IRC Logs for #crux Wednesday, 2018-03-07

ryuobtrfs... i always thought of it as a joke. "butter fs"00:05
ryuoor "better fs"00:05
ryuoever.00:05
ryuoeven*00:05
*** pedja has joined #crux00:09
pedjais there a CONFIG_CHAOSMONKEY=y option in kernel config that I am not aware of?00:11
pedjaor ext4 hates me that much?00:12
ryuoOui. Waiting for my new laptop... long wait.00:12
joacimwhich one are you getting?00:13
joacimthinking about getting a quad core upgrade for my t42000:13
joacima quad core and the gpu heatsink to keep it cool00:14
ryuooh.00:23
ryuoi'm getting something simpler.00:23
ryuohttp://h10032.www1.hp.com/ctg/Manual/c0577816600:23
ryuoit has an IPS display if you get the 1920x1080 display.00:24
ryuohttp://www8.hp.com/us/en/thin-clients/mobile/mt21.html00:24
ryuoIt's basically a rebranded ProBook.00:24
joacimthat looks sweet00:25
ryuoAnd the version I ordered ships with a Linux derivative.00:25
ryuoA rarity among HP products.00:25
ryuoi could only find it from the HP web store.00:27
jaegerLooks like a decent laptop00:27
ryuothe resellers generally didn't stock the model I wanted.00:27
joacim128GB SSD?00:27
ryuoyea... it's small I know.00:27
ryuoi can upgrade it later though.00:28
joacimusually for that price, for celeron/atom stuff like that, someone like asus would give you about 32GB eMMC00:28
joacimso that SSD sounds great00:28
ryuoI ordered the $489 model from their store.00:28
ryuoIt has the higher DPI screen.00:28
ryuoi'm thinking of using containers on my laptops going forward.00:29
joacim"Both slots are customer accessible / upgradeable"00:29
ryuowhich one?00:29
joacimRAM00:30
ryuoAh.00:30
ryuoyea. from what i could tell00:30
ryuothey're both under the service door.00:30
ryuosame with the M.2.00:30
ryuothis will be my first laptop without conventional SATA.00:31
jaegerI've had pretty good luck running linux on HP laptops and workstations in the past at the previous job... couple of elitebook models, one probook, and a few Z workstations00:31
ryuoseems a bit odd.00:31
jaegerAll worked pretty well, though their UEFI implementations were all over the place00:31
ryuojaeger: yea... i've sworn off HP consumer but I'm still open to business products.00:31
ryuosame with HPE.00:32
jaegerI love m.2, so much nicer with no cables00:32
jaegerNot that cables are a concern with a laptop, but in general00:32
ryuojaeger: isn't SATA also cableless in laptops?00:32
joacimsome norwegian store is selling it for 2500 USD...00:32
ryuojoacim: the mt21?00:32
ryuoThat's insane.00:33
jaegerhence the "not that cables are a concern with a laptop" comment :) But I like m.2 in desktops as well00:33
ryuothey also have a higher end mt43 if you don't mind AMD.00:33
joacimyeah. looks like a great laptop, but the local prices arent that great00:33
ryuohttp://www8.hp.com/us/en/thin-clients/mobile/mt43.html00:33
ryuoweird.00:34
ryuowhy is it so cheap here?00:34
joacimthe price is better. only 1500 USD =)00:34
ryuolmao00:34
ryuomakes you wonder if importing would be a smarter move.00:34
joacimat that price range, i would prefer an elitebook/latitude/t-series00:35
ryuoyea, it's weird.00:35
joacimwhen i buy from norwegian retailers, i get 5 years "warranty" through our consumer laws00:35
ryuois that why it's so expensive?00:36
ryuosheesh. not even a choice.00:36
ryuoi've heard of a similar problem in Australia where00:36
joacimelitebooks are much cheaper tho. generally start at less than 1000 USD00:36
ryuopeople would go to the US to buy shit because it was cheaper overall due to overzealous consumer warranty laws.00:37
joacimi dont think our prices are that much higher than US prices, when considering the 25% sales tax00:37
ryuoi guess it made sense. They were buying stuff like furniture that seldom goes bad if done well.00:38
joacimi think it could be a good thing, since stores would have to stop selling items that are known to go bad often00:39
joacimthe one thing that often go bad on hp laptops are their chargers00:39
ryuoConsumer HP?00:39
joacimthe chargers works well, but the barrel plugs center pin gets bent easilly00:40
joacimi think they use the same barrel plug on all of their consumer, prosumer, and business lines00:40
ryuoprosumer...?00:40
ryuolol00:40
joacimbut getting a new charger for 50 dollars once in a whole isnt that bad00:40
joacimand you just have to be careful about plugging it in00:41
joacimif it doesnt go in all the way, check the pin, it might be bent00:41
ryuoyou know, it's funny. i only found the mt21 because i decided to see what showed up under Laptops if i chose to restrict it to "Linux".00:41
joacimhaving a 2-4 mm gap of silver showing is normal tho00:41
ryuoonly one was left; the mt21. lol00:42
ryuoOn Newegg even.00:42
joacimi think there are developer edition xps laptops, but i think those are more expensive than windows alternatives00:42
ryuothose are special order.00:42
joacimi think some retailers here have them00:43
ryuoah.00:43
ryuoi meant generally you only find Windows preinstalled junk.00:43
joacimyeah00:44
joacimi think it's more common to find gnu/linux installed on laptops from computer reuse type places00:44
ryuowhy does it seem like even high end models are stuck with 1366x768?00:44
ryuoit's weird.00:45
joacimi guess they install mint or whatever when the windows license is too old (vista), or it doesn't work00:45
*** brian|lfs has quit IRC00:45
joacimyeah. a low end asus e403 has a high resolution screen, but the much more expensive elitebook comes with a 1366x768 TN panel00:46
ryuoIt's like there's an oversupply of 1366x768 TN displays.00:46
joacimthere are 1920x1080 or better IPS options tho00:46
ryuoyea, it's why i did a special order to ensure i got the exact model I wanted.00:46
joacimon a laptop, i dont mind a slower cpu or less ram, but i still want a nice screen00:47
ryuolol00:47
ryuojoacim: i can't stand anything based on Atom though anymore.00:47
ryuothey SUCK so much.00:47
jaegerthat's one reason I picked the xiaomi laptop, good screen00:47
ryuoeven the entry level Celerons that are based on the Core line are vastl superior.00:47
joacimim curious about atom for stuff like file servers and router boards00:48
ryuothat's probably a tolerable use case.00:48
joacimi dont know how atom those 8 core chips are, that they use on mini ITX boards00:48
ryuolol. i found a troll product.00:49
ryuohttps://www.newegg.com/Product/Product.aspx?Item=9SIA77G726354000:49
joacimif theyre just atom in name00:49
ryuojoacim: Atom SoCs come in a variety of flavors. What's the board?00:49
joacimthats about 6 years old now iwnt it00:49
joacimim thinking about the boards asrock rack and supermicro makes. the ones intended for fileservers00:50
*** brian|lfs has joined #crux00:50
joacimi see people use them for small low powered esxi home labs too00:50
ryuoThey work, but you could also consider ARM alternatives.00:50
ryuoI have an old Dreamplug i'm intending to make use of in the near future.00:50
ryuoit's still got decent IO ports despite being from 2011.00:51
ryuoand being armv500:51
ryuoOnly Debian and Arch support it from what I've heard.00:51
joacimthere is this tiny arm board with a single laptop-style SATA connector on the side00:52
joacimcomes with a case that can hold a 2.5" drive00:52
ryuoNot bad...00:52
joacimhttps://ameridroid.com/products/odroid-hc1-0003a00:53
v0latileWhy don't you considera lenovo? I have an x1 and it has a very good screen and its not expensive00:54
joacimthey are very expensive ere00:56
joacimhere00:56
ryuoLenovo... fine I guess, though i'm hesistant to trust them after the crap they've done with their UEFI.00:57
ryuohttp://www.digitaljournal.com/technology/lenovo-used-hacker-technique-to-install-unwanted-software-on-pcs/article/44086100:58
joacimhttp://cusl2.radified.com/cusl2.htm01:01
joacimi like finding relics online01:01
joacimshame what happened to falconflys 3dfx site01:01
joacimbut there are several archives of it now01:01
ryuoAncient? Ha.01:02
ryuohttp://putnamelectronics.com/products.html01:02
ryuo:)01:02
joacimreminds me of https://rc2014.co.uk/01:04
v0latileYes, the security of the Lenovo products is horrible, but they are quite good in quality, I have had HP and ASUS and I have always had to take it to guarantee because something fails01:07
brian|lfsAsus has been good for me01:25
ryuoi'm moving away from this X200. it's freezing randomly when i play videos and such.01:51
*** tilman_ has joined #crux02:01
ryuotilman_: is xmms2 still a thing?02:05
*** tilman has joined #crux02:07
*** tilman_ has quit IRC02:07
*** brian|lfs has quit IRC02:33
*** brian|lfs has joined #crux03:10
brian|lfspulling my hair about to say screw it just use the appleimage03:13
brian|lfsor appimage rather03:13
*** _________mavric6 has quit IRC03:25
*** _________mavric6 has joined #crux03:27
v0latileHey guys, I've been fixing the application I had done some time ago to deal with orphaned dependencies when removing packages, maybe they can try it and tell me if it works well, I've tried some ports and it works great03:28
v0latileI put everything in a .c file03:28
v0latilehttps://pastebin.com/SAxDgS9Y03:28
*** dlcusa has quit IRC03:37
ryuov0latile: do you have a phobia of char* or something?03:45
ryuobleh. strncat?03:46
v0latileI know, that is a code from 201203:46
ryuoare you implying you've improved then?03:47
ryuoHm.03:48
v0latileWhat's the problem with strncat? I'm not expert so...03:49
ryuofirst, the problem with strncpy.03:49
ryuohttp://pubs.opengroup.org/onlinepubs/007908799/xsh/strncpy.html03:49
ryuoIf the array pointed to by s2 is a string that is shorter than n bytes, null bytes are appended to the copy in the array pointed to by s1, until n bytes in all are written.03:50
ryuo If there is no null byte in the first n bytes of the array pointed to by s2, the result will not be null-terminated.03:50
ryuo1) strncpy writes nulls pointlessly if the buffer is bigger than necessary.03:50
ryuo2) strncpy is not guaranteed to null terminate every string it copies into the buffer.03:51
ryuohttp://pubs.opengroup.org/onlinepubs/007908799/xsh/strncat.html03:51
ryuoThe strncat() function appends not more than n bytes (a null byte and bytes that follow it are not appended) from the array pointed to by s2 to the end of the string pointed to by s1.03:51
ryuonotice it doesn't talk about the size of the destination buffer. only the amount to copy.03:51
ryuoit's rather unintuitive to use correctly.03:52
ryuoso, proper usage of strncat requires a lot of care.03:52
ryuoi generally advise people use snprintf instead.03:52
ryuoit's brain dead simple to use it consistently and correctly.03:53
ryuosnprintf(buf, sizeof(buf), "%s", "Hello!");03:53
ryuoneed appending?03:53
ryuosnprintf(buf+strlen(buf),sizeof(buf)-strlen(buf), " %s", "World!");03:54
ryuoalternately, you can use the non-standard strlcat and strlcpy.03:55
ryuov0latile: for future reference, initializing an array can be shortened to this if you only want 0s:03:56
ryuochar foo[32] = {};03:56
ryuothe only downside to snprintf really is it had some overhead, but i still think correct code is more importance than a neglible performance hit.03:57
ryuohas*03:58
v0latileryuo: I understand, that's interesting, I always thought that strncpy was a safe function04:06
ryuov0latile: think again. C can be very dangerous.04:07
ryuothe standard C functions for string/memory copying need to be used with care.04:07
ryuoif you have no reason to use them, you should use alternatives that are easier to use correctly.04:07
ryuoas for the read-only functions like strchr, there's no real problems as long as your strings are null terminated.04:09
v0latileYep thanks, I understand, I will read about this, some other recommendation? :-)04:09
ryuoYes... use a macro for doing allocations.04:09
Worksterand then there are a heap of obsolete but still present glibc functions that should never be used in new code ever :D04:09
ryuoideally, you'd also handle allocation failures.04:09
ryuobut that can be a nightmare too.04:10
ryuoinvariably you often end up writing allocation code for every major type you devise.04:10
ryuojust to make it manageable.04:10
ryuoAssuming they're meant to be allocated via malloc, calloc, or realloc.04:10
ryuothis is the macro I generally use for allocations.04:11
ryuo#define alloc(T,N) ((T*)malloc(sizeof(T)*(N)))04:11
ryuoit does two things.04:11
Worksterno sure what this actually does, does it do something like remove leaf ports that have no dependencies on?04:11
ryuo1) handles allocation size computations that is good enough for 99.99% of use cases.04:12
ryuo2) enables the compiler to perform type checking on the type you're allocating.04:12
ryuo(when it gets assigned to a pointer)04:12
ryuov0latile: other than that, using C correctly largely comes down to experience. you have to think in terms of actually memory layout and such and less in terms of high level concepts.04:13
ryuoin general, don't overcomplicate code if you can help it. if appropriate, use other libraries.04:14
ryuothe fact that it's C generally makes performance issues non-existent.04:14
ryuobut, still algorithmic complexity can bite you in the ass in any language.04:14
ryuoerr04:15
ryuoalgorithmic efficiency04:15
ryuosorry, tired.04:15
ryuoa sorting algorithm of quadratic time will be awful in every language.04:15
ryuobut, you don't need to write sorting algorithms for arrays in C. just use qsort.04:16
ryuov0latile: another thing to keep in mind. when you specify an array as a function argument, it's not actually an array. it's a pointer.04:16
ryuoat best it's a hint that your function expects an array type instead of a scalar to be behind the pointer, but it's impossible to tell for sure from the called function.04:17
ryuosome people swear by glib2. it's a very versatile library for C programs and is also used with a lot of software.04:19
ryuoPart of the whole GObject ecosystem.04:19
ryuoIncludes GTK+ and all the other languages that have bindings to it.04:20
ryuov0latile: oh... yea... avoid functions like gets that write into a buffer but don't take a size argument.04:22
ryuomost of the time they're trouble.04:22
ryuothough some assume a minimum size bufer and are safe if supplied with one.04:22
ryuoother functions that maintain private state across function calls (strtok) can be trouble.04:23
ryuothey're non-reentrant.04:23
ryuoso, if you call a function that uses strtok while you're in the middle of using it...04:23
ryuoall hell breaks loose04:24
ryuostrtok_r is an alternative that avoids this problem.04:24
ryuofunny enough though, srand and rand don't have this problem.04:27
ryuoeven though private state is used.04:28
v0latileryuo: I understand, I see that there is still a lot to learn, thanks for taking the time04:29
v0latileI will read about the topics you tell me and I will try to make it better04:29
ryuov0latile: i've been using C since 2008. it's a real pain.04:29
ryuoi'm considering picking up Perl when I can.04:29
ryuobut basically, you should understand every function you're calling.04:30
ryuoi guess that's true regardless but especially in C.04:30
v0latileI started in 2010, but the work took me a long time away from this and from linux, it's just that I reinstalled it again04:30
ryuoC is painful to use, but also generally has the least surprises when you're experienced.04:30
ryuoit doesn't do a whole lot without being told to.04:31
ryuoC++ on the other hand...04:31
ryuolol04:31
jaegerThat's why I like C... and I'm not a prolific programmer04:31
ryuojaeger: i've been thinking CRUX could use a small library to reduce code duplication for like reading ports...04:32
ryuoerr04:32
ryuono04:32
ryuoi mean the package database04:32
jaegerI was intended to make a libpkgdb or similar with my rewrite, for what that's worth04:32
jaegersomeday (tm)04:32
ryuomaybe even make pkgadd and friends into library functions.04:33
ryuoand then bindable from languages like Perl.04:33
jaegers/intended/intending/04:33
ryuomaybe even a system for extracting metadata from ports. :)04:34
jaegerI also want to add a few metadata things to the package database04:35
ryuoAh, that would require some breaking changes.04:35
jaegerlike explicit vs. implicit installation, dependencies in the DB, etc.04:35
ryuoWould you change to something like sqlite?04:35
jaegerI hadn't decided yet, but was considering json04:36
jaegersqlite would be an option04:36
ryuowell, considering all the relationships it might be best.04:36
jaegerI like the ease of parsing text-based formats04:36
ryuoyou could still export data to JSON or XML.04:36
jaegerbut there are advantages to both04:36
ryuoXML is rather appealing.04:38
ryuoespecially XSLT...04:38
brian|lfsI don't get this04:38
ryuo?04:38
brian|lfsdid we just update pyqt504:38
brian|lfspast few days04:38
brian|lfstake that back only tsaop has it and df04:41
brian|lfsand alan and kde4 repo04:41
brian|lfswell I doubt cura is usiing Python2 for anything04:44
brian|lfsbut thats the only thing I see different between me compiling pyqt5 or just extracting the files from the atch one and using theirs04:44
brian|lfsunless there on an older or newer qt504:45
*** abenz_ has joined #crux04:47
jaegerI may try to pick the c pkgutil stuff back up again soon, would like to at least04:47
mayfrostwhere can one start to learn about pkutil internals?04:49
jaegerhttps://crux.nu/gitweb/?p=tools/pkgutils.git;a=summary <-- in the source04:50
mayfrostthanks04:52
jaegerwelcome05:00
brian|lfsahhh shouldn't mess with stuff over tired I figured it out05:03
brian|lfsor maybe not05:04
*** SiFuh has joined #crux05:24
*** SiFuh_ has quit IRC05:25
*** SiFuh has quit IRC06:06
*** SiFuh has joined #crux06:18
*** LinAck has joined #crux06:42
*** j_v has joined #crux07:22
*** LinAck has quit IRC07:26
Romsterryuo, is all this on a learn programming wiki or something?07:27
*** SiFuh has quit IRC07:28
*** SiFuh has joined #crux07:28
*** SiFuh has quit IRC07:43
*** SiFuh has joined #crux07:45
*** SiFuh has quit IRC07:59
*** SiFuh has joined #crux08:05
*** workodera has joined #crux08:23
*** SiFuh has quit IRC08:30
*** SiFuh_ has joined #crux08:30
*** SiFuh has joined #crux08:55
*** SiFuh_ has quit IRC08:56
*** john_cephalopoda has joined #crux09:06
*** SiFuh has quit IRC09:11
*** SiFuh has joined #crux09:12
*** SiFuh_ has joined #crux09:19
*** SiFuh has quit IRC09:20
*** SiFuh_ has quit IRC09:39
*** SiFuh has joined #crux09:40
frinnstno matching key exchange method found. Their offer: diffie-hellman-group1-sha109:51
frinnstno matching cipher found. Their offer: aes128-cbc,3des-cbc,aes192-cbc,aes256-cbc09:51
frinnstcisco are so modern09:51
john_cephalopodafrinnst: Be glad they didn't offer rot13 ;)10:01
SiFuhThis APC UPS is so noisy, I am going to remove the speaker on this also.10:18
*** fun- has quit IRC10:19
*** fun has joined #crux10:19
frinnstw t f10:32
frinnstPlease ensure the Honeywell Download manager tool is installed on your PC prior to trying to download files.  This tool is required for downloading files. The Download Manager installation file can be downloaded by clicking here10:32
frinnsthttp://hsmftpsoft.blob.core.windows.net/downloader/HoneywellSoftwareDownloaderDM.Setup.msi10:32
frinnstWELL THANKS10:32
*** mutin-s has joined #crux10:32
frinnstwhat a fucking joke10:33
*** mutin-sa has quit IRC10:35
*** groovy2shoes has joined #crux10:38
abenz_Make sure you install the included bloatware by choosing Recommended during package selection.10:50
abenz_The bloatware included with this useful tool gives a better user experience.10:51
*** jaeger_ has joined #crux10:55
*** jaeger_ has quit IRC10:56
*** nogagplz has joined #crux11:05
TimB_https://pastebin.com/JQzciBTM <- does somebody have an idea why building anything from compat-32 fails when linking? :^11:14
TimB_I can build glibc-32 though..11:15
cruxbot[opt.git/3.3]: mailx: update to 14.9.911:23
cruxbot[opt.git/3.3]: libxml2-python: update to 2.9.811:23
cruxbot[opt.git/3.3]: libxml2: update to 2.9.811:23
TimB_nevermind, I think I misconfigured my buildflags in pkgmk.conf11:26
TimB_that was it :)11:28
frinnstdont rice your build flags! :-)11:33
TimB_I just swapped -march to haswell, but I did it solely for 32 bit builds AND removed -m32, which was definitly the wrong thing to do :D11:41
abenz_I finally figured out why the kernel threw all those objtool mismatch errors11:43
abenz_its because of the processor optimization11:43
abenz_https://github.com/graysky2/kernel_gcc_patch11:44
j_vabenz_: does it seem like a noticable improvement using those patches?12:00
abenz_from a user experience perspective, barely12:01
abenz_there's some test data comparisons on the github page above..12:01
*** v0latile_ has joined #crux12:01
*** v0latile has quit IRC12:01
j_vmight be interesting to give a whirl on next point release12:07
frinnstand there's a reason why its not included upstream12:15
frinnstwonder if those patches have been submitted upstream and what the reaction was12:16
j_vNight be interesting to looking to it. I'll do some digging after my coffee's done brewing.12:18
j_vMight*12:18
abenzI've been running with march=native kernel builds for a while12:18
abenzthe objtool mismatch errors only appeared recently (4.14 I think)12:19
abenzbut they are cosmetic, the kernel builds and runs fine12:19
j_vI haven't had any of those recently, but I'm building kernels with gcc 7.3.012:20
*** pyl has joined #crux12:20
j_vit has the retpoline additions that gcc 6 doesn't (or didn't last I checked).12:21
cruxbot[xorg.git/3.3]: xorg-xf86-video-ati: update to 18.0.012:31
cruxbot[xorg.git/3.3]: xorg-xdriinfo: update to 1.0.612:31
cruxbot[xorg.git/3.3]: xorg-mkfontscale: update to 1.1.312:31
*** pedja has quit IRC12:33
RomsterSiFuh, try having 5 ups's all go off at the same time when the power goes out. removing the speaker wont prevent all the noise. hum/buzz12:49
frinnstthat 50hz buzz is so soothing12:55
frinnstI want an analog clock with PoE and sntp support12:56
frinnst:D12:56
frinnstbit pricy tho12:56
frinnsthttps://www.direktronik.se/globalassets/_product-images--pdf/1.-natverk/1.1-lan/01--natverkstidklockor/19-0300.jpg12:56
*** pedja has joined #crux13:11
Romsteri've been thinking of that frinnst, why not hack up a small computer for the job13:21
john_cephalopodaBuy a cheap, analog clock, put a RasPi behind it, done.13:25
frinnstSupermicro SC848 4U 24x3.5" 32-core Xeon 32GB ASR-72405 2x1600W13:28
frinnst*do not click buy*13:28
frinnst~1000EUR13:29
frinnstsooooooo inviting13:29
pedjaimpressive willpower13:30
RomsterACTION watches as frinnst melts all the snow around his place with them space heaters13:30
frinnst:D13:31
joacimthe hotspot can be seen from the iss13:31
frinnsthttp://www.supermicro.com/products/chassis/4U/848/SC848E16-R1K62B13:33
frinnstfuck it13:34
frinnstI need to buy a new sofa instead13:34
joacimsome machines can also double as a sofa13:35
Romsterwell at least keep the sofa warm13:37
*** SiFuh has quit IRC13:51
*** SiFuh has joined #crux13:51
john_cephalopodaThe packing material of those things can be pretty comfy.14:15
abenzfrinnst: what generation CPU? wouldn't your ryzen offer comparable performance at a fraction of power use?14:25
abenzI saw a bench where a threadripper was faster than a 64core opteron quad socket of a few years back14:25
joacimhaving actual 64 cores is handy for some things tho15:01
joacimeven if the processor is a bit slower in total15:01
AnselmoI always had heard/thought that there were usually no cases where it was better to have say, 64 2GHz cores, if you had the option of one 128GHz core15:07
frinnsthttps://ark.intel.com/products/64607/Intel-Xeon-Processor-E5-4620-16M-Cache-2_20-GHz-7_20-GTs-Intel-QPI15:07
Anselmojust that . . . .128GHz cores arent a thing really... . .15:07
frinnstdepends on the workload i guess15:07
Anselmobut cant a single core do as many operations per time, without the overheard of managing multiple cores et cetera?15:09
joacimwith virtual machines, you can dedicate cores to it15:10
frinnstnuma is a bit of a pain tho15:10
joacimi assume there are some io related benefits of using multiple cores too15:10
Anselmohmm15:12
*** SiFuh has quit IRC15:13
*** SiFuh has joined #crux15:13
pedjaLevel1Tech tested kdenlive on several machines, 8-20 cores, iirc15:19
pedjaor 4-20+, I forget.anyway, the point of diminishing returns for it is ~10 cores, it just doesn't get any significantly faster15:22
pedjaso, as is always in tech, 'it depends' :)15:23
Anselmoha15:24
pedjaone of the interesting observations they made is that Ryzen/Threadripper systems were more smoother to use during the kdenlive benchmarks15:26
pedja'smoother' is subjective, granted15:29
Anselmomeaning. . . . even while under some amount of load, the system isnt substantially slowed15:29
pedja8 core Ryzen smoother than 10 core Intel :)15:30
Anselmoand here I am with just 15 cores on four machines15:33
Anselmoor, well. . .thats not that few I suppose x-x15:34
pedjamy ancient smartphone has double the cores than my desktop, so you're fine :)15:37
pedjaI think joacim and jaeger are neck to neck in 'how many cores I have to play with' category15:39
Anselmoif I really need I can borrow a few other computers and probably get to 35 or something without much trouble:P15:40
jaegerI think he has me beat in terms of most cores on a single system15:53
joacimhow about in one room?15:54
joacimright next to my desk, i have 16+8+4+2+1+1 cores15:54
joacimnot counting hyperthreading15:54
jaegerI think I have 18 in here, not counting hyperthreading15:55
joacim8 cores from two quad core LGA1366 Xeons15:55
jaeger38 if I add all the rpis, I guess :D15:55
joacimthis is also the room i hoard in15:56
joacimthats a lot of RPIs15:56
joacimcluster?15:56
jaegerthe pi3 is quad-core, right?15:56
Anselmommhmm15:56
jaegerI have 5 of those in a little cluster15:56
Anselmoha15:56
joacimi just have the one. thinking about getting one of the odroid boards to attempt making an appliance style nas15:57
joacimsee if i can use it for automatic backups of windows clients15:57
Anselmoin this room there are at least 32 cores I think, but many arent mine . . .15:57
joacimwhen i get good internets at home, i think i'll ask someone if they're willing to have a backup server of mine at their place15:58
Anselmoha15:58
AnselmoI have no idea how many Tb of hard disks are in this room, but I am pretty sure its over 30 x-x15:58
joacimfile server sits on 24TB in total15:59
joacim8TB in my desktop15:59
Anselmoand I remember when once  Ihad all my data on an 8Gb flash drive16:01
jaegerI mostly have smallish data storage in my machines now... 256GB or 512GB SSDs16:01
jaegereverything large can live on the NAS16:02
joacimi want my 4TB drives to go in a backup server16:03
joacimjust need a few more for RAIDZ I think16:03
Anselmojaeger: similar to my arrangements16:03
Anselmothought my desktop has 16Gb internal storage. .16:04
Anselmomy laptop's 120Gb ssd seems needlessly large, I take less than 20Gb Gb on it all together, and could keep it tidier probably16:10
joacimi mostly got the 128GB SSD for my laptop because the smaller ones werent that much cheaper16:13
AnselmoI picked this one because I desperately needed a working computer and it looked like the best option in the store I was at :P16:15
AnselmoI should figure out what nation the keyboard is for. . .16:17
Anselmoweird, norway it looks like,16:18
joacimwhich is next to the L?16:22
joacimØ or Æ?16:22
joacimif Ø, it is norwegian16:23
AnselmoI cannot type them, but its the J, K, L, AE, O(with line throught it)16:23
joacimDanish16:24
john_cephalopodaj k l æ ø16:24
joacimÆ is just a AE ligature16:24
joacimØ is OE, Å is AA16:24
joacimwell, in english, Æ is just a AE ligature16:25
Anselmohttp://www.keysourcechina.com/Lenovo-ThinkPad-T430-X230-Danish-Keyboard-Backlit.html looks to be the same layout, which is norway/danish it looks like16:26
AnselmoI wonder what brought it to Romania. . .. .16:26
Anselmoalways interesting to think about the prior existances of things16:27
joacimi've seen norwegian laptops listed from germans on ebay16:27
joacimso i guess it might end up at recycling centers or something16:27
joacimthere are a lot of eastern and central europeans gathering electronics from dumpsters, so i guess they sell them in other countries16:28
Anselmohaha16:30
Anselmothis one is sorta scratched but doesnt look like dumpster material. . . .16:30
Anselmoor maybe I need to look in more dumpsters16:31
pedjajoacim, not just electronics :) local flee markets have a lot of other stuff too 'imported' from Western Europe16:36
*** workodera has quit IRC16:36
Anselmoand stores ''german friends'' :316:37
Anselmoor just big GERMAN IMPORT signs in the window and whatnought16:37
pedjaAustria and Germany, yeah :)16:37
pedja'made in Germany' is synonymous with 'quality' in these parts of the world for decades16:39
john_cephalopodaOf all the "Made in ..." labels, "Made in Germany" is the most recognized one, followed by "Made in Switzerland".16:40
joacimAnselmo: you need to look in more dumpsters =)16:40
joacimpeople throw out a lot of nice stuff16:40
joacimthere is a made in norway thing going on. but a lot of it is just "designed in norway" bs16:41
joacimstill made in china16:41
pedjawell, 'God created Earth, everything else is made in China'16:41
john_cephalopodaThere are a lot of hair products that advertise with the slogan "Swiss formula"16:41
joacimwe also have bottled water from norwegian glaciers16:42
joacimi have a feeling theyre just from lakes and ground water, like tap water16:42
pedja:)16:42
Anselmoha16:42
Anselmojoacim: I shall try to make a point of it then :P16:43
joacimsome of our glaciers disappeared recently, so i guess you can blame the bottled water industry16:43
pedjajoacim, 'Peckham Spring' :)16:44
joacimsigned up for a trial with azure. and now i get three emails every single day16:54
joacimthey really want me to dislike their service16:55
j_vIf history suggest anything about MS, then I think that it is possible that they thrive on customers disliking their services.17:13
joacimi notice a lot of people complain about ms, but they never try alternatives17:19
pedjaif they buy Canonical, will customers love them more or less?17:19
joacimbuying canonical would be a huge brownie point17:20
pedjaMS Ubuntu :)17:20
joacimi dont think itll end well for ubuntu tho17:21
Anselmoperhaps many of the people who would hate them more, are already at the point of saturation when it comes to hate for MS . . .17:21
joacimbut i guess existing forks get new users, and thre will be more new ones too17:21
*** v0latile_ has quit IRC17:34
cruxbot[core.git/3.3]: vim: update to 8.0.158718:12
*** onodera has joined #crux18:21
*** SiFuh has quit IRC20:12
*** SiFuh has joined #crux20:12
*** dlcusa has joined #crux20:44
cruxbot[opt.git/3.3]: fontconfig: updated to 2.13.021:16
cruxbot[opt.git/3.3]: nss: updated to 3.3621:16
cruxbot[opt.git/3.3]: nspr: updated to 4.1921:16
john_cephalopodaThere is AXE/LYNX (deodorant) with "minerals from the Himalaya".21:23
frinnstsounds very useful21:34
frinnst:D21:34
dbrookemy sofa cost more than this computer (and is from norway)21:40
*** trothr has joined #crux21:45
john_cephalopodaYou can get very cheap sofas from the country next to Norway ;)21:53
dbrookeyeah, but this came assembled21:55
dbrookeactually, not entirely true, but the delivery people fitted the legs21:56
dbrookewell, feet really http://www.ekornes.co.uk/stressless-sofas/stressless-wave-lowback/wave-2s22:00
AnselmoI havent a sofa x-x22:05
john_cephalopodaI have to sleep on my sofa.22:09
john_cephalopoda(Well, it's a convertible sofa, but about 10 cm shorter than me and only 90 cm wide.22:10
joacimif your sofa is old and from norway, it might have been built by my grandfather =)22:11
dbrookeonly about 8 years old22:15
dbrookehttps://www.youtube.com/watch?v=-BppOWoK2Fc22:15
*** aragornII has joined #crux22:45
aragornIIhello22:45
aragornIII'm trying to install blender but I keep getting an error regarding libwebp.22:48
Anselmowhat does your error say ?22:51
aragornIIlibwebp.so.7 not found.22:57
aragornIII have libwebp installed from contrib22:59
jaegerlooks like contrib/libwebp contains libwebp.so.6, not .723:02
jaegerso maybe you need a newer version23:02
Anselmoor just to update your system23:04
Anselmodid you change anything after building blender?23:05
aragornIII updated everything so I figured I would rebuild blender after it wouldn't start from the terminal23:07
aragornIII just did a ports -u and there's quite a few updates today so I'll see how that goes.23:10
*** onodera has quit IRC23:11
Anselmothat sounds sensible23:12
*** SiFuh has quit IRC23:46
*** SiFuh has joined #crux23:47
*** aragornII has quit IRC23:52

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!