Threadwell

Recent Notes

DemetriSpanos
Oct 06, 2026
NOTimothyLottes
Oct 06, 2026
bonzajplc
Oct 06, 2026

jonathan_blow

219 items with this tag.

Sep 08, 2026
Jonathan_Blow
May 29, 2026
Usually I would expect this to be lame.
Aug 14, 2025
I got a bit colorful in this reply, but let me describe why this drives me so crazy, especially now in 2025:
May 19, 2025
Some people might read this and think, that's just not my world, I am stuck in this world where software breaks all the time and everything I build is disposable.
Mar 31, 2025
For years I have used dsound.dll to play audio on Windows, because if you have your own mixer etc then the more-recent APIs don't really do anything better.
Dec 24, 2024
I have a moral dilemma in the design of the programming language.
Oct 16, 2024
Source engine is correct here.
Jan 04, 2022
Re-doing things, that didn't come out well enough the first time, is one of the most effective methods of making good things.
Dec 15, 2021
@cmuratori I was thinking today about how almost all new crypto efforts are like "level 2" and DAOs and stuff, and I realized this is like why software has exploded into higher levels of abstraction -- to do something at "level 1" you have to actually know what's going on
Sep 20, 2021
This weekend I gave a speech saying that software technology has mostly stopped advancing, and stalled sometime in the mid-1990s.
Jul 03, 2021
The "excuse factory" Casey discusses in this video is something I see all the time, not just about performance, but also functionality, or how many people it takes to program something.
Jun 21, 2021
I think I'm a good programmer, then I go do something that is way more complicated and messy than what is required, and I don't figure that out until 4 years later, when it's obvious.
May 29, 2021
RW Hamming dropping the pills.
Jun 01, 2026
@Jonathan_Blow If my Microsoft is behind it, I don't want to touch whatever It is.
Jun 01, 2026
@Jonathan_Blow That is not how a GPU works?
Jun 01, 2026
@Jonathan_Blow When I see this, my first instinct is -- accelerated inference chips to compete with Apple Silicon in their strategy to move LLMs to local
Jun 01, 2026
@Jonathan_Blow Guess not too far
May 31, 2026
@Jonathan_Blow
May 31, 2026
@Jonathan_Blow what
May 31, 2026
@Jonathan_Blow
May 31, 2026
@Jonathan_Blow It's most likely referring to Nvidia's new ARM-based CPU
May 31, 2026
@Jonathan_Blow wtf are you even saying gng,boots to the gpu? What do you imagine the benefit would be.
May 31, 2026
@Jonathan_Blow Doubt it lol but it would be cool to see something radically different, its been a while.
May 31, 2026
@Jonathan_Blow Heard some rumors people from Europe, actually agree with them.
May 31, 2026
@Jonathan_Blow You still need a scheduler, so you still need a cpu.
May 31, 2026
@Jonathan_Blow This sounds like a PC that boots straight into Copilot and doesn't let you out.
May 31, 2026
@Jonathan_Blow Do you mean like for the challenge, or do you have specific use case needs this would enable?
May 31, 2026
@Jonathan_Blow Or an operating system which digitizea GPUs.
May 31, 2026
@Jonathan_Blow If you had a GPU at the center of your computer and you did general purpose computing on it, that would make it the CPU.
May 31, 2026
@Jonathan_Blow > A PC that boots straight to the GPU and has no CPU.
May 31, 2026
@Jonathan_Blow These former crypto/NFT/metaverse influencers really don't understand how computers actually operate.
May 31, 2026
@Jonathan_Blow its ARM cpu from mediatek with nvidia APU in laptop running windows on arm.
May 31, 2026
@Jonathan_Blow Microsoft doesn't have the chops to pull that off.
May 30, 2026
@Jonathan_Blow My suspicion is that it’s a Strix-like
May 30, 2026
@Jonathan_Blow Man you’re a f***ing moron.
May 30, 2026
@Jonathan_Blow Dude image that we wilo have gone from cpus with buikt in gpus to gpus with built in cpu
May 30, 2026
@Jonathan_Blow They're just going to announce an end to consumer hardware while they force all compute processing into their data centers and charge you a subscription fee for remote access from a cheap dumb terminal.
May 30, 2026
@Jonathan_Blow It's just a Spark that runs windows, maybe in laptop form factor.
May 30, 2026
@Jonathan_Blow It just means ARM running on Windows 11 using a Nvidia SoC.
May 30, 2026
@Jonathan_Blow That would make the GPU a CPU though
May 30, 2026
@Jonathan_Blow This is what it is: offload computation to cloud
May 30, 2026
@Jonathan_Blow It's almost certainly an ARM based integrated SoC with both GPU and CPU - a Windows equivalent of Apple silicon.
May 30, 2026
@Jonathan_Blow Take the perf hit in sequential processing, implement manual speculation and result discrimination and just go massively into pure parallel computing? Basically SoC for 10k that just needs a monitor, a keyboard and a mouse?
May 30, 2026
@Jonathan_Blow @stupidtechtakes just take this rock and bash me over the head with it
May 30, 2026
@Jonathan_Blow Arm is involved on the tweet sequence too, so I would expect unified memory and built-in tpu or something AI-centric
May 30, 2026
@Jonathan_Blow Jon, this is most likely NVIDIA N1 and N1X chips.
May 30, 2026
@Jonathan_Blow That would be cool, but I suspect it's just a new ARMv9 platform that NVIDIA has been developing for two years.
May 30, 2026
@Jonathan_Blow No CPU? So everything will be vector based?
May 30, 2026
@Jonathan_Blow Pretty sure this is their new ARM based SOC baloney that is supposed to be as powerful as a 5070.
May 30, 2026
@Jonathan_Blow Translation: "I'm a fucking idiot who has no idea how a computer works"
May 30, 2026
@Jonathan_Blow Are these the longitude and latitude to some stupid event no one outside of the current ai bubble will care about? I would check but I don't want to be right.
May 30, 2026
@Jonathan_Blow I bet is a DGX Spark Copilot edition.
May 30, 2026
@Jonathan_Blow Does that mean as an application developer the idea of a “main thread” would totally disappear? The centrality of a central processing unit is a really nice mental model.
May 30, 2026
@Jonathan_Blow Boots straight to the GPU?
May 30, 2026
@Jonathan_Blow That would make that GPU a CPU by definiton😁
May 30, 2026
@Jonathan_Blow Are you making the argument that most computer tasks can be parallelized to run well on the GPU, or that GPUs can be viably modified to deal with single thread tasks decently enough that separate hardware for that is not worth it?
May 30, 2026
@Jonathan_Blow It's definitely not that...
May 30, 2026
@Jonathan_Blow GPUs are awesome at some things, suck at others.
May 30, 2026
This just wouldn't work?? There is a very good reason why CPUs and GPUs are different and separate, they handle workloads in entirely different ways and both fill each other's weak spots.
May 30, 2026
@Jonathan_Blow So a system on a chip? But the entire concept of a PC is the ability to swap parts over efficiency?
May 29, 2026
@Jonathan_Blow That sounds awful to me.
May 29, 2026
@Jonathan_Blow It's this:
May 29, 2026
@Jonathan_Blow @rianflo We’re building an RTOS on the GPU.
May 29, 2026
@Jonathan_Blow GPUs are extremely dumb.
May 29, 2026
@Jonathan_Blow Most likely another handheld, sadly
May 29, 2026
@Jonathan_Blow Pero sigue siendo un CPU con integrada, nada mas que es ARM.
May 29, 2026
@Jonathan_Blow N1X chip probably
May 29, 2026
@Jonathan_Blow It’s that new Nvidia ARM based APU that @mooreslawisdead has talked about
May 29, 2026
@Jonathan_Blow Explain as you would to a child….
May 29, 2026
@Jonathan_Blow Bolt Graphics is claiming to have a cheap GPU they'll be launching on Risc V that can be ran as a computer with performance better than the 5090.
May 29, 2026
@Jonathan_Blow +1, here's hoping.
May 29, 2026
@Jonathan_Blow windows cannot get into a new era as long they cannot have the balls to abandon x86 and legacy compatibility
May 29, 2026
@Jonathan_Blow https://videocardz.com/newz/nvidia-teases-new-era-of-pc-ahead-of-n1-and-n1x-laptop-chip-announcement
May 29, 2026
@Jonathan_Blow It's probably going to be some internet box, where you don't own your own memory, processing, power, or storage.
May 29, 2026
@Jonathan_Blow which also means your gpu will have to run windows as your trying to fully utlise the gpu for other things
May 29, 2026
@Jonathan_Blow there are ARM and x86 cpus in the works with RTX chiplets.
May 29, 2026
@Jonathan_Blow It's a pc that requires you to login to xbox battle pass on boot
May 29, 2026
@Jonathan_Blow My money is on this being some ARM-based SoC.
May 29, 2026
@Jonathan_Blow One can dream, in an ideal world this would have a publicly available well defined spec that would allow a swathe of alternative OS projects to be built.
May 29, 2026
@Jonathan_Blow What would that actually mean, in terms of hardware? I assume you don't mean that the GPU would have a CPU on package, because that's really no different than a Strix Halo or Core Ultra, etc.
May 29, 2026
@Jonathan_Blow My guess is a remote desktop/OS.
May 29, 2026
@Jonathan_Blow Please NVIDIA don’t be lame.
May 29, 2026
the part that makes this tricky is the memory model.
May 29, 2026
@Jonathan_Blow Seamlessly and irrevocably integrated with AI!
May 29, 2026
@Jonathan_Blow no CPU no Harddrive....
May 29, 2026
@Jonathan_Blow Where do Microsoft put the subscription button then? 🙈
May 29, 2026
@Jonathan_Blow This was not even exciting not because of NVIDIA, which is okay...
Aug 19, 2025
@Jonathan_Blow I feel like it's almost screenshot Saturday
Aug 15, 2025
@Jonathan_Blow I like how people are obsessed with "minimal" APIs, libraries and game engine's.
Aug 15, 2025
@Jonathan_Blow for most cases it comes down to UI needs.
Aug 15, 2025
I still have flashbacks to implementing a CEF browser (Chromium based) and having to deal with the absolute horror that their input wrappers are.
Aug 14, 2025
@Jonathan_Blow are you talking about SDL here or the web browser wrapper on top of SDL?
Aug 14, 2025
@Jonathan_Blow How bad can it be?
May 24, 2025
@Jonathan_Blow Making your own tools vs learning how to use someone else’s tool, usually takes the same amount of time, so it’s almost always better to just write your own tools.
May 21, 2025
@Jonathan_Blow It's really bad and getting much worse.
May 21, 2025
@Jonathan_Blow @TheGingerBill Another thing that contributes to resistance to writing your own is the constant fear mongering from other programmers on the internet that you can’t make it as good as the library.
May 21, 2025
@Jonathan_Blow How do you see these principles apply for languages with runtimes and built-in data structures with syntactic sugar that are heavily optimized at compile and runtime? Like JavaScript and C#.
May 20, 2025
@Jonathan_Blow Also you are less likely to get stuck in the "we have to release a patch but they require the current platform SDK which requires the newest engine SDK - of which we have 5 plugins that don't support that version yet" :|
May 20, 2025
@Jonathan_Blow The paragraph about spam really hits hard.
May 20, 2025
@Jonathan_Blow You should be friends with time-based media conservators, aka me.
May 20, 2025
I've said this for a long time: you can either pull in dependencies willy-nilly and start off with a rapid pace of development which gets slower and slower as time goes by and ultimately you find you can't do certain things at all, or you can build your own libraries which takes time up front, but as time goes by your productivity keeps growing exponentially, and you can do anything you want because you have total control.
May 20, 2025
@Jonathan_Blow This is the epitome of a slippery-slope argument.
May 20, 2025
@Jonathan_Blow And if your OS provided the major dependencies and you eliminated the minor dependencies, welcome to the 60s.
May 20, 2025
@Jonathan_Blow The paragraph about limiting the contact surface for major deps reminded me of C/C++ header/source split compilation model.
May 19, 2025
@Jonathan_Blow A practical example is burgerlib, used over an ~30 year timespan:
May 19, 2025
@Jonathan_Blow I like the discussion around dependencies and them making builds fail(among other things).
May 19, 2025
@Jonathan_Blow Wait, you haven't taken the signed-distance-function pill?
May 19, 2025
@Jonathan_Blow @ID_AA_Carmack Love your post! Interestingly, almost all of it also applies to management consultants.
May 19, 2025
@Jonathan_Blow finally we agree on something, programmers should be careful with dependencies.
May 19, 2025
@Jonathan_Blow I was trying to explain this to my colleagues, but they explicitly insist on adding more dependencies
May 19, 2025
@Jonathan_Blow I once started a project on XNA but it got discontinued, so I switched to OpenTK but it was so buggy that I found myself reimplementing all the features one by one - inputs, sound, graphics etc - until I had my own framework built on top of Win32 & OpenGL.
May 19, 2025
@Jonathan_Blow I don't know how making things from scratch became so taboo, this term is like a boogeyman in the broad software industry.
May 19, 2025
@Jonathan_Blow Do you regard standard library as external dependency?
May 19, 2025
@Jonathan_Blow I frequently find that making my own small tool for something simple is far cheaper in the long run anyway, because having to figure out somebody else’s bloated crap and work with it over and over again is a heavy burden.
May 19, 2025
@Jonathan_Blow Isn't the amazing capability that drove this industry forward, that you do'nt need to constantly reinvent the wheel?
May 19, 2025
@Jonathan_Blow The warning against hostile APIs that you mention Apple and Microsoft use to achieve lock-in also applies to companies with too much proprietary tech with regard to the careers of their employees
May 19, 2025
@Jonathan_Blow I've had folks ask me why I like to build everything myself.
May 19, 2025
@Jonathan_Blow In the web world there's a common wisdom against using "thunk" files, where you centralize the interface towards an external library.
Apr 02, 2025
@Jonathan_Blow If you have your own mixer, what’s wrong with plain waveOut-API?
Apr 02, 2025
@Jonathan_Blow I use WASAPI, it's not bad for low level control to build a sound engine on it.
Apr 02, 2025
@Jonathan_Blow I had to go to XAudio2 just recently, but I have no decent sample code yet.
Apr 02, 2025
@Jonathan_Blow @KodiTV uses WASAPI and DirectSound, maybe some helpful code there
Apr 02, 2025
@Jonathan_Blow Just checked and my clean install from 3 months ago has `C:/system32/dsound.dll` and it appears to be PATHed correctly.
Apr 01, 2025
@Jonathan_Blow Somehow, whenever I look for ways to do something one Windows I'm looking at Wine source code.
Apr 01, 2025
@Jonathan_Blow MME will still be around when Microsoft invents and abandons it’s next multimedia baby.
Apr 01, 2025
@Jonathan_Blow Casey and Martins like Wasapi
Apr 01, 2025
@Jonathan_Blow We also have our own mixer.
Apr 01, 2025
@Jonathan_Blow I’m probably the outlier here, but why not just use waveOut…? It may be deprecated, but it’s definitely not going away anytime soon.
Apr 01, 2025
@Jonathan_Blow Xaudio is pretty simple, but has high latency.
Apr 01, 2025
@Jonathan_Blow i'm using wasapi, very similar to dx from what i remember.
Apr 01, 2025
@Jonathan_Blow There's an unfortunate design decision in XAudio2, where if an audio device is removed (you unplug your headphones), you have to free all resources and reinitialize the API.
Apr 01, 2025
I use waveOutOpen()/waveOutWrite() family, from winmm.dll.
Apr 01, 2025
@Jonathan_Blow Avoid Microsoft Media Foundation!
Mar 31, 2025
@Jonathan_Blow What do you mean in a non-findable way anymore ? Is it in a custom path instead of the classic System32 / SysWOW64 ?
Mar 31, 2025
@Jonathan_Blow I used WASAPI recently (first time), it was every bit the crapola I expected it to be, but no worse, and I would use it again because it's there
Mar 31, 2025
@Jonathan_Blow Have you looked into SoLoud?
Mar 31, 2025
@Jonathan_Blow SDL has audio output functionality, seems to work pretty well from what I can tell.
Mar 31, 2025
@Jonathan_Blow OpenAL?
Mar 31, 2025
I asked Grok for you: For XAudio2, Microsoft’s got decent sample code in their docs—check out the "XAudio2 Basic Audio" sample (http://learn.microsoft.com/en-us/windows/win32/xaudio2/xaudio2-sample-programs).
Dec 28, 2024
@Jonathan_Blow @rflaherty71 FWIW I implement SIMD for Dlang.
Dec 26, 2024
@sasuke___420 @corsix @lemire @rflaherty71 Yeah, AVX512 seems like kind of a waste of time right now unless one is targeting a specific machine for specific reasons.
Dec 26, 2024
I’ve never looked at a real-world optimizer.
Dec 26, 2024
@Jonathan_Blow @rflaherty71 Are you aware of fasmg? It's an assembler engine, meaning that even the instruction sets are libraries (literally just include 'http://avx.inc').
Dec 26, 2024
LLVM does what you want, you just need to learn what the MC layer is all about.
Dec 26, 2024
@Jonathan_Blow @rflaherty71 I'm going to admit my radical opinion here, I think high level languages have been a mistake in general and that all you really want at all is a high level assembler.
Dec 26, 2024
@Jonathan_Blow @rflaherty71 I think its a question of modularity in that long term plan sounds like totally replacing LLVM.
Dec 26, 2024
@Jonathan_Blow @rflaherty71 Why are your metadata tables so large?
Dec 25, 2024
@Jonathan_Blow @rflaherty71 I'm working with my personal language with some similarities in philosophies.
Dec 25, 2024
@Jonathan_Blow @rflaherty71 Id consider being able to compile to C, then you are not dependent on any particular implementation and you are guaranteed to support all platforms.
Dec 25, 2024
@Jonathan_Blow @rflaherty71 It might be worth looking at SPIRV-V as it provides cross-platform RTL that can be converted to platform-specific insns with a lightweight driver/converter.
Dec 25, 2024
@Jonathan_Blow @rflaherty71 With incomplete knowledge and opinions, I believe in the monolithic compiler.
Dec 25, 2024
@Jonathan_Blow @rflaherty71 Feels like the inevitable path down to creating a bespoke IL.
Dec 25, 2024
@Jonathan_Blow @rflaherty71 Could you keep the tables in userspace, but mark them with a compiler directive saying “use this table for codegen”? The compiler would then parse it out in earlier phases for use in the backend.
Dec 25, 2024
I think ffmpeg is one of the large & prolific open source projects that maintains huge segments of hand written assembly code, and they maintain separate assembly code for arm (aarch64) and x86.
Dec 25, 2024
(My last serious work on compilers was a project course way back in 2008, where we got as far as doing a few basic optimizations on SSA, allocating registers and doing instruction tiling for x86.
Dec 25, 2024
@Jonathan_Blow @rflaherty71 I definitely don't see a reason to externalize the assembler if you're going to need it to eventually replace the backend code generation.
Dec 25, 2024
@Jonathan_Blow @rflaherty71 It’s interesting that I’ve seen this exact issue (re:llvm) come up in several domains - there is a monolithic solution that is really 10 pieces combined.
Dec 25, 2024
@Jonathan_Blow @rflaherty71 I do lots of asm/intrinsics.
Dec 25, 2024
@Jonathan_Blow @rflaherty71 As the solution for this exact dillema, in my lang I'm thinking about exposing inline IR to the user.
Dec 25, 2024
@Jonathan_Blow @rflaherty71 Can popcount and SIMD functions be implement using inline assembly? It sounds like they're built into the compiler right now rather than the standard library.
Dec 24, 2024
@Jonathan_Blow @rflaherty71 Suggestion: focus on compile speed and small size, aka simple code generation (easy to maintain), but not necessarily runtime performance optimal.
Dec 24, 2024
@Jonathan_Blow @rflaherty71 I wonder if this is this why MSVC did away with inline assembly in favor of intrinsics? Just the hassle and maintenance effort involved with assembly generation?
Dec 24, 2024
@Jonathan_Blow @rflaherty71 I know a few compiler people and they also hate dependencies, one told me he puts things into modules so the backend and frontend sees as little as possible.
Dec 24, 2024
The costs of dropping SIMD are low in the space of game development.
Dec 24, 2024
IIRC your language lets you run code at compile time as part of the compiling process so why not have the assembler be at that level instead of in the compiler itself and let the compile time code gen be able to output machine code directly & let the assembly instructions be part of the standard library for each specific platform instead?
Dec 24, 2024
@Jonathan_Blow @rflaherty71 I feel like you sort of answered your own question.
Dec 24, 2024
@Jonathan_Blow @rflaherty71 Is it too late to target some subset of C for the IR and optionally let users pick their own toolchain to build the binary if they want to do something exotic?
Dec 24, 2024
@Jonathan_Blow @rflaherty71 Please consider the data communication cost of having more than one binary.
Dec 24, 2024
@Jonathan_Blow @rflaherty71 For graphics APIs you can have a mapping for the lowest common denominator with virtually no overhead; is this similar to what you had in mind for intrinsics?
Dec 24, 2024
@Jonathan_Blow @rflaherty71 I think you should keep the assembler in tree because you know you’re gonna need it for instruction selection.
Dec 24, 2024
If the plan in the long term is to be off LLVM, you’ll need all that code anyway so that is fairly fixed.
Dec 24, 2024
@Jonathan_Blow @rflaherty71 In my experience writing optimized numerical routines the "generic builtins" approach doesn't work well because the underlying data shape and algo is also influenced by the available instructions and bitwidths of the simd impl.
Dec 24, 2024
We have two practical variable-length SIMD instruction sets: SVE/SVE2 and RISC-V.
Dec 24, 2024
@Jonathan_Blow @rflaherty71 I would prefer a cool assembler and "intrinsics" could be implemented with it in userspace.
Dec 24, 2024
As an embedded dev working with ARM and some DSPs, I vote for keeping the assembler inside the compiler but stripped down.
Jan 05, 2022
@Jonathan_Blow Good points.
Jan 04, 2022
@Jonathan_Blow perhaps a topic for your next lecture?
Jan 04, 2022
@Jonathan_Blow My college writing teacher said almost every class that a shitty first draft was required for a good final paper
Jan 04, 2022
@Jonathan_Blow It's particularly frustrating when talking about science.
Dec 22, 2021
@dneto1969 @redgpus @cmuratori @Jonathan_Blow @johnkslang Oh, come on.
Dec 16, 2021
@cmuratori @Jonathan_Blow I'm selling an NFT of this conversation.
Sep 21, 2021
@Jonathan_Blow Are we talking about low level tech or just the tools used to debug and visualize problems in code?
Sep 21, 2021
@Jonathan_Blow Not sure how well they fit into your category, but, coverage guided fuzzing, dynamic/static race detection, Valgrind/ASan/MSan, TLA+, Paxos/Raft, CRDT-s
Sep 21, 2021
@Jonathan_Blow How about other engineering fields? It seems most of them had periods of significant advancements during the 20th century; but appear to be not advancing at that rate today.
Sep 20, 2021
@Jonathan_Blow is there anyone/any org doing work you think to advance the state of programming ? Anything you think is interesting / worth taking a look at?
Sep 20, 2021
@Jonathan_Blow curious if programmers will ever start using waveform debuggers in a widespread way like logic engineers do
Sep 20, 2021
@Jonathan_Blow Big part due to corporations taking over the software industry.
Sep 20, 2021
@Jonathan_Blow As a younger programmer - I think the gravitation back towards these archaic CLI tools mainly comes from the fact that the newer ones have so much hardware overhead that vastly outpaces their human convenience
Sep 20, 2021
@Jonathan_Blow What about crash reporting systems like crashpad paired with aggregation systems like Mozzila's Soccoro and not actively horrible debuggers like msvc? I'd say the state is much better now than back in the day.
Sep 20, 2021
@Jonathan_Blow hmm, if anything I could say about current state of programming is we still program in "2D programming" and we should move to "3D".
Sep 20, 2021
@Jonathan_Blow I guess the main improvement is that now you can ship with the same codebase to multiple platforms without rewriting it (React, Unity).
Jul 04, 2021
@BuschnicK @Jonathan_Blow Maybe the problem here is that science and art does not mean the same for both of you.
Jun 24, 2021
@Jonathan_Blow Test-driven development, done properly, helps enforce the YAGNI principle (You Ain't Gonna Need It, so don't implement it).
Jun 23, 2021
@Jonathan_Blow Once I realized how much being a programmer is like being a blacksmith, things became much clearer.
Jun 22, 2021
@Jonathan_Blow ditto
Jun 22, 2021
@Jonathan_Blow Trial and error is the oldest, and most effective, problem solving method known to man.
Jun 22, 2021
@Jonathan_Blow Code is a snapshot in time primarily based on requirements and knowledge AT THE TIME.
Jun 22, 2021
@Jonathan_Blow if it makes you feel better, i lost the source code for the latest version of a tool i wrote 6 years ago, and now i've discovered an annoying bug in it.
Jun 22, 2021
@Jonathan_Blow In my experience “what is required” often requires an omnipotent oracle, and even then is transient.
Jun 22, 2021
@Jonathan_Blow So what you're saying is you built something that works AND learnt from it? Sounds like a good programmernto me.
Jun 22, 2021
@Jonathan_Blow Jai post reactive? I think its okay to over estimate yourself this is only way to grow.
Jun 22, 2021
@Jonathan_Blow This recently happened to me in a 2y timeframe.
Jun 22, 2021
@Jonathan_Blow I'm not sure that good programmers produce good code.
Jun 22, 2021
@Jonathan_Blow I usually get my wakeup call when I browse my old code from the 90s...
Jun 22, 2021
@Jonathan_Blow The fight between fun code and boring practical code.
Jun 22, 2021
@Jonathan_Blow It's even better when it happens live on stream.
Jun 22, 2021
@Jonathan_Blow We stand on the shoulders of giants.
Jun 22, 2021
@Jonathan_Blow Brute force (as in not more complex than necessary) is usually the best you can do at any point in time.
Jun 22, 2021
@Jonathan_Blow Had the same, but now I have reduced turnaround time for that, around 1-12 months, in the last few years.
Jun 21, 2021
@Jonathan_Blow Nobody can know everything.
Jun 21, 2021
@Jonathan_Blow And then you think "It's been four years since I wrote this trash code, I've grown alot since then" and the cycle continues
Jun 21, 2021
@Jonathan_Blow I’m the best programmer because I never realize my mistakes
Jun 21, 2021
@Jonathan_Blow I was writing a function that iterated over cells in a world map.
Jun 21, 2021
@Jonathan_Blow I recently spent a weekend writing a feature my database already had, and didn't find out until I was writing documentation for the open source library.
Jun 21, 2021
@Jonathan_Blow it's hard to imagine 'thinking you're good at something', all the people who are great at something i know of think they fuckin suck at it
Jun 21, 2021
@Jonathan_Blow Maybe you were good, and now you are simply better.
Jun 21, 2021
@Jonathan_Blow I thought I was good too...
Jun 21, 2021
@Jonathan_Blow This is why I bang on about not trying to write the code right first time.
May 29, 2021
@AndriyTyurnikov @TylerFCloutier @Jonathan_Blow It's a sphere, you can get 1 in one dimension and 0 in all the others...

Created with Quartz v5.0.0 © 2026

GitHub