Hi boin,
nice to see you here! Thanks for joining the discussion.
I do not dislike XML to the degree
others might, but the new HID syntax is great! Impressed with the progress.
Overall, the feedback is much more positive than I expected. It's
reassuring to see that this change is perceived as an improvement, not
as a mere whim of mine. ;-)
Is our software supply chain for this
system secure?". I'm no security specialist, my role is mostly
automation/observability/etc. I do not see Genode fitting into this
(yet) but my knowledge is limited.
The discussion touched topics like trusted/authenticated boot and
assured system integrity already, e.g., the postings of Alexander and
Johannes (J03). I definitely want to see the use cases of Johannes
covered in 2026.
Regarding (software) supply-chain security, I think that Genode's ports
and depot tools provide a solid starting point for a systematic
assessment. E.g., I think a tool for creating the "bill of material" for
a given pkg archive (a list of all 3rd-party software downloads that go
into the pkg) would not be difficult to implement. But as I don't see
that as urgent, I wouldn't put the topic on the road map of this year.
When I hear "printf-like functionality" I'm reminded of the Benno Rice
talk "What Unix Cost Us", where he laments the amount of snprint
required to use the USB Linux APIs/usb_f_fs api. It's not necessarily my
opinion but more like "Benno Rice's opinion".
Hah! Mea culpa. This was just a clumsy attempt to express that HID is
easy to generate. I'm indeed not a friend of format strings myself,
which is the reason we removed them from the Genode framework 10 years ago.
[1]
https://genode.org/documentation/release-notes/16.08#Cultivation_of_the_new_text-output_API
What little I've learned about Sculpt so far, I really like this. I
noticed if I want to change something in Sculpt that is in the codebase,
it's a full recompile of that component. Learning the Sculpt codebase
itself, on top of learning the rest of Genode, looks like a large task.
I agree. Good that your are reinforcing this point.
> What are your plans for 2026?
Part one of my plan is accomplished: get Sculpt on a bare-metal system.
I would like to build and publish a package that I can fetch within a
Sculpt system. Get a full write-up of how to do it, publish to a blog,
then keep working from there.
You are probably aware, but let me drop the link to the "Applications"
book [2], which wants to be your guide along the way.
[2] https://genode.org/documentation/genode-applications-25-04.pdf
I've been doing lots of "Modern C++" (C++23, linux io_uring +
coroutines) and seeing about building from-scratch Wayland bindings for
it. I've thought about putting this peanut butter in the Genode
chocolate, get some form of Wayland compositor proxy that can talk to
nitpicker. I can not estimate how much work that would be, whether I
could even accomplish it.
Interesting! I sense that this undertaking would potentially benefit
from the VFS-GUI line of work I mentioned earlier.
Cheers
Norman
--
Dr.-Ing. Norman Feske
Genode Labs
https://www.genode-labs.com · https://genode.org
Genode Labs GmbH · Amtsgericht Dresden · HRB 28424 · Sitz Dresden
Geschäftsführer: Dr.-Ing. Norman Feske, Christian Helmuth
_______________________________________________
users mailing list -- [email protected]
To unsubscribe send an email to [email protected]
Archived at
https://lists.genode.org/mailman3/hyperkitty/list/[email protected]/message/LVARWKR5FMZTOVQZDLA5OYQKDAH7FVEO/
