<midfavila>
when a window is xembeded in another window it becomes a child window of the host window
<midfavila>
iirc
<midfavila>
it's the same principle behind reparenting window managers
<midfavila>
when the wm reparents an x client you don't actually have access to that x client's window id
<midfavila>
you have access to the id of the window that contains that client
<midfavila>
if you try to use xdotool or similar to affect the reparented/xembed'd x client, it wont work well
<midfavila>
i don't know the actual X11 jargon here but that's just what i've experienced from tinkering on my own
<sad_plan>
makes sense. I did notice that if I run lsw to list all window ids, itll only show tabbed (plus other windows if you have any open), but not the child wids
<sad_plan>
have you ever figured a way around it?
<midfavila>
sometimes you can list the children of an x client
<midfavila>
that'll let you manually fiddle with things
<midfavila>
i don't recall specifically how to do that though
<midfavila>
it's been ages since i've fiddled with stuff like that
<midfavila>
anyway i gotta bike home
<midfavila>
gimme 30-45 minutes
<sad_plan>
hm. lsw doesnt list them atleast. maybe someone in #suckless knows
<sad_plan>
ait
<sad_plan>
man lsw
<sad_plan>
ffs
<sad_plan>
midfavila: with xwininfo, I was able to get one of the child wids. and with that I was able to interact with said window using wmutils through the shell
<sad_plan>
it was not very practical
<midfavila>
back
<midfavila>
yeah it's a pain in the ass
<midfavila>
you could always write a set of shell routines to access and manage children that are xembeded
<midfavila>
xembedded
<midfavila>
whatever
<midfavila>
unrelated side note i got these kickass korean chips at the local asian grocer
<midfavila>
they're called turtle chips
<midfavila>
they're almost like... pastry dough kinda texture, but crispy
<midfavila>
each chip has multiple layers
<midfavila>
these ones are truffle flavoured
<midfavila>
big recommend
macewentoo has quit [Read error: Connection reset by peer]
<sad_plan>
I could do that. ill look more into it.
<sad_plan>
dope
solaare has joined #kisslinux
riteo has joined #kisslinux
riteo has quit [Ping timeout: 252 seconds]
riteo has joined #kisslinux
<midfavila>
oh shit its riteo
<midfavila>
workstation should be here soonish
<midfavila>
once that arrives i'm gonna port the main kiss rootfs to aarch64 and put it up on my git
<midfavila>
can mirror it on github idc
<midfavila>
is there anyone running builds for firefox still?
<midfavila>
i dont pay for electricity and my machine is plenty beefy so i can run CI on it while im at work and stuff
<sewn>
aw man my school year is over
<sewn>
what do I do now
<sewn>
one more year and I suddenly have to be in college
geekthattweaks has quit [Quit: Connection closed for inactivity]
fultilt has quit [Quit: WeeChat 3.7.1]
<sad_plan>
sewn: learn some new skill, or research a topic youre interested in
<sad_plan>
i have an idea. implement tls 1.3 in bearssl
<sewn>
im not a networking enthusiast
<sewn>
how about you do that so you can use bearssl lol
<sad_plan>
midfavila: there were a binary repo, but I dont think its been updated for a really long time now
<sad_plan>
I already use bearssl, but lacking tls 1.3 is an issue as of next year. curl is dropping support for it appearantly
<sad_plan>
or all ssl libraries actually which doesnt have tls 1.3
<sad_plan>
so its no exclusive to bearssl
<sad_plan>
but yes, I would probably attempt to implement it if I had the appropriate skillset. which I dont
<sewn>
ok no need for tls 1.3 then
<sewn>
if curl doesnt have it it doesnt exist
<sad_plan>
what? bearssl doesnt seize to exist because curl stopps supporting it
<sewn>
sad_plan: oh wait do you mean curl is dropping support for bearssl
<sewn>
i thought you were talking about tls 1.3 in curl
avs_origami has joined #kisslinux
<sad_plan>
yes, curl will be dropping support for bearssl among others, which does not support tls 1.3, by may next year.
<sad_plan>
no, that would be silly if curl dropped support for tls 1.3
<riteo>
yes mid it was me
<riteo>
this morning I ran `ping boscaiolo` since it looks like power goes out daily and look at that, it indeed was powered off
avs_origami has quit [Remote host closed the connection]
<riteo>
I don't want to make an auto-booting jig but it looks like I might need one
avs_origami has joined #kisslinux
<riteo>
I might also go super-crazy and make a PoE parser with a relay or sth if I get real obsessed with it
<sad_plan>
riteo: is it your hardware thats borked, or does italy really have that unstable power?
<riteo>
but I'll have to wait a month anyways since I'll have this damn exam soon
<riteo>
sad_plan: the whole power grid's a mess
<riteo>
you know, heat coming
<sad_plan>
hm, I dont recall having any issues with that when Ive visited italy. maybe I was just lucky
<riteo>
it usually is way more stable but in summer it messes up real bad, at least here around naples
<sad_plan>
I see
<riteo>
sad_plan: it really depends on the place ig
<riteo>
I don't live in a particularly popolous aread
<riteo>
area*
<sad_plan>
ive been in naples, and dont recall having issues there. neither in rome, or verona either. but I suppose I was just lucky
<riteo>
kinda medium-small village kind of thing
<sad_plan>
I wasnt there for very long though, so I guess that might come into play here aswell
<riteo>
might be, could as well be related to my local area
<riteo>
last year I could admire a giant-ass generator just for my local grid being placed temporarily
<riteo>
could really be the specific grid I live onto lol
<sad_plan>
atleast they take some steps to leviate the grid issue :p
<riteo>
no idea either how it goes in "actual" naples
<sad_plan>
as opposed to just shrug it of as someone elses problem
<riteo>
i'm in province if that makes sense
<riteo>
sad_plan: oh yeah indeed, at least they try lol
<sad_plan>
yeah, almost like a county, right?
<riteo>
yeah
<sad_plan>
figured
<riteo>
italy's weird
<sad_plan>
i know.
<sad_plan>
do you eat croisants with nutalla and a teeny tiny cup of coffee for breakfast too, or is this just a hotel thing?
<sad_plan>
I remember I asked for milk and bread with, I dunno, jam. but got croisants and coffee, or juice
<riteo>
yup, that's a thing
<riteo>
actually, more of a bar thing
<sad_plan>
do all of you just eat that for breakfast?
<riteo>
like if you go to have breakfast at a bar that's what you usually ask for
<sad_plan>
as opposed to more healthy alternatives I mean
<riteo>
nah there are
<riteo>
but if you go to the bar you usually ask for a croissant and a coffe/cappuccino
<riteo>
there are also other sweet breads but nothing exactly healty
<riteo>
at home we usually eat milk, bread, jam and so on
<riteo>
do you folks have those weird crunchy slices of "bread"?
angerisagift has quit [Read error: Connection reset by peer]
<riteo>
we call them "fette biscottate"
<riteo>
they're not really normal bread and they're really crunchy OotB
<sad_plan>
so you do eat regular bread and drink milk after all. I always got these wierd looks when I asked for it
<riteo>
yup
<riteo>
where did you ask that?
<sad_plan>
in scandinavia we got something for it. i dont recall the english name for it.
<riteo>
mh. Well we put jam usually on those things for breakfas
<sad_plan>
in the hotel. he asked what we wanted for breakfast. and I said bread on, ham or something, aswell as misk
<riteo>
yeah but _where_, what region
<riteo>
every region is basically a different country, culturally
<riteo>
campania, lazio, molise, lombardia ecc
<sad_plan>
crispbread was what I was thinking of btw
<sad_plan>
when we got to rome or verona, I kinda just gave up on asking for anything else, as I figured you italians wouldnt give me what I wanted anyway :p
<riteo>
unless it was a bar, no idea why they refused to give you jam and bread
<riteo>
if it was a bar it makes sense though, bars don't have that
<riteo>
sad_plan: you aren't exactly wrong on us not giving foreigners unusual stuff :P
<sad_plan>
I dunno. maybe it had to do with the hotel of choice.
<sad_plan>
exacly :p
<riteo>
but also keep in mind that every region eats and speaks in very different ways
angerisagift has joined #kisslinux
<riteo>
italy's unification didn't happen too long ago, relatively speaking, and everybody's really proud of the local culture
<sad_plan>
absolutely. right
<sad_plan>
awesome, wmutils builds right out of the box with cproc
<riteo>
nice
<sad_plan>
now if only their opt repo would build statically linked.
* sad_plan
squints
<riteo>
why is software so obtuse to build statically?
<riteo>
does nobody really use static stuff?
<midfavila>
p much
<sad_plan>
^
<riteo>
I still fail to understand why folks are so obsessed with dynamic linking, then proceed to flatpak everything and do the same with extra steps
<midfavila>
dynamic linking is in vogue because you only need to keep one copy of a library in memory at a time
<midfavila>
but yes flatpak to solve the problem of dynamic linking is funny
<riteo>
midfavila: that makes sense only on paper
<riteo>
optimizing compilers and good static libraries strip out most of the library when unused
<riteo>
by good static libraries I mean the ones that ship lots of tiny files inside, which is suppose is most of them? No idea on the actual details admittely
<riteo>
but from what I understand there are very little downsides to static linking
<midfavila>
the problem is that libraries are so fucking huge these days that static linking isnt super viable for most programs
<midfavila>
imho this is a problem of said programs
<riteo>
indeed
<midfavila>
but imagine if every single instance of every single GUI program you ran had to keep a complete and unique copy of every single library that program depended on in memory at once
<riteo>
not complete though
<sad_plan>
theres also the issue about linking several libraries that also links to several of the same libraries
<sad_plan>
like statically link mesa as an example. all fun and games
<midfavila>
well, okay, sure, not complete-complete
<midfavila>
just the symbols in use
<midfavila>
but that's still a lot of redundancy
<riteo>
sad_plan: I'm not sure what you mean
<riteo>
midfavila: I suppose it depends on what a lot means
<riteo>
because if it's just "kinda a lot" then it becomes a race to which overhead is the worst and what kind of overhead is preferred where
<sad_plan>
riteo: if you build a library, which links to say zlib, which is rather usuall. this library also links to 5 other libraries, which also links to zlib. now the final product will have 6 libz.a initially
<riteo>
sad_plan: wait, do transitive dependencies stay between linkings?
<riteo>
to be clearer, wouldn't the library's zlib be "embedded" inside this library?
<sad_plan>
afaik know, yes
<riteo>
it would be optimized and it would be "indistinguishable" from the actual library. I don't think that the symbols get exported so what's the issue here?
<sad_plan>
for static linking youd have to include libs of libs. which is what makes static linking a pain at times
<sad_plan>
symbol resolution if often the issue I encounter atleast
<riteo>
I fail to understand why, if the libs are statically linked themselves. Unless I'm missing something
<riteo>
or are they just code pieces
<riteo>
wait I'm not sure now
<riteo>
fuck yes
<riteo>
yeah that makes sense, sorry for the confusion
<sad_plan>
static linking is a mess at times to say the least :p
<midfavila>
also sad_plan re: curl if you want to continue to use bearssl you might want to look into the port of libtls to it
<midfavila>
i'm developing all of my software to target libtls so at the least anything i write will be usable with bearssl
<sad_plan>
midfavila: I already use that on oasis. more or less atleast
* midfavila
thumbs-ups
<riteo>
sad_plan: do you think that a (reasonably) statically linked and modern system would be too hard?
<sad_plan>
midfavila: curl doesnt actually link to it though
<sad_plan>
riteo: well, oasis is 100% static. im not sure wether youd call netsurf modern by any stretch, but its possible
<riteo>
I was planning to try and make my distro (once this exam stuff ends) at least mostly statically linked, especially as most annoying stuff would be delegated to flatpak, which does an astounding job at keeping their silly mouths shut (yes I'm still talking about applications)
<sad_plan>
if you dont care for mesa, and can avoid it, then yes
<riteo>
mh that'll be a fun experiment
<sad_plan>
but most browsers do link to mesa, and mesa can be problematic
<riteo>
I might delegate also that to flatpak at this point
<riteo>
it's astoundingly good at its thing and if something needs flatpak it's too big to be practically build anyways
<sad_plan>
if we can get mesa statically linked, maybe we can statically link webkit too, thus statically link surf aswell
<riteo>
I wonder about the performance difference between a juicy browser and an hypotetical LTO-linked static version
<riteo>
perhaps there are better programs for benchmarking such a thing, perhaps some libre game?
<sad_plan>
lto should be faster. but by how much, I have no clue
<sad_plan>
im not overly convinced by its benefits tbh
<midfavila>
riteo if you want to avoid mesa you can use tinx
<midfavila>
:3
<midfavila>
i'm pretty sure it doest use mesa
<riteo>
midfavila: sorry, I MUST GAME
<riteo>
UGHHH I'M GAMING
<midfavila>
since its rendered directly the kernel framebuffer
<sad_plan>
midfavila: still need mesa for webkit :c
<riteo>
gamer juice flows through my veins
<midfavila>
UGH I GOTTA CONSOOOOOOOOOOOOOOOOOOOOOOOOOOOOM MID
<riteo>
indeed
<midfavila>
sad_plan use links
<riteo>
I'm addicted to DRM-riddled reddit-grade battery-draining programs
<riteo>
only the best HD indie classic gem reloaded with DLCs thank you
<sad_plan>
well, if I was content with just using links, then yes. that would be ideal.
<riteo>
gaming is my passion
* riteo
looks at their huge list of bought-and-never-played games
<riteo>
although wait
<riteo>
I think flatpak handles mesa too
<riteo>
oh lord that'd be unholy still but the fact that's it's technically feasible is... unsettling
<riteo>
the bare minimum just to run flatpak lmao
<riteo>
that would be too much even for me. I think I'll stick to my hardware accelerated wayland compositor which does indeed use mesa
<riteo>
I'm really happy that this place is active again BTW, this is the year of #kisslinux
<midfavila>
my boss just got into a screaming match with his son over capitalism lmao
<midfavila>
the workers are rising up and seizing the means of floor production
<sad_plan>
riteo: try oasis, and build flatpak but use velox. it only uses libdrm. so no mesa needed
<sad_plan>
probably piss poor performance though
<midfavila>
this is kisslinux
<midfavila>
we arent exactly going to be wildly performant
<midfavila>
like as a rule
<midfavila>
clear linux this is not
<sad_plan>
some people have attempted to get more performance out of kiss, so that they can play games better. or build faster. w/e theyr goal is
<midfavila>
yeah but i think its a fair bet to say most KISS users are concerned with space and not time :p
avs_origami has quit [Remote host closed the connection]
<sad_plan>
time is probably our least concern :p
<midfavila>
>my main hobby is troubleshooting obscure compiler errors that literally nobody else has ever had
<sad_plan>
something like that
<riteo>
actually I think that the priority here is simplicity and perhaps even more manutenibilty
<riteo>
the infamous bus factor zero
<midfavila>
riteo when i say space i mean complexity
<riteo>
oh all right sorry
<sad_plan>
space!=complexity midfavila
<midfavila>
yes i'm aware but generally more complex algorithms can achieve greater performance in terms of time if not in memory-space or in implementation complexity
<sad_plan>
I suppose so
<riteo>
mh yeah that's fair
<midfavila>
ionno just my pre-caffeine brain take
<midfavila>
im still in bed, mentally
<riteo>
anyways, IMO, overall, it's simpler to use flatpak than building something humongous if it's absolutely needed
<riteo>
for example
<midfavila>
i just say fuck it and use qemu ngl
<riteo>
that'd be more complex under most aspects I think
<sad_plan>
its easier, but not better
<riteo>
in my head, simplicity is a tradeoff between different factors
<midfavila>
its a more complete separation from flatpak and similar
<midfavila>
compared to*
<midfavila>
fegg
<midfavila>
language is hard
<riteo>
that's a double-edged sword though
<midfavila>
plus i already use qemu for other stuff
<riteo>
I'd actually like hardware acceleration or native windowing
<midfavila>
kvm and X11 forwarding
<riteo>
since most of this stuff is some big chungus GUI heavy "application"
<midfavila>
>big chungus
<midfavila>
whiplash
<riteo>
midfavila: I'm a wayland user
<sad_plan>
you can have qemu connect you running X session
<midfavila>
smells like a wayland problem
<riteo>
midfavila: yes I do indeed find that use funny
<midfavila>
use the based chad x11
<riteo>
midfavila: it's still going to render on my CPU and waypipe's a thing
<midfavila>
wayland no run xaw
<midfavila>
xaw good
<midfavila>
not running xaw bad
<midfavila>
therefore wayland bad
<midfavila>
qed
<riteo>
I'm sorry, me no like xaw
<midfavila>
me no like you
<midfavila>
>:c
<riteo>
this is understandable
<midfavila>
xaw is objectively based and comfypilled
<riteo>
I mean, I don't think it'd be _that_ hard to build an API-compatible xaw and run it on wayland
<riteo>
it's just a library, the fact that it uses the X server as a drawing "library" is an implementation detail
<riteo>
it's less cool though I suppose
<midfavila>
honestly i'm going to just drop GUIs on linux if x11 ever gets to a point where its not usable
<riteo>
that's perfectly understandable too
<midfavila>
get really fucking good at curses programming
<riteo>
it's not like the average KISS Linux experience is very gui based either
<midfavila>
anyway with xaw itd probably be tricker to extricate all the x bits
<midfavila>
since its based on xt
<midfavila>
as opposed to qt/gtk/etc
<riteo>
fair, I don't know most about xaw
<riteo>
but it'd be pretty cursed anyways
<midfavila>
its the totally temporary do-not-use-seriously-guys demonstration toolkit in the standard x distribution
<midfavila>
that people have been using daily since the 80s
<midfavila>
very small, very simple
<midfavila>
the entire distribution is like 40k lines of C and half of that is just for the overgrown text widget
<riteo>
nice
<midfavila>
has all your basic WIMP elements
<midfavila>
and its very easy to extend because its based on xt
<riteo>
but how big is the rendering logic in the x server
<midfavila>
for xt? minimal
<riteo>
is xt an x server
<midfavila>
its literally "draw a line from X,Y to X,Y", "fill this region", "bit-blit this bitmap", and so on
<midfavila>
no
<midfavila>
xt is a library that implements basic drawing primitives on top of xlib/xcb
<midfavila>
and the mechanisms to make widgets
<riteo>
ah that makes sense
<midfavila>
when you write a xaw application youre actually using xt
<midfavila>
xaw is just a collection of xt routines
<midfavila>
in essence
<riteo>
are the native drawing routines of the X server even used
<midfavila>
so theres no like, xaw api
<riteo>
like, anywere
<midfavila>
yes
<midfavila>
any xaw or motif or xt-based application
<midfavila>
but yeah theres no like, xaw api
<riteo>
mh I see
<midfavila>
you use the Xt api to spawn a widget tree and you populate that tree with xt widgets e.g. xaw motif fltk and so on
<riteo>
I read spam at first lol
<riteo>
anyways, fair
<riteo>
perfectly understandable for you to like xaw, it's all personal preference, that's why I'm going to make another opinionated linux distro
<riteo>
because I like fragmentation
<riteo>
Chaos is my second name
<sad_plan>
I cant help but feel like people creating a new distro is because they just didnt like w/e defaults other distros had, and simply wanted to track everything through git
<sad_plan>
atleast thats how I feel when I started using kiss, and now w/e monster Ive created with my oakiss
<midfavila>
i just disagree with upstream transitioning to wayland
<midfavila>
and to openssl
<midfavila>
and i was already constantly switching to suckless utils so
<midfavila>
i also might study this and xaw and implement my own toolkit in the future
<riteo>
sad_plan: actually I'd be switching to a completely different package manager
<riteo>
and in my head 90% of a distro is the package manager
<riteo>
midfavila: is that a windows clone? It looks oddly familiar
<midfavila>
no
<midfavila>
its a toolkit that was developed for use with fvwm95
<midfavila>
it copies the look and feel of win95 but its its own thing
<riteo>
ah all right
<sad_plan>
riteo what about when the distro has no package manager?
<sad_plan>
I personally would love nothing more than to port everything I want/need to oasis' buildsystem, and at that point no longer need a package manager
avs_origami has quit [Remote host closed the connection]