be 12 |
| 07:07 |
erlehmann |
now we do have a problem here, the default font size is 16pt |
| 07:08 |
BoostedMenchers |
so multiple=12, size=13, antialiasing off? |
| 07:08 |
erlehmann |
v-rob, but if the dpi was wrong, the default should be changed to 12pt |
| 07:08 |
BoostedMenchers |
oh antialiasing makes a bit of a difference |
| 07:08 |
v-rob |
For Unifont size 12 with the PR, it shouldn't be necessary to disable anti-aliasing |
| 07:08 |
v-rob |
The PR changes the default from 16 to 12 |
| 07:08 |
v-rob |
for font size |
| 07:08 |
BoostedMenchers |
wow it's thinner with that off |
| 07:08 |
erlehmann |
well that image is surely blurry http://menche.us/files/font2.png |
| 07:09 |
erlehmann |
BoostedMenchers can you give a non-blurry image? |
| 07:09 |
BoostedMenchers |
yeah that has antialiasing on |
| 07:09 |
BoostedMenchers |
which settings? |
| 07:09 |
BoostedMenchers |
so multiple=12, size=13, antialiasing off? |
| 07:09 |
v-rob |
Yeah, see, that's a problem. Your screenshot says the font size is 12, so that means my PR apparently doesn't work properly on Linux. That's problematic. |
| 07:10 |
BoostedMenchers |
https://menche.us/files/font3.png |
| 07:10 |
erlehmann |
uh |
| 07:10 |
BoostedMenchers |
make sure to zoom because the browser will smooth out the PNG when it's zoomed out |
| 07:10 |
v-rob |
Actually, do you have a custom GUI scaling or screen density scaling factor applied? That might cause a difference. |
| 07:10 |
BoostedMenchers |
I'll set X's DPI to 96 |
| 07:10 |
|
Sven_vB joined #minetest |
| 07:11 |
BoostedMenchers |
https://menche.us/files/font4.png |
| 07:11 |
erlehmann |
BoostedMenchers can you post the screenshots including the exact settings and OS and DPI at which you took the screenshot to the PR? |
| 07:11 |
BoostedMenchers |
ok |
| 07:12 |
BoostedMenchers |
see it is a lot smaller now |
| 07:12 |
BoostedMenchers |
that's what I'm used to seeing |
| 07:12 |
erlehmann |
oh LOL |
| 07:12 |
|
ShadowNinja joined #minetest |
| 07:12 |
erlehmann |
BoostedMenchers, the slashes |
| 07:12 |
BoostedMenchers |
I just did `xrandr --dpi 96` before relaunching mt for that shot |
| 07:13 |
BoostedMenchers |
those |
| 07:13 |
BoostedMenchers |
aren't straight lines |
| 07:13 |
BoostedMenchers |
lol |
| 07:13 |
|
valhalla joined #minetest |
| 07:13 |
v-rob |
Enable antialiasing. Disabling it can cause artifacts. It's a trade-off |
| 07:13 |
v-rob |
Artifacts such as the slashes |
| 07:13 |
BoostedMenchers |
I just set font-size to 12 |
| 07:13 |
|
evlc joined #minetest |
| 07:13 |
BoostedMenchers |
no more artifacts, antialiasing still off |
| 07:14 |
v-rob |
Cool |
| 07:14 |
BoostedMenchers |
https://menche.us/files/font5.png |
| 07:14 |
v-rob |
So, I apparently need to do more DPI experimentation |
| 07:14 |
BoostedMenchers |
err wait |
| 07:14 |
BoostedMenchers |
nvm |
| 07:14 |
erlehmann |
ok so this is evidence that the patch does not work right on linux? https://mister-muffin.de/p/QVe2.png |
| 07:14 |
BoostedMenchers |
it was on again |
| 07:14 |
erlehmann |
the 3 is broken too |
| 07:14 |
erlehmann |
but the slashes are funny |
| 07:15 |
erlehmann |
v-rob as i said, i am willing to help. i want crisp fonts, *how* i get them is not that important. |
| 07:15 |
v-rob |
That's what disabling antialiasing does for Windows as well. As said, you get crisp fonts no matter what with disabled antialiasing, but you get artifacts |
| 07:15 |
v-rob |
Turn antialiasing back on |
| 07:15 |
BoostedMenchers |
funny, with antialiasing off it looks *better* with font_require_multiple=13 |
| 07:15 |
BoostedMenchers |
at least the slashes |
| 07:16 |
v-rob |
OK, listen up! Antialiasing looks **better** for sizes that are not multiples of 12, but it looks **worse** for non-multiples of 12. That's how it works |
| 07:16 |
|
chilledfrogs joined #minetest |
| 07:17 |
BoostedMenchers |
ok |
| 07:17 |
BoostedMenchers |
lol |
| 07:17 |
v-rob |
Err, I mean looks worse for multiples of 12 |
| 07:17 |
BoostedMenchers |
that seems counterintuitive on its face |
| 07:17 |
BoostedMenchers |
wonder what happens when you throw subpixel filtering and hinting into this! |
| 07:18 |
|
Teckla joined #minetest |
| 07:19 |
v-rob |
This is all in relation to getting crisp pixel fonts, of course. For normal fonts, none of that applies, and you leave antialiasing on. |
| 07:20 |
BoostedMenchers |
I find myself unable to form an opinion regarding what could be considered a sane default here |
| 07:21 |
erlehmann |
i would call the setting crisp_font_base_size AND make it local to every font |
| 07:22 |
erlehmann |
or pixel_font_base_size_in_points or whatever ;) |
| 07:23 |
erlehmann |
v-rob something about having no antialiasing producing holes irritates me. why is that the case? |
| 07:23 |
|
hlqkj joined #minetest |
| 07:23 |
BoostedMenchers |
what about a snap_to_multiple boolean setting |
| 07:24 |
erlehmann |
snap to multiples of what |
| 07:24 |
v-rob |
Here's a chart of antialiasing enabled and disabled for multiples of 12: https://user-images.githubusercontent.com/31123645/145352031-70aa856c-b646-463c-9b4e-a695260a21f3.png |
| 07:24 |
BoostedMenchers |
that sets the font size to the closest multiple of 12 to the font_size setting |
| 07:24 |
v-rob |
That's what the font_require_multiple does |
| 07:24 |
v-rob |
Yeah, it's just badly named |
| 07:25 |
BoostedMenchers |
it is not a boolean |
| 07:25 |
BoostedMenchers |
ok |
| 07:25 |
erlehmann |
it should not be a boolean |
| 07:25 |
BoostedMenchers |
does it just overwrite font_size |
| 07:25 |
BoostedMenchers |
ok |
| 07:25 |
BoostedMenchers |
different pixel fonts might have different values for which they are designed? |
| 07:25 |
erlehmann |
v-rob i do not understand why the slashes look so bad bottom left |
| 07:26 |
erlehmann |
BoostedMenchers yes. with bitmap fonts (PNG+XML) you got that “for free” because, well, the size of the bitmap font has to be known to cut out the glyphs from the spritesheet |
| 07:26 |
erlehmann |
i.e. it does not scale, ever |
| 07:26 |
BoostedMenchers |
ok |
| 07:27 |
BoostedMenchers |
but Unifont is made to look pixelated but is TrueType |
| 07:27 |
BoostedMenchers |
and that's the root of the problem? |
| 07:27 |
v-rob |
Here's the scoop on disabling antialiasing: What it does is disables font smoothing. You will not have smooth fonts if you disable it, period. This is good-ish for Unifont at non-sizes of 12. _However_, it's just an approximation and tries to do its best. As you can see, that doesn't work out too well for the slashes. |
| 07:27 |
erlehmann |
unifont is a bitmap font that is forced to use trootype |
| 07:27 |
erlehmann |
truetype format |
| 07:27 |
BoostedMenchers |
huh what |
| 07:27 |
BoostedMenchers |
ok |
| 07:28 |
erlehmann |
ttf is a … difficult … format if you want to support unicode, because you are limited in the number of glyphs |
| 07:28 |
v-rob |
Now, ignore anti-aliasing for a moment. font_require_multiple is all you need to emulate bitmap fonts: it's a way of forcing integer scaling. |
| 07:28 |
erlehmann |
so you need multiple fallback ttfs to support unicode |
| 07:29 |
v-rob |
Bitmap is... and even more difficult format for Unicode because they almost always don't support complex scripts, ligatures, other glyph substitution, etc. |
| 07:29 |
v-rob |
== Problem |
| 07:29 |
erlehmann |
“bitmap” is not a format |
| 07:29 |
v-rob |
Bitmap formats |
| 07:29 |
erlehmann |
uh? |
| 07:29 |
BoostedMenchers |
openbsd's Xorg set ships with *.pcf.gz formatted fonts |
| 07:29 |
v-rob |
E.g. the one sfan5 made |
| 07:29 |
BoostedMenchers |
for the bitmap ones |
| 07:30 |
BoostedMenchers |
hmm there's 100dpi and 75dpi bitmpas |
| 07:30 |
BoostedMenchers |
bitmaps |
| 07:30 |
v-rob |
Still, if my PR works out properly (and I might just remove the antialiasing and stick with multiples) then there's no need for bitmap fonts anyway |
| 07:30 |
erlehmann |
v-rob i am solely bitching about the difficulties unifont as a bitmap font has supporting unicode via ttf |
| 07:31 |
erlehmann |
v-rob only if all the other breakage is fixed as well |
| 07:31 |
BoostedMenchers |
hooo boy that's a lot of files in there… I guess each bitmap font needs a whole series of gzipped PCF files to cover a decent set of glyphs |
| 07:32 |
BoostedMenchers |
but… *should* antialiasing introduce artefacts with multiples of 12? |
| 07:32 |
v-rob |
There's no point MT has to support bitmaps, especially if TTF can do everything bitmap can. |
| 07:32 |
v-rob |
OK, forget antialiasing. Pretend it never existed. |
| 07:32 |
BoostedMenchers |
I assume it'll default to on |
| 07:33 |
BoostedMenchers |
ok |
| 07:33 |
v-rob |
yeah. |
| 07:33 |
v-rob |
I'll remove it from the PR, it's really not helpful. |
| 07:33 |
v-rob |
Just use the multiples setting |
| 07:33 |
erlehmann |
v-rob right now if you have a pixel font defined in 5.4.1 (i.e. disable freetype at runtime in the GUI settings) and then update to 5.5-dev you get a) a non-working game b) a wrong error message |
| 07:33 |
|
epoch joined #minetest |
| 07:34 |
v-rob |
Whoop-de-doo. Change the error message or something. It's not a reason to keep bitmaps. |
| 07:34 |
erlehmann |
v-rob, i believe “game does not start because a feature was ripped out without deprecating it and/or automatically fixing it” is bad |
| 07:34 |
v-rob |
Like, one line change |
| 07:34 |
erlehmann |
i made an issue about the error message being wrong but sfan5 closed it as a duplicate i think |
| 07:34 |
v-rob |
How many people use bitmap fonts? Like, approximately you and nobody else |
| 07:34 |
BoostedMenchers |
I kinda like Terminus |
| 07:35 |
BoostedMenchers |
… not sure I'd use it |
| 07:35 |
erlehmann |
i wonder where the upvotes os the PR come, is it me and my 3 alt accounts? |
| 07:35 |
v-rob |
Probably people who support Unifont and don't realize that TTF can do it |
| 07:35 |
v-rob |
That's all it talks about in the PR description |
| 07:35 |
erlehmann |
i am a contributor to unifont, just so you know |
| 07:35 |
MTDiscord |
<Jonathon> given at least one of those is a person who upvotes basically everything and used to spam emotestill vrob called them out |
| 07:35 |
BoostedMenchers |
lol /usr/local/share/fonts/terminus has 238 gzipped pcf files |
| 07:35 |
v-rob |
If someone makes a PR to rip out bitmap fonts, then that PR can change the error message/fix people's configurations |
| 07:36 |
erlehmann |
if you have unifont and you see emoji, i have made like hundreds of them |
| 07:36 |
v-rob |
I'm not sure how that's relevant |
| 07:36 |
BoostedMenchers |
not sure how I'd even use Terminus with minetest, or if it supports that bitmap format |
| 07:37 |
v-rob |
Unifont is cool, and I'm making a PR to make Unifont TTF work and make bitmap fonts unnecessary. That's kind of the idea here. |
| 07:37 |
erlehmann |
v-rob i retract my offer of helping, at least short-term. i absolutely hate the kind of “oh, you complain, that must mean only YOU use the thing” argument. years of software development have shown be that for each bug filed you have approx. 10 to 100 people (depending on the size of the project) who never file bugs even when they run into them several times a week. |
| 07:37 |
v-rob |
It makes our code simpler and makes supporting Unicode possible regardless of font |
| 07:37 |
erlehmann |
and it's kind of a standard response by lazy devs too |
| 07:37 |
erlehmann |
shown be → shown me |
| 07:38 |
v-rob |
Nobody complained when the feature flat out didn't work in the past. It seems likely that no one was using it. |
| 07:38 |
erlehmann |
uh, but it did work? |
| 07:38 |
v-rob |
I forget the exact versions, but it didn't for many 0.4 versions |
| 07:38 |
BoostedMenchers |
whoa I crashed it |
| 07:38 |
BoostedMenchers |
2021-12-08 23:38:37: ERROR[Main]: FontEngine: attempt to use font size 0 |
| 07:38 |
erlehmann |
how lol |
| 07:39 |
v-rob |
Yeah, I don't like how FontEngine just crashes on failure. It could be a little more graceful. |
| 07:39 |
BoostedMenchers |
I set a 32x64 font, then went to set font_require_multiple to 32, and it crashed |
| 07:39 |
erlehmann |
BoostedMenchers to paraphrase v-rob, “How many people use that font size? Like, approximately you and nobody else” |
| 07:40 |
BoostedMenchers |
oooh |
| 07:40 |
erlehmann |
LOL ;) |
| 07:40 |
BoostedMenchers |
font_size was still 13 |
| 07:40 |
erlehmann |
hahahahahahaha |
| 07:40 |
BoostedMenchers |
and the closest multiple of 32 to 13 |
| 07:40 |
BoostedMenchers |
is 0 |
| 07:40 |
v-rob |
Please, don't be toxic. I don't say ridiculous things. |
| 07:40 |
erlehmann |
v-rob i was of jokings |
| 07:40 |
v-rob |
ok |
| 07:40 |
erlehmann |
v-rob you should prob fix the 0 multiple thing though ;) |
| 07:40 |
v-rob |
yes |
| 07:41 |
BoostedMenchers |
and I'll use a smaller font |
| 07:41 |
erlehmann |
<v-rob> If someone makes a PR to rip out bitmap fonts, then that PR can change the error message/fix people's configurations |
| 07:41 |
erlehmann |
yes i agree |
| 07:41 |
erlehmann |
but no one has done the work |
| 07:41 |
erlehmann |
and again, stuff needs to be DEPRECATED if anyone is using it before removing it |
| 07:41 |
v-rob |
After my PR, the work can actually be done since bitmaps are then redundant |
| 07:41 |
erlehmann |
only if you tell me how i remove the font coloring |
| 07:42 |
v-rob |
Like I said, talk to the other devs |
| 07:42 |
v-rob |
If they think it's good, I'll add it to the PR |
| 07:42 |
erlehmann |
i do not youbt that freetype can do everything better. but it *has* to actually do that. |
| 07:43 |
v-rob |
I don't know if disabling a feature games may rely on is really a "feature", so that's why I'm neutral here |
| 07:43 |
erlehmann |
i have seen how taking away working code in the name of maintainability works in other projects. first, claiming “we'll fix it eventually” is a lie, most of the time. second, you lose access to the feature. third, people usually remove what they themselves do not use. |
| 07:43 |
v-rob |
E.g. a team battle game where nametags of opposing teams are different colors. I don't knoww |
| 07:43 |
|
AwesomeAdam54321 joined #minetest |
| 07:43 |
erlehmann |
it's not about nametags. and also the game so far provides a method to shut that off. |
| 07:44 |
erlehmann |
it's about clowns who make every letter a different color. i have problems with contrast perception. |
| 07:44 |
erlehmann |
so it is only about chat for me |
| 07:44 |
v-rob |
Is this a CSM or server mod problem generally? |
| 07:44 |
erlehmann |
also, TTF can surely do that! i do not doubt |
| 07:45 |
erlehmann |
it is a general problem with fonts for me if the font color is the color of the sky. |
| 07:45 |
erlehmann |
but the sky can be blue or black or red or the color of tv static at least in mineclonia |
| 07:46 |
erlehmann |
if letters have different colors at least a few of them match the sky color too closely |
| 07:46 |
v-rob |
Font outlines would be the ideal solution, of course. But again, other devs--I will make color disableable if they agree to it. |
| 07:47 |
erlehmann |
i kinda dislike the “removing a feature requires nothing, but putting it back in requires a big discussion” |
| 07:47 |
v-rob |
I understand that |
| 07:47 |
erlehmann |
how about next time stuff is discussed before someone removes something they don't use themselves? |
| 07:47 |
|
CWz joined #minetest |
| 07:47 |
BoostedMenchers |
probably somewhat offtopic to current discussion but selecting a .pcf.gz bitmap font is somewhat catastrophically broken https://menche.us/files/lol-boom.png |
| 07:48 |
BoostedMenchers |
I'll stick with the OTF |
| 07:48 |
v-rob |
I think bitmap fonts breaking was an unforseen consequence of the Irrlicht ripping. That could have been done better. |
| 07:48 |
BoostedMenchers |
which still looks a bit "funny" with antialiasing off |
| 07:48 |
erlehmann |
it was entirely foreseen according to sfan5 |
| 07:48 |
BoostedMenchers |
I wanted the pixelly look one gets from antialiasing being off, though :< |
| 07:49 |
v-rob |
Well, I might be wrong. I wasn't directly involved myself. |
| 07:49 |
erlehmann |
i remember i accused him of accidentally breaking it and he said no it was deliberate? |
| 07:49 |
erlehmann |
i would have to look it up again |
| 07:49 |
erlehmann |
v-rob and just to be clear, i agree with deprecating bitmap fonts. but removal should only happen once freetype can actually do stuff. |
| 07:49 |
v-rob |
Yup |
| 07:50 |
v-rob |
Hopefully, my PR can make that work. |
| 07:50 |
erlehmann |
to be fair the whole irrlichtmt situation is kind of a mess with not requiring proper code reviews or reasonably-sized commits and not having it as a submodule |
| 07:50 |
erlehmann |
but that's another thing |
| 07:50 |
v-rob |
Yeah, I stay in my realm. GUIs |
| 07:51 |
erlehmann |
and by “not requiring proper code reviews” i mean if someone gives me a PR with 200k lines removed, i'll probably call them and scream at them for an hour and not approve it the next day without checking any details |
| 07:51 |
erlehmann |
IMO that should have been many small patches, so they can be individually reviewed and tested |
| 07:52 |
erlehmann |
v-rob regarding GUIs, what are your plans to prevent unicode attacks? |
| 07:52 |
erlehmann |
like, say, zalgo text |
| 07:52 |
BoostedMenchers |
I have the weirdest font issue |
| 07:52 |
BoostedMenchers |
well a weird font issue |
| 07:52 |
erlehmann |
good, because you are talking with ppl who are majorly into font issues |
| 07:52 |
erlehmann |
like, at leats waist-deep |
| 07:52 |
erlehmann |
at least |
| 07:52 |
BoostedMenchers |
when I launch minetest from a terminal, Unicode strings render correctly |
| 07:53 |
BoostedMenchers |
when I launch minetest from any application launcher, it's all "Invalid UTF-8 string." |
| 07:53 |
erlehmann |
lul |
| 07:53 |
erlehmann |
check locale |
| 07:53 |
erlehmann |
v-rob in case you do not know http://eeemo.net |
| 07:53 |
v-rob |
Honestly, I probably won't do anything about zalgo text. There's already too many ways for servers to go "ha ha sucker" to the client. Say, render tons of text at a font size of 500. |
| 07:54 |
BoostedMenchers |
LANG is en_US.UTF-8 |
| 07:54 |
erlehmann |
v-rob, ok but what if some asshole zalgos the chat |
| 07:54 |
erlehmann |
> There's already too many ways for servers to go "ha ha sucker" to the client. |
| 07:54 |
erlehmann |
they all need to be removed |
| 07:55 |
v-rob |
One thing at a time, I say. Once we get big font issues out of the way, then we can focus on the smaller issues. |
| 07:55 |
erlehmann |
i think it is much more likely that a player will drop zalgo than a server operator |
| 07:55 |
v-rob |
True. |
| 07:55 |
erlehmann |
oh i just wanted to know if you have a plan |
| 07:55 |
BoostedMenchers |
can I write a CSM that reads from /dev/urandom then just drops it raw into chat |
| 07:55 |
v-rob |
I think mod security won't let you |
| 07:55 |
erlehmann |
recently on oysterity we had some griefer (Mr_Anderson) spam duped echest at spawn |
| 07:56 |
erlehmann |
BoostedMenchers try using a hacked client like waspsaliva |
| 07:56 |
AwesomeAdam54321 |
BoostedMenchers: If you could, it would never actually send the chat if it doesn't stop |
| 07:56 |
erlehmann |
to fight anderson we did two things |
| 07:56 |
AwesomeAdam54321 |
*chat message |
| 07:56 |
erlehmann |
first, the dupes were not perfect copies, you now can not place duped echests |
| 07:57 |
erlehmann |
second, fleckenstein provided a patch so that 10k echests do not lag the game anymore |
| 07:57 |
BoostedMenchers |
welp it's midnight here … my blood caffeine content is probably still high |
| 07:58 |
erlehmann |
i think Mr_Anderson is our most valuable bug-finding asset, with a heavy pronounciation on “ass” |
| 07:58 |
v-rob |
oof |
| 07:58 |
erlehmann |
i mean heavy emphasis |
| 07:58 |
erlehmann |
he likes to crash servers |
| 07:58 |
erlehmann |
so we fix it |
| 07:58 |
erlehmann |
we are playing a game, but half of it is not in minetest, but in git |
| 07:59 |
v-rob |
Delightful person. |
| 07:59 |
erlehmann |
well, i recently trolled him |
| 07:59 |
erlehmann |
he started killing noobs at spawn a few hours ago |
| 07:59 |
erlehmann |
so i started minetest with --random-input |
| 07:59 |
v-rob |
:DDD |
| 08:00 |
erlehmann |
and auto respawn |
| 08:00 |
erlehmann |
the guy is so focussed on killing and griefing he can not distinguish a real player from a literal random walk |
| 08:00 |
erlehmann |
he left after some time hihiu |
| 08:00 |
v-rob |
Nice |
| 08:01 |
erlehmann |
btw, --random-input should be customizable with a seed i think |
| 08:01 |
erlehmann |
for testing |
| 08:01 |
erlehmann |
and also i think it should be slightly adjusted, because it moves so fast right now that it rarely digs a node |
| 08:01 |
erlehmann |
but otherwise i like it |
| 08:02 |
v-rob |
Anyhow, the discussion went all over the place, but I did learn a few things that I can do to improve the PR. |
| 08:02 |
v-rob |
I hop I didn't come off as rude or slighting at any point since that wasn't my intent |
| 08:02 |
v-rob |
*hope |
| 08:03 |
erlehmann |
v-rob i think the “only you use that point” is the rudest thing you said (i think it never really means “only you use it” literally, it means “i do not care for people relying on this, in particular: you”). otherwise, i appreciate how you are trying to fix the font rendering and DPI issues. |
| 08:04 |
erlehmann |
i learned a few things too |
| 08:04 |
v-rob |
Sorry about that one then, I'll try not to say things like that again |
| 08:04 |
erlehmann |
thx |
| 08:05 |
erlehmann |
BoostedMenchers v-rob which font formats does freetype take actually? i was not aware of pcf support |
| 08:05 |
erlehmann |
maybe some of them have the base size included |
| 08:05 |
erlehmann |
i mean bitmap formats must have |
| 08:05 |
erlehmann |
then the patch of v-rob could read it from there |
| 08:05 |
v-rob |
If it's available from the Freetype API, then yeah |
| 08:08 |
erlehmann |
i mean that would be one good way to solve it |
| 08:10 |
BoostedMenchers |
TTF and OTF both work |
| 08:11 |
BoostedMenchers |
that PCF wasn't complete garbage, maybe it just read something wrong |
| 08:15 |
erlehmann |
can you load BDF? |
| 08:15 |
erlehmann |
BDF definitely has a size property |
| 08:15 |
erlehmann |
PCF should have too |
| 08:17 |
BoostedMenchers |
unsure |
| 08:17 |
BoostedMenchers |
I don't have any |
| 08:23 |
|
calcul0n joined #minetest |
| 08:26 |
erlehmann |
BoostedMenchers, are you on linux? |
| 08:27 |
erlehmann |
BoostedMenchers, better question, have access to an xserver? |
| 08:27 |
BoostedMenchers |
openbsd |
| 08:27 |
BoostedMenchers |
yeah it uses Xorg |
| 08:27 |
erlehmann |
try the fixed font |
| 08:27 |
erlehmann |
Xorg will not start without it |
| 08:27 |
erlehmann |
so you must have it |
| 08:27 |
erlehmann |
(mandatory fonts are fixed and cursor) |
| 08:28 |
erlehmann |
uh hmmm |
| 08:28 |
erlehmann |
i have none |
| 08:29 |
erlehmann |
BoostedMenchers http://unifoundry.com/pub/unifont/unifont-14.0.01/font-builds/unifont-14.0.01.bdf.gz |
| 08:30 |
BoostedMenchers |
2021-12-09 00:30:24: ERROR[Main]: FontEngine: Cannot load '/home/menche/unifont-14.0.01.bdf.gz'. Trying to fall back to another path. |
| 08:31 |
erlehmann |
uh |
| 08:31 |
erlehmann |
what |
| 08:31 |
erlehmann |
the PCF version working? http://unifoundry.com/pub/unifont/unifont-14.0.01/font-builds/unifont-14.0.01.pcf.gz |
| 08:32 |
BoostedMenchers |
yup |
| 08:32 |
BoostedMenchers |
renders fine too |
| 08:32 |
BoostedMenchers |
wait |
| 08:33 |
BoostedMenchers |
renders fine *in some places |
| 08:34 |
BoostedMenchers |
https://menche.us/files/crunch.png |
| 08:34 |
BoostedMenchers |
it thinks it's very short |
| 08:35 |
BoostedMenchers |
oh |
| 08:35 |
BoostedMenchers |
renders perfectly when I set font_size to 12 |
| 08:35 |
BoostedMenchers |
when it was scrunched like that it was 16 |
| 08:35 |
BoostedMenchers |
huh |
| 08:35 |
BoostedMenchers |
antialiasing is disabled and there are no visual artefacts |
| 08:36 |
BoostedMenchers |
https://menche.us/files/workingfont.png |
| 08:38 |
BoostedMenchers |
switching to the TTF version of unifont and leaving all other settings as-is results in slightly icky slashes, as before |
| 08:40 |
BoostedMenchers |
as in https://menche.us/files/font4.png |
| 09:11 |
|
calcul0n_ joined #minetest |
| 09:12 |
erlehmann |
BoostedMenchers, hey comparign pcf and ttf version of unifont is great |
| 09:37 |
|
delta23 joined #minetest |
| 10:27 |
|
basxto joined #minetest |
| 10:36 |
|
Fixer joined #minetest |
| 11:21 |
|
Thomas-S joined #minetest |
| 11:21 |
|
Thomas-S joined #minetest |
| 11:36 |
|
AristotIe joined #minetest |
| 12:04 |
|
Fleckenstein joined #minetest |
| 12:24 |
|
tech_exorcist joined #minetest |
| 12:29 |
|
tech_exorcist joined #minetest |
| 12:29 |
|
troller joined #minetest |
| 12:30 |
|
evlc joined #minetest |
| 12:46 |
|
kamdard joined #minetest |
| 12:54 |
|
TempestMancer joined #minetest |
| 12:54 |
|
specing joined #minetest |
| 13:05 |
|
GNUHacker joined #minetest |
| 13:13 |
|
TempestMancer joined #minetest |
| 13:30 |
|
tech_exorcist joined #minetest |
| 13:36 |
|
definitelya joined #minetest |
| 13:39 |
|
tech_exorcist joined #minetest |
| 13:49 |
|
ronoaldo joined #minetest |
| 14:02 |
erlehmann |
BoostedMenchers i can load unifont as pcf and it renders fine. i might have found my solution for crisp pixel fonts, but it needs testing. |
| 14:04 |
|
powerjungle joined #minetest |
| 14:08 |
|
powerjungle joined #minetest |
| 14:37 |
|
Teckla joined #minetest |
| 14:51 |
|
evlc joined #minetest |
| 14:56 |
|
AristotIe joined #minetest |
| 15:13 |
|
appguru joined #minetest |
| 15:33 |
|
GNUHacker joined #minetest |
| 15:50 |
|
v-rob joined #minetest |
| 15:52 |
|
Fixer joined #minetest |
| 15:53 |
|
Taoki joined #minetest |
| 15:58 |
|
tech_exorcist joined #minetest |
| 16:07 |
|
cation joined #minetest |
| 16:13 |
|
sparky4 joined #minetest |
| 16:23 |
|
GNUHacker joined #minetest |
| 16:42 |
|
Hawk777 joined #minetest |
| 17:06 |
|
Fixer_ joined #minetest |
| 17:10 |
|
Extex joined #minetest |
| 17:10 |
|
epoch joined #minetest |
| 17:12 |
|
v-rob joined #minetest |
| 17:33 |
|
garywhite joined #minetest |
| 17:33 |
|
garywhite joined #minetest |
| 17:41 |
|
CWz joined #minetest |
| 17:55 |
|
v-rob joined #minetest |
| 18:15 |
|
Talkless joined #minetest |
| 18:18 |
|
evlc joined #minetest |
| 18:20 |
|
sparky4 joined #minetest |
| 18:29 |
|
fluxionary joined #minetest |
| 18:49 |
|
troller joined #minetest |
| 18:57 |
Krock |
Is this IdleRPG ? |
| 19:01 |
|
Verticen joined #minetest |
| 19:10 |
|
troller joined #minetest |
| 19:23 |
|
___nick___ joined #minetest |
| 20:07 |
|
___nick___ joined #minetest |
| 20:09 |
|
___nick___ joined #minetest |
| 20:31 |
|
SuperMiner joined #minetest |
| 20:34 |
|
v-rob joined #minetest |
| 20:40 |
|
Verticen joined #minetest |
| 20:56 |
|
programmerjake joined #minetest |
| 21:08 |
|
organizedglobals joined #minetest |
| 21:13 |
|
tinnh[m] joined #minetest |
| 21:19 |
|
freshreplicant[m joined #minetest |
| 21:19 |
|
zehka[m] joined #minetest |
| 21:26 |
|
mister-e[m] joined #minetest |
| 21:29 |
|
Noclip[m] joined #minetest |
| 21:33 |
|
plate[m] joined #minetest |
| 21:38 |
|
rogerm[m] joined #minetest |
| 21:41 |
|
iliekprogrammar[ joined #minetest |
| 21:42 |
|
delta23 joined #minetest |
| 21:49 |
|
unexploredtest[m joined #minetest |
| 21:50 |
|
wsor4035 joined #minetest |
| 22:00 |
|
ThomasPayne[m] joined #minetest |
| 22:09 |
|
v-rob joined #minetest |
| 22:15 |
|
exe_virus[m] joined #minetest |
| 22:24 |
|
troller joined #minetest |
| 22:25 |
|
ThomasPayne[m] joined #minetest |
| 22:50 |
|
freshreplicant[m joined #minetest |
| 22:50 |
|
wsor4035 joined #minetest |
| 22:50 |
|
organizedglobals joined #minetest |
| 22:50 |
|
iliekprogrammar[ joined #minetest |
| 22:50 |
|
rogerm[m] joined #minetest |
| 22:50 |
|
zehka[m] joined #minetest |
| 22:50 |
|
exe_virus[m] joined #minetest |
| 22:50 |
|
mister-e[m] joined #minetest |
| 22:50 |
|
Noclip[m] joined #minetest |
| 22:50 |
|
unexploredtest[m joined #minetest |
| 22:50 |
|
tinnh[m] joined #minetest |
| 22:50 |
|
plate[m] joined #minetest |
| 22:50 |
|
programmerjake joined #minetest |
| 22:51 |
|
Verticen joined #minetest |
| 23:00 |
|
Sven_vB joined #minetest |
| 23:18 |
|
fluxionary joined #minetest |
| 23:42 |
|
Hawk777 joined #minetest |
