I loved Utrecht so much last year, Mara! It felt like the most humane city I've ever been in.
I won't be able to make it to the conference this year. But I hope I do make it next year and get the chance to thank you once again in person for the work you put in it :)
To all my Rust frens: I will sadly not be able to attend RustWeek this year 🥲
On the other hand, it is for a positive reason! I got hired for a full time job.
Not doing Rust (as far as I'm aware), but it is safety critical related 😁 so maybe Rust is in its future.
Hoping for RW 2027 🌱
Have fun :)
4 panel comic. Panel 1: Naolin is writing something on a sign. Panel 2: Naolin looks up to the viewer and stop writing. Panel 3: Naolin puts the marker down and turns the sign around. Panel 4: The panel says: "Respect my transhomies or I'm gonna identify as a fucking problem."
Hey. :)
Trans rights.
Holy shit waffle, that's awesome!
"irreversible side effects of HRT" all of life is irreversible. i cannot go back a single second in time
(b) & (c): I don't even care, man. Keep yourself safe first.
Also if we view writing as an extension of one's brain, then it's pretty clear that what you're doing as you write these docs, is basically just thinking clearly about things.
Not sharing 100% of what's in your mind is super reasonable.
Lmao, fair enough. I guess we all suck at predicting the future x3
(I also got the rust taking off right! Dunno, my gut was like "this is the kind of thing we've been needing since the 70s" x'D)
(Not rhetorical question)
Would you say that Windows still has a future? The quality of the OS has been plummeting in the past couple of years, faster than I've ever seen MS manage to do.
Is that what the x86 VMs take to run on M-macs? Damn
(Sort of related: reminded me of when I dared to add some docs to the Swift compiler. My old computer took about 14 hours to build the compiler. In the end I made a bot that sent me updates about the build's state, to avoid going crazy 😂)
Damn x'D
I wonder if a more generic version of this scheme could work.
Like a trait for MIR opts, where the default impl just runs the opt and returns true if it changed the input.
More complex opts could override the default impl and return true iff all of their steps changed the input.
And then we check (for each opt) whether that opt did anything:
- For opt O(•) in <same sequence of MIR opts>
-- // We take the vector corresponding to this opt
-- opt_changes = made_changes[:, O(•)]
-- assert!(opt_changes != [0....0])
I've butchered the syntax for things, but I hope it makes sense
- For MIR program p in <list of programs>
-- p' = p
-- For MIR opt O(•) in <sequence of MIR opts the compiler always runs>
--- p'' = O(p')
--- // This is a map that stores
--- // whether an opt did something
--- // or not to program p
--- made_changes[p, O(•)] = p'' != p'
The test I'm thinking of would be something like the following, lmk if it makes sense.
Get well soon ❤️🩹
Mm, if it triggers in in lower-opt setups, then perhaps it's good to leave it. Maybe it's there to reduce compilation times?
(Now that I think about it... we could have some kind of test to see that the MIR opt passes don't do nothing, right?)
If you haven't already, you should take a look at Daniel Bernstein's blog: blog.cr.yp.to
And if you care about having proper (ie not NSA bullshitted) post-quantum cryptography protocols, you might want to take a look at how the IETF has been failing at giving us that: blog.cr.yp.to/20260405-vot...
Millions of American voters wanted to be villains. They came to believe it’s the key to success, while goodness is weak, woke, loser stuff, done only to signal.
Offered the option of unvarnished villainy, they enthusiastically took it. Seeing others suffer more made them feel big.
And here we are.
Nooooooo :'(
I think this is correct. I'm assuming here that Swift still behaves as it did 9 years ago (the last time I played with Swift was back then), but from what I recall from convos with its maintainers, it was really easy to trigger badly behaved paths in the compiler.
Now, if you're seeing this several times a day... maybe something's fucked 😆
OTOH your recent projects have been quite avant garde, so I wouldn't be surprised if they were consistently making the compiler struggle 😁
I don't know, actually. Is there such a thing as a language design that can let a program always be type-checked in PTIME?
I mean, apart from languages with trivial, useless type systems... no, I don't think there is such a thing. I would be very surprised if there were.
Can Elon Musk buy Wikipedia?
Wikipedia Editor @molly.wiki explains why it's impossible.
@wikipedia.org is run by a nonprofit with no owner to buy it from, unlike Twitter. Its content is written by thousands of editors whose opinions all carry equal weight.
Full ep, out now.
"i have two Microsoft Outlooks, and neither one of those are working."
The Astronauts truly representing us all up there.
If there is anything hormonal with soy, via phytoestrogens en.wikipedia.org/wiki/Phytoes..., then oat would most certainly have nothing to do with that :3
And if oat happened to affect the hormonal profile significantly, it would be on a completely different manner to soy.
TLDR: yeah
I really don't know what to say, other than: I'm genuinely thankful to have found you here in the internet. I dunno, we have barely talked, yet my gut knows to tell me when I've found a good person.
You're not just cool; you're also very kind. I want your light to shine and keep shining.
🫶🏼
Maybe internalized blame / guilt or something? Very common in millennials like you and me.
Gambatte, the struggle is real
It's fine, fam. Stay safe :)
Well, what a better time than now to show you this :p
m.youtube.com/@cs618progra...
For opts that depend on liveness and opts based on values known at compile time (like constprop and value range analysis), this is one of the best online resources I found when I was learning about it :)
