<ttmrichter>
Well, wow. I didn't realize the HackRF is only an 8-bit ADC/DAC. That seems an odd design choice.
<ttmrichter>
And double-wow. I'm looking at a picture of a LimeSDR and ... they're still using Cyclone IVs? :o
<ttmrichter>
The Mini uses a Max 10.
boru` has joined #openocd
boru has quit [Killed (NickServ (GHOST command used by boru`))]
boru` is now known as boru
tsal has quit [Ping timeout: 268 seconds]
tsal has joined #openocd
thinkfat has joined #openocd
thinkfat_ has quit [Ping timeout: 244 seconds]
cp- has joined #openocd
<cyrozap>
ttmrichter: "they're still using Cyclone IVs" Wait till you hear about all the devices still being built with the Spartan 6 :P
<cyrozap>
But the LimeSDR was crowdfunded back in 2016 (5 years ago now), so it's not _too_ odd.
Haohmaru has joined #openocd
<ttmrichter>
I think it's more a comment about the almost moribund nature of FPGAs.
<ttmrichter>
By being so horribly, and badly closed off they don't attract a lot of the kinds of people who drive adoption. :(
<cyrozap>
But yeah, the 8-bit ADC/DAC and High-Speed USB are kind of limiting on the hackRF. Especially the USB--max throughput in the best case is _maybe_ 320 Mbps, and that's if there's literally nothing else on the bus. That's where the max samplerate comes from--you can send 20M 16-bit complex samples in 320 Mbps.
<ttmrichter>
Hobbyists and the generally intrigued can get reasonably priced, stupidly powerful MCUs for a song and a dance (other than this year, I mean).
<ttmrichter>
There are no good FPGAs in that kind of space.
<cyrozap>
I think the bigger issue is that the only "cheap" FPGAs are the really super old ones, because modern stuff is simply soooo expensive.
<ttmrichter>
And the tooling for it, even if you want to pay the price, is soooooooooooooooooooo bad.
<ttmrichter>
Like they set the intern the task of making it.
<cyrozap>
More like "team of offshore contractors" :P
<cyrozap>
Interns are too expensive :)
* ttmrichter
laughs.
<ttmrichter>
OK, fair cop.
<ttmrichter>
But it's really reached the point where I know people who won't even consider an FPGA for a project because of the raw stupidity surrounding them.
<ttmrichter>
Like how you can't get a datasheet that's even suited to ASSESSING the part (not to mention USING it!) without an NDA.
<ttmrichter>
And you can't get the NDA without committing to certain sales levels.
<ttmrichter>
And even when you do get an NDA signed the datasheet still hides all the information you'd need to use the part without their crap software.
<ttmrichter>
It's like buying an MCU and being told "no, you can't learn these parts of the ISA".
<ttmrichter>
(Not that I'm pointing fingers at ARM here but ... I am.)
<ttmrichter>
Only instead of being ISA parts for the JVM which nobody cares about anymore, it's ... branch instructions you're not allowed to learn.
<cyrozap>
Or it's like, being given datasheets for an SoC that don't explain how the peripherals work, and only to "just use the BSP blobs"? Because that's how the Android phone world does it.
<olerem>
lol
zjason` is now known as zjason
<ttmrichter>
Yeah, pretty much exactly like that, cyrozap.
<ttmrichter>
Though there's a reason for that in the Android world that goes beyond corporate greed.
<ttmrichter>
They may actually be facing regulatory burden. :-/
akaWolf has quit [Ping timeout: 272 seconds]
akaWolf has joined #openocd
<cyrozap>
ttmrichter: Regulations don't prevent the manufacturer from telling the OEM how the DRAM controller works. Or the TRNG. Or the power management CPU core (with a custom ISA). Or the custom (not licensed) video decoder hardware.
<cyrozap>
And in most cases, OEMs do get access to (partial) modem source code, since they need to customize it to account for small hardware differences between the product and the reference design.
<cyrozap>
I see a lot of people throw up their hands and say "it's because of 'regulations', so it can't be helped", even when that's 100% not the case.
<cyrozap>
The simple fact is, writing docs takes time and effort, or in other words, _money_, and so companies can often save money by simply delivering a BSP and not writing docs. It isn't any more complicated than that.
<karlp>
^^^^
<Haohmaru>
o hai karlp, you're here?
<karlp>
yes?
<Haohmaru>
left freenope yet?
<karlp>
when they killed themselves by nuking services, yes, of course.
<braunr>
ttmrichter: actually, a young person from the team who worked on the newer ST apis (like HAL and LL in STM32 Cube) applied at my company, and had a horrible technical score
<braunr>
ttmrichter: he felt like an intern to me
<zmatt>
lol wtf did freenode do now
<jybz>
hi
<jybz>
zmatt: what is freenode doing today ?
<braunr>
jybz: a new era of irc of course
<zmatt>
well I can't connect anymore due to sasl authentication failure, apparently they decided to switch from ircd-seven to InspIRCd and from Atheme services to Anope, and did not bother to migrate *any* nickserv or chanserv registrations in doing so
<jybz>
Totaly new and fresh, virgin of all registration and history
<jybz>
Oh ok, I thought something happend today again.
<zmatt>
no I only just noticed
<jybz>
Yes, I jump into my ZNC configuration to disable SASL in order to reconnect
<jybz>
Do the registration process of all channels I'm in charge, forward all of them in a single one, kicking out, and all people joining the unique remaining room, announcing that we are not on freenode anymore.
Haohmaru has quit []
Hawk777 has joined #openocd
akaWolf has quit [Ping timeout: 272 seconds]
akaWolf has joined #openocd
akaWolf has quit [Ping timeout: 244 seconds]
akaWolf has joined #openocd
Steffann has joined #openocd
Steffanx has quit [Ping timeout: 244 seconds]
ipatch has quit [Quit: WeeChat 3.1]
roxfan has joined #openocd
akaWolf has quit [Ping timeout: 240 seconds]
roxfan has quit [Remote host closed the connection]