I personally don't like this at all. This means that regex engines that try to generate optimized matching code for an expression can end up generating suboptimal code if you don't want alternative order to matter, since the engine needs to keep that invariant, except in the case when it can prove that the alternatives won't overlap, and a later one can be checked in constant time. If both are true, it is legal to reorder them to do the constant time check before the big complicated wildcard-filled alternative.

But personally, I have never written a regex where I actively cared about the alternative evaluation order. I've used some other people made where order is important but never written one myself.

I'd love to be able to tell the engine "feel free to swap the evaluation order of my alternatives while optimizing", but few if any such engines offer that as a feature.

Now I get that PEGs have commutivity problems are that are different from regexes', which make the issue worse, but that doesn't mean regexes do things right either.

Compilers keep taking more and more advantage of inferring that a values in variables cannot be `x`, because if it were than some previous usage would have been UB. When people file bugs to complain, the compiler authors point at the spec which allows them to assume that UB behavior never happens, so the compiler behavior is legal. The only counterargument is if the compiler has chosen to document some specific behavior for this UB (possibly only with specific flags enabled) in which case the compiler testing that scenario as proof of impossibility is indeed a bug (when the required flags are set).

The "custom os" part could also be done easilly enough with the correct approach.

Specifically systemd has a less-known feature known as system extensions intended for basically exactly these sort of scenarios. These system-extensions are basically disk images containing files in /usr and/or /opt that can be dynamically overlaid on the existing filesystem (the intent is that these are purely additive). Systemd also intends that all os provided configuration live in /usr, with /etc existing only for machine specific or admin applied configuration. (And which should enabling overriding anything specified by the package or OS.)

System extensions when used default /opt and /usr to be read-only, but you can enable mutability if you having write routing directories or symlinks in the right spot.

So for userland this whole os profiles things could literally just be a set a of system-extensions, a distinct /etc folder, and distinct set of write redirect directories for each. An initramfs can simply bind mount the /etc directory, and add the correct write redirection symlnks before systemd starts. Rolling back a profile is simply wiping its write redirection and /etc folders. If you also want each to potentially have distinct device trees and/or customized kernels, that would need additional bootloader work on top, but nothing that feels too extreme.

Now in reality, since not everything support systemd style configuration, these OS profiles would probably need to construct an initial /etc by copying files from a base-os template, and then copying in anything included in the system extension images (which can have these as systemd will ignore such folders), but that is straightforward enough.

As far as I can tell, the only reason that the explanations even resemble human speech is that AV and AR start off based on a trained language model. If we instead trained the same model architecture from scratch as AV and AR, they would eventually converge to some round trip format for activations, but it probably would be completely unintelligible and look only like human speech in so far as many of the tokenizer's tokens look like words or word fragments.

This whole process seems to rely on the fact that the text AR's output will still strongly favor output sentences that seem to make sense, rather than contradicting learned facts, etc. So it will favor mapping activations to plausible sounding text in ways where patterns can consistently hold across most of the training data. There absolutely is a risk that it will learn the wrong things for certain activation subpatterns like swapping concepts especially if none of the training data included a set of activation sub patterns that would help distinguish them the right way around.

%%{init: {"theme": "base", "themeVariables": { "pie1": "#FF5733", "pie2": "#33FF57", "pie3": "#3357FF", "pieStrokeColor": "#000000", "pieStrokeWidth": 3, "pieOpacity": 0.8 }}}%%

This looks like it works on this site too.

The problem is that the only safe way for the compiler to generate non-boxed unions would require non-overlapping fields for most value types.

Specifically the CLR has a hard rule that it must know with certainty where all managed pointers are at all times, so that the GC can update them if it moves the referenced object. This means you can only overlap value types if the locations of all managed pointers line up perfectly. So sure, you can safely overlap "unmanaged" structs (those that recursively don't contain any managed pointers), but even for those, you need to know the size of the largest one.

The big problem with the compiler doing any attempt to overlap value types is that if the value types as defined at compile time may not match the definitions at runtime, especially for types defined in another assembly. A new library version can add more fields. This may mean one unmanaged struct has become too big to fit in the field, or that two types that were previously overlap compatible are not anymore.

Making the C# compiler jump though a bunch of hoops to try to determine if overlapping is safe and even then leaving room for an updated library at runtime to crash the whole things means that the compiler will probably never even try. I guess the primitive numeric types could be special cased, as their size is known and will never change.

It is tricky to meaningfully expose a dollar cost equivlent value for subscribers in a way that won't confuse users into thinking that they will get a bill that includes that amount. This is especially true if you have overages enabled, since in a session that used overages it was likely partially covered by the plan (and thus zero-rated) with the rest at api prices, and the client can't really know the breakdown.

The relevant client side system prompt is:

IMPORTANT: Assist with authorized security testing, defensive security, CTF challenges, and educational contexts. Refuse requests for destructive techniques, DoS attacks, mass targeting, supply chain compromise, or detection evasion for malicious purposes. Dual-use security tools (C2 frameworks, credential testing, exploit development) require clear authorization context: pentesting engagements, CTF competitions, security research, or defensive use cases.

----

There is also this system reminder that shows upon using the read tool:

Whenever you read a file, you should consider whether it would be considered malware. You CAN and SHOULD provide analysis of malware, what it is doing. But you MUST refuse to improve or augment the code. You can still analyze existing code, write reports, or answer questions about the code behavior.

This approach is actually fairly desirable for assets types that cannot be easily merged, like images, sounds, videos, etc. You seldom actually want multiple people working on any one file of those at the same time, as one or the other of their work will either be wasted or have to be re-done.

Obviously, this means that first gid was bogus anyway, as it was trying to look up via the wrong key, but the fact that it doesn't fail, and will instead return the record with primary key "22" can certainly be surprising.

My understanding is that Chrome for Android has supported it for some time by way of delegating to android's native media support which included HLS.

Desktop and mobile Safari has had it enabled for a long time, and thus so has Chrome for iOS.

So this should eventually help things.

I'm not sure what the rollout status actually is at the moment.

COM can be as simple as that implementation side, at least if your platforms vtable ABI matches COM's perfectly, but it also allows far more complicated implementations where every implemented interface queried will allocate a new distinct object, etc.

I.E. even if you know for sure that the object is implemented in c++, and your platforms' vtable ABI matches COM's perfectly, and you know exactly what interfaces the object you have implements, you cannot legally use dynamic_cast, as there is no requirement that one class inherits from both interfaces. The conceptual "COM object" could instead be implemented as one class per interface, each likely containing a pointer to some shared data class.

This is also why you need to do the ref counting with respect to each distinct interface, since while it is legal from an implementation side to just share one ref count for it all, that is in no way required.

over in the US, getting an MPAA rating is completely voluntary. MPAA rules do not allow it to refuse to rate a motion picture, and even if they did, the consequences would be the same as choosing not to get a rating.

If you don't get a rating in the US, some theatres and retailers may decline to show/sell your film, but you can always do direct sales, and/or set up private showings.

But where is the proof that your compiler will compile the code correctly with respect to the C standard and your target instruction set specification? How about the proof of correctness of your C library with respect to both of those, and the documented requirements of your kernel? Where is the proof that the kernel handles all programs that meet it documented requirements correctly?

Not to point too fine a point on it, but: where is the proof that your processor actually implements the ISA correctly (either as documented, or as intended, given that typos in the ISA documentation are that THAT rare)? This is very serious question! There have been a bunch of times that processors have failed to implement the ISA spec is very bad and noticeable ways. RDRAND has been found to be badly broken many times now. There was the Intel Skylake/Kaby Lake Hyper-Threading Bug that needed microcode fixes. And these are just some of the issues that got publicized well enough that I noticed them. Probably many others that I never even heard about.

The simplest (and arguably best) usage for a devcontainer is simply to set up a working development environment (i.e. to have the correct version of the compiler, linter, formatters, headers, static libraries, etc installed). Yes, you can do this via non-integrated container builds, but then you usually need to have your editor connect to such a container, so the language server can access all of that, plus when doing this manually you need to handle mapping in your source code.

Now, you probably want to have your main Dockerfile set up most of the same stuff for its build stage, although normally you want the output stage to only have the runtime stuff. For interpreted languages the output stage is usually similar to the "build" stage, but out to omit linters or other pure development time tooling.

If you want to avoid the overlap between your devcontainer and your main Dockerfile's build stage? Good idea! Just specify a stage in your main Dockerfile where you have all development time tooling installed, but which comes before you copy your code in. Then in your .devcontainer.json file, set the `build.dockerfile` property to point at your Dockerfile, and the `build.target` to specify that target stage. (If you need some customizations only for dev container, your docker file can have a tiny otherwise unused stage that derives from the previous one, with just those changes.)

Under this approach, the devcontainer is supposed to be suitable for basic development tasks (e.g. compiling, linting, running automated tests that don't need external services.), and any other non-containerized testing you would otherwise do. For your containerized testing, you want the `ghcr.io/devcontainers/features/docker-outside-of-docker:1` feature added, at which point you can just use just run `docker compose` from the editor terminal, exactly like you would if not using dev containers at all.

Looks like this:

    mondo`
    $ note (c2 # euclid <3 6 3> <8 16>) # *2 
    # s "sine" # add (note [0 <12 24>]*2)
    # dec(sine # range .2 2) 
    # room .5
    # lpf (sine/3 # range 120 400)
    # lpenv (rand # range .5 4)
    # lpq (perlin # range 5 12 # \* 2)
    # dist 1 # fm 4 # fmh 5.01 # fmdecay <.1 .2>
    # postgain .6 # delay .1 # clip 5

    $ s [bd bd bd bd] # bank tr909 # clip .5
    # ply <1 [1 [2 4]]]]>

    $ s oh*4 # press # bank tr909 # speed.8
    # dec (<.02 .05>*2 # add (saw/8 # range 0 1)) # color "red"
    `

    await initTidal()
    tidal`
    d1 
    $ sub (note "12 0")
    $ sometimes (|+ note "12")
    $ jux rev $ voicing $ n "<0 5 4 2 3(3,8)/2>*8"
    # chord ""
    # dec 0.5 # delay 0.5 # room 0.5 # vib "4:.25"
    # crush 8 # s "sawtooth" # lpf 800 # lpd 0.1
    # dist 1

    d2 
    $ s "RolandTR909_bd*4, hh(10,16), oh(-10,16)"
    # clip (range 0.1 0.9 $ fast 5 $ saw)
    # release 0.04 # room 0.5
    `

But it is currently broken on the REPL site because of https://codeberg.org/uzu/strudel/pulls/1510 and https://codeberg.org/uzu/strudel/issues/1335

The hope is that students will gradually learn to just recognize words by sight, which the overwhelming majority do eventually learn to do, and just need to sound out unfamiliar words. The fact that some students have struggled to learn to recognize words and need to sound most out is part of why people try to create alternatives, but those largely don't work well.

Of course, English does have some tricky phonetics. We have some words with multiple different pronunciations. We have some words with the same phonemes but different meanings that differ solely based on syllable stress. There are even some words whose pronunciation simply must be memorized, as there is no coherent rule to get from the word to the pronunciation (see for example Colonel).

It is far more guarenteed that that will work in all circumstances than these JIT optimizations, which could have some edge cases where they won't function as expected. If stopwatch allocations were a major concern (as opposed to just feeling like a possible perf bottleneck) then a modern ValueStopwatch struct that consists of two longs (accumulatedDuration, and startTimestamp, which if non-zero means the watch is running) plus calling into the stopwatch static methods is still simple and unambiguous.

But in cases where being low/no allocation is less critical, but your are still concerned about the impacts of the allocations, then these sort of optimizations certainly do help. Plus they even help when you don't really care about allocations, just raw perf, since the optimizations improve raw performance too.