| lkcl | programmerjake, as octavius has dropped off IRC for tonight, he's not aware that this is going to be done by BTC and then *you* would put in the RfP once we've got the MoU signed for this one | 00:18 |
|---|---|---|
| lkcl | or wise. yes, if BTC costs are ridiculous at the moment. | 00:18 |
| lkcl | however see https://bugs.libre-soc.org/show_bug.cgi?id=774 because i suggested some changes and additions to the equipment | 00:19 |
| lkcl | it's more like GBP 300, not GBP 250 | 00:19 |
| lkcl | or thereabouts. needs recalculating basically. | 00:19 |
| programmerjake | i was under the impression that we were using our existing fpga fund, rather than the nlnet grant. since that isn't ready quite yet | 00:58 |
| programmerjake | I also expanded the introduction you started: https://git.libre-soc.org/?p=libreriscv.git;a=commitdiff;h=466cd6f330ac5f18047377cf2ed8126506223c4d | 00:59 |
| programmerjake | lkcl ^ | 01:01 |
| *** alMalsamo is now known as lumberjack123 | 04:04 | |
| lkcl | programmerjake, yes, correct: we're using the FPGA BTC fund. but octavius is under the (mistaken) impression that *he* can put in an RFP to *NLnet*... right now. | 05:29 |
| lkcl | ah brilliant, i got interrupted so had to save it fast. | 05:30 |
| *** alMalsamo is now known as lumberjack123 | 12:32 | |
| octavius | programmerjake, when you talk about BTC fees, did you mean conversation to currencies? | 13:10 |
| octavius | Fees to send from one wallet to another are usually pretty small. I can hold the wallet funds until I need them for later, and spend the fiat I already have. | 13:11 |
| octavius | Otherwise if wise more convenient for you, I could look into it (probably need to give up a lot of personal info though) | 13:12 |
| lkcl | octavius, BTC transactions require a fee to be paid to BTC Miners | 13:32 |
| lkcl | it's their "incentive" to engage (at all) in the actions required to keep the entire network afloat | 13:32 |
| lkcl | the problem is that that cost varies. a few years ago it went as high as USD 70. | 13:33 |
| octavius | I know, I read the paper XD | 13:33 |
| octavius | And been using it | 13:33 |
| lkcl | so if you transferred $0.01 worth of BTC, you paid $70.01 | 13:33 |
| lkcl | everyone says "BTC is great" until you realise it has centralised "blacklisting" and is rate-limited | 13:34 |
| octavius | The fees I found within the electrum wallet are pretty low. Last month sending to a friend was about 0.5USD | 13:36 |
| octavius | Oh, don't get me wrong, BTC is a terrible "crypto" currency, the chain is completely transparent and public | 13:36 |
| octavius | monero is much more interesting | 13:36 |
| octavius | The main problem I've noticed is the centralisation of cryptocurrency (like btc and eth), which defeats the purpose. However even with these shortfalls, BTC is a robust method of moving value from wallet to wallet. Conversion to fiat is a different matter though. | 13:39 |
| octavius | (and big data tracking...) | 13:43 |
| lkcl | there's someone who helps out by donating us in DECRED | 13:43 |
| octavius | Interesting, never heard of it before. Thanks for introducing it :D | 13:52 |
| lkcl | tplaten is doing a LibrePlanet talk about microwatt on orangecrab | 14:08 |
| lkcl | https://libreplanet.org/2022/program/ | 14:08 |
| lkcl | that's in 15 minutes! | 14:10 |
| tplaten | and more applications, including a proposed VR headset that uses Libre-SOC. Yesterday I got godot3.4 with OpenXR running on my Talos II. | 14:12 |
| lkcl | oooOoo :) | 14:12 |
| lkcl | where's it online? | 14:12 |
| lkcl | or do you have to register? | 14:12 |
| lkcl | found it | 14:13 |
| lkcl | https://libreplanet.org/2022/live/ | 14:13 |
| lkcl | i love the music :) | 14:14 |
| tplaten | So if I connect an I2C IMU to the OrangeCrab, I can do head tracking, possibly even inside the maze game. | 14:15 |
| tplaten | https://codeberg.org/librevr | 15:24 |
| tplaten | http://platen-software.de/tobias/LP2022/MazeGame.mkv | 15:30 |
| lkcl | tplaten, that was a pain, the jumping about and stalling, but the replay should be steady | 15:33 |
| tplaten | And for be the bbb did not work on my Talos II, I had to switch to my Laptop | 15:35 |
| tplaten | Here I have connected the OrangeCrab to my Talos II: http://platen-software.de/tobias/LP2022/slimeVR_debian.png | 15:35 |
| tplaten | Currelty having a look at https://media.libreplanet.org/u/libreplanet/m/free-software-game-restoration/ | 15:37 |
| tplaten | When on the microwatt verilator_trace branch I get a "Cell has missing pin: wb_sel_i" error. | 18:41 |
| lkcl | octavius, you happen to have any 0.1mm crimp-headers? https://www.amazon.co.uk/HALJIA-2-54mm-Dupont-Jumper-Connectors/dp/B06WWB66WL/ | 18:44 |
| lkcl | i'm finding that the cables i've set up (about 10in) are too long to run 50 mhz over, i'm going to have to drastically shorten them | 18:45 |
| lkcl | Arty A7 PMODs you're supposed to just plug into the sockets | 18:45 |
| lkcl | but putting into VERSA_ECP5 or orangecrab that's not going to work, they have rows of 0.1 pin headers (arduino style) | 18:46 |
| tplaten | If I copy external_core_top.v to microwatt it seems to be working, so that I can simulate libre-soc. | 18:46 |
| lkcl | cool! | 18:46 |
| lkcl | is this in microwatt_verilator branch? | 18:47 |
| lkcl | you remembered to set EXTERNAL_CORE=true in the Makefile? :) | 18:47 |
| tplaten | For flashing the Linux-Kernel to the OrangeCrab I used 10 cm cables | 18:47 |
| lkcl | yyeah but that's a 100 khz JTAG bus, with serial-only | 18:47 |
| lkcl | this is an 8-bit parallel bus @ 50 to 150 mhz | 18:48 |
| lkcl | where the clock lines have to be exactly on-phase @ 90 degrees | 18:48 |
| tplaten | Yes, I am on the verilator branch | 18:49 |
| lkcl | that's fantastic, then, if you got verilator sims running | 18:50 |
| tplaten | Yes, I will be working on that first, then orangecrab. | 18:51 |
| lkcl | here is the annoying hack to get broken compatibility with microwatt from 12 months ago | 18:53 |
| lkcl | https://git.libre-soc.org/?p=soc.git;a=blob;f=src/soc/simple/issuer.py;h=5f556d3e2aa765e413c8396683998079b180e7c1;hb=5e392ebd3ba9f1568080abe3cde0f38e71e8fb51#l388 | 18:53 |
| lkcl | this was fixed right throughout the microwatt codebased by paul mackerras | 18:53 |
| octavius | lkcl, I think you mean 0.1" (2.54mm) ;) | 19:00 |
| lkcl | yes | 19:01 |
| octavius | I do have them yeah. Do you need a set? Once the parts arrive I could make up a set for you | 19:01 |
| lkcl | but the cable length needs to be really short | 19:01 |
| octavius | I guess you want several sets for the pmods | 19:01 |
| lkcl | it's probably best if i do it, i may have to try different lengths | 19:01 |
| octavius | I even have different colour wire. Do you have any colour schemes in mind? :D | 19:02 |
| lkcl | but in the meantime i've found a verilog sim of hyperram | 19:02 |
| lkcl | heh | 19:02 |
| lkcl | https://libre-soc.org/HDL_workflow/HyperRAM/?updated | 19:02 |
| lkcl | orangecrab needs to be added to that | 19:02 |
| lkcl | and after verilog models work i think i'll try the Arty A7 because that's what these PMODs are designed for | 19:03 |
| lkcl | such a lot of frickin hassle just to get linux running :) | 19:03 |
| octavius | Yes, so many abstraction layers... | 19:06 |
| lkcl | tell me about it | 19:07 |
| octavius | btw, for the orangecrab with ethernet pmod, what kind of test might we do? telnet? | 19:08 |
| lkcl | pffh, haven't even thought that far ahead. | 19:08 |
| lkcl | there is likely some GMII verilator tests somewhere | 19:09 |
| lkcl | but ultimately we want a linux kernel driver | 19:09 |
| lkcl | which given that ETH MACs are a standard is not that hard | 19:09 |
| tplaten | I have been thinking about USB ethernet on the orangecrab. | 19:09 |
| lkcl | tplaten, that means you need USB | 19:10 |
| lkcl | (in HDL) | 19:10 |
| lkcl | yes this is literally that low-level | 19:11 |
| lkcl | anything you might take for granted has to literally be implemented from the ground up | 19:11 |
| octavius | lkcl, do you need an orangecrab? I'll start ordering some of the parts now. | 19:12 |
| lkcl | therefore in the case of USB2, because it requires an Analog PHY, we can't do that, so need a USB UTMI adapter | 19:12 |
| lkcl | https://libre-soc.org/shakti/m_class/ULPI/ | 19:12 |
| lkcl | i've got one here already | 19:12 |
| octavius | sure | 19:12 |
| lkcl | just one of those ETH MACs | 19:12 |
| lkcl | and hm just for fun one of these | 19:13 |
| lkcl | https://www.waveshare.com/usb3300-usb-hs-board.htm | 19:13 |
| octavius | Found one on amazon https://www.amazon.co.uk/Waveshare-USB3300-USB-Board-MIC2075-1BM | 19:13 |
| tplaten | and USB3 and thunderbolt are even more complex | 19:14 |
| lkcl | ah watch out there's a company selling knock-off clones of waveshare parts | 19:14 |
| lkcl | one of them is based in India, and waveshare has warned that their parts are sub-standard | 19:14 |
| octavius | Ah, ok | 19:15 |
| octavius | do they have a uk distributor? I only found robotshop.com | 19:15 |
| lkcl | very unlikely | 19:16 |
| octavius | Ok, I'll order the USB board and lan8720 boards off there then instead | 19:17 |
| octavius | so you need one usb, and two lan8720 boards? | 19:17 |
| lkcl | yes | 19:18 |
| octavius | Sure. I'll grab a usb one for myself as well (just so we have someone else to test with) | 19:20 |
| octavius | lkcl, tomorrow (sunday), I'll make the waveshare order. 3 lan8720 boards and 3 usb3300 boards (one of each for me, two for just in case :) ) | 23:26 |
| lkcl | octavius, great | 23:27 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
