<re_irc>
<@lachlansneff:matrix.org> It’s be impossible to solder this stuff by hand haha
<re_irc>
<@thalesfragoso:matrix.org> therealprof: That gives me flashbacks from when I tried to make a driver for those OV cameras, end up scrapping the Linux kernel because it requires some magic values written to magic registers...
starblue1 has joined #rust-embedded
starblue has quit [Ping timeout: 248 seconds]
fabic_ has joined #rust-embedded
fabic_ has quit [Quit: Leaving]
fabic has joined #rust-embedded
tokomak has joined #rust-embedded
fabic has quit [Ping timeout: 245 seconds]
<re_irc>
<@hargonix:matrix.org> thalesfragoso: Hey I did that as well! xD
<re_irc>
<@jamwaffles:matrix.org> That's the best I've got for this hellchip I have to work with
<re_irc>
<@jamwaffles:matrix.org> Still doesn't pass DRC for Aisler but 🤞🤞🤞
<Lumpio->
Looks sus
<Lumpio->
Probably not designed for PCB fabricators that can't do really tiny traces or tented vias or something
<Lumpio->
The chip, that is
<re_irc>
<@jamwaffles:matrix.org> Yeah, the eval board def uses tented vias
<re_irc>
<@jamwaffles:matrix.org> Such a pain, but I got a similar layout to above working in my previous revision of this board
<re_irc>
<@jamwaffles:matrix.org> Just such a pain in the arse
<Lumpio->
Ordered from Aisler before btw? I'm curious about it
<Lumpio->
How fast is it really compared to China if you're ordering to another EU country
<Lumpio->
(China being e.g. JLCPCB)
<re_irc>
<@thejpster:matrix.org> My general rule is "if you can't see the legs, it's no good" ;)
<re_irc>
<@thejpster:matrix.org> QFN and BGA bad. DIP and SOIC good :)
<re_irc>
<@adamgreig:matrix.org> hehe, i'd take a nice 0.8mm bga over a 0.5mm tqfp any day
<re_irc>
<@jamwaffles:matrix.org> thejpster: I really want the QFN version of this IC but it's not in production yet or something weird like that :(. 0.5mm BGA sucks
<re_irc>
<@therealprof:matrix.org> Lumpio-: Shipping takes only 2-3 work days without extra cost, fabrication takes way longer than usually in China.
<re_irc>
<@jamwaffles:matrix.org> Lumpio-: I've quite enjoyed it so far. Not having to do the gerber dance is pretty nice, just straight KiCAD file upload. Production takes a bit longer, but shipping to the UK is usually pretty fast
<re_irc>
<@jamwaffles:matrix.org> This is also a production design so there's a lil bit of paranoia in my choice of Aisler over JLC
<re_irc>
<@therealprof:matrix.org> But in the end you have your boards way quicker than from China unless you're willing to fork out half a fortune for expedite shipping. Bigger problem here is: shipping cost is subject to customs and (at least before the recent new rules) customs really liked picking up my board shipments from China for...
<re_irc>
... closer inspection and extra fuss + costs.
<re_irc>
<@therealprof:matrix.org> It's like Oshpark in the US: Not super cheap but very nice quality, good service, reasonably quick and you don't have to deal with cross continental shipping.
<Lumpio->
I think I've paid about 20€ for ~5 day shipping from JLCPCB
<Lumpio->
I wanna give Aisler a try though next time when I need only a couple of boards
<re_irc>
<@therealprof:matrix.org> Sounds a bit under but in the right ball park. I just notice Elecrow has a new reasonably priced 7 day DDP shipping option for $9.44.
<re_irc>
<@ryan-summers:matrix.org> I used Aisler a while back and was very unimpressed by their timelines. They quoted a reasonable 2-4 week ship time, but then my production run just got continually pushed back. I think in the end it was easily 2x as long as expected
<re_irc>
<@ryan-summers:matrix.org> Which was all before the covid situation as well
<re_irc>
<@therealprof:matrix.org> I don't particularly like JLCPCB, the entry price is very low and catchy but as soon as you add options the price jumps up like crazy. Just adding ENIG makes the service unattractive compared to e.g. Elecrow but if you want to go even beyond and do e.g. castellations you have to pay through the nose...
<re_irc>
<@ryan-summers:matrix.org> It was super low volume (like 10 copies), so my guess is that they just had lower than expected hobbyist orders to assembly into a single run
<re_irc>
<@ryan-summers:matrix.org> But I was used to OSHpark from the USA, where they do that as a business model and I never had any timeline slips with them
<re_irc>
<@ryan-summers:matrix.org> Other than that, no complaints about quality/price
<re_irc>
<@therealprof:matrix.org> Never ran into that problem, also they daily update their service status so it should be easy to estimate what you'd get currently: https://aisler.net/help/service-status/status
fabic has joined #rust-embedded
fabic has quit [Ping timeout: 240 seconds]
fabic has joined #rust-embedded
Darius has quit [Ping timeout: 272 seconds]
tokomak has quit [Ping timeout: 240 seconds]
Darius has joined #rust-embedded
<re_irc>
<@jamesmunns:matrix.org> Aisler was pretty quick for me (over the past year), blitz service is usually "order on a Wednesday, arrive on Mon/Tues"
<re_irc>
<@henrik_alser:matrix.org> jamesmunns: Same here
<re_irc>
<@henrik_alser:matrix.org> Oshpark to Sweden is like 4 weeks in customs 🙃
<re_irc>
<@henrik_alser:matrix.org> How come Mouser can get here from US in 2 days?
<re_irc>
<@jamesmunns:matrix.org> Digikey does it too
<re_irc>
<@jamesmunns:matrix.org> 1. their warehouse is like... across the street from an international airport
<re_irc>
<@jamesmunns:matrix.org> 2. they do this shit (international logistics and customs) all day erry day
<re_irc>
<@jamesmunns:matrix.org> I can usually get digikey parts (from the US, to DE) faster than I can from literally anyone other than Amazon.
<re_irc>
<@jamesmunns:matrix.org> You just gotta spend 50 EUR lol
<re_irc>
<@henrik_alser:matrix.org> jamesmunns: Haha yes same here! Just sondering why some companies pass customs in no time while others’ take weeks?
<re_irc>
<@jamesmunns:matrix.org> paperwork is all in order, then probably a "recognize the box on sight, know the paperwork will be in order" at the customs office :p
<re_irc>
<@newam:matrix.org> my mailman knew which neighborhood he was going to whenever he saw a digikey box 😀
<re_irc>
<@henrik_alser:matrix.org> jamesmunns: That must be it
fabic has quit [Ping timeout: 252 seconds]
<re_irc>
<@therealprof:matrix.org> Yeah, if you do the extra steps and have all the paperwork at the ready it just flies through.
<re_irc>
<@therealprof:matrix.org> The keyword is DDP.
<re_irc>
<@therealprof:matrix.org> Or Incoterms DDP.
<re_irc>
<@henrik_alser:matrix.org> Compact Disc, wonder if the acronym will change meaning to Continous Deployment even in the music industry? Could be a fun way to release music!
<re_irc>
<@jamesmunns:matrix.org> btw, Technology Connections has a really good series on the history of audio
<re_irc>
<@jamesmunns:matrix.org> Like, from alexander bell, to the grammaphone, to records, to tape, to CDs
<re_irc>
<@jamesmunns:matrix.org> and explains the tech and innovations in each one
<re_irc>
<@jamesmunns:matrix.org> Also digikey: lol 0 stock of any part
<re_irc>
<@henrik_alser:matrix.org> 🤣
<re_irc>
<@jamesmunns:matrix.org> Also DAMN, 25 EUR for a 512KiB part.
<re_irc>
<@almindor:matrix.org> i had good look with aliexpress orders from china, shipped in < 2 weeks somehow
<re_irc>
<@almindor:matrix.org> faster than from US (I'm in Canada) and cheaper shipping too, I wonder if it's due to their bulk of exports
<re_irc>
<@dirbaio:matrix.org> would there be a way for PACs to not depend on cortex-m-rt? 🤔
<re_irc>
<@dirbaio:matrix.org> because the recent cmrt 0.7 release is breaking
<re_irc>
<@dirbaio:matrix.org> which would then require a breaking bump in all PACs because PACs depend on cmrt
<re_irc>
<@dirbaio:matrix.org> which in turn would require a breaking bump in all HALs
<re_irc>
<@adamgreig:matrix.org> PACs can depend on >0.6,<=0.8 so it's not a breaking bump
<re_irc>
<@dirbaio:matrix.org> oh what you can depend across multiple major versions like that?
<re_irc>
<@adamgreig:matrix.org> yea
<re_irc>
<@adamgreig:matrix.org> so long as it's compatible with all the bits you need of course
<re_irc>
<@dirbaio:matrix.org> how does that work then?
<re_irc>
<@adamgreig:matrix.org> cargo can resolve any compatible version
<re_irc>
<@dirbaio:matrix.org> if I have a PAC that requires "0.6 or 0.7" and another crate that requires "must be 0.7" then cargo is guaranteed to use 0.7 for the pac?
<re_irc>
<@adamgreig:matrix.org> yes
<re_irc>
<@adamgreig:matrix.org> and same if the other crate requires 0.6
<re_irc>
<@dirbaio:matrix.org> huh
<re_irc>
<@adamgreig:matrix.org> obviously still broken if one crate needs 0.6 and one needs 0.7
<re_irc>
<@adamgreig:matrix.org> the PACs do two things with c-m-rt: activate the 'device' feature if their own 'rt' feature is enabled (annoying, silly, should be fixed), and re-export the 'interrupt' macro (required due to how it's implemented, perhaps we could fix in the next c-m-rt release?)
<re_irc>
<@adamgreig:matrix.org> they could re-export the macro from the cortex-m-rt-macros crate instead but that doesn't really solve the problem
<re_irc>
<@adamgreig:matrix.org> next release of stm32-rs will have `>0.6.13,<=0.8` for this reason, allowing projects to upgrade to c-m-rt 0.7 once they've upgraded pac
<re_irc>
<@adamgreig:matrix.org> of course the pac comes with other changes that means HALs will need work to update
<re_irc>
<@adamgreig:matrix.org> so end projects probably can't use a new c-m-rt until the pac is released and the hal is updated and released too
<re_irc>
<@adamgreig:matrix.org> it would be nice if the PACs were more decoupled from c-m-rt
<re_irc>
<@dirbaio:matrix.org> the "device" feature thing could maybe be done with weak symbols
<re_irc>
<@dirbaio:matrix.org> if they were usable in rust >_<
<re_irc>
<@dirbaio:matrix.org> but the interrupt macro hack, I have no idea
<re_irc>
<@dirbaio:matrix.org> fun, the `links` key was just added in 0.7 so you still get the ugly error if you have 0.6+0.7
<re_irc>
<@adamgreig:matrix.org> the latest 0.6 also has the links key
<re_irc>
<@adamgreig:matrix.org> (0.6.15)
<re_irc>
<@dirbaio:matrix.org> ah oops for some reason I had 0.6.15
<re_irc>
<@adamgreig:matrix.org> nearly gave me a heart attack there lol
<re_irc>
<@adamgreig:matrix.org> good edit
<re_irc>
<@dirbaio:matrix.org> 🙈
<re_irc>
<@adamgreig:matrix.org> cargo is not very proactive about updating minor releases like that
<re_irc>
<@adamgreig:matrix.org> gotta 'cargo update' occasionally or bring something in that explicitly depends on the newer version
<re_irc>
<@dirbaio:matrix.org> so embassy-nrf is stuck until the nrf pacs update
<re_irc>
<@dirbaio:matrix.org> embassy-stm32+stm32-metapac work fine with cmrt 0.7 🚀
<re_irc>
<@adamgreig:matrix.org> nice!
<re_irc>
<@dirbaio:matrix.org> the `>0.6.13,<=0.8` trick doesn't seem to be working though...?
<re_irc>
<@adamgreig:matrix.org> how so?
<re_irc>
<@adamgreig:matrix.org> oh sorry you need
<re_irc>
<@adamgreig:matrix.org> >=0.6.13,<0.8
<re_irc>
<@adamgreig:matrix.org> uh
<re_irc>
<@adamgreig:matrix.org> lol
<re_irc>
<@adamgreig:matrix.org> probably put 0.6.15 there though
<re_irc>
<@dirbaio:matrix.org> ah yes it works
<re_irc>
<@dirbaio:matrix.org> I had `>0.6,<=0.8` which failed if I tried to force using 0.6.15
<re_irc>
<@dirbaio:matrix.org> with `>=0.6.15,<=0.8` it works
<re_irc>
<@dirbaio:matrix.org> YES
<re_irc>
<@adamgreig:matrix.org> you want `<0.8` not `<=0.8`
<re_irc>
<@dirbaio:matrix.org> ah oops
<re_irc>
<@firefrommoonlight:matrix.org> It would be nice if PACs had no dependencies
<re_irc>
<@firefrommoonlight:matrix.org> I pondered that Q, re cortex-m dep a few days ago. Thx for the concise explanation
<re_irc>
<@dirbaio:matrix.org> lol defmt's panic-probe also depends on cmrt
<re_irc>
<@dirbaio:matrix.org> cmrt everywhere
<re_irc>
<@dirbaio:matrix.org> dependency hell
<re_irc>
<@adamgreig:matrix.org> why :/
<re_irc>
<@adamgreig:matrix.org> you can see why crates just give up and keep pushing "minor" releases with or without breakages, lol
<re_irc>
<@adamgreig:matrix.org> I mean, on balance, it's nice that everything keeps working and people follow semver and you basically don't get surprises or bitrot
<re_irc>
<@adamgreig:matrix.org> sure makes doing breaking changes a pain though :P
<re_irc>
<@dirbaio:matrix.org> adamgreig: no idea lol, it seems unused