// #eevblog — 2018-12-12
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
00:00:22 <DHess_US> Devtron, the tail current sets the transconductance and output drive.
01:37:11 <IoTPanic> @bba WS2812s require 5v correct?
01:37:25 <bba> IoTPanic, yes
01:38:05 <IoTPanic> I usually put a diode in front of the LEDs VCC, it lowers their logic requirment from 3.4v I beileve to 2.8v
01:38:13 <IoTPanic> Then you can easily use 3.3v logic
01:38:22 <IoTPanic> I have a project now doing the same thing
01:38:38 Quit: Devtron ([email protected]) Quit: AustNet Chat Network - WebIRC
01:38:40 <bba> I know this method, but I will lose brightness
01:39:42 <IoTPanic> Ah, okay. Give me a bit, gotta bring kitten to vet, then if you want Ill take a peek in the datasheet again
01:42:13 <bba> I need to place order soon, so it could be that I will just pick any random voltage shifter; otherwise I will have to wait for 1 more week before I can place next order
01:56:28 Join: Bonny ([email protected]) to #EEVblog
01:56:59 Part: Bonny ([email protected]) from #EEVblog
02:02:18 * BurtyB just uses a 74hct245 to give them 5v
02:17:39 Quit: ChanOP ([email protected]) serviceshub.austnet.org services.austnet.org
02:27:23 Join: eddie ([email protected]) to #EEVblog
02:32:08 <aram> Ampera: using ssh I was able to network boot a SPARC server off a VM from my laptop, while I was in a different country.
02:32:20 <aram> in order to fix it
02:32:37 <aram> network diagram: https://xw.is/wiki/Boot_SPARC_server_behind_NAT_over_the_Internet_from_VM_also_behind_NAT
03:28:13 * IoTPanic agrees with @BurtyB
03:28:32 <IoTPanic> I am trying to make this dumb RAK833 work and being dumb
03:29:22 <IoTPanic> If I see "failed to start concentrator" again Imma gonna have an aneurysm
04:31:01 <Halcyon> Damn it's early *yawns*
04:46:45 <Ampera> yeah, aram, ssh is great.
05:00:36 Join: WebChatUser ([email protected]) to #EEVblog
05:13:56 <Ampera> I'm actually quite glad Windows 10 comes with OpenSSH out of the box, so I don't have to use PuTTY anymore.
05:30:16 <IoTPanic> It does? It didnt when it came out
06:00:54 <rust_collector> I think they added it this summer, iIrc
06:11:47 <Ampera> It does now, works identically to the OpenBSD version (because it is the OpenBSD version)
06:21:21 Quit: WebChatUser ([email protected]) Quit: AustNet Chat Network - WebIRC
06:53:20 <DHess_US> I remember when SPARC pizza box systems were all of the rage.
06:53:48 <DHess_US> That sure didn't last long when faced with such poor performance for a given price.
06:54:35 <coppice> SPARC pizza boxes were at the heart of EDA for quite a long time
06:55:24 <DHess_US> Yep. And 68000 systems were before them.
06:56:17 <DHess_US> The people I knew who had them were late 80's era Unix hackers.
07:07:04 Join: ChanOP ([email protected]) to #EEVblog
07:07:04 Mode: (by ChanOP) +o ChanOP
07:07:05 Mode: (by ChanOP) +o ServalSpots
07:08:06 Quit: gremlinz (~j@localhost) Changing hosts
07:08:06 Join: gremlinz (~j@localhost) to #EEVblog
07:08:17 Quit: gremlinz (~j@localhost) Changing hosts
07:08:17 Join: gremlinz (~j@localhost) to #EEVblog
07:41:12 <IoTPanic> What was so special about SPARC?
07:42:29 <coppice> for a time it was wiping the floor with every other processor, but its approach ran out of steam
07:43:36 <coppice> there are a few different processor architectures which work well over a certain range of silicon geometries, but fall apart as things shrink
07:56:22 <IoTPanic> Got it. Neat bit o' computing history
07:57:18 <IoTPanic> This RAK833 is shit, I am getting pissed
07:58:18 <IoTPanic> When I buy something "easy to hook up!" and end up reading PCI standards it is NOT going well
07:59:31 <aram> the best part about SPARC is that there is open source verilog for it based on the actual production chips at that time.
07:59:39 <aram> and it's also pretty open in general.
07:59:56 <aram> the EEA uses SPARC for this reason.
08:00:07 <aram> they rolled their own chips based on T2 design.
08:09:44 <coppice> Several groups showed interest when Sun open sourced some SPARC design material, but Sun only opened the design because it was already dying. Its hard to get much momentum like that
08:40:31 <IoTPanic> I wasent aware Sun was in the CPU business at all at any point
08:41:29 <coppice> Have you never heard of the SPARC family?
08:42:40 <coppice> Sun developed the original SPARC and tried to get others involved. Fujitsu designed and manufactured SPARC parts. TI fabbed Suns designs for them. I think there were a few other people involved.
08:42:55 <IoTPanic> I have not, I was after that time
08:43:41 <coppice> Sun started out making workstations based on 68k family CPUs, but moved to the SPARC in the late 80s
08:44:43 <IoTPanic> Ah, why was it called a Pizza box?
08:45:20 <coppice> because that it what they looked like
08:47:55 <IoTPanic> Could I run it on an FPGA?
08:48:53 <IoTPanic> I mean, I assume you could assuming documentatioin
08:48:58 <IoTPanic> documentation**
09:28:16 Quit: Crazor ([email protected]) Client exited
09:28:58 Join: Crazor ([email protected]) to #EEVblog
09:30:02 Quit: Crazor ([email protected]) Client exited
09:38:28 Join: Guest|42051 ([email protected]) to #EEVblog
09:39:41 Join: HexicPyth ([email protected]) to #EEVblog
09:39:52 Quit: Guest|42051 ([email protected]) Quit: http://www.kiwiirc.com/ - A hand crafted IRC client
09:54:43 <DHess_US> IoTPanic, SPARC is a classic example of what *not* to do in a RISC processor.
09:55:35 <DHess_US> MIPS and PowerPC have a bunch of gotchas like that also.
09:55:54 <DHess_US> They have features which seem great but become liabilities at higher performance levels.
09:56:50 <DHess_US> They were called Pizza Boxes because they had a desktop PC with SPARC which was like half the height of most PCs then.
09:57:40 <DHess_US> https://en.wikipedia.org/wiki/SPARCstation#%22Pizzabox%22_systems
10:16:26 <coppice> Most RISC processors turned out to be paths you shouldn't take. 5 years ago it seemed like RISC was a total loser for high throughput, but ARM seems to be achieving some interesting results at the moment
10:18:12 <IoTPanic> Ah, I'd want to run it just cause. I really want to get into CPU architecture stuff on the side.
10:53:52 <DHess_US> ARM was the least RISC of the RISCs.
10:54:09 <DHess_US> SPARC - Register Windows
10:54:17 <DHess_US> MIPS - Branch Delay Slots
10:54:31 <DHess_US> Anything which cannot handle unalaigned accesses
10:54:55 <DHess_US> Anything which lacks subregister wirth operations - Alpha
10:55:40 <DHess_US> Double Indirect Addressing - 68K
10:55:58 <DHess_US> Leave it to 68K to make even x86 look good.
10:56:32 <DHess_US> Pentium Pro is what killed SPARC
10:56:41 <DHess_US> They never caught up or even stayed even.
10:58:04 <coppice> Pentium Pro started the decline in SPARC's value, but it was later on, when the Pentiums started hitting 1GHz, that SPARC actually started falling behind on pure performance
10:58:59 <coppice> When I had both Pentium Pro and Alpha machines, the Alpha running Unix was WWWWAAAAAYYYYY faster than the Pentium Pro running Unix
11:07:40 <IoTPanic> Would you guys recommend starting with an FPGA and go that route if I wanted to play around with cores? If so, would you happen to have an FPGA preference?
11:08:38 <bba> IoTPanic, did you have a chance to look at WS2812's datasheet?
11:09:04 <IoTPanic> I did not, I just recommended to go with the guy I tagged in my response's
11:09:38 <IoTPanic> If you're making a one off, just grab a cheap level shifter that can go fast enough
11:11:02 <IoTPanic> Or if you want something really small you can sometimes hack in a tiny opto isolator but just because that works with some personnel projects with me doesn't mean you should do it if you can get a level shifter
11:12:32 <bba> the question is how to choose the right one, when I do the parametric search I get thousand different ones; that's why I was asking if I need to pick the one that has tri-state capability and push-pull
11:14:04 <bba> e.g. I have picked this one: https://www.digikey.com/product-detail/en/nexperia-usa-inc/74LVC1T45GW125/1727-4560-1-ND/2530939
11:21:11 <IoTPanic> Well, honestly, you don't need anything bi-directional since the LEDs dont talk back, with a resistor you can just you a bi or uni state, and man, if it goes fast enough it's good enough. The LEDs are $0.04 made in china and dont really care
11:21:38 <IoTPanic> Would not be concerned with much spec wise. If it has it great
11:22:20 <IoTPanic> What you found is probably overkill but it would work very nicely
11:26:00 <IoTPanic> I love Digikey added that stacked/scrolling switch, I guess they had enough of people like Mike bitching
11:26:15 <T3sl4> it doesn't stay though :/
11:26:53 <IoTPanic> Wow.
11:27:03 <IoTPanic> Cant even put it in a cookie or somethin
11:27:10 <IoTPanic> Or at LEAST a GET variable
11:27:15 <T3sl4> or the cookie is set too short
11:27:27 <T3sl4> i never stay logged in either, they probably have it set shorter than a day
11:27:45 <IoTPanic> Oh.
11:27:51 <IoTPanic> Why? XD
11:28:04 <T3sl4> ¯\_(ツ)_/¯
11:28:36 <IoTPanic> That hurts me physically, at work we're expected to load user's sessions just like how they left it, no matter the computer
11:29:37 <IoTPanic> However, it does all load in one big JSON reply and planned for it so I guess thats not a fair comparison
11:31:38 <bba> IoTPanic, ok, thanks, I will give it a try
11:32:10 <IoTPanic> @bba you ever have to get those in quantity, buy from china, they are SOO cheap
11:32:19 <IoTPanic> LCSC has them
11:34:24 <bba> it will be much more complicated given that I will have to fill outl import forms, pay different taxes and so on
11:36:18 <IoTPanic> Ah
11:45:10 Join: Guest|30958 ([email protected]) to #EEVblog
11:49:06 Quit: Guest|30958 ([email protected]) Quit: http://www.kiwiirc.com/ - A hand crafted IRC client
12:00:22 <IRCStats> mIRCStats update for #EEVblog started
12:00:29 <IRCStats> New stats available for #EEVblog! Check out the stats at http://www.irc-stats.org/austnet/eevblog
12:07:20 <IoTPanic> The eevblog stats crashed Windows
12:07:31 <IoTPanic> Is there nothing that cant crash windows
12:16:34 <T3sl4> crashed your browser crashed windows*
12:32:07 <IoTPanic> No, windows crashed
12:32:16 <IoTPanic> Due to video scheduler error
12:32:28 <IoTPanic> WIndows doesn't like Vega as much as linux does
12:36:08 <T3sl4> ew
12:36:23 <IoTPanic> To the error or vega?
12:37:25 <T3sl4> the error
12:37:40 <T3sl4> i dont know vega, i assume it's as ugly as everything else inside
12:43:49 <IoTPanic> Its AMDs new graphics solution. So it's a weird situation where my laptop has a SoC with a Ryzen 7 CPU, Vega 12 GPU, and 12GB of RAM, 2 to 4 GB of which is given to the GPU as needed
12:45:52 <T3sl4> ¯\_(ツ)_/¯
12:53:48 <Ampera> Actually, out of all of the issues I've had with Windows NT, stability has not been one of them.
12:54:47 <Ampera> Full system crashes I think only happened a very very small number of times, with the only significant instance being a bluescreen upon running ext2fsd, which went away over time.
12:55:35 <Ampera> Security hasn't been an issue on NT for me in quite a while, but I'm not all that cavalier with what I install/download/run
13:00:47 * T3sl4 bitchslaps Ampera
13:04:38 * Ampera moans
14:06:55 Quit: bba ([email protected]) Quit: Leaving
14:29:41 Join: darksaboteur ([email protected]) to #EEVblog
14:52:50 Quit: byteturtle ([email protected]) Ping timeout: 121 seconds
15:00:56 Join: byteturtle ([email protected]) to #EEVblog
15:20:29 Join: ThunderSqueak ([email protected]) to #EEVblog
15:24:12 <timk> why are you calling it nt? what is this like 1997
15:25:40 <IoTPanic> Its the kernel name?
15:27:10 <IoTPanic> It encompasses multiple distros
15:27:59 <timk> fair call
15:34:52 <timk> DHess_US: i tried for ages to get wifi working on my sparcstation in netbsd using a pcmcia wifi card and an sbus-pcmcia bridge card
15:35:03 <timk> never did figure it out though :(
15:46:58 <DHess_US> Wifi has always been a problem for BSD. The usual recommendation is to just use a separate ethernet to wifi bridge or AP.
15:47:21 <DHess_US> The wifi chipset makers don't provide the needed drivers or documentation.
16:02:51 <Ampera> wlan is even an issue with Linux quite often. It's pretty much the one thing where Linux drivers are often inferior to NT drivers.
16:03:09 Quit: HexicPyth ([email protected]) EOF from client
16:24:37 <DHess_US> For a while there was a shim to use Windows wifi drivers on Linux.
16:25:01 <DHess_US> There was debate about whether that was a good idea because it woudl discurage developement of native Linux drivers.
16:25:59 <steve30> I've used Wireless OK on linux. Can't say I've had much experience though as my Linux machine normally uses ethernet
16:28:48 Quit: steve30 ([email protected]) Quit: <Baljem> steve30: by the way, I had the most bizarre dream the other day that someone came in here and seriously flipped their shit over your quit message and why aren't the ops /doing/ something about it
16:32:53 <Ampera> I mean, as with much on Linux, it's gotten better over time.
16:33:30 <Ampera> That, however, is the paradox with Wine. If it ever gets perfect, then there's little incentive to make Linux native versions of anything, but if it doesn't get good, there's a lot of software that will never become linux native
18:12:44 <Halcyon> Evening all
18:19:34 <Ampera> that's your opinion
18:19:48 <Ampera> how do halcyon
18:35:03 <Halcyon> Not too bad, how about you?
18:40:59 Quit: ThunderSqueak ([email protected]) Quit: Leaving
19:17:15 <Ampera> A mix between fairly happy and incredibly existentially depressed.
19:49:22 <Halcyon> So a wide range as usual
20:09:09 <Ampera> I have am cold.
20:09:44 <Halcyon> Nawww :(
20:15:33 <Ampera> I blame you.
20:15:45 <Ampera> you invited me to IRC and I got a virus from it.
20:16:03 * Halcyon makes Ampera chicken soup
20:17:00 * Ampera would walk 5,000 miles and walk 5,000 more just for that soup.
20:18:23 <Halcyon> You really are sick :(
20:18:53 <Ampera> imma sick fuck who likes a quick fuck
20:19:15 <Halcyon> Lies
20:20:23 <Halcyon> I'm going to go prepare tomorrows lunch and watch YouTubes in bed
20:21:14 <Ampera> prepare tomorrow's lunch in bed.
20:41:47 Quit: darksaboteur ([email protected]) Ping timeout: 121 seconds
21:08:38 Join: steve30 ([email protected]) to #EEVblog
21:08:43 Mode: (by ChanOP) +o steve30
21:32:42 <IRCStats> mIRCStats update for #EEVblog started
21:32:48 <IRCStats> mIRCStats update for #EEVblog complete! Check out the stats at http://www.irc-stats.org/austnet/eevblog
23:07:26 Join: coppice_ ([email protected]) to #EEVblog
23:07:35 Mode: (by ChanOP) +o coppice_
23:08:19 Quit: coppice ([email protected]) Ping timeout: 121 seconds
Search this day's buffer: