**** BEGIN LOGGING AT Wed Jan 13 02:59:57 2021
Jan 13 03:00:16 <buzzmarshall> when i got the bbb i seen the channel here and just log in when i get home
Jan 13 03:00:36 <set_> Smart. Now you can teach me stuff!
Jan 13 03:00:40 <buzzmarshall> i still working mon-wed so its usually just at nite i sit and watch
Jan 13 03:00:48 <set_> Oh.
Jan 13 03:00:54 <set_> Yea. That is the slowest part.
Jan 13 03:00:59 <set_> Sometimes.
Jan 13 03:01:01 <buzzmarshall> the rest of the week tho i work outta the house so i am on here longer
Jan 13 03:01:09 <set_> Oh.
Jan 13 03:01:11 <set_> Okay.
Jan 13 03:01:16 <buzzmarshall> im kinda semi-retired
Jan 13 03:01:25 <set_> Talk and chat way. Retired!
Jan 13 03:01:48 <set_> I am alone and looking for girls to flirt w/ me and to have their dogs NOT piss on my boot.
Jan 13 03:01:55 <buzzmarshall> haha
Jan 13 03:01:59 <set_> That happened.
Jan 13 03:02:02 <set_> The other day.
Jan 13 03:02:07 <set_> It was miserable.
Jan 13 03:02:28 <set_> I played it off like my boots were waterproof which I thought they were.
Jan 13 03:02:38 <set_> Oops. Nope. Wet toes!
Jan 13 03:02:53 <set_> I should have called it quits and kicked her dog.
Jan 13 03:03:07 <buzzmarshall> lol
Jan 13 03:03:25 <set_> Thank you, Thank you. I would like to thank the academy.
Jan 13 03:03:37 <set_> Ut oh!
Jan 13 03:03:58 <set_> I am a side show for now.
Jan 13 03:04:01 <set_> I need a break.
Jan 13 03:04:25 <set_> bbl or brb = undecided = have fun
Jan 13 03:04:54 <buzzmarshall> no probs... have fun whatever it is your doing...
Jan 13 03:05:24 <buzzmarshall> i am jetting soon as i got one more day but i will be bak tormorrow
Jan 13 03:06:05 <set_> otay
Jan 13 03:07:18 <buzzmarshall> this compile is just about done so i think i will just say nite and go check it and be back tomorrow
Jan 13 03:07:22 <buzzmarshall> take care
Jan 13 03:49:38 <GenTooMan> not fully here
Jan 13 03:55:21 <MattB000ne> what?
Jan 13 05:40:29 <slicktux> o/
Jan 13 14:28:52 * bradfa applauds the V, seems like a cool next board jkridner[m] :)
Jan 13 14:29:14 <mattb0000ne> whats the V
Jan 13 14:30:16 <bradfa> mattb0000ne: https://twitter.com/beagleboardorg/status/1349362203850321921
Jan 13 14:31:17 <mattb0000ne> also can someone take a peek at my uEnv.txt
Jan 13 14:31:20 <mattb0000ne> whoah
Jan 13 14:31:44 <mattb0000ne> based on how hot the AI runs I will be able to make lunch on that thing
Jan 13 14:31:45 <mattb0000ne> lol
Jan 13 14:32:09 <mattb0000ne> computation and heat source for a flat iron grill
Jan 13 14:34:33 <mattb0000ne> looks really cool
Jan 13 14:34:41 <mattb0000ne> my uEnv.txt https://pastebin.com/9jWKPpBT
Jan 13 14:35:13 <mattb0000ne> my overlay https://pastebin.com/dUUy3q74
Jan 13 14:35:31 <mattb0000ne> this worked before on my last beagle, I am setting up again but I cannot get the overlay to take
Jan 13 14:35:41 <mattb0000ne> it loads per the bootloader
Jan 13 14:37:04 <zmatt> yeah I recall this was caused by AM335X-PRU-UIO-00A0.dtbo creating a new node and redefining the &pruss label instead of configuring the existing &pruss for uio
Jan 13 14:37:41 <zmatt> and apparently gets applied last, hence the &pruss to which your overlay applies ends up not being used
Jan 13 14:38:11 <zmatt> https://github.com/mvduin/overlay-utils/blob/master/uio-pruss.dtsi
Jan 13 14:38:17 <mattb0000ne> i new it was something
Jan 13 14:38:24 <mattb0000ne> thanks
Jan 13 14:42:32 <mattb0000ne> we are back!!!!
Jan 13 14:42:37 <mattb0000ne> woot thanks zmat
Jan 13 14:43:54 <mattb0000ne> whats this I see on the V USB3.0 ports
Jan 13 14:44:01 <mattb0000ne> I wonder what this means for the AI
Jan 13 14:44:24 <mattb0000ne> seems like a it would cannibalize the market for that.
Jan 13 14:44:40 <mattb0000ne> unless this is way more expensive and the term affordable needs air quotes
Jan 13 14:44:56 <mattb0000ne> how can I get one?
Jan 13 14:44:59 <mattb0000ne> lol
Jan 13 14:45:27 <bradfa> mattb0000ne: the AI with the am57xx is quite a different beast from the V, in my mind at least
Jan 13 14:46:43 <mattb0000ne> i do not know much about the chipsets can you expand on that
Jan 13 14:47:34 <zmatt> what's the
Jan 13 14:47:35 <zmatt> V ?
Jan 13 14:48:08 <mattb0000ne> https://twitter.com/beagleboardorg/status/1349362203850321921
Jan 13 14:49:06 <zmatt> the beaglebone family has never been particularly interesting/relevant when evaluated purely on "mudane" criteria/specs for generic computer systems (things like cpu speed, ram, storage, networking)
Jan 13 14:50:42 <zmatt> oh, how weird
Jan 13 14:52:56 <zmatt> what a strange product to have under the beagleboard.org umbrella ... it seems very much unlike anything previously made
Jan 13 14:53:44 <zmatt> also the SoC name has literally zero google results apart from third-party sites that transcribed the images on the BeagleV page... is it a typo?
Jan 13 14:55:01 <zmatt> no "JH7100" appears to exist
Jan 13 14:59:24 <zmatt> I hope it's just due to a typo in the SoC part number, and in fact the SoC has the same kind of availability, documentation, and support we've come to expect from TI SoCs.  otherwise this project seems like an embarrassment to beagleboard name to me
Jan 13 15:01:29 <zmatt> it'll just be yet another random SBC based on some obscure and poorly supported chinese SoC
Jan 13 15:02:19 <zmatt> the fact that StarFive's website is not found by google (had to find the link on their LinkedIn page) and so far doesn't load for me is... not encouraging
Jan 13 15:02:34 <zmatt> bradfa: any comment on this?
Jan 13 15:03:52 <bradfa> zmatt: I only know what's on the beagleboard.org site, sorry, no extra info from me.  But it feels a lot like the original Beagle Board or the xM in spirit from the info that is available.
Jan 13 15:07:30 <zmatt> those still used SoCs that were well-documented and widely available to anyone... and they're still manufactured and stocked today (even the OMAP3530 of the original beagle)
Jan 13 15:23:52 <MattB0ne> I think they should enforce the backwards compatibility at least. if it were me I would not want to scare off black users
Jan 13 15:24:32 <zmatt> ehm, this is a completely different architecture
Jan 13 15:25:04 <zmatt> it has nothing in common with other beagles, not even the cpu architecture
Jan 13 15:33:48 <MattB0ne_> that is my point I would not want to have to start from scratch
Jan 13 15:45:28 <Konsgn> hey, why would it scare off black people more than any other race?.....
Jan 13 15:45:49 <Konsgn> also, what if it ran qemu for compatibility?
Jan 13 16:25:44 <Konsgn> woa BeagleV, just saw it on twitter, I'm game, looks sweet
Jan 13 16:34:51 <MattB0ne_> capes though?
Jan 13 16:35:08 <zmatt> this isn't a beaglebone
Jan 13 16:37:23 <zmatt> it doesn't have the form-factor or anywhere near the amount of I/O... it doesn't seem likely it has any interesting (non-bog-standard) I/O capabilities like recent TI SoCs do, but no SoC documentation seems to be available so we'll see I guess
Jan 13 16:57:28 <Konsgn> haha, yea it has pretty much nothing to do with the current hardware.
Jan 13 16:58:05 <Konsgn> seems to be more targeted to ai applications
Jan 13 17:13:19 <ayjay_t> why do i get doregged :-O
Jan 13 17:13:25 <ayjay_t> deregistered*
Jan 13 17:15:32 <zmatt> Konsgn: without documentation and pricing it seems a bit early to get enthausiastic :P    and what I found while attempting to research the components so far hasn't given a great impression mostly w.r.t. openness: https://pastebin.com/vu3e6fBk
Jan 13 17:16:09 <Konsgn> 150$ for the 8gb ram
Jan 13 17:18:22 <Konsgn> i will say i really hope it is not related to the gryfalcon ai chips. I hate those guys, spent 80 dollars for a brick usb stick. now they want more money to allow me to use tensorflow and not their proprietary crap.
Jan 13 17:27:48 <zmatt> hmm, I didn't see any price information on the webpage, where did you get that?
Jan 13 17:31:20 <humpelstilzchen[> heise.de says 120$ for 4GB, 150$ for 8GB RAM
Jan 13 17:31:23 <humpelstilzchen[> (german only)
Jan 13 17:31:46 <humpelstilzchen[> aah original source: https://www.cnx-software.com/2021/01/13/beaglev-powerful-open-hardware-risc-v-linux-sbc-targets-ai-applications/
Jan 13 17:35:56 <ayjay_t> v excited for this 8GB and v happy they broke the form factor
Jan 13 17:36:35 <Konsgn> I got linked from tweeter here: https://abopen.com/news/open-hardware-beaglev-at-149-is-the-most-affordable-risc-v-linux-computer-yet/
Jan 13 17:36:47 <Konsgn> zmatt ^
Jan 13 17:37:50 <Konsgn> hahah i just realized there is no way to confirm that the form is actually from beagle and not an info harvesting scheme
Jan 13 17:37:54 <Konsgn> oops
Jan 13 18:17:49 <NishanthMenon> pdp7: is bb.org on mastadon or pleroma?
Jan 13 18:18:51 <pdp7> We should but have not yet.  What instance do you recommend?
Jan 13 18:20:28 <NishanthMenon> pdp7: you folks could host one on bb.org, perhaps? Not sure I'd give good answers for https://linuxrocks.online/@nm/105548759957509752
Jan 13 18:21:06 <pdp7> I like the domain!
Jan 13 18:21:53 <NishanthMenon> Hehe.. yah.. little server in houston, which we donate to keep running. Nice host..
Jan 13 18:23:23 <pdp7> This is the official form to register interest: https://docs.google.com/forms/d/e/1FAIpQLSff2hpW19aiB7SyzA4NK8zhDGgDje26C_z4lpwzJ-Hmk8aJiQ/viewform
Jan 13 18:23:50 <pdp7> We especially want developers for the initial batch of boards this Spring
Jan 13 18:24:26 <pdp7> Before the full distribution availability in September
Jan 13 18:58:42 <mattb0000ne> pdp7: who was the target user when designing the board?
Jan 13 19:02:16 <jay85> I use beaglebone black, I would like to source a 4 pin header that fits the holes beneath 5V power supply connector. Any part number suggestions?
Jan 13 19:02:33 <zmatt> mattb0000ne: people who want a RISC-V board in their hands
Jan 13 19:03:04 <zmatt> jay85: that's just a 4x4 header... note however that that battery header is not supported
Jan 13 19:03:46 <jay85> standard 4*4 headers have pins with diameter too large, I am trying to source one that I don't have to jam into the board for it to fit
Jan 13 19:03:56 <zmatt> and by "not supported" I mean "pretty much unusable without hardware patch"
Jan 13 19:04:36 <zmatt> I've never had problems sticking normal headers into it
Jan 13 19:04:46 <jay85> part number?
Jan 13 19:05:53 <mattb0000ne> RISC-V vs arm
Jan 13 19:06:03 <mattb0000ne> is that the big driver
Jan 13 19:06:10 <mattb0000ne> with RISC-V being???
Jan 13 19:07:25 <zmatt> jay85: actually I don't think I've ever put 2x2 headers into them, I've always used two 2x1 headers that were lying around (optionally cut from longer header strips)
Jan 13 19:09:04 <jay85> And the pin itself wasn't too wide for the PCB holes? PCB measures 30mil and most pins I can find are 30+mil
Jan 13 19:10:40 <zmatt> uhh no? the holes seem normal size to me... the same male headers that fit into the P9/P8 expansion headers also fit in those holes, and the female headers I can find here actually seem to have smaller pins
Jan 13 19:14:14 <zmatt> jay85: so not sure about part number, it's just whatever was lying around randomly
Jan 13 19:14:37 <zmatt> regardless, if you intend to use the battery header, see warning here: https://elinux.org/BeagleBone_Power_Management#Battery
Jan 13 19:15:14 <zmatt> (also beware that the tps65217 is a fairly crappy li-ion charger in that it gives almost no info about the charging process and has no battery monitoring)
Jan 13 19:17:51 <zmatt> mattb0000ne: https://en.wikipedia.org/wiki/RISC-V
Jan 13 19:18:53 <zmatt> mattb0000ne: the main benefit is that it's an open standard (while the ARM architecture requires licensing)
Jan 13 19:19:13 <zmatt> (this is of little importance to developers though)
Jan 13 21:07:04 <jkridner[m]> /msg tbr didn't realize you were a big 802.15.4 fan. have you seen https://github.com/jadonk/beagleconnect ?
Jan 13 21:07:19 <jkridner[m]> grr.... silly Riot/Element
Jan 13 21:13:50 <mattb0000ne> jkridner[m]: Hey do you have any fixes for the USB3 port on the BBAI other than forcing to host. I saw your solution on the TI forums but it is not working for my hub
Jan 13 21:16:19 <jkridner[m]> no, that’s it. just force host and it should work.
Jan 13 21:16:33 <zmatt> jkridner[m]: no
Jan 13 21:16:38 <jkridner[m]> power ok? board doesn’t give type-c power
Jan 13 21:16:46 <zmatt> jkridner[m]: https://github.com/beagleboard/beaglebone-ai/issues/56
Jan 13 21:16:47 <jkridner[m]> no?
Jan 13 21:16:52 <jkridner[m]> k
Jan 13 21:17:42 <zmatt> jkridner[m]: it cannot work unless the bbai's usb spec violation caused by forcing host mode happens to combine in some fortuitous way with a hub's usb spec violation
Jan 13 21:18:08 <zmatt> for details, see that issue I just created
Jan 13 21:18:48 <jkridner[m]> k. thanks. I've used it with a number of docking stations, but it seems they might play nice with this violation.
Jan 13 21:19:57 <zmatt> it's possible, since this spec violation by the hub (unconditionally acting in device data role) is probably not harmful in practice
Jan 13 21:20:07 <zmatt> but it's not something you can rely on...
Jan 13 21:20:20 <mattb0000ne> I am going to get the one you referenced in your solution for now
Jan 13 21:20:40 <jkridner[m]> :-(
Jan 13 21:20:45 <mattb0000ne> hopefully it works but may want to have a list of currently usable hubs
Jan 13 21:21:16 <zmatt> you can omit the "currently", fixing it requires hardware change
Jan 13 21:21:56 <zmatt> a Type-C controller capable of performing data role swap is required
Jan 13 21:22:19 <mattb0000ne> oh I take it we missed the boat for the A2 boards
Jan 13 21:22:23 <mattb0000ne> doh
Jan 13 21:22:47 <zmatt> yeah it replaced the Type-C controller, but unfortunately the replacement also doesn't support it
Jan 13 21:24:02 <zmatt> jkridner[m]: I'll admit this stuff definitely was not straightfoward to figure out... I'm not surprised it was missed during design
Jan 13 21:26:12 <zmatt> Type-C is kind of a complicated mess
Jan 13 21:26:23 <zmatt> well, usb in general really
Jan 13 23:19:43 <set_> buzzmarshall: Sorry about yesterday. I have bad days, assuming you forgive me. I will play nice for now on.
Jan 13 23:19:56 <set_> Risc-V!
Jan 13 23:20:24 <set_> Hey...is that like one big PRU?
Jan 13 23:21:17 <zmatt> no
Jan 13 23:21:40 <set_> Okay.
Jan 13 23:21:42 <set_> No issue.
Jan 13 23:21:54 <set_> You guys hit the books again?
Jan 13 23:21:57 <set_> Nice!
Jan 13 23:22:20 <set_> I saw something about the BBAI having a second version...is that out already?
Jan 13 23:22:46 <zmatt> nope
Jan 13 23:22:53 <set_> Oh. Okay.
Jan 13 23:22:54 <set_> No issue.
Jan 13 23:23:02 <set_> Staying in my corner again...
Jan 13 23:23:18 <set_> I guess I will build this dang bot again. Sheesh.
Jan 13 23:39:59 <ds2> risky-v is as much of a big PRU as the boards are as much of a turing machine :D
Jan 13 23:42:34 <set_> Oh.
Jan 13 23:42:37 <set_> So, nope.
Jan 13 23:43:33 <set_> I just found some neat notes from a Linux course I once was going to use to get a cert.
Jan 13 23:43:34 <set_> Nice!
Jan 13 23:44:25 <set_> networking at its smoothest...time to update.
Jan 14 00:50:36 <buzzmarshall> hey there... no probs alls good...
Jan 14 00:50:59 <buzzmarshall> just was busy reflowing acouple of macair boards for a buddy
Jan 14 01:03:55 <set_> Oh. Okay.
Jan 14 01:07:15 <set_> I am trying to configure a OpenSSH server on my BBB for no known reasoning.
Jan 14 01:07:19 <set_> It is fun, aw!
Jan 14 01:07:40 <set_> I am messing w/ the sshd_config file now.
Jan 14 01:08:09 <set_> I restarted the daemon for sshd and now, it hangs and then lets me sign in under root as another user.
Jan 14 01:08:35 <set_> Once signed into root, it hangs for a long time.
Jan 14 01:18:12 <buzzmarshall> ya somethings off as it should let you sign on pretty quick once it accepts the user/pass
Jan 14 01:19:02 <buzzmarshall> on debian 10 i didnt even have to touch the config files as by default it just worked
Jan 14 01:19:30 <buzzmarshall> only thing ive really altered sofar was turning on the x11 forwarding stuff
Jan 14 01:20:21 <buzzmarshall> one thing tho i do use tho is ip's not host names
Jan 14 01:20:52 <buzzmarshall> i can use hostnames if i want as my inside lans got shared hosts on all the machines in my rack
Jan 14 01:21:54 <buzzmarshall> but i find ip's just always work were as sometimes depending what i am doing the host names seem to not work right even tho they do properly resolve
Jan 14 01:22:54 <set_> Right.
Jan 14 01:22:57 <set_> I have done that too.
Jan 14 01:23:17 <set_> I have the x11 forwarding done too.
Jan 14 01:23:21 <buzzmarshall> personally i never use dhcp either but thats just my thing
Jan 14 01:23:32 <set_> I am learning slowly.
Jan 14 01:23:42 <buzzmarshall> i always use static ip's
Jan 14 01:23:46 <set_> Aw.
Jan 14 01:24:18 <buzzmarshall> ive got over 20 machines in my house network and its just more sane to use static numbers on my lan
Jan 14 01:24:32 <set_> oH.
Jan 14 01:24:33 <buzzmarshall> specially my build farm
Jan 14 01:24:39 <set_> Farms!
Jan 14 01:25:17 <buzzmarshall> lol... ya ive got about 7 old dell 2950's maxed out that i use to build software
Jan 14 01:25:27 <set_> Oh.
Jan 14 01:25:28 <set_> Ha.
Jan 14 01:25:44 <buzzmarshall> when i was doing the amlogic/rockchip dev i was maintaining about 20 differnt boxes
Jan 14 01:26:37 <buzzmarshall> there old but each have a pair of quads and 64g of ram which works good for me
Jan 14 01:26:50 <set_> Of ram. Nice.
Jan 14 01:26:55 <set_> Quick access...
Jan 14 01:27:19 <buzzmarshall> anyways my point is i use static ip's as then everythings always at the same place
Jan 14 01:27:30 <buzzmarshall> just makes management easier
Jan 14 01:27:46 <zmatt> buzzmarshall: I just leave everything on dhcp and connect by .local hostname :P
Jan 14 01:27:59 <buzzmarshall> so with the bbb i did the same thing so i am not sure if doing it hostname wise makes it any differnt
Jan 14 01:28:06 <set_> do you use xeyes?
Jan 14 01:28:16 <buzzmarshall> nope
Jan 14 01:28:16 <zmatt> manually configuring static IPs seem like a relic from the past in most cases
Jan 14 01:28:32 <set_> xeyes for x11 forwarding?
Jan 14 01:28:50 <zmatt> set_: "use" xeyes?  xeyes is just silly demo app
Jan 14 01:28:54 <zmatt> it has no use
Jan 14 01:28:55 <buzzmarshall> i just use one common host file and my routers run them as well as my switches
Jan 14 01:29:05 <set_> Oh.
Jan 14 01:29:14 <set_> Okay. I wonder why this old note says to use it.
Jan 14 01:29:19 <buzzmarshall> i use ssh mostly from the command line or putty somethimes
Jan 14 01:29:38 <set_> Oh.
Jan 14 01:29:46 <set_> I get it. It is for testing the x11 forwarding.
Jan 14 01:29:48 <buzzmarshall> some of them with desktops i will use xrdp on them
Jan 14 01:30:16 <buzzmarshall> but the stuff in the rack is mostly all headless and server builds
Jan 14 01:30:36 <buzzmarshall> so ssh works good for me tho i am sure theres better newer ways
Jan 14 01:31:01 <zmatt> buzzmarshall: putting static hostname definitions on multiple devices sounds like a pain.. just make sure you have local dns (which you generally get for free when using a typical consumer router whose local dns is populated by the dhcp server)
Jan 14 01:31:29 <zmatt> (or use mDNS/avahi, but devices like switches and routers may not support that)
Jan 14 01:31:51 <buzzmarshall> for awhile i was playing with dhcp and .local but for some reason when i had power issues if the whole system didnt come up in the proper order it would act up at times
Jan 14 01:32:59 <buzzmarshall> this craps all in my basement at home and when power goes it sometimes it can be a pain getting everything to work right again as it seems to take a few days for the lan to relearn something
Jan 14 01:33:36 <buzzmarshall> static ip just seems to make that all go away and was what i cut my teeth on so i kinda relapsed
Jan 14 01:33:38 <buzzmarshall> lol
Jan 14 01:33:51 <zmatt> that doesn't make sense
Jan 14 01:34:39 <buzzmarshall> i gave up trying to make sense of what the issue was
Jan 14 01:34:39 <zmatt> like, if everything loses power, they'll obtain new dhcp leases and thereby re-inform the router of their dns name
Jan 14 01:35:12 <buzzmarshall> well not everything goes down as ive got about 80 terabytes of storage in san's units
Jan 14 01:35:59 <buzzmarshall> those are ups'd so when the power flickers and kills the rest of the system they stay up or go down sanely depending on how long the powers out
Jan 14 01:36:22 <zmatt> I don't see how that matters
Jan 14 01:36:42 <buzzmarshall> either  way something with dhcp and the mix of machines don't seem to like it
Jan 14 01:36:54 <buzzmarshall> honestly i can't either
Jan 14 01:37:41 <zmatt> like, the only way I can imagine that results in extended loss of hostname is if the router hands out dhcp leases, does not bother to save these persistently, and then the router is power cycled but not all other devices are
Jan 14 01:37:47 <set_> Okay. It works.
Jan 14 01:37:57 <set_> xeyes. Useless! Yes!
Jan 14 01:38:29 <zmatt> and even that's assuming you didn't configure fixed leases for devices in the router to effectively give them static IP/hostname without having to mess with the individual devices
Jan 14 01:38:40 <buzzmarshall> zmatt... ya i kinda think that might be something to do with it
Jan 14 01:38:51 <zmatt> but the problem in that case is the router neglecting to persist dhcp leases
Jan 14 01:39:06 <zmatt> using avahi would also avoid this problem
Jan 14 01:39:11 <buzzmarshall> i just found the static instantly solves it and stopped playing around trying to figure it out
Jan 14 01:39:42 <buzzmarshall> and dhcp kinda came after i learned so maybe im doing something not quite right with it
Jan 14 01:39:50 <buzzmarshall> thats another posibility
Jan 14 01:41:20 <buzzmarshall> when i started networking stuff i was the 1st dial up provider in the area outside the university and had over 200 9600 baud modems and static was the only way back then
Jan 14 01:41:26 <zmatt> lol
Jan 14 01:41:40 <buzzmarshall> so most of my networking knowledge is built up ontop of that
Jan 14 01:42:00 <zmatt> I mean, fair.  it's not like I haven't used static ips on a LAN in the past... but emphasis on past ;)
Jan 14 01:42:07 <buzzmarshall> dhcp back then for terminal boxes and modems didnt have dhcp
Jan 14 01:42:09 <buzzmarshall> lol
Jan 14 01:42:32 <buzzmarshall> i had 2 class c address pools back then as numbers where plentiful
Jan 14 01:42:48 <zmatt> the days of plenty
Jan 14 01:43:49 <buzzmarshall> i shared a feed with the UofW here in town... fractinal framerelay till isdn came out
Jan 14 01:44:20 <buzzmarshall> there werent even any graphical browsers then
Jan 14 01:44:30 <buzzmarshall> lynx was the thing and uucp
Jan 14 01:44:59 <zmatt> it's still kinda amazing how at the quadrennial dutch hacker camps (https://en.wikipedia.org/wiki/Quadrennial_Dutch_hacker_convention) you'd actually get a public, internet-accessible, unfirewalled IP address from their DHCP whenever you connected to the network
Jan 14 01:45:14 <zmatt> (and also somewhat scary)
Jan 14 01:45:37 <buzzmarshall> but what you say in regards to dhcp i am sure your right and its something im not seeing and doing right
Jan 14 01:45:58 <buzzmarshall> ya for sure the nets become a spooky place these days
Jan 14 01:46:44 <zmatt> yeah, and like... even if you want static assignments, I'd just assign those in the dhcp/dns server instead of messing with the network config of the individual devices
Jan 14 01:46:59 <zmatt> like, I want to be able to pick up a device, connect it to a different network, and just have it work
Jan 14 01:47:18 <zmatt> this since is something I do rather frequently
Jan 14 01:47:23 <zmatt> *since this
Jan 14 01:48:11 <buzzmarshall> actually i was thinking about maybe sticking some dns server daemon on one of the machines in the rack and looking more into that
Jan 14 01:48:28 <buzzmarshall> as your right it would make things much easier
Jan 14 01:49:09 <buzzmarshall> i just kinda fell of the rails messing the bbb and have kinda let the rack and server stuff sit while i play
Jan 14 01:49:16 <zmatt> like I said, typical home routers do this sort of stuff for you... hand out dhcp leases and provide local dns to resolve hostnames of the clients who obtained dhcp leases
Jan 14 01:49:35 <set_> bind!
Jan 14 01:49:42 <buzzmarshall> the only thing i really keep going it it is the storage boxes as all my media devices in the house pull from them
Jan 14 01:49:46 <zmatt> but if you want to do it yourself, dnsmasq on linux also provides this dhcp+dns combination
Jan 14 01:50:27 <buzzmarshall> cool thanx for the tip i will look more into that
Jan 14 01:52:31 <buzzmarshall> for awhile i was using dns server in a couple of co-located servers to pipe to my home lan but then my residential provider started giving me grief so i had to stop piping those streams back into my home
Jan 14 01:53:15 <buzzmarshall> i cant get static ip at home here where i am and they actively swap my ip when the see to much crap coming into the home
Jan 14 01:53:57 <zmatt> the nice thing about using avahi .local addresses is that it works regardless of network infrastructure (router, dhcp server, local dns)... even if you just connect two IPs with an ethernet cable
Jan 14 01:54:07 <zmatt> thanks to link-local IPv6
Jan 14 01:54:36 <zmatt> *connect two devices with an ethernet cable
Jan 14 01:54:37 <zmatt> lol
Jan 14 01:54:50 <zmatt> weird brainfart
Jan 14 01:54:50 <buzzmarshall> now thats definately a area that other then being aware of what it is i no very little about
Jan 14 01:55:37 <buzzmarshall> that was something i tried shutting down inside my lan while trying to figure out the .local goofyness
Jan 14 01:55:51 <zmatt> why would you do that /o\
Jan 14 01:56:06 <zmatt> IPv6 makes things so much better
Jan 14 01:56:07 <buzzmarshall> not knowing how it works mostly
Jan 14 01:56:24 <buzzmarshall> grasping at straws trying to solve my internal problems
Jan 14 01:56:55 <buzzmarshall> and most of what i had read online at the time about ipv6 seemed to say for inside the house you don't really need it
Jan 14 01:57:33 <buzzmarshall> in the end i just reset everything back to the factory or install state which turns it all back on i quess
Jan 14 01:57:37 <zmatt> I'd rather dump IPv4 if I could :P
Jan 14 01:58:06 <buzzmarshall> if i was more up to speed with the newer ways i probably would as well
Jan 14 01:58:46 <buzzmarshall> as ive said tho before one tends to go with what one knows till it don't get you where you want to go
Jan 14 01:58:49 <buzzmarshall> :)
Jan 14 01:59:14 <zmatt> the main difference is that IPv6's large address space generally avoids the need for centralized assignment of individual addresses.  Like, DHCPv6 does exist but it's kinda obscure, rarely used, and Android clients don't even support it
Jan 14 02:00:17 <buzzmarshall> interesting... i had no idea android doesn't support it
Jan 14 02:00:22 <zmatt> normally an IPv6 is split into the prefix (the "where") and the node id (the "who").. the latter is chosen by the individual device (based on MAC address, or randomized if you have "privacy extensions" enabled)
Jan 14 02:01:11 <buzzmarshall> personally i hate android so i don't use it on anything as im not a big fan of java but didn't know it doesn't support piv6
Jan 14 02:01:20 <zmatt> it supports IPv6 just fine
Jan 14 02:01:24 <zmatt> it doesn't support DHCPv6
Jan 14 02:01:40 <buzzmarshall> aw..k...brainfart now on my part
Jan 14 02:01:48 <zmatt> intentionally, since they think DHCPv6 is moronic and against the whole idea of IPv6
Jan 14 02:02:55 <zmatt> normally your router will just announce the prefix of your network
Jan 14 02:03:08 <zmatt> devices append their individual node id, and voila
Jan 14 02:04:12 <buzzmarshall> neat
Jan 14 02:04:37 <buzzmarshall> someday i quess i should make some time to learn a few new things in that regard
Jan 14 02:05:22 <buzzmarshall> im one of those that once i get into something everything gets dropped till i get bored and then go back around the circle
Jan 14 02:05:49 <buzzmarshall> but i will definately add that to the que for the next time around
Jan 14 02:07:43 <buzzmarshall> back in a bit... old lady seen i bagged the newest oak island and is about to take over my tv down here
Jan 14 02:08:13 <zmatt> buzzmarshall: here's an example, snippet from "ip addr" (with numbers changed a bit for privacy) trimmed to just the relevant bits,  and reformatted to highlight how the node id is MAC-derived, and the link-local and public IP address only differ by prefix:  https://pastebin.com/pk0b8Jk8
Jan 14 02:09:07 <zmatt> the link-local (fe80::) one is always present, even before link up... the public one is added when the router has been discovered (through the neighbour discovery protocol, which is IPv6's replacement for ARP)
Jan 14 02:11:37 <zmatt> buzzmarshall: since IPv6 always has a self-assigned link-local IPv6 address, devices can always communicate with each other when connected via ethernet... no need to first get an address assigned, no possibility to have "incompatible" network configuration, unless some idiot turned IPv6 off ;)
Jan 14 02:11:52 <zmatt> and with avahi that also means you can find them by their .local hostname
Jan 14 02:12:52 <zmatt> and the length and ugliness of IPv6 addresses has the additional benefit of encouraging people to let go of the desire to ever deal manually type/configure them, instead of just using hostnames (whether using avahi or otherwise) ;)
Jan 14 02:25:38 <zmatt> buzzmarshall: another cool thing: note how the public prefix ends in :1 .. that's because our router actually gets a /48 from the provider (which is typical).  from that it allocates the :1/64 subnet for its LAN, but if you have internal sub-networks that need their own IPv6 range, they can simply obtain one from the internet router using "prefix delegation" (if enabled), no need to configure static ...
Jan 14 02:25:44 <zmatt> ...routes and such
Jan 14 02:31:55 <zmatt> actually I'm not sure /48 is typical... I'm also reading about people getting /56
**** ENDING LOGGING AT Thu Jan 14 02:59:57 2021