ThePhD

🎇 Somehow Is A C Expert ✨ Smooches Unicode-kun 💕 The C Standard Cannot Be Replaced And Will Never Be Destroyed. 🧟‍♀️ Pixel Art is Cool ✨

Public Key

npub19pz4w2q5uvs9jz8aydmgk9pgtgpmdy6380aekuy8gdvjfvztcx0qnu862l

NIP-05 Address

thephd_at_pony.social@mostr.pub

Profile Code

nprofile1qqszs32h9q2wxgzepr7jxa5tzs595qakjdgnh7umwzr5xkfykp9ur8scas0ny

Author Public Key

npub19pz4w2q5uvs9jz8aydmgk9pgtgpmdy6380aekuy8gdvjfvztcx0qnu862l

Show more details

Last Notes

2025-04-17 15:36:25

An important separator in what makes a good array abstraction from an awful one is whether your compiler understands it at a native level or not.
Right now, C and C++ don't really cut the mustard there. I wonder if I can improve that situation for C, at the very least...

2024-08-15 03:25:00

- reply

@npub1aku…xjax The GNU ecosystem if it was good

2024-08-06 00:14:36

A good W since switching to an M1 bouncing between Asahi Linux Fedora and the typical MacOS Sonoma;
I forgot my charger today, but I was still able to: listen to video, compile code, and browse the web/respond to e-mails for a good 6-7 hours before it hit 18% charge!
#asahilinux #asahi #macos #fedora

2024-07-19 17:10:07

Finally, a global computer problem that's not rooted in C for a change.

2024-07-18 05:56:55

C23 is, as far as I'm able and aware, completely done.
Congratulations to WG14. Let's get cooking for C2y (C29, maybe/probably.)
https://cdn.pony.social/media_attachments/files/112/805/950/861/993/804/original/b418b1c3a032cadd.png

2024-07-16 14:06:49

The C standard has been restored and is back on for being published as C23. My bio continues to remain true.
https://www.iso.org/standard/82075.html
But for a brief moment, we were all free and chaos reigned....

2024-07-15 16:00:02

All behavior is now undefined.
The back and forth on the edits for C23 has resulted in the publication deadline passing before getting a copy of C23 ISO-approved. Thus, ISO have automatically cancelled ISO/IEC 9899:2023.
The Nasal Demons are coming. Your code has nowhere to run. Your code cannot hide.
And neither can you.
https://www.iso.org/standard/82075.html

2024-07-11 15:53:57

... Oh. Uh. How badly would everyone hate me if I wrote a paper that allowed
T foo[2];
and
T[2] foo;
to mean the same thing as far as type declarations go in C?

2024-07-08 02:50:58

Hot C take:
no string library anyone writes is going to be good enough for general-purpose use without a real slice type.

2024-06-18 14:56:55

Yeah, finally got a ""working"" solution and my conclusion is VLAs/VMTs are not ready for general purpose deployment across any industry. Specialized use is fine but this stuff is bloody awful and requires several language improvements before use. You can make the situation better with macros but otherwise these are just not the things we should be building any library on top of; might be decent for specific, targeted, data-only-flows-inwards use, though.
https://godbolt.org/z/x5f5fPfjK

2024-05-22 12:59:13

It's time.
RAII in C, and why nobody's getting it right with the increasingly "simple" juggling that keeps getting tossed to me like table scraps for a dog.
The Pasture | Why Not Just Do Simple C++ RAII in C? | https://thephd.dev/just-put-raii-in-c-bro-please-bro-just-one-more-destructor-bro-cmon-im-good-for-it

2024-05-20 16:16:33

Not lost on me that for some of these proposals and ideas I keep seeing pop up for C, everyone's just kind of trying to figure out the magic set of syntax they can write that will let them escape the consequences of the system.
Versus C++, where they just Add Something To Do It.

2024-05-12 15:31:39

One day, things'll calm down and people will be normal.
One day.
https://cdn.pony.social/media_attachments/files/112/428/831/303/176/072/original/5c2395e58f570f28.png

2024-04-23 17:05:33

I should post positive C content.

2024-03-25 19:16:52

PSA since it hit my inbox:
You do not need to put your "legal name" (????) on a Standards Paper. You can submit a paper from Little Blorbo and make your company name "self". Whether or not you'll want to live down being Little Blorbo for the rest of C's history is on you.

2024-03-12 16:40:22

Remember, everyone:
PREFIX YOUR MACROS
PREFIX YOUR MACROS
PREFIX YOUR MACROS
PREFIX YOUR MACROS
PREFIX.
YOUR.
MACROS!!!!!!
https://gitlab.freedesktop.org/mesa/mesa/-/issues/10803

2024-02-14 18:50:01

Didn't think "writing over the edge of an array is bad and we should not do that if we can help it without relying on Undefined Behavior Sanitizer or Address Sanitizer" was going to be something I had to write about in 2024 but...
... here we are??

2024-02-07 19:34:59

Y'all. This coffee machine broke down in this place and from the back entryway the most butch lass decked in heavy tool-laden belts, hand-repaired headphones, a weighty backpack and the most relaxed work fit just walked in. And.... and.
Lesbians, I get it.

2024-01-19 19:49:33

I should make TikTok series where I post one video every two weeks or something doing a brief 2 minute braindump of a C feature.

2024-01-14 17:45:36

Been trying to formulate an e-mail about this that I want to send back to somebody. I'm going to delete this paragraph from the e-mail but it's something that bothers me so much about working on C:
”… the community never grasps proposals or anything related to it, or reads the paper trail, so suddenly they're shocked or surprised when things go into C or change C. And then they advocate at the literal last minute (or post-last-minute). Even now, most of the feedback I get comes in later and later, and often it's sent as private e-mails that I can't publish or point to in order to say "hey, look, here's this thing we've done that has had genuinely good impact". Sometimes, the folks in these groups never communicate at all; if I don't literally dive into their Discord/Slack/mailing list/personally reach out, it never gets to me. Or anyone, for that matter; it's just a bottle of sentiments that never go anywhere. I cannot begin to explain to you the number of times I've heard feedback from e.g. [REDACTED] Employees or people in the [REDACTED] Division or some old, ancient Really Smart Assembly Guy at [REDACTED] happy with the work some of us are doing, but as like a small e-mail or a passing thought in a DM. And believe me, I appreciate that e-mail/DM/message a whole hell of a lot and it gives me motivation to keep fighting! But it also puts a lot of pressure on me to keep being the "Emissary" on behalf of this sea of people that are reasonably powerful and could change the flow of the tide with a single public post.... anywhere. But they don't. So I have to keep shouldering a ton of this work on my back, and every time it just looks like I'm some stupid upstart ruining C and not carrying the wishes of so many people on my back; and it's STUPIDLY frustrating.“

2024-01-08 05:56:13

Some folks are sleeping on this part of C, but this is now well-defined and working code in the latest C23 thanks to @npub1atm…30ak , @npub1xxl…yqct , and @npub1085…yrya:
https://godbolt.org/z/Gd4j6nnv3
https://cdn.pony.social/media_attachments/files/111/718/771/363/092/879/original/89d79dd54d5f3c6e.jpg

2024-01-06 18:19:32

Once again, hats off to C people because programming solely inside of a Macro is a fucking awful experience. I have no idea how you all keep doing this and making software that doesn't continually and perpetually snap in half.

2023-12-31 15:16:48

If I have someone try to explain to me that integer promotion or lvalue conversion are simple one more time, I think I'm gonna just commit Eternal Sleep.

2023-12-14 17:49:02

"Why are you working on C?"
I wrote this in an e-mail to people, but I figured I'd copy part of the response here since people keep asking.
…
     We have people writing critical software. They are not migrating to new software anytime soon (modulo regulation-based incentives). But they have serious problems. Everything from vulnerabilities that are used by nation-state actors to quell dissidents, to not being able to change a typedef like intmax_t because the functions tied to it are baked into specific named symbols in an invisible way (ABI), to constantly seeing people's names getting butchered by Airlines, Databases, and Governments because they're using software that relies on the C locale and mangles names.
     These are C problems. Not C++ problems. Not Java problems. Not Rust problems.
     C problems.
     My job is to solve C problems. That's the motivation. That's the coherent plan. When we stop having long-term, 20-to-40+ year problems, with 30+-year implemented existing practice that we never standardize despite it solving a wide variety of problems, that's when I'll stop writing C proposals.

2023-12-09 13:28:05

Whoever wrote this damn document: https://www.regulations.gov/comment/ONCD-2023-0002-0020
You sound exactly like the C people when C++ came along, and it's actually embarrassing.

2023-11-10 17:05:10

https://www.open-std.org/jtc1/sc22/wg14/www/docs/n3176.pdf
...
(* deep breath *)
I need to write not only a paper, but a full-blown series on the history of C. The historical revisionism and prescription of some magical properties to C89 or whatever without mentioning the vast landscape of extensions is.
Ugh.

2023-11-10 16:08:47

I don't think it'll ever sit well with me that we straight up argued "typeof will encourage low-level code that forces people to make mistakes". Like I didn't even respond well because it was such a baffling fucking assertion. I also couldn't even get my brain to realize that people were taking "typeof is hard to teach" as, like, a serious argument. I don't think I've ever been this discombobulated by an actual veteran C++ person just trying to do their damnedest to shit on C.
Shell-shocked since that damn conversation.

2023-11-08 22:01:35

It would help me a lot if I was a lot less fucking stupid.

2023-11-03 18:23:19

Got Stun Locked Again, so it's time to dump a rant I made elsewhere:
C is not a language for direct control of the hardware. C is a langauge that is coded to the semantics of an Abstract Machine specified in a document. It's no more capable of hardware control that e.g. Rust or Zig. What it has is a wide variety of pre-existing implementations that allow you to touch that hardware, but most of that control was programmed in an assembly language or worked into the hardware/firmware by somebody.
C is not more suitable for hardware and your computer is not a PDP-11. (This is part of why "Nobody writes ISO C" is a thing. C, the language K&R made, the language that got standardized, at any point in its lifetime, was never good enough for a kernel. It just let people coordinate stuff in the cheapest way possible and accepted extensions.)
The part that's extra nutso is that the last 20 years were people who held this exact belief -- that C was just a thin layer over the hardware -- get bodied, over and over again. Compiler vendors gave them the big middle finger every time they said "wait, no, UB is for hardware!", and compiler vendors traded in UB for (sometimes negligible) speed ups.
To still believe C is "for the hardware" in today's day and age where GCC will literally run your for loop for eternity because you tried to access an array out-of-bounds and it found out about it, or Clang will solve fermat's last theorem due to a loop, is magic shroom thinking. You can access the hardware just as good with Java by using the Pointer class, you can hit the same register that the shitty ISA Manual fucking lied about by writing the same integer address into the Pointer class and dumping out a 2-byte integer to the right place.
C is not magic and you're not improving its design by insisting it is, for the love of God start Evaluating Your Tools Properly. Yer an engineer, not a fucking wizard, Harry.

2023-11-02 17:47:14

The FSF should no longer exist.

2023-11-01 16:16:51

Taking this Secure Coding in C++ class is making me so sad.
So much of our insecurity could be handled so much better if we just..... stopped writing functions that were full of bullshit by-default. :(

2023-10-25 16:44:37

- reply

HEY @npub1085…yrya GUESS WHAT YOU'RE AWESOME AND THIS IS MORE THAN JUST A LITTLE JOKE THING AND IT'S MADE ME SUPER HAPPY AND I HOPE YOU HAVE SEVERAL LONG DECADES OF GREAT FORTUNE AND GOOD WILL AND EVERYTHING WORKS OUT BECAUSE THIS HAS PROBABLY MADE MY WHOLE DAMN WEEK AFTER THE CRAP (SOME OF IT LITERAL) I HAVE BEEN THROUGH!!!!!!!!!!

2023-10-25 16:43:26

- reply

AND THIS PERSON JUST SAID IT WAS A LITTLE JOKE THING THEY WERE SENDING ARE YOU KIDDING ME THIS IS FAR FROM A LITTLE JOKE THIS THIS IS AWESOME IT IS PRECIOUS THIS LITTLE DUDE WILL WATCH OVER THE C STANDARD FOR THE REST OF TIME AND ENSURE ALL BRING THE BEST AND BRIGHTEST TO IT!!!

2023-10-25 16:41:30

- reply

The book, THE LITTLE BOOK WITH THE K&R ORIGINALS on it!! This creature is TRULYREADY TO GO BE THE EDITOR OF THE C PROGRAMMING LANGUAGE!!!
https://cdn.pony.social/media_attachments/files/111/296/647/521/774/540/original/4c7e20be151a9407.jpg

2023-10-25 16:32:55

YA'LL. YAAAAAAA'LLLLL SOMEONE MADE ME A FUCKING GIFT AND I HAVE HAD IT FOR TWO (2) DAYS AND I AM STILL GEEKING OUT ABOUT IT WDHUGFGIQWFhieGgbuiQWFhiWQvbuqwefg!!!!
LOOK AT THEM. LOOK. AT. THEM!!!!!!!!!!!!!
https://cdn.pony.social/media_attachments/files/111/296/563/401/280/656/original/8cff3bef61029652.jpg
https://cdn.pony.social/media_attachments/files/111/296/563/878/924/904/original/903e81ac454342ff.jpg
https://cdn.pony.social/media_attachments/files/111/296/564/303/963/980/original/c7cd13370b87073c.jpg
https://cdn.pony.social/media_attachments/files/111/296/564/849/952/214/original/cdf11462e133bc96.jpg

2023-10-22 05:52:23

- reply

One thing I will always be grateful for #embed, std::embed, and std::optional for showing me was that people absolutely do not measure shit. And people will gleefully talk with absolute confidence about things they've never tested, and assumptions they've never actually tried.

2023-10-22 05:46:14

Me, watching people in the comments debate just using objcopy to "get the data in faster" when this table exists and was already published on my blog:
https://thephd.dev/embed-the-details#results
https://cdn.pony.social/media_attachments/files/111/277/082/046/317/908/original/0de88e057d59a6e9.png

2023-10-18 17:10:35

Hot Take: C++ lambda captures are the best control mechanism for "function with data" that we've seen, and every other solution barely comes close to the level of flexibility and power it gives.

2023-10-07 20:33:06

Let me work on your programming Language.
I am
normal
and can be Trusted with Language design.
https://cdn.pony.social/media_attachments/files/111/195/633/958/506/231/original/0db4dbb592a71df3.png

2023-10-05 17:57:00

It's not lost on me that C and C++ toolchains, loaded with static analyzers, could barely reach ASIL B and just barely scratched ASIL C, but Rust got to ASIL D on its first publicly-released verified toolchain.
Just getting our asses spanked over and over and over, meanwhile we have people writing proposals (derogatory) about safety and security and giving keynotes where they barely understand the material instead of inviting actual experts to give us their industry-hardened opinions.

2023-09-29 00:21:08

- reply

@npub19zq…u4m5 As far as I know, only in a few propretiary compilers. Everyone else is taking their time, because doing the optimized version is a bit harder than what most people expect.

2023-09-29 00:07:27

#embed (naive version) is done and polished and ready to be submitted to clang, but I won't submit it until I finish the polished speedy version, because I can't trust implementers
https://github.com/ThePhD/llvm-project/commit/4f81b166f3b896a711a730d3f727b6e7a829ebf1

2023-09-06 18:52:50

As I have no intention of being publicly smeared but private apologized to, I am writing this post as an immortal (for as long as my instance lasts, anyways) record of my response to an e-mail I have just received from David Tolnay. My response is brief and enclosed just below:
“Dear David,
I have no intention to have a (private) video call to resolve an intensely public and, at this point, explosively volatile situation that never needed to be. I have also pulled out of the work and we will not be engaging with the Rust Specification work or anything related to compile-time reflection, so there's no reason to apologize at this point because the (metaphorical) vase is already broken.
As I stated over 3 months ago when this first kicked off, the only thing (even if it was published anonymously, again, before this ever blew up) was a technical critique of the work presented. If you do not have a technical critique to offer, then I would like to not hear from you or anyone else on the subject material for the next few years at a minimum.
JeanHeyd”
As ever, I am consistent: the damage is done, but several people have spoken up that they would like to continue the work. If you fervently believe this was a fuck up, you will write out why the compile-time direction Shepherd Oasis wasted tons of money supporting me through and collaborated on by a dozen+ different people who are titans in their respective focuses is Bad, Actually so we can put the idea to rest and the Rust Project can go do more productive things.
It would also probably help if that response did not take a peer's work, label it a bungling failure, and then use it as a projection field to say the work would never be ready, too.
That would also be helpful.

2023-09-05 03:00:01

You know what makes this whole thing extra foolish? There's supposed to be a technical rebuttal. Somewhere in that statement, some allusion to what was wrong with the compile-time reflection work. What we got instead?
A bunch of excuses.
I break my back and pick brains and do a crapload of heavy lifting and back it with technical reasons and I pull in experts and everything and they can just tank it by saying... "I have some bad vibes and I wasn't sure some people okayed it."
These places are never about Merit.
All of them were holding water for these apparent enormous technical objections. These well-founded, deeply grounded reasons why these things couldn't work.
Buncha bureaucracy and a whole lot of nothing! Just.
Wouldn't it be nice. If we evaluated.
The #*$%#ing work.

2023-09-04 05:13:17

When someone decides to make up entire bungles of horse shit in their "explanation" and "apology" ( https://archive.ph/Utaz1 ) about the weird shit they did.
I want off this fucking ride.
"Postpone to 2024"? Present at a "tier 2 conference"? My reasoning is contained directly in my blog post ( https://thephd.dev/i-am-no-longer-speaking-at-rustconf-2023#-huh ). If you're going to fucking lie about me at least try and be fucking accurate about it:
“It is also deeply confusing and ultimately insulting for them not to contact me beforehand and simply ask me if I would disclaimer my work to make it clear that they did not explicitly endorse this direction. Multiple times before the RustConf schedule and program was released, I made it obscenely clear that there was not going going to be an RFC for the work I was talking about (“Pre-RFC” is the exact wording I used when talking to individuals involved with Rust Project Leadership), that this might bias folks, and whether or not it would be okay to do this. Individuals in contact with me both inside and outside RustConf leadership made it abundantly clearly that this topic was perfectly fine. Furthermore, they had already met to discuss my work before hand, so at no point should anyone be confused about what my intentions and goals are.”
That someone would try to use game-of-telephone bullshit to make it seem like I deserved to have my keynote outright removed (which was the ACTUAL SUGGESTION given to RustConf, and they changed it to a DOWNGRADE to prevent bigger fallout from Triplett and Tolnay's CLEAR oversteps and, apparently, outright lies here!), or that I was in agreement with such a removal, when I literally took the required time to make a proper assessment of the situation and respond like an ADULT, is some WILD nonsense!
Is this a common Rust Project occurrence? Do people just do this ALL THE TIME? WHAT IS GOING ON OVER THERE?!
Edit: added some fucking alt-text because I'm so fucking annoyed at this bullshit.
https://cdn.pony.social/media_attachments/files/111/005/142/803/015/620/original/96774221e2768faa.webp

2023-09-03 18:23:22

- reply

.... So this is why nobody wanted to mention his name. Why nobody wanted to tell me what the real objections were. This is the Strong Man behind a lot of those kinds of Rust behaviors, huh?
I look forward to a well-written rebuttal for compile-time reflection, then. My blog post included potential examples of automatically-generated, macroless serde code, after all. Maybe that's why he got pissed off enough to try to cut me at my legs. After all,
why address things technically, when you can just dagger somebody in the back under protection from the whole Rust Project who won't say your name?

2023-09-03 18:08:52

https://lobste.rs/s/sgoshm/serde_is_no_longer_shipping_precompiled#c_5cye7q
Case in point: dtolnay was also the one thus far unnamed by anyone speaking for the project person who was involved in ThePHD’s talk being downgraded from a keynote.
.... Excuse me?

2023-09-03 05:13:52

C is a demonic language made by demonic people. why can't i return arrays? nobody knows. why can't i return strings? oh because you can't return arrays. why does the program segfault when i try to use modulo on a double? well you see …
Preach that truth! — https://cohost.org/trainsfemme/post/2705000-my-c-homework
Someone did write a paper to fix this for C++. As usual, we tiptoed around wanting to handle that problem and the person got busy doing other stuff after the C++ people redirected them to C, and God Knows C doesn't want to address this legacy wart!
https://wg21.link/p1997

2023-08-27 04:12:43

"parsing C is simple" these people are not real.

2023-08-24 04:09:51

All the shit I'm getting up to in C, for the next release.
Full papers:
https://thephd.dev/_vendor/future_cxx/papers/C%20-%20More%20Modern%20Bit%20Utilities.html
https://thephd.dev/_vendor/future_cxx/papers/C%20-%20Initialized%20const%20Integer%20Declarations.html
https://thephd.dev/_vendor/future_cxx/papers/C%20-%20Transparent%20Aliases.htmlSkeleton papers (abstract only):
https://thephd.dev/_vendor/future_cxx/papers/C%20-%20Comprehensive%20Statement%20Expressions.html
https://thephd.dev/_vendor/future_cxx/papers/C%20-%20Improved%20__attribute__((cleanup))%20Through%20defer.html
https://thephd.dev/_vendor/future_cxx/papers/C%20-%20Non-owning%20Sized%20Arrays.html
https://thephd.dev/_vendor/future_cxx/papers/C%20-%20Functional%20Functions.html
https://thephd.dev/_vendor/future_cxx/papers/C%20-%20A%20Measured%20Approach%20to%20_Operators.html

2023-08-18 18:19:27

People read a book whose last published edition was 35 years ago, peel through a rationale document that is now just over 20 years old, and then proceed to Very Confidently tell me what C is supposed to be.
I'm having a Normal One right now.

2023-07-13 03:07:57

- reply

@npub19zq…u4m5 👋 Happy to answer most WG14 & related questions!

2023-07-09 20:45:36

It's done. N3149 is going to be the last standard the Committee has direct control over: now we're just dealing with National Bodies and ISO. Here's the report (it's not long), and a link to the (ZipCrypto passworded) zip:
https://www.open-std.org/jtc1/sc22/wg14/www/docs/n3150.htm
https://www.open-std.org/jtc1/sc22/wg14/www/docs/n3149.zip
It's out of my hands now (until ISO asks for a million different edits), but this process will take 4-6 months and then an ISO standard will pop out on the other side.
See you all for the next revision of C!

2023-06-29 17:34:10

(Apologies for the Hellsite Share, but this is actually normal stuff, I promise, no CW needed!)
https://i.imgur.com/ceBouSW.png
So this tweet (https://twitter.com/muk0rk/status/1674354587036184576) got shared, and I actually know a little bit about this topic in particular! It turns out... ⤵
https://cdn.pony.social/media_attachments/files/110/628/674/069/382/374/original/3efee58aa8c6629e.png

2023-06-18 02:40:54

Everyone thinks U.B. is just about Nasal Demons and Compiler Elision, but as @npub1d5n…kqyl knows too and I find out in the thick of it, well. It can be far, far worse...
🎨: KIINGKIISMET ! (https://kiingkiismet.com/)
https://cdn.pony.social/media_attachments/files/110/562/894/388/702/150/original/c7dde4fe91ffef0f.png

2023-05-28 18:30:49

Made the mistake of keeping up with the Bad Place and got introduced to this fucking Banger of a tweet on @npub1uqk…v098's post.
I regret to inform you that despite:
fixing widespread industrial problems, including with the oldest and most popular language in existence;
writing full frameworks people study to get good at code;
authoring and drafting full-fledged fixes to the entire encoding situation for C and C++; and,
editing fucking ISO/IEC 9899:2023/4,
I am apparently a diversity hire, who's only here to be a drain on the Smart Rust People's resources.
Anyways I've blocked the person, because they wouldn't want to be reading stuff from a "diversity hire" like myself, would they?
On god I need to finish my own programming language, even if I'm the ONLY USER I will NEVER have to look these pricks in the eye EVER AGAIN.
https://cdn.pony.social/media_attachments/files/110/447/727/710/352/814/original/7b9138ede2a6ce30.png

2023-05-26 22:01:04

I deserve better than this opaque, institutional bickerings from a mature, open-source programming language.
The Pasture | I Am No Longer Speaking at RustConf 2023 | https://thephd.dev/i-am-no-longer-speaking-at-rustconf-2023

2023-01-23 23:39:26

@npub17xj…0938 @npub12dh…q6qy Yeah, plus I mean. These are the people that brought us seatbelts. This is not the fight any programming language ever wants to get its hands on, and Consumer Reports is just throwing the gloves down.