From paul at boddie.org.uk Mon Dec 2 15:22:13 2019 From: paul at boddie.org.uk (Paul Boddie) Date: Mon, 02 Dec 2019 16:22:13 +0100 Subject: [Arm-netbook] EOMA68 Computing Devices Update: Measurements and a Hypothesis Message-ID: <2142097.Xevr7QOWls@jeremy> Hello again, I saw the latest update when it was issued a few days ago, with these details being particularly important: "PCBs are etched with acid, and when the tracks or pads are particularly close together (such as with BGA balls), not enough acid gets in to eat away the copper, and (in a very indirect way) the BGA pads end up being far larger than they should be. This, in turn, means that when the IC is put on the PCB and heated up, the BGA balls (which are made of solder) will spread out much farther, and potentially even spread so far that they contact each other and short out." https://www.crowdsupply.com/eoma68/micro-desktop/updates/measurements-and-a-hypothesis >From what it says, it rather sounds like same (or a similar) problem is being encountered as the one which happened with the Ingenic JZ4775 boards: "When the first 6 samples were done, only one was found to be working: for some reason the DDR3 BGA solder balls had spread and caused short circuits. One of the samples was still with the factory, and that too was found to have shorts. Mike has been extremely helpful, and this is the first time that he's been involved with BGA, where his uncle's factory has been the one that assembled the samples." http://rhombus-tech.net/ingenic/jz4775/news/EOMA68-jz4775_XRay_Photos/ At that point in time, I don't remember any more news about resolving the matter, and with the campaign being finalised the priority was no longer to get these boards working. But perhaps one of these situations informs the other, or maybe they inform each other somehow. Thanks in any case for keeping us updated! Paul From lkcl at lkcl.net Mon Dec 2 20:05:55 2019 From: lkcl at lkcl.net (Luke Kenneth Casson Leighton) Date: Mon, 2 Dec 2019 20:05:55 +0000 Subject: [Arm-netbook] EOMA68 Computing Devices Update: Measurements and a Hypothesis In-Reply-To: <2142097.Xevr7QOWls@jeremy> References: <2142097.Xevr7QOWls@jeremy> Message-ID: On Mon, Dec 2, 2019 at 3:23 PM Paul Boddie wrote: > Hello again, > > I saw the latest update when it was issued a few days ago, with these > details > being particularly important: > > "PCBs are etched with acid, and when the tracks or pads are particularly > close > together (such as with BGA balls), not enough acid gets in to eat away the > copper, and (in a very indirect way) the BGA pads end up being far larger > than > they should be. This, in turn, means that when the IC is put on the PCB > and > heated up, the BGA balls (which are made of solder) will spread out much > farther, and potentially even spread so far that they contact each other > and > short out." > > > https://www.crowdsupply.com/eoma68/micro-desktop/updates/measurements-and-a-hypothesis > > From what it says, it rather sounds like same (or a similar) problem is > being > encountered as the one which happened with the Ingenic JZ4775 boards: > yehyeh. except i didn't know about the PCBs not actually matching the gerber files, back then. > > "When the first 6 samples were done, only one was found to be working: for > some reason the DDR3 BGA solder balls had spread and caused short > circuits. > One of the samples was still with the factory, and that too was found to > have > shorts. Mike has been extremely helpful, and this is the first time that > he's > been involved with BGA, where his uncle's factory has been the one that > assembled the samples." > > http://rhombus-tech.net/ingenic/jz4775/news/EOMA68-jz4775_XRay_Photos/ > > At that point in time, I don't remember any more news about resolving the > matter, and with the campaign being finalised the priority was no longer > to > get these boards working. i did actually partly get them up and running, but i'd put a 24mhz XTAL on instead of 48mhz, and the SD/MMC wasn't having it. i tried fixing that in software (doubling the PLL frequencies for the SD/MMC) but couldn't get it up and running. with it only being single-core 1ghz MIPS32 i dropped the investigation. But perhaps one of these situations informs the > other, or maybe they inform each other somehow. > i'd forgotten about it, so thank you for the reminder. > > Thanks in any case for keeping us updated! > no problem paul. l. From maillist_arm-netbook at aross.me Sat Dec 7 07:13:14 2019 From: maillist_arm-netbook at aross.me (Alexander Ross) Date: Sat, 7 Dec 2019 07:13:14 +0000 Subject: [Arm-netbook] EOMA68 Computing Devices Update: Measurements and a Hypothesis In-Reply-To: References: <2142097.Xevr7QOWls@jeremy> Message-ID: i noticed the update too, right on que, was wondering how things where. :D Great that mikes helping. the project is fortunate to have someone like that. noticed your ut demo vids about the motorbike to to e-bike upgrade kit. :) but when i searched that bike model on ebay, all i got seamed to be classic old bikes not the modem looking bike in the video. Thanks again for efforts and the info :). From lkcl at lkcl.net Sat Dec 7 08:54:14 2019 From: lkcl at lkcl.net (Luke Kenneth Casson Leighton) Date: Sat, 7 Dec 2019 16:54:14 +0800 Subject: [Arm-netbook] EOMA68 Computing Devices Update: Measurements and a Hypothesis In-Reply-To: References: <2142097.Xevr7QOWls@jeremy> Message-ID: On Saturday, December 7, 2019, Alexander Ross wrote: > i noticed the update too, right on que, was wondering how things where. :D > Great that mikes helping. the project is fortunate to have someone like > that. yes, very. > > noticed your ut demo vids about the motorbike to to e-bike upgrade kit. > :) but when i searched that bike model on ebay, all i got seamed to be > classic old bikes not the modem looking bike in the video. look up "sur ron" on youtube, internet, and endless-sphere. > > Thanks again for efforts and the info :). :) -- --- crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68 From doark at mail.com Fri Dec 13 03:09:31 2019 From: doark at mail.com (David Niklas) Date: Thu, 12 Dec 2019 22:09:31 -0500 Subject: [Arm-netbook] Dear luke, what do you think of North American PCB fabs? Message-ID: <20191212220931.3a2c9dac@Phenom-II-x6.niklas.com> Hi luke, Today I got an email from allaboutcircuits and it was not an article (what I subscribed for), but an advertisement for a "deal" that Advanced Circuits (www.4pcb.com) was running. I don't have a thing to make, but thought of you and wondered if such companies really do give good deals and are nice to work with. They advertise "free" software that's not GPL so I'm unimpressed but other prospective customers would be. The "deal" was 50% off fist and second orders up to $250 and for businesses, up to $500 off PCB orders with a promo-code (Of course, nothing is "free" and they'll have to make it back somehow). You see, at some point in time (I'm guessing 2+ years at least), I'm going to have to not roll my own and instead use a real PCB fab so I'm keeping my eyes open. Thanks, David From lkcl at lkcl.net Fri Dec 13 04:21:20 2019 From: lkcl at lkcl.net (Luke Kenneth Casson Leighton) Date: Fri, 13 Dec 2019 04:21:20 +0000 Subject: [Arm-netbook] Dear luke, what do you think of North American PCB fabs? In-Reply-To: <20191212220931.3a2c9dac@Phenom-II-x6.niklas.com> References: <20191212220931.3a2c9dac@Phenom-II-x6.niklas.com> Message-ID: hi david i generally found them to be very expensive, relatively speaking, even the lower-entry ones. eurocircuits on the other hand, which uses factories in hungary, was extremely competitively priced for 2-layer PCBs with 12mil clearances. the advantage of local PCB manufacturing: if you need it now, you need it now, and you can get it now. l. From doark at mail.com Sun Dec 29 02:34:37 2019 From: doark at mail.com (David Niklas) Date: Sat, 28 Dec 2019 21:34:37 -0500 Subject: [Arm-netbook] Once upon a time you suggested building an opensource router Message-ID: <20191228213437.72de778d@Phenom-II-x6.niklas.com> Hi luke, I don't think that you created one because it would have not been capable of 1000Mb ethernet, though most people in the US have 100Mb or less internet. The company friendly elec has decided to do so: https://www.friendlyarm.com/index.php?route=product/product&product_id=274 A brief overview: H3/H5 powered 512MB RAM. Advertised as gigabit capable, but not actually able to do this with just TX: 230Mbps and RX 334Mbps on the second port (the first runs at almost the correct speed). The minimal config without wall brick PSU and PSU cable (USB) goes for $19.99. Thought you would be interested, David From lkcl at lkcl.net Sun Dec 29 03:06:55 2019 From: lkcl at lkcl.net (Luke Kenneth Casson Leighton) Date: Sun, 29 Dec 2019 11:06:55 +0800 Subject: [Arm-netbook] Once upon a time you suggested building an opensource router In-Reply-To: <20191228213437.72de778d@Phenom-II-x6.niklas.com> References: <20191228213437.72de778d@Phenom-II-x6.niklas.com> Message-ID: On Sunday, December 29, 2019, David Niklas wrote: > Hi luke, > I don't think that you created one because it would have not been capable > of 1000Mb ethernet, though most people in the US have 100Mb or less > internet. > The company friendly elec has decided to do so: > https://www.friendlyarm.com/index.php?route=product/product&product_id=274 > > A brief overview: > H3/H5 powered 512MB RAM. > Advertised as gigabit capable, but not actually able to do this with just > TX: 230Mbps and RX 334Mbps on the second port (the first runs at almost > the correct speed). how the heck did they manage _not_ to achieve full gigabit throughput on a router? moo? :) > > Thought you would be interested, > David > > only in a horrified-fascinated kind of way :) did they wire up a USB2-ETH IC or something? those numbers sound like they sourced a GbE USB2 chip. USB2 of course maxing out at 480mbps. l. -- --- crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68 From doark at mail.com Sun Dec 29 03:44:48 2019 From: doark at mail.com (David Niklas) Date: Sat, 28 Dec 2019 22:44:48 -0500 Subject: [Arm-netbook] Once upon a time you suggested building an opensource router In-Reply-To: References: <20191228213437.72de778d@Phenom-II-x6.niklas.com> Message-ID: <20191228224448.12f4cdbe@Phenom-II-x6.niklas.com> On Sun, 29 Dec 2019 11:06:55 +0800 Luke Kenneth Casson Leighton wrote: > On Sunday, December 29, 2019, David Niklas wrote: > > > Hi luke, > > I don't think that you created one because it would have not been > > capable of 1000Mb ethernet, though most people in the US have 100Mb > > or less internet. > > The company friendly elec has decided to do so: > > https://www.friendlyarm.com/index.php?route=product/product&product_id=274 > > > > A brief overview: > > H3/H5 powered 512MB RAM. > > Advertised as gigabit capable, but not actually able to do this with > > just TX: 230Mbps and RX 334Mbps on the second port (the first runs at > > almost the correct speed). > > > how the heck did they manage _not_ to achieve full gigabit throughput > on a router? moo? :) What is moo for? I didn't get you cow joke. I want to though. > > > > Thought you would be interested, > > David > > > > > only in a horrified-fascinated kind of way :) Now to tell the tail of Franken Debian... :D > did they wire up a USB2-ETH IC or something? those numbers sound like > they sourced a GbE USB2 chip. USB2 of course maxing out at 480mbps. > > l. Spot on. David From lkcl at lkcl.net Sun Dec 29 04:51:12 2019 From: lkcl at lkcl.net (Luke Kenneth Casson Leighton) Date: Sun, 29 Dec 2019 12:51:12 +0800 Subject: [Arm-netbook] Once upon a time you suggested building an opensource router In-Reply-To: <20191228224448.12f4cdbe@Phenom-II-x6.niklas.com> References: <20191228213437.72de778d@Phenom-II-x6.niklas.com> <20191228224448.12f4cdbe@Phenom-II-x6.niklas.com> Message-ID: On Sunday, December 29, 2019, David Niklas wrote: > > > > > how the heck did they manage _not_ to achieve full gigabit throughput > > on a router? moo? :) > > What is moo for? I didn't get you cow joke. I want to though. moo as in "i don't understand", said in a funny way. i thought they were using RGMII GbE PHYs, which would be really bad not to reach 1000 mbos. integrated USB2-GbETH, now it makes sense not to reach 1000mbps. l. -- --- crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68 From kaklik at mlab.cz Sun Dec 29 13:46:05 2019 From: kaklik at mlab.cz (=?UTF-8?Q?Jakub_K=C3=A1kona?=) Date: Sun, 29 Dec 2019 14:46:05 +0100 Subject: [Arm-netbook] Once upon a time you suggested building an opensource router In-Reply-To: <20191228213437.72de778d@Phenom-II-x6.niklas.com> References: <20191228213437.72de778d@Phenom-II-x6.niklas.com> Message-ID: This should be related to actual discussion: https://www.turris.cz/en/mox/overview/ ne 29. 12. 2019 v 3:34 odesílatel David Niklas napsal: > Hi luke, > I don't think that you created one because it would have not been capable > of 1000Mb ethernet, though most people in the US have 100Mb or less > internet. > The company friendly elec has decided to do so: > https://www.friendlyarm.com/index.php?route=product/product&product_id=274 > > A brief overview: > H3/H5 powered 512MB RAM. > Advertised as gigabit capable, but not actually able to do this with just > TX: 230Mbps and RX 334Mbps on the second port (the first runs at almost > the correct speed). > The minimal config without wall brick PSU and PSU cable (USB) goes for > $19.99. > > Thought you would be interested, > David > > _______________________________________________ > arm-netbook mailing list arm-netbook at lists.phcomp.co.uk > http://lists.phcomp.co.uk/mailman/listinfo/arm-netbook > Send large attachments to arm-netbook at files.phcomp.co.uk From lkcl at lkcl.net Sun Dec 29 17:44:23 2019 From: lkcl at lkcl.net (Luke Kenneth Casson Leighton) Date: Mon, 30 Dec 2019 01:44:23 +0800 Subject: [Arm-netbook] Once upon a time you suggested building an opensource router In-Reply-To: References: <20191228213437.72de778d@Phenom-II-x6.niklas.com> Message-ID: On Sunday, December 29, 2019, Jakub Kákona wrote: > This should be related to actual discussion: > https://www.turris.cz/en/mox/overview/ i really like that design. they clearly thought about that, a lot. l. -- --- crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68 From lkcl at lkcl.net Tue Dec 31 08:42:10 2019 From: lkcl at lkcl.net (Luke Kenneth Casson Leighton) Date: Tue, 31 Dec 2019 16:42:10 +0800 Subject: [Arm-netbook] [OT only slightly] STEAM Camp OpenSourceEcology 22jan2020 for 9 days Message-ID: apologies for the cross-list post, the topic is very likely to be of direct or indirect interest, please do share elsewhere, however for questions please do reach out directly to Marcin from OSE rather than crosspost replies! the OpenSourceEcology Project is running a hands-on real-time collaboration and training course for 9 days that happens to have one of its camps in Ghent, Belgium, only around an hour by train away from FOSDEM and running up to overlap with FOSDEM only on its last 2 days, Feb 1st and 2nd, 2020. https://www.opensourceecology.org/steam-camp-january-2020/ Marcin's Ted Talk is here: https://www.ted.com/talks/marcin_jakubowski_open_sourced_blueprints_for_civilization?language=en it is the beginning of a practical real-world bootstrap process for a modern civilisation and actually uses and puts into use everything that we love about FOSS and FOSH. participants will get to actually *make* a 3D Printer, a CNC milling machine, then use those to fabricate an Arduino PCB which... you get the idea. a battery operated welder will also be one of the hands-on projects. Alasdair, an idea just occurred to me: can we discuss privately, they are doing "remote" collaboration: is there room / time for OSE to book a workshop by remote link, for two reasons: one in case FOSDEM attendees are interested to see what is going on, but secondly, if anyone who attends STEAM would like to keep in touch whilst attending FOSDEM? Can we discuss off-list(s)? best, l. -- --- crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68 From lkcl at lkcl.net Tue Dec 31 13:12:02 2019 From: lkcl at lkcl.net (Luke Kenneth Casson Leighton) Date: Tue, 31 Dec 2019 21:12:02 +0800 Subject: [Arm-netbook] [OT only slightly] STEAM Camp OpenSourceEcology 22jan2020 for 9 days In-Reply-To: References: <5987455.PU5xYybd5s@dizzy-7> Message-ID: guuuys i _did_ say don't crosspost. did i say don't crosspost? i did say, "don't crosspost". armnetbooks is ok because it's such a weirdly eclectic and topic-tolerant list. fosdem mailing list for visitors? mmmm... no. l. On Tuesday, December 31, 2019, pavel at noa-labs.com wrote: > Are 1um masks that expensive (compared to the rest of setup)? Or is it has > more to do with turnaround times? > > Can you share details of a light source and resist? One important thing > will be to pick light source that will not burn the LCD matrix, and will > not be absorbed by the glass much. Making a custom made LCD with special > materials might not be that cheaper that sticking with masks for the > beginning. > > It will be truly great if one can just replace the mask with LCD with no > extra work > > Sent from my Huawei Mobile > > > -------- Original Message -------- > Subject: Re: [Libre-silicon-devel] [OT only slightly] STEAM Camp > OpenSourceEcology 22jan2020 for 9 days > From: David Lanzendörfer > To: Libre-RISCV General Development ,Linux on small ARM machines , > Libresilicon-developers at list.libresilicon.com,fosdem at lists.fosdem.org, > libresilicon-developers at list.libresilicon.com > CC: Alasdair G Kergon ,Marcin Jakubowski > > > Hi > My idea was to use an LCD plus a lens system in order to build a mask less > stepper unit for manufacturing semiconductors. > This way the costs for the mask sets would fall away and you would suddenly > be able to provide something like a pooling service for semiconductors. > > What do you think? > > Something worth combining your brain power over in those said 9 days? > > Cheers > -lev > > On Tuesday, 31 December 2019 4:42:10 PM HKT Luke Kenneth Casson Leighton > wrote: > > apologies for the cross-list post, the topic is very likely to be of > direct > > or indirect interest, please do share elsewhere, however for questions > > please do reach out directly to Marcin from OSE rather than crosspost > > replies! > > > > the OpenSourceEcology Project is running a hands-on real-time > collaboration > > and training course for 9 days that happens to have one of its camps in > > Ghent, Belgium, only around an hour by train away from FOSDEM and running > > up to overlap with FOSDEM only on its last 2 days, Feb 1st and 2nd, 2020. > > > > https://www.opensourceecology.org/steam-camp-january-2020/ > > > > Marcin's Ted Talk is here: > > https://www.ted.com/talks/marcin_jakubowski_open_ > sourced_blueprints_for_civi > > lization?language=en > > > > it is the beginning of a practical real-world bootstrap process for a > > modern civilisation and actually uses and puts into use everything that > we > > love about FOSS and FOSH. > > > > participants will get to actually *make* a 3D Printer, a CNC milling > > machine, then use those to fabricate an Arduino PCB which... you get the > > idea. a battery operated welder will also be one of the hands-on > projects. > > > > Alasdair, an idea just occurred to me: can we discuss privately, they are > > doing "remote" collaboration: is there room / time for OSE to book a > > workshop by remote link, for two reasons: one in case FOSDEM attendees > are > > interested to see what is going on, but secondly, if anyone who attends > > STEAM would like to keep in touch whilst attending FOSDEM? Can we discuss > > off-list(s)? > > > > best, > > > > l. > _______________________________________________ > Libresilicon-developers mailing list > Libresilicon-developers at list.libresilicon.com > https://list.libresilicon.com/mailman/listinfo/libresilicon-developers > > -- --- crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68 From doark at mail.com Tue Dec 31 19:04:47 2019 From: doark at mail.com (David Niklas) Date: Tue, 31 Dec 2019 14:04:47 -0500 Subject: [Arm-netbook] Libre RISC-V -- I mean OpenPower M-Class GPU update Message-ID: <20191231140447.243978ef@Phenom-II-x6.niklas.com> Not that I haven't complained about RISC-V's seeming, now proven, lack of openness from the beginning, but did you ever find out why they restricted access to people with UC Berkeley associations? It can't be: the license, newbie complains, IP theft concerns, etc., I give up. I would have asked them point blank, myself and put it in the update. It's counter productive to their stated goals. Thanks, David From lkcl at lkcl.net Tue Dec 31 19:10:58 2019 From: lkcl at lkcl.net (Luke Kenneth Casson Leighton) Date: Wed, 1 Jan 2020 03:10:58 +0800 Subject: [Arm-netbook] Libre RISC-V -- I mean OpenPower M-Class GPU update In-Reply-To: <20191231140447.243978ef@Phenom-II-x6.niklas.com> References: <20191231140447.243978ef@Phenom-II-x6.niklas.com> Message-ID: On Wednesday, January 1, 2020, David Niklas wrote: > Not that I haven't complained about RISC-V's seeming, now proven, lack of > openness from the beginning, but did you ever find out why they > restricted access to people with UC Berkeley associations? it wasn't "deliberate", it's just pure academic arrogance. "we're better than everyone else, we have to filter out bullshit, we know and trust these people, we haven't time to "vet" other sources of knowledge and expertise, and if they don't sign *our* agreement then regardless of what Trademark Law actually says they can go f### themselves because we know better" > > It can't be: the license, newbie complains, IP theft concerns, etc., I > give up. I would have asked them point blank, myself and put it in the > update. It's counter productive to their stated goals. several people independebtly know that. l. -- --- crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68 From doark at mail.com Tue Dec 31 21:30:27 2019 From: doark at mail.com (David Niklas) Date: Tue, 31 Dec 2019 16:30:27 -0500 Subject: [Arm-netbook] Libre RISC-V -- I mean OpenPower M-Class GPU update In-Reply-To: References: <20191231140447.243978ef@Phenom-II-x6.niklas.com> Message-ID: <20191231163027.2304f5d0@Phenom-II-x6.niklas.com> On Wed, 1 Jan 2020 03:10:58 +0800 Luke Kenneth Casson Leighton wrote: > On Wednesday, January 1, 2020, David Niklas wrote: > > > Not that I haven't complained about RISC-V's seeming, now proven, > > lack of openness from the beginning, but did you ever find out why > > they restricted access to people with UC Berkeley associations? > > > it wasn't "deliberate", it's just pure academic arrogance. "we're better > than everyone else, we have to filter out bullshit, we know and trust > these people, we haven't time to "vet" other sources of knowledge and > expertise, and if they don't sign *our* agreement then regardless of > what Trademark Law actually says they can go f### themselves because we > know better" Maybe I have not been clear, how does signing an agreement or joining UC "vet" people? I always thought that my code/schematics vetted me as a fool or a wise man. Thanks, David From lkcl at lkcl.net Tue Dec 31 23:02:51 2019 From: lkcl at lkcl.net (Luke Kenneth Casson Leighton) Date: Tue, 31 Dec 2019 23:02:51 +0000 Subject: [Arm-netbook] Libre RISC-V -- I mean OpenPower M-Class GPU update In-Reply-To: <20191231163027.2304f5d0@Phenom-II-x6.niklas.com> References: <20191231140447.243978ef@Phenom-II-x6.niklas.com> <20191231163027.2304f5d0@Phenom-II-x6.niklas.com> Message-ID: On 12/31/19, David Niklas wrote: > Maybe I have not been clear, how does signing an agreement or joining UC > "vet" people? > I always thought that my code/schematics vetted me as a fool or a wise > man. it's not that simple when it comes to collaboration where it really matters if incompatibilty is hosed. trademark law - actually Certification Mark Law - is the "safety barrier" behind which interoperability actually works. the problem is: it is *essential* that the Trademark holder be FRAND - Fair, Reasonable and Non-Discriminary, if they wish to actually continue to hold that Trademark. The RISC-V Foundation have been BLATANTLY unfair, unreasonable and discriminatory, for a protacted, consistent and persistent time-period, spanning SEVERAL YEARS. l. From lkcl at lkcl.net Tue Dec 31 23:28:29 2019 From: lkcl at lkcl.net (Luke Kenneth Casson Leighton) Date: Tue, 31 Dec 2019 23:28:29 +0000 Subject: [Arm-netbook] [Libre-silicon-devel] [OT only slightly] STEAM Camp OpenSourceEcology 22jan2020 for 9 days In-Reply-To: References: <5987455.PU5xYybd5s@dizzy-7> Message-ID: (removing lots of unnecessary ccs, particularly the FOSDEM visitor list) On 12/31/19, Marcin Jakubowski wrote: > Luke, any thoughts on you stepping up and coordinating the maskless stepper > module of this crazy evolution? as we (privately) discussed yesterday, there are 4 levels of NDAs: HDL tools, VLSI tools, Cell Libraries and Foudries. i'm waaay overloaded by dealing with level 1 heavily and interacting with people at level 2 and 3. david and the libre-silicon team are dealing with level *4* heavily, level 3 in a big way, and interacting with people at levels 2 and 1. they're the best people for this. l.