00:06 <ukky> SiFuh: So, if you have found cursed hammer, does the curse hammers back black magician? Second option would be to carry hammer with you until you find black magician and ask them nicely to take their curse back.

07:43 <SiFuh> ukky: This is Malaysia. :-P So the reddit post is kind of useless. I gave the hammer to a friend.

08:21 <SiFuh> ukky: I don't know how that woman from Matt's Off-Road was able to spray paint inside the cabin of the Morvair Wrecker without wearing a face mask. I still get an unwell feeling when I spray with a face mask on. And I am using three layers. In the open.

10:09 <SiFuh> ukky: I took the chainsaw to the STIHL shop. After I told him it is my friends chainsaw and I want to have it service. I started to look at all the tools. I asked him which do you prefer the Husqvarna or the STIHL. He sells both. He says STIHL because easier to service and repair. Then as I was walking around the first thing he said was these are Chinese tools (Pointing at two shelves), and everything

10:10 <SiFuh> else is not. HAHAHAHAHA

10:53 <SiFuh> ukky: I also noticed HiKOKI for sale. They have rotary hammer drills. They look awesome.

11:48 <farkuhar> Hmm, function signatures must have changed somewhere in the dependency tree, between gimp 2.10.38 and gimp 3.0. Trying to build the older gimp against the latest libraries, I hit the "incompatible pointer types" error (expected 'gdouble *' but argument is of type 'gfloat *').

11:50 <farkuhar> Good thing I'm not a regular user of the GIMP. If I needed photo editing capabilities for everyday work, all these compilation errors would be intolerable.

11:56 <farkuhar> The context of stenur's latest post to the mailing list is lost on me. I can't find in the libdbusmenu-gtk2 Pkgfile anything that would explain why he's reminding us of the circumstances when shell variables get expanded in a heredoc, versus when they are printed literally.

12:26 <farkuhar> I wonder why jue still felt it necessary to have a separate line for setting default values. https://git.crux.nu/tools/rc/commit/cb977d81e634a457bf8c8d4201570b275b858047 Consolidating the ${LANG:=C.UTF-8} into the printf command itself should have the same effect.

13:02 <ppetrov^> farkuhar, well, i do need gimp every now and then... of course inkscape is mych more important for me

13:02 <ppetrov^> btw, did you investigate pkg-get's issue with showing the dependencies without [i] boxes?

13:03 <farkuhar> ppetrov^: yes, I think I know why it's not annotating the dependencies any more. I can restore the older output format for you.

13:07 <ppetrov^> "for me"?

13:07 <ppetrov^> if you think i am the only user of pkg-get or that it's not an important feature, please, don't bother

13:11 <farkuhar> I'm happy to add back the feature. It's like the recent Handbook revisions that resulted in the loss of the Package Guidelines section. Somehow a cut-and-paste operation went awry, and the old functionality was lost.

13:26 <ppetrov^> :)

13:26 <ppetrov^> please

13:29 <farkuhar> In the meantime, do you get informative output from `prt-get depends gimp`? Obviously the [u] and [i] indicators might differ because prt-get and pkg-get have different ideas about the available version, but the empty box for not-yet-installed dependencies should still work.

13:31 <ppetrov^> prt-get works fine, but i have never seen a [u] box there

13:31 <ppetrov^> am i missing sth?

13:32 <ppetrov^> there's a update for gegl, a dependency of gimp, but i see it as a regular [i], not [u]

13:35 <farkuhar> Hmm, I thought prt-get also had a separate indicator to tell you which packages were updatable. Maybe I'm misremembering.

14:22 <farkuhar> ppetrov^: Thanks for pointing out the discrepancy in their respective outputs. You've now given me a new feature to add to prt-get too.

15:00 <farkuhar> Hmm, prtget.cpp has two distinct subroutines: printDepends and printDependTree. (The first one does double-duty, handling both `quickdep` and `depends`.) It was easy enough to add the [u] indicator to the first subroutine. Then I realized there was a crazy amount of code duplication in printDependTree, which will only get worse if I add the [u] indicator there.

15:03 <farkuhar> It would be better to refactor the code and let one subroutine handle all three output formats: quickdep, depends, and deptree. That's how it works in my Perl rewrite, which made it rather trivial to distinguish between [i] installed and [u] updateable.

19:19 <SiFuh> zorz: https://i.snipboard.io/O7rsaU.jpg https://i.snipboard.io/6ShTE7.jpg Written in ANSI C.

19:41 <zorz> ooo SiFuh is nice

19:41 <zorz> but now i use runnit

20:06 <SiFuh> zorz: Now that it appears to be functioning as it should, (I haven't been able to test the big scary message for fsck), I can remove all those printf messages that were informing me of every step and I can put the services onto a single line again. Merge it all into one file.

20:09 <zorz> SiFuh: is great!

20:09 <SiFuh> zorz: Yeah I know.

20:11 <SiFuh> zorz: Besides, I find it nicer to write in ANSI C than shell scripting. Code looks cleaner, it's just more logical

20:12 <zorz> you took out the seed thing?

20:12 <zorz> no need for seed

20:12 <SiFuh> I commented it out on rc.shutdown.c and rc.c

20:12 <SiFuh> zorz: Depends.

20:12 <zorz> great

20:12 <zorz> virtuaialisation

20:12 <SiFuh> zorz: CONFIG_CRYPTO_RNG CONFIG_CRYPTO_RNG=y CONFIG_CRYPTO_RNG2=y CONFIG_CRYPTO_RNG_DEFAULT=y If you have this enabled in the kernel, then you won't need to bother with creating a seed.

20:13 <SiFuh> It seems to only be needed if you have those options disabled and/or you are running an embedded system, or certain virtual systems

20:13 <zorz> yes after kernel 5.6

20:13 <SiFuh> If you don't have those options enabled in the kernel then you will probably need to generate a seed.

20:14 <zorz> they are on on mine

20:14 <zorz> and keep only bold kolors...

20:14 <zorz> fuck the gay things

20:15 <SiFuh> I was thinking to actually leave a comment in the rc.c and rc.shutdown.c mentioning that this can be uncommented if you don't use RNG in your kernel.

20:15 <SiFuh> zorz: I don't have any colours or bold. Screw that shit.

20:16 <zorz> yeap i agree

20:16 <SiFuh> zorz: Well, bold doesn't work on older terminals anyway.

20:17 <zorz> now i need to spend sometime witn 1 2 3

20:18 <zorz> but runit works great

20:19 <zorz> SiFuh: https://0x0.st/8yRG.txt

20:21 <SiFuh> Wow, just what I always wanted to see. Log files.

20:23 <zorz> yes

20:33 <ukky> SiFuh: reddit post was regarding drilling the wall when you don't know where live wire is.

20:34 <ukky> SiFuh: never seen HiKOKI products in hardware stores here

20:35 <SiFuh> ukky: I was looking at the reviews. Almost always flogs the other drills. I think this is what I will buy when I do get a Rotary Hammer Drill.

20:38 <SiFuh> ukky: HiKOKI was formerly named Hitachi until 2018.

20:42 <ukky> SiFuh: this looks decent: https://www.hikoki-powertools.com/products/powertools/ac-drill/dm20v/dm20v.html

20:43 <SiFuh> ukky: My history has gone but this was the last review I watched. https://www.youtube.com/watch?v=LSFkntg7qF8

20:43 <SiFuh> It is the battery version. But I won't buy a battery version.

20:45 <ukky> chuck in dm20v is bad (keyless)

20:45 <SiFuh> Yeah but that uses the grooved bits.

20:47 <SiFuh> They are becoming more and more common because they don't slip like with the standard old style keyless chucks. And I am with you on a keyed chuck. I hate keyless chucks. But for the grooved bits, it is okay.

21:00 <zorz> ukky: this is the 2 https://0x0.st/8y73.sh

21:00 <zorz> i cnahnged the runlevel to 2 from 1 as you said

21:02 <ukky> zorz: yes, that's where rusvchdir should be used.

21:02 <SiFuh> zorz: You were booting into multi user mode in runlevel 1?

21:03 <ukky> zorz: but now you have another problem. You should move execution of /etc/rc.local into /etc/runit/1

21:04 <ukky> SiFuh: he meant '1' as in '/etc/runit/1'

21:05 <SiFuh> ukky: But isn't that basically runlevel 1?

21:06 <ukky> SiFuh: runit has no runlevels at all. Only boot-up, run-services, and shutdown stages.

21:07 <SiFuh> ukky: I though it had /etc/runit/2 as well which is the equivalent of runlevel 2 where local should be. Hmmmmmmm

21:07 <ukky> SiFuh: though runit can select which set of services to start

21:09 <ukky> SiFuh: rc.local will work from /etc/runit/2 too, but that would be weird to look at.

21:09 <SiFuh> ukky: https://smarden.org/runit/runlevels

21:09 <SiFuh> ukky: Yeah well I remember using /etc/runit/2/local

21:10 <ukky> SiFuh: term 'runlevel' for runit is not related to the same term of sysvinit.

21:12 <SiFuh> https://smarden.org/runit/runit.8

21:13 <SiFuh> I just don't see why rc.local would be in stage 1 and not in stage 2

21:16 <SiFuh> Anyway, doesn't matter. runit isn't the original CRUX way, so I will let it be

21:17 <ukky> /etc/runit/1 executes one-time tasks at system startup. /etc/runit/2 starts services and when /etc/runit/2 crashes, it will be restarted. Thus, rc.local will be restarted.

21:20 <zorz> yes one one guys

21:21 <zorz> again reboot

21:21 <zorz> lol

21:34 <SiFuh> ukky: zorz must have screwed up. :-)

21:40 <zorz> yeap we are good ukky this is the 1 https://0x0.st/8yhO.sh

23:47 <ukky> zorz_: instead of printf "%s\n" "The system is coming up. Please wait." you could just do printf 'The system is coming up. Please wait.\n'

23:51 <ukky> zorz_: . /etc/runit/rc.conf should be [ -r /etc/runit/rc.conf ] && . /etc/runit/rc.conf