I see, thanks for clarifying that further. As far as I've got with the quest for free hardware is ARM Cotex A7 (not affected by spectre) on the allwinner boards with the Mali GPU, which seems like an OK solution if 3d acceleration won't be used or the mali free driver project works out.
I'll take a look a the T2 and see if that's viable for me on the desktop, thanks for the info. ‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐ On March 10, 2018 8:46 AM, taii...@gmx.com taii...@gmx.com wrote: > I still have not received an answer from the FSF about if purism will be > > allowed to fraudulently market their products at libreplanet, they avoid > > the question as if I never asked it whilst answering my other questions. > > What a shame - blobbed firmware is not "open source firmware" and the > > "LibreM" is NOT libre hardware. > > On 03/10/2018 02:51 AM, Chillfan wrote: > > > Since Purism have created a fully libre distribution > > You listen to their marketing too much :'\[ > > They haven't created anything it is literally just a debian clone with > > minor modifications to make it seem unique to the layman. > > I like the edited roadmap better. > > https://goblinrefuge.com/mediagoblin/u/onpon4/m/what-purism-s-road-to-fsf-ryf-endorsement-chart-should-look-like/ > > > it seems like that would be the reason for having a seat there. Not that > > I'm saying you're wrong here - it's important that hardware with binary > > only blobs are not endorsed by anyone as free hardware. > > > > It's true that they disable ME in their laptops according to all accounts > > anyway. > > No it isn't - the ME kernel and mask ROM still runs they simply take the > > NSA/intel at its word when they say the HAP bit shuts off ME post kernel > > boot. > > Remove the ME ROM from purisms blobbed coreboot and the system will shut > > off after 30 minutes, physically disconnect the ME processor and the > > system won't work at all thus not disabled. > > They have plenty of shills that would tell you otherwise, which is why > > you should talk to a real hardware engineer about this and similar > > matters instead. > > > Afaik it's not fully possible to remove intel ME yet, just disable it (or > > so we think) and tinker with the partitions. By definition that > > disqualifies them as producing free or libre hardware or even being certain > > ME isn't still doing something. > > Removed would be physically disconnect whereas disabled would be simply > > not running like the pre-core series thinkpads where the ME ROM is not > > included but the system is still allowed to boot and stay on past 30 > > minutes. > > I doubt intel will allow ME cleaner's nerfing mechanism to work for > > long, I seriously can't believe purism thinks they can make a successful > > company on something that could change instantly. > > > Also I haven't heard of Novena/Talos until now despite being on the lookout > > so you may have a point there. Afaict it's only crowd funding or pre order > > so far, what's the TLDR? Any others around? > > Both are real shipping products for once. > > The TALOS 2 board/cpu costs less than a non-free intel xeon system with > > comparable performance and features - the T2 has libre firmware for both > > the board and bmc plus zero hardware code signing enforcement including > > cpu microcode. > > The TALOS 2 was on display at FOSDEM, go look up the pics :D the massive > > heatsinks are for the high end 24 core CPU (4 SMT threads per core) > > POWER9 has the general stuff power users have come to expect from a > > modern workstation such as POWER-KVM, IOMMU with IOMMU-GFX, CAPI, PCI-e > > 4.0 and IBM OpenBMC. > > If you are familiar with the facebook OpenBMC that was recently ported > > to the KGPE-D16 and KCMA-D8 boards IBM's is different in that it is more > > advanced with more features. > > Unfortunately for some reason the tech media only issues glorified press > > releases these days they don't do real research anymore - I have a few > > systems running raptor's libre firmware and even I had no idea it > > existed until recently. _______________________________________________ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
