[00:04:01] you gotta get the return header when you do the upload [00:05:40] so append the -i flag when uploading? [00:05:52] s/append/use/ [00:15:42] yessir [00:18:18] honestly we should provide something like a kiss-debug which allows users to upload build logs to 0x0 or something... [00:19:54] I could see the usefullness of that for sure. people giving logs is an absolute mess at times [00:24:57] would be pretty neat to just do kiss b foo | gimmiethelinkalready [00:25:24] however, im pretty sure people would find a way to miss this regardless.. [00:26:45] oh absolutely [00:27:48] does anyone have a counter on how many times someone has popped in to ask about busybox diff error messages when building a kernel lmfao [00:29:16] not sure I would bother to count that high. especially anything related to kernel not booting. lol [00:30:32] oh, I figured it out. it doesnt seem like Im able to get the token if the file already exists. wierd [00:35:09] it makes sense tho; there's no way of knowing if you're the person who made the original request [00:35:16] don't worry it'll only be up for like, a year lmfaooooo [00:38:21] youre right. only identifier is probably the ip, which doesnt neccessarly have to be unique anyway so.. [00:38:36] lol, yeeah, luckily, it wasnt that bad [01:50:32] hm, this is a new one. autotools generates a configure script with errors on tinyx now. love that things are subtly breaking at times [02:04:43] how do I know if I am using tinyx? [02:05:28] how would you not know? if you havent installed tinyx, youre likely not running tinyx [02:05:42] but a dead giveaway is wether or not you can bind anything to the superkey [02:05:57] the performance difference is also pretty huge [02:06:04] tinyx is a turd in comparisson [02:06:54] tinyx is faster? [02:07:23] other way around [02:08:19] tinyx doesnt have any acceleration, so its rather slow. it only runs on the framebuffer [02:08:29] oh I installed xorg [02:09:18] figured as much. youd have to purposefully install tinyx. [02:09:34] oh makes sense [04:14:55] dilyn: sad_plan: ooh I heard sff [04:14:59] what's your case [04:20:12] sewn: sl3m its called. its by custom_mod [04:20:57] holy fuck that's tiny [04:21:14] yep [04:21:32] why is this more expensive than my 10.8L case???? [04:23:48] nfc-systems tries to brag about the smallest sff gaming computer. theirs is 3.39L. whereas sl3m is 2.97L to be precise [04:24:08] wym? custom_mods cases is pretty affordable imo [04:26:17] it's expensive because it's (not a joke) small volume and harder to machine [04:26:58] its also niche. its often less costly to procude something 10M times, as oposed to 1k [04:28:29] however, I dont belive nfc systems s4t or custom_mods cases is particularly expensive. ive seens less niche cases with higher pricetags. custom_mod's cases is manually made, as is nfc's s4t [04:42:55] sad_plan: my S400v2 was 72 dollars [04:43:55] hm, wierd. sbase recently pushed some new commits, which altered their sort. this seems to break kiss' pkg_fix_deps() function. it runs sort -uk1,1 on the depends files, and seems to truncate the space if a dependency is set as makedep. [04:44:00] but it doesnt always happen [04:44:14] sewn: I see. thats not too shabby pricewise [04:47:46] atleast I'm not forced to use an nvidia GPU hehehe [04:48:26] neither am i. I could use dilyns gpu. it does fit. I know of a couple others aswell which should fit, but I havent doublechecked it [04:48:51] but yes, ive for sure locked myself out of hellova lot of gpus :p [04:50:47] if theres nothing funky about the dimensions ive found for the rtx a2000, it should fit aswell. altough this is also nvidia [05:08:03] bweh [05:17:46] yeah my card would work, it's low profile [05:17:53] iirc doesn't even need pcie power lmfao [05:32:48] what's the point of having a card like that tho [05:35:50] low power, low need [05:35:56] offloading specific workloads, simple graphics [05:40:46] wouldn't the igpu suffice then [05:43:59] igpu eats RAM [05:47:03] just download more ram [05:49:01] kek [05:49:08] but yes ram is a big reason, also efficiency [05:49:17] can an igpu do opencl well? [05:49:34] I see [05:49:41] what about an soc [05:49:57] SoCs tend to be heavily optimized [05:50:13] for instance they'll embed an NPU to do task offloading, or a RT core for RT compute [05:50:37] like the new jetson boards are great for CUDA stuff but piss for general compute imho [05:51:08] ofc depends on the SoC; cheaper boards are less optimized and also bad at generic compute [05:51:18] I see [05:51:21] Bummer [05:52:19] a SoC could do KISS stuff relatively well, for instance my sipeed licheerv can run a web browser; it's laggy, but it can run it. that's 512MB ram, 1GHz cpu (single core??) -- the hard part is getting the software (binary distributions shine here) [08:39:56] what is it about kiss that gives me stutters in my driving game wtf [16:00:50] skill issue [16:02:05] :3c [16:31:22] very helpful [16:46:44] sewn: What is your timer frequency? zcat /proc/config.gz | grep 'CONFIG_HZ_' [16:48:24] same kernel [16:48:35] different userland causes the issue [16:48:47] I was using kiss kernel to boot chimera [16:49:15] anyway its 1000 [16:52:14] and what is the preemption model? zcat /proc/config.gz | grep 'CONFIG_.*PREEMPT' [16:57:24] voluntary [17:00:19] There is one more responsive preemtion mode to try: CONFIG_PREEMPT [17:02:56] changing the wlroots renderer to vulkan removed the stuttering [17:06:16] I'll try that tho [17:07:51] yes it makes no sense [17:42:33] nope it didn't help [17:42:37] anyway I'm removing kiss