amarioguy (Arminder Singh)

Once the Windows kernel is running and I hit the first bugcheck - then Windows driver bringup can start (and I'll also need to start work on that UEFI module that I keep deferring...)

We're still not *quite* at the Windows kernel yet, but we're really close, right now what's happening is that Windows is jumping into a bunch of 0's for some reason (this has happened in the past, my assumption is that I need to trap another MSR as last time not trapping MSRs was what caused this)

A computer monitor with a Mac Studio attached to it booting a custom debugging UEFI build, which is currently running the Windows OS loader and loading boot.wim into the ramdisk it initialized prior.

figured I'd give a brief update on where I'm at with AppleWOA stuff, most updates have been localized to the Discord/Telegram but I'll give a briefer here so everyone outside those bubbles are on the same page

I've finally brought PCIe up, and Windows is now able to at least start bootstrap off USB

Apple silicon windows wise, been working on PCIe still, just realized I needed to bring up the PCIe root ports and links myself, and set up the MSIs (message signaled interrupts) so that will be smth I need to do this week

Hey, to everyone who follows me on here, sorry I haven’t been too active here lately! I’ve had some pretty exciting IRL things to do and not much time for social media as of late, but I’m still here don’t fret lol, just been busy

#windowsbeta #windowsinsider

hashtags since I want to make sure this question has some outreach, don't normally do this
anyone who officially beta tested windows vista and 7 are also very welcome to respond with feedback experiences during these dev cycles!

hey had a question for anyone who did beta testing back in the 2000s for whistler to windows 7 officially (as in you were registered with MS during any of those dev cycles), does anyone remember how they used to submit feedback, and more importantly how well it was responded to?

#Windows #windev

Casual windows error screen on a Mac Studio with only the best framebuffer @xenopanther.bsky.social

hello 👋

Personal thought: feel like this is the decade where many people are going to realize the value of IRL bonds and relationships. Some of the relationships I’ve made online have absolutely been fantastic ofc, just that I also feel that it should never *replace* IRL connections

@zac.onl hi zac :)

pretty fine, continuing to reconsider my presence on the bird site every passing second

i'm just thoroughly torn on this technology lol, would like to see if anyone else shares similar doubts to me, if anyone disagrees or if anyone is even interested in even discussing the future of confidential computing lol

End of long thread lol (11/11)

it seems to only really be something that benefits special applications? I'm not sure at this specific point in time what would actually benefit an end user by embracing confidential computing? (genuine doubt here, i'm open to many perspectives in this regard) (10/11)

(even just to search for errata that the manufacturers themselves may accidentally let slip)

Next, other than hyper-specific cloud applications, what benefit will confidential by default *actually* provide to end users? (9/x)

(which is a reasonable assumption but there have been many cases where not trusting that some computing thing works as it should has been a very wise call) which isn't promising with regards to independent auditability of these chips (8/x)

And right now based on the current language and tone they're using, it feels an awful lot like the silicon/cloud vendors are asking us to just trust that the chips do as they say they do (7/x)

(not in production as obviously the data needs to be protected in production at all costs, that's the point of confidential, but in a research environment). (6/x)

We don't want to lose the security advantages confidential offers but given how impactful CPU bugs can be to consumers and enterprises who essentially trust these chips to do what they say, it's important to keep them auditable to some degree (5/x)

Static key for debugging enabled VMs? Specially fused chips only? Will there be a way to independently audit these black boxes at all or will this remain strictly within the purview of silicon vendors? (4/x)

how will researchers, particularly those who research hardware/CPU/cryptographic bugs or those who make debuggers, be able to independently audit confidential VMs as they're running? Will that be an NDA only capability? (3/x)

My skepticism about confidential computing comes from a few ideas: One, given the idea that if VMs and their CPU state by default in the future will become encrypted black boxes (in the world where confidential becomes mandatory and not just an option) (2/x)

It's going to be an "interesting" few years for the computing industry as it reckons on how it'll deal with the whole confidential computing trend and the industry's hope it becomes the new norm. On my personal thoughts, I'm generally ok with the idea, but I have some concerns and skepticism (1/x)

Definitely a bit of an interesting feeling when you’re working on a power management subsystem (PSCI) for a platform that was never meant to use it, and your target platform just reboots without any warning when you try to test the code haha

hey so i'm on bluesky now, that's fun