*** tpb has joined #tomu | 00:00 | |
*** rohitksingh has quit IRC | 00:34 | |
*** rohitksingh has joined #tomu | 00:37 | |
*** dwg has quit IRC | 01:41 | |
*** dwg has joined #tomu | 01:49 | |
*** ademski has quit IRC | 02:10 | |
*** xkapastel has quit IRC | 02:18 | |
*** andi- has quit IRC | 02:47 | |
*** andi- has joined #tomu | 02:56 | |
xobs | MadHacker: I redid the address detection to pull it from the Revision field: https://github.com/im-tomu/fomu-flash/commit/5a59794d91d95ab281a689e65cd5c4b37a34b82d | 07:04 |
---|---|---|
tpb | Title: rpi: better handling of cpu revisions · im-tomu/fomu-flash@5a59794 · GitHub (at github.com) | 07:04 |
MadHacker | Ah right, good shout. | 07:33 |
MadHacker | I'll grab that. | 07:33 |
MadHacker | Yep, works fine. Thanks for that. :) | 07:38 |
xobs | I provided the factory with two options for SPI flash, and now they've gone and put both on the board. So I'm getting the test jig to accept two SPI IDs now, and making sure Fomu Flash supports both. | 07:38 |
MadHacker | Heh. Two options and taking both is just being awkward. Maybe issues sourcing either in volume? | 07:39 |
xobs | Could be! Or maybe one was cheaper. Still, it's a supported configuration. And better to have more supply chain options. | 07:40 |
MadHacker | As you say. | 07:40 |
MadHacker | Periphally relatedly, I got 300 PCBs delivered yesterday from JLPCB for an LED project; first time I've seen a PCB mask being fixed with tippex where they've scratched it or something. :D | 07:41 |
MadHacker | White solder mask and they apparently felt it was OK to fix it with correction fluid. :D | 07:41 |
xobs | Creative. | 07:42 |
MadHacker | I don't know whether to be impressed or horrified, TBH. | 07:42 |
xobs | Both seem appropriate. | 07:42 |
MadHacker | https://www.jamiecraig.com/wp-content/uploads/2019/09/b9e70a6a-6a43-43da-8be3-f723a8355047-4523-0000035936b8c224_file-1024x768.jpg | 07:46 |
xobs | There, fixed. | 07:46 |
xobs | But if they're the PCB house I'm thinking of, that really seems in-line with their operating practices. Cheap and fast. | 07:47 |
MadHacker | The mask seems to have misbehaved on the vias in general - some have full-on bubbles on top of them - and I think this is just where there was bare copper showing. | 07:47 |
MadHacker | Oh, absolutely. This was *dirt* cheap. | 07:47 |
MadHacker | 300 PCBs, panelised, and a framed stencil for £57, shipped. | 07:47 |
MadHacker | The quality's crappy but the turnaround time and price were both great. | 07:48 |
MadHacker | Usual triangle - cheap, fast, good, pick two. | 07:48 |
*** rohitksingh has quit IRC | 08:03 | |
*** ademski has joined #tomu | 12:57 | |
*** olasd has quit IRC | 13:09 | |
*** halosghost has joined #tomu | 13:29 | |
halosghost | hey there everyone | 13:29 |
halosghost | any word on the fomu shipping timeline? | 13:29 |
xobs | halosghost: you can email [email protected] and ask them. I thought they had stock. | 13:31 |
*** xkapastel has joined #tomu | 13:33 | |
*** CarlFK has quit IRC | 13:47 | |
halosghost | mm, I may; haven't received mine or gotten an email notification | 13:53 |
*** emeb has joined #tomu | 13:53 | |
halosghost | (about when I should expect it) | 13:55 |
halosghost | and since it's been shipping for a month or so, I was hoping to have gotten mine :P | 13:55 |
halosghost | sent | 13:56 |
* halosghost is so excited to mess with it | 13:56 | |
halosghost | my current plan is to try and implement some basic posit hardware on it | 14:48 |
halosghost | and then I can have a posit coprocessor of sorts | 14:48 |
halosghost | would be so cool | 14:48 |
halosghost | I should probably start with qnums | 14:48 |
halosghost | much easier | 14:48 |
halosghost | and also quite cool | 14:49 |
*** futarisIRCcloud has quit IRC | 15:22 | |
*** xkapastel has quit IRC | 15:53 | |
*** rohitksingh has joined #tomu | 15:55 | |
acathla | halosghost, what's "posit hardware"? | 17:31 |
halosghost | acathla: posits are an encoding of fractional numbers into binary that is meant to be a competing standard to floating-point arithmetic | 17:32 |
halosghost | acathla: where floating-point acceleration hardware is quite complex to implement, posit hardware is apparently much more approachable | 17:33 |
halosghost | that's one of the purported benefits of posits over floating-point | 17:33 |
halosghost | somewhat similarly, qnums are a dramatically simpler binary format for fractional numbers | 17:35 |
halosghost | though, at least at first blush, qnums are dramatically less widely usable than posits | 17:35 |
acathla | Okay, thanks | 17:36 |
halosghost | you're quite welcome; let me know if you want links to documentation or anything :) | 17:37 |
*** CarlFK has joined #tomu | 17:43 | |
*** rohitksingh has quit IRC | 17:47 | |
*** rohitksingh has joined #tomu | 18:02 | |
*** rohitksingh has quit IRC | 19:58 | |
*** rohitksingh has joined #tomu | 20:00 | |
halosghost | apparently, there was a light delay, but the next batch of fomus are expected to start shipping in the next week or so | 20:04 |
halosghost | o/ :) | 20:04 |
*** halosghost has left #tomu | 20:04 | |
*** rohitksingh has quit IRC | 20:19 | |
mithro | felix_: it would be awesome if someone got flashrom working well enough that fomu-flash could be retired | 20:24 |
felix_ | yep, the combination of the rather broad device support of flashrom and the additional features of fomu-flash would be awesome to have in one tool | 20:43 |
felix_ | will probably need enough qspi support for one or two flashrom programmer types to be able to switch spi flashes from quad to normal spi mode to use that for the currently not existing programmer to be used with my qspimux | 20:48 |
*** ademski has quit IRC | 21:10 | |
*** rohitksingh has joined #tomu | 22:00 | |
*** futarisIRCcloud has joined #tomu | 23:04 | |
*** rohitksingh has quit IRC | 23:14 | |
*** rohitksingh has joined #tomu | 23:37 | |
*** js has quit IRC | 23:49 | |
*** js has joined #tomu | 23:52 |
Generated by irclog2html.py 2.13.1 by Marius Gedminas - find it at mg.pov.lt!