Elias Mårtenson

Lisp, Emacs, APL and a bunch of other stuff.

From Sweden, living in Singapore.

I always work on a bunch of projects. My current major ones are:

A graphical frontend to Maxima: https://github.com/lokedhs/maxima-client

Kap: An APL-based programming language: https://codeberg.org/loke/array

#lisp #commonlisp #apl #retrocomputing #linux #kap #climaxima #emacs #atari #fedi22

Matrix: Elias Mårtenson (npub1qhm…qhxt):dhsdevelopments.com
Retro computers: C64, Atari ST
Blog: https://blog.dhsdevelopments.com

Public Key

npub1qhmvul058af20389edx4p2vnuxs68d9cehk544e4hj0l0rkhwuasn3qhxt

NIP-05 Address

[email protected]

Profile Code

nprofile1qqsqtakw0h6r7548cnjukn2s4xf7rgdrkjuvmm226u6me8lh3mthwwcprpmhxue69uhhyetvv9ujumt0d4hhxarj9ecxjmntgphwtl

Publishing to

relay.momostr.pink

Author Public Key

npub1qhmvul058af20389edx4p2vnuxs68d9cehk544e4hj0l0rkhwuasn3qhxt

Show more details

Last Notes

2026-06-27 22:43:14 UTC

I have a weird Arch Linux problem.
My mouse likes to completely hang whenever the USB power save mode kicks in (which happens after 2 seconds on inactivity).
I followed the instructions here to disable auto suspend for the mouse, and it works after I plug the mouse in (the udev rule is applied when the 'add' action is triggered)
https://wiki.archlinux.org/title/Power_management#USB_autosuspend
However, it's still activated if the mouse is connected when then computer is booted. So now I have to unplug and replug the mouse immediately after login to have to correct settings applied.
Surely there is a simple way to fix this?
#linux #arch #archlinux

2026-06-25 11:40:30 UTC

- reply

if the source data is just altitude data it would be an interesting exercise to do it from scratch. I'd actually like to give it a try.

2026-06-21 15:07:44 UTC

- reply

Are you coming to sommarhack, and if so, are you bringing this thing?

2026-06-16 16:16:03 UTC

- reply

I have to admit the maths is over my head, although I'll read the linked article anyway, since I do know some of the concepts so perhaps there is something for me to learn.
Anyway, I was curious if there is some aspect of regular QM that is easier to describe using this formulation? That is to say, is there something that suggests this is a better (or shall I say, more effective) description of reality?

2026-06-10 13:41:33 UTC

- reply

my concern is bigger. My opinion wouldn't change even if they were absolutely capable of doing everything the fanboys claim they will be able to do.
I have seen first hand just how quickly people's competence drops, and in particular how openly averse to learning even the basics they become. I've experienced it myself, and the only way to stop it was to completely eliminate these tools from my life.
I'm sure there are plenty of people able to use these systems responsibly. But far too many people are not, which likely includes myself, and it's a time bomb.

2026-05-22 01:47:59 UTC

- reply

if you multiply it together it's 234 W. Somewhat excessive for SD cards, but not enough for welding.

2026-04-29 16:56:25 UTC

- reply

yes, you make a good point. I think the jury is still out on the actual impact. Other than the increase in CVE's we've already seen (I'll leave it as an exercise to the reader to decide how that impacts the actual security posture of a system).

2026-04-29 15:26:41 UTC

- reply

attackers only need to succeed once. Defenders need to succeed every time.
When using a broad system with lots of capabilities, but also stochastic and unpredictable, guess for which side it's the most useful?

2026-04-23 17:14:57 UTC

So I was watching some presentations at Black Hat today, and while the content was pretty good, more than one presenter had created slides with diagrams that had that classic llm look.
You know, the one with way too much text and pictures, arrows that point in weird directions, and boxes full of text. And which doesn't make sense one you actually look at it. Other than the headline, the things that are included in the diagrams have no actual meaning. And when they do, the meaning is wrong.
I really should have taken photos of them. They were actually pretty funny in their ridiculousness.
In several cases it was clear that they never even looked at these generated diagrams, or is there no way to actually edit them after they have been generated?
Which tool are yet all using? They all have the same feel to them, so it's definitely the same one.
The Black Hat judges vet the proposals, but clearly they don't see the slides because this is embarrassing. I can't believe the judges panel would have accepted them if they had seen this.

2026-03-31 09:50:29 UTC

I'm just going to leave this here, with no further comment.
https://files.functional.cafe/media_attachments/files/116/323/164/430/131/248/original/13daf8814070b875.jpg

2026-03-22 23:43:49 UTC

- reply

a GPU inside a vm really only helps if you have a specific usecase, like games. The problem is that even just web browsing requires a GPU these days to not be a depressing experience.

2026-03-22 17:07:10 UTC

- reply

sadly had to stop using Qubes OS because I need the GPU. Annoying, because it's really a phenomenal system.

2026-03-17 16:47:14 UTC

I have a physics (or perhaps maths) question that one been thinking about for a while, and I'm hoping someone here could shed some light on it.
Why is it that everything in the universe appears to fit in a scale that is on the order of roughly 10-80 to 1080? In maths, you can easily come up with problems where the solution goes way beyond what exponential notation can describe (Grahams number is one such example) but in the real world, everything seems to fit in 100 orders of magnitude.
Is there something fundamental that causes this? I know this might be more philosophy than physics, but surely people have thought about this?
#physics #maths

2026-03-09 10:06:16 UTC

- reply

do you have to sign up, or do you just go there?

2026-03-09 09:23:36 UTC

George demands that you look at him.
#mastocats #cat
https://files.functional.cafe/media_attachments/files/116/198/483/477/440/108/original/e1e6baa781c80c0a.jpg

2026-03-07 07:15:34 UTC

- reply

what. Just what.
What?
If you were not a trustworthy person, I'd accuse you of spinning tall tales.

2026-03-01 14:33:56 UTC

- reply

that took a while. I hope you find some good food. 😀

2026-03-01 11:56:50 UTC

- reply

are you taking the bus? I was thinking of doing that to visit the ramadan bazaar there, but now I'm getting scared.

2026-03-01 11:44:14 UTC

- reply

you're back in Singapore? How long this time?

2026-02-17 06:00:52 UTC

- reply

My consumption of present-day gaming discourse is limited to a few gaming podcasts and whatever I catch on here, so I'm definitely not tuned in.
But I've never seen that level of hype about that. Now I really want to see what kind of universe these people live in. Do you have some links?

2026-02-17 05:52:55 UTC

- reply

Are you sure that people claim it's great? At most I've heard people say'it was an interesting idea, they sure were trying new things back then' . And 'that mario tennis game wasn't bad, it's a pity it was on the virtual boy'.
Of course, you're always going to have the usual nintendo fanboys claiming that everything they do is great (I mean, I've seen people praise Metroid Dread as being a great game) but I'm not sure even they are saying it was a great system. At most they'll say it could have been great if the technology was up to the task of embodying their vision.

2026-02-13 16:22:29 UTC

I started working on making the Swing UI for Kap useful.
The world of UI development has gone backward for the last... well... 20 years? And I've come to realise that using any UI framework that originated on this side of the millennium is actively worse than using an older one.
It's remarkable how refreshing it is to use a GUI API which is designed with actual GUI's in mind, and not an overengineered document model.
How did it get so wrong?
#programming #gui

2026-02-08 16:22:09 UTC

- reply

was there something about the review that you suspect is the reason they did it?

2026-02-03 07:54:03 UTC

- reply

where can I sign up for this? Or can I just show up?

2026-01-29 01:17:16 UTC

- reply

the more I think about this the less sense it makes.
Do people not know what international means? Do people buy a ticket to France without knowing where it is? Did the passport check not tip them off?
For whom was that message directed?

2026-01-23 08:47:08 UTC

- reply

BMX simulator isn't bad. https://en.wikipedia.org/wiki/BMX_Simulator

2026-01-17 06:58:26 UTC

- reply

I listened to your acoustic version before the I heard the result.
When I did, I had not idea anyone could make anything sensible out of it (sorry if I'm making assumptions, but I think you are, like me, an amateur in this area).
The result, however, is way more impressive than it has any right to be.
Which band is this? I'd love to hear other things they've done.

2025-12-29 00:02:42 UTC

- reply

I'd say barely active. As someone who used to work with Solaris at Sun back in the day, it's sad to see.

2025-12-25 04:57:48 UTC

- reply

there is one case where it matters. If the username is an email address, and the site allows for SAML authentication.
In that case, it has to check the domain of the email address before it knows what site to redirect to for authentication.
Microsoft at least used to not do that, and did indeed have the username and password fields on the same page, and what happened was that sometimes you could end up writing your password into the wrong site before it had time to redirect.
Now, if the site does not allow for multiple IDP's, then there is no excuse.

2025-12-24 11:27:26 UTC

- reply

if I'm not mustaken, in finlandssvenska it's julgubben.

2025-12-24 07:49:09 UTC

- reply

isn't it like that in most of Europe? I thought celebrating on the 25'th was an English invention?

2025-12-23 11:52:34 UTC

- reply

I started reading it, and although I'm certainly not qualified to critique it, I think I understand the gist of it.
I found this to be absolutely fascinating. I mean, sure, when you think about the paths the ball can take, you kinda get it, but it's still interesting.
My question is, the proof does not rely on chaotic systems, but if you implement a turing machine that simulates a chaotic system, does that make the entire system chaotic?

2025-12-21 07:29:31 UTC

- reply

one could make the argument that this isn't a chiptune though. It's much more than that. There are even analogue, physical instruments there (the printer and disk drives).
Not taking anything away from how impressive it is. It's that using this as proof of an accusation of 8-bit sound lacking soul (do people even say that? It's just wrong) is not the best answer to such accusation.

2025-12-17 06:57:46 UTC

I notice people posting about trying to get a job, so here's hoping that an appropriate candidate reads this.
My employer is looking for someone with devops skills (Linux, databses, networking and development) who's willing to work in Singapore.
The job involves supporting customers (mostly banks) running financial software.
Let me know if you want to know more.
Not sure what hashtag to use for this, so feel free to boost.

2025-11-25 17:12:32 UTC

Why you shouldn't assume machine translation proprely represents what was said.
https://files.functional.cafe/media_attachments/files/115/611/441/256/417/268/original/260bed6d8c031620.jpg

2025-11-04 14:09:53 UTC

@npub1tgg…qwzh Do you know if odfdom-java is maintained? It has a bunch of outdated dependencies, but there seems to be at least some activity in the source repository.
There is also something odd with the versioning, where there is a 1.0.0-beta from 2019, but the latest release is from 2023 is 0.12.0. This causes dependency check tools to be somewhat confused.
Or is there perhaps an alternative library I should be using instead to connect to a running instance of Libreoffice?
#libreoffice

2025-10-25 17:26:58 UTC

- reply

is there more information about them? Do you know if it was common for them to be that young? And why were they in a list of the members of parliament?

2025-10-25 04:11:50 UTC

- reply

thanks for the information about the clerks, I was equally confused about them.

2025-09-27 18:10:09 UTC

- reply

Kolla adressen på google maps, och detta är vad man ser.
https://files.functional.cafe/media_attachments/files/115/277/599/069/480/420/original/626688fc906e4842.png

2025-09-10 13:35:22 UTC

- reply

I found the proposal.
This is actually the chess symbol, and the argument is that some authors prefers to write it with the infinity symbol above instead of below the equals sign.
https://www.unicode.org/L2/L2024/24018-one-chess-symbol.pdf
So arguably it isn't a maths symbol.

2025-09-10 12:44:43 UTC

- reply

ok, this just begs the question: in what contexts could this be used?

2025-09-09 14:47:15 UTC

- reply

I have to admit that the product theyre offering seems appealing though. Are they legit, and just spending a lot of money on marketing, or are they actually crap compared to the competition (how much competition is there)

2025-09-07 03:26:43 UTC

- reply

what do they serve? Indian chinese food is something you have a very hard time finding anywhere outside India, but it's amazing. It's basically Chinese style cooking but with Indian flavours.

2025-08-31 16:29:47 UTC

- reply

I was just about to reply to you with a question about negative mass neutrinos, but since it was raised here, I'll reply to this post instead.
I was reading the paper, and came across the part quoted below at the later part of the paper.
After pointing out that this leads to the possibility of negative neutrino mass, they say this:
> Taken at face value, these results raise important concerns, either pointing to possible undetected observational
> systematics in the DESI BAO data (and/or other probes involved in the analyses) or suggesting the emergence of new
> physics beyond the ΛCDM model, adding to a longstanding series of tensions challenging the validity of the standard
> cosmological framework. Given the current uncertainty, with analyses supporting each direction [102, 110–121], both
> possibilities warrant serious consideration.
Do you agree with this assessment?

2025-08-31 08:42:45 UTC

- reply

thanks! I did indeed send the question before you posted the third part. It answers most of my questions. I'll see if i can understand the paper now. 😀

2025-08-31 08:20:44 UTC

- reply

so are they doing this so that they get neutrinos with a specific flavour (sincey presumably have a specific flavour when emitted, and the wavefunction hasn't spread put much when being received)?
Will this help us understand the absolute mass of the neutrinos, and not just their masses relative to each other?

2025-08-30 15:11:21 UTC

It sure must be comfortable to be a US American. Everything is so simple, there are categorical answers to everything that ails you and everything is black and white.
Most people outside the US have learnt to simply not say anything when someone from over there proclaims some universal truth.

2025-08-29 18:34:57 UTC

- reply

it never dropped the experimental label so no one was supposed to be using it for production anyway.

2025-08-23 18:03:19 UTC

- reply

It was indeed the waiting for the results to come in that took time. If I skip that, and just batch read all the results afterwards, the performance compared to a single write is very close.
A single write of an almost 60 MB stream took 453 ms. With 1024 byte block size, i.e. split up into almost 59000 APPEND calls, the entire operation, including reading the results at the end took just 474 ms. That's remarkably close.

2025-08-23 15:39:22 UTC

- reply

or wait a minute, should I just keep writing, and not wait for a confirmation after each write?

2025-08-23 15:21:12 UTC

- reply

OK, I ran it through a profiler, and each request spends over 40 ms blocking on SocketDispatcher.read0 which is the underlying native function in the JVM which reads from a network port. So it would seem that that is what it's doing.
Is there anything else I can do on the Valkey side to figure this out?

2025-08-23 14:16:33 UTC

- reply

Interesting, but sadly that gave me an empty list. That suggests that things should be fine on the Valkey side, no?
I'm going to instrument my code so I can see what's actually going on.

2025-08-23 14:11:55 UTC

- reply

I'm not sure I have given you enough information to take a look. 🙂
Is there some way I can get statistics from Valkey to tell me what is slow here? Perhaps it's not coming from Valkey itself, but maybe I'm doing something badly? (although it should all be pretty straightforward, so I'm not sure what that would be)

2025-08-23 14:04:38 UTC

- reply

I implemented something similar to this solution, but it's not ideal.
I have a few issues with this though. I set the block size pretty low, to ensure the APPEND trick works well, which it does, but the overhead for the APPEND call is not great. A 55 kB stream (i.e. roughly 50 separate APPENDS) turns a 50 ms operation into 2 seconds.
Sure, it'll be much less bad when I create the buffer size, but this suggests that while the idea works, it's not perfect.
https://codeberg.org/loke/array/commit/c41c9daa5aceb46b4dbfcab1ac34389646d3be5e

2025-08-22 17:35:50 UTC

- reply

Tell me more! What is this?

2025-08-22 10:55:42 UTC

- reply

oh, I see. That's pretty smart. Is the move a zero copy operation?

2025-08-22 09:18:27 UTC

- reply

in my case, we're looking at millions. I guess I'll have to do some tests. 😀

2025-08-22 08:47:21 UTC

- reply

I could use append, but I don't want the data to be visible until the entire stream has been written.
I just tested, and it does seem that a series of APPEND in a MULTI translation might do what I need.
Is there a limit to the number of APPEND I can have in a translation?

2025-08-22 02:00:49 UTC

- reply

If I try to summarise it, the language uses lazy evaluation on everything by default. Thus, you may have an array of some number of elements that I want to perist. However, I don't know anything about the contents of those elements, because the functions that compute them haven't even been executed yet.
This means that even if I wanted to, there is no way to find out how many bytes the serialised form of this array is.
Currently, I serialise the data into a buffer, and then send that buffer to Valkey in a single SET call.
This is jot ideal, because if I could write the data directly to Valkey, there would be zero copying involved. Also, there would be no need to hold on to an array in menory that could get pretty large.
I have not found a way to do this in Valkey, since wire protocol requires a size prefix that tells Valkey how many bytes it need to read.
And even if this existed, I'd need one more feature. Since in my ideal solution, the write to Valkey starts before the code that generates this data has even finished evaluating, so there also needs to be a way to abort the Valkey write if the evaluation yields an error.
Although the last one can probably be solved using transactions.

2025-08-21 18:09:18 UTC

- reply

It's an array-based language that is based on APL: https://kapdemo.dhsdevelopments.com/
It's quite niche, and while it would be neat to add it, I'm not sure it's popular enough. 🙂
I'm working on making Valkey the backend for general data persistence in the language, but I've just started with it.
Speaking of, what is the correct place to talk about (and suggest improvements to) the wire protocol? I could improve performance significantly if there was a way to write a string to a key without knowing the size beforehand (there is a super interesting technical discussion why that is, but I'll stop myself from babbling too much about it right now).

2025-08-21 17:43:42 UTC

- reply

I wish I could go. I've become a bit of a valkey fan recently, as I've been implementing builtin support for it in my programming language. 😀

2025-08-20 07:36:30 UTC

- reply

I didn't know about this game, and had no idea that it was this early.
I don't think a description of microtransactions in games can be complete without mentioning Gauntlet. Was this the game that perfected it? Was there any earlier game that did it as blatantly and successfully as that one?

2025-08-18 07:23:44 UTC

- reply

Probably. I think good pentesters know how to handle this, and will take this into considering when finalising the report with the customer.
The problem is that there are some bad incentives around bounties though. There is a financial interest in the pentester to get the severity as high as possible. And even if they don't get a monetary award, the testers reputation will improve if they can show they've found a lot of severe vulnerabilities.
Good testers working for serious customers don't have this problem, because then the incentive is to provide a useful and actionable report to a customer, and in my experience those reports tend to be useful.
But there is also the problem that the CVSS scoring system doesn't take this into consideration, so when a formal scoring is done, the fact that the there is an assumption that the system is misconfigured isn't reflected in the final score, at least not to the degree it should.

2025-08-18 06:49:32 UTC

- reply

Suppose one wanted to honour such reports, what on earth is the curl library supposed to do to prevent this?
I can't even think of a theoretical way to do that, and I don't think the reporter did so either. I suspect they just ran a fuzzer and never even understood what had happened.

2025-08-17 09:11:10 UTC

- reply

DOS better than Multics? I think you're going to have to explain that one? The rest, I can see why you'd say that (even though I might not agree)

2025-08-17 04:36:49 UTC

- reply

what operating systems are worse than DOS? 😀 I have a hard time thinking of any. Not counting the very early 50's experiments before the concept even existed.

2025-08-02 07:53:10 UTC

- reply

perhaps I should write a real fediverse library in Cobol.
Would be neqt using SCREEN SECTION to display a post editor.

2025-07-28 09:02:28 UTC

- reply

what's the context? Anything in particular that happened?

2025-07-10 13:02:42 UTC

- reply

is it called Meonv?

2025-07-10 10:21:14 UTC

- reply

the intended word was 'about'.
The swipe style input in the Samsung keyboard is pretty awful.

2025-07-10 10:13:16 UTC

- reply

hah, you are 100% correct. 😀

2025-07-10 09:59:28 UTC

- reply

I knew spy this pub, since some time ago I stumbled across this video. It's about the hike rather than the pub. https://youtu.be/0WVOSmWKbgw

2025-07-09 13:17:10 UTC

I'm visiting my country of origin (Sweden) and went on a bike ride. I love this kind of moss (pr is it lichen?). I don't know what it's called, but perhaps the Fediverse can help?
#mosstodon
https://files.functional.cafe/media_attachments/files/114/823/464/744/505/840/original/9c0cdef9d2051a31.jpg
https://files.functional.cafe/media_attachments/files/114/823/464/810/329/562/original/c98e4c68a2cae63e.jpg

2025-06-28 17:02:07 UTC

- reply

The same as the one I saw? https://functional.cafe/@loke/114755766240121761

2025-06-21 15:04:40 UTC

- reply

Indeed they are. And sometimes I know the person making the post very well, and I concluded that one of two things must be true. Either:<li>These people go insane as soon as they open the linkedin page (perhaps by pressing Control-Shift-Alt-Windows-L in windows), or:</li><li>They are always like this, but are just good at hiding it.</li>
I don't know which one worries me the most.

2025-06-21 14:57:25 UTC

For some reason, I decided to log into linkedin for the first time in, I don't know... a decade?
My intent was to actually post something ironically vacuous, perhaps a subtle jab at llm fanboys, or something like that.
*I. Closed. The. Browser*
Seriously, nothing I could post ironically could even get close to the average post made in what appear to be seriousness.
I'm sure one can master that hellhole and become 'succesful', but I'm not sure I want to be successful if it means spending time reading that flood of drivel.

2025-06-18 15:59:36 UTC

- reply

I am absolutely terrible at coming up with names, but I am still intent on suggesting one for you (without help, computerised or otherwise). What will the topic of the podcast be?

2025-06-11 18:30:28 UTC

I needed to log in to my gmail account just now and was met with a new type of 2FA check. I could no longer just press ok on my phone but instead I was shown a QR code.
Well, I scanned it and I got a confirmation screen. I clicked OK, and the authenticator hung. The only thing I could do was press cancel.
I tried multiple times with the same result.
Thankfully I had also activated a yubikey with my google account, so I was able to use that.
But if I hadn't had that, I'd be screwed, I think?
What is this thing, and why wasn't I warned this was going to happen?
#infosec

2025-06-08 06:34:22 UTC

- reply

I checked a map where Portland is. All is this the famed oregon train?

2025-06-07 06:12:24 UTC

- reply

is this the one which runs steamos?

2025-06-05 13:14:55 UTC

Does anyone have a template for the most bland corporate web page possible? I know what I want to create, but I don't have the skills to actually design the thing.
I have this idea for a joke, or art project, or whatever you want to call it.
I want to create an "Enterprise" marketing page for my programming language (the actual web page is [here](https://kapdemo.dhsdevelopments.com/ )).
I have a pretty good idea what it will contain: A lot of corporate jargon, and an explanation of what the solution is which doesn't actually explain what it is.
You know, like most corporate websites where you have to go to the WIkipedia page to understand what they're actually trying to sell).
I've read so many of these, and I get so infinitely frustrated by all of them, that I want to try to make my own just to see how bad you can make them.

2025-05-26 14:57:53 UTC

- reply

that thread is interesting. It cannot possibly be very easy to get a machine into this state, can it? Otherwise all Freebsd servers on the internet would be dosed (no extra d needed) already.

2025-05-23 16:42:46 UTC

- reply

This discusses improvements to the algorithms, which interesting of course. But how close are we to actually being able to run these algorithms on actual hardware?

2025-05-22 08:30:15 UTC

I just came across a really weird problem with my Linux laptop. I'm plugging in the charger, and the touchpad becomes laggy. When I unplug the charger, it's responsive again.
Anyone saw this issue before?
#linux

2025-05-11 13:49:10 UTC

Here's one of those interesting posts by @npub12cu…d38p which talks about some interesting properties of hexadecimal numbers: https://mathstodon.xyz/@gregeganSF/114489111257308697
I wanted to play around with this myself, and since I used Kap to do it, I just wanted to share what this looks like. Here's the original list of numbers, computed from first principles by splitting up each number into a binary array, performing the shifts, grouping the bits in groups of 4, encoding each group in binary and then encoding each group in base 16: [example](https://kapdemo.dhsdevelopments.com/clientweb2/#c=aW86dG9IZXjCqCAxNuKKpSAy4oql4o2JIOKKgyg2NOKNtDEgMCAwIDApIOKKhiDiioMgKOKNszUpIOKMvcKoIOKKgiAoNjTijbQyKSDiiqQgMHhhZjlkMDNjNmJlMjE1NzQ4 )
It could have been shorter if I had encoded the entire number in binary first, but I didn't do that because what I actually wanted to do was to work with the groups of 4 bits to show the symmetry where each group adds up to 8. Here's the code to do this: [example](https://kapdemo.dhsdevelopments.com/clientweb2/#c=Ky9bMV0g4oqDICg2NOKNtDEgMCAwIDApIOKKhiDiioMgKOKNszUpIOKMvcKoIOKKgiAoNjTijbQyKSDiiqQgMHhhZjlkMDNjNmJlMjE1NzQ4 )
The output is as follows:<code> +/[1] ⊃ (64⍴1 0 0 0) ⊆ ⊃ (⍳5) ⌽¨ ⊂ (64⍴2) ⊤ 0xaf9d03c6be215748 ┌→──────┐ ↓8 8 8 8│ │8 8 8 8│ │8 8 8 8│ │8 8 8 8│ │8 8 8 8│ └───────┘ </code>
If you change the value to a number which doesn't have the symmetry mentioned in the post, the array will not be all the same number.
#maths #kap #apl

2025-04-20 15:00:52 UTC

What does enshittification look like? Well, most of you know what it looks like, but here's another example.
Today I was biking in an area I had never been to before, going on a bike route that takes the route of an old railroad line.
I needed to go back where I came from, so I asked google maps for a bike route back. It recommended a 43 minute route that began with going in the opposite direction, and then snaked its way via small side roads to where I needed to go. The bike road I was on wasn't even properly marked.
I then when to OsmAnd (openstreetmap for android) and it gave me a 30 minute route that started by turning back on the same road I was, back for about 10 minutes and then pretty much straight back.
I think I know why this is. Google Maps no longer exists to make navigation easier. For users, it's become little more than a vehicle to provide connections to businesses. Taking it a bit more abstractly: Its real goal it's all about location-based advertising, and bikers don't make them any money.

2025-04-13 04:45:03 UTC

- reply

it's X68000, isn't it?

2025-04-12 14:30:07 UTC

- reply

How are those cars even allowed? Is this in Belgium or Netherlands?

2025-04-03 16:38:41 UTC

- reply

I paged down page, after page, after page, and with maybe 3 exceptions, every. single. fix modifies code comments, change the character case of html tags.
Here is one of hundreds of examples: https://github.com/apache/openoffice/commit/8c3681a51823f1c4a400590c771b000ee885c29d
What on earth is going on? Surely a human isn't assigned to make a few of these changes every single day? Or maybe the contract allows them to afford this?

2025-04-03 15:05:44 UTC

- reply

I don't understand the openoffice project.
They claim to maintain it, and if you look at the repository, there are updates all the time, but pretty much every single one of them are reformatting, some typo fixes, changes to formatting of documentation, etc.
You have to scroll way down before you find a single fix that actually changes the code.
https://github.com/apache/openoffice/commits/trunk/
Just read the commits. It's so very strange.
Why are hey even spending the time doing this? The only explanation I have is that they have some contract where someone is paying them money for it to be updated, and they do the absolute bare minimum to avoid breaking the contract.
This, unfortunately, has had an extremely harmful effect on the open source community, who are best represented by libreoffice here.

2025-03-28 08:31:48 UTC

- reply

Yes, indeed. All fractions are reduced.
That explains it, thanks a lot.

2025-03-28 05:29:36 UTC

- reply

Is the code correct? I wrote a small FRACTRAN interpreter, and your program never ends.
Here is a [link](https://kapdemo.dhsdevelopments.com/clientweb2/#c=KDIqNSkgeyBj4oaQ4o21IOKLhCBk4oaQ4o21LMKvMSDii4QgLSAoeyDijbXDl2RbMeKNs+KNqG1hdGg6ZGVub21pbmF0b3IgY8OX4o21XSB94o2je+KNujwwfSkg4o26IH0gMTLijbQzcjcgMXIxMSAxNHIzIDExcjUgMXIzIDQzN3IyMjEgM3IyIDM1cjIzIDI5cjE5IDIzcjcgMTNyMjkgNDQ4NXI1Mjc= ) to my small hack written Kap that runs your algorithm.
Now, it may be that my tiny implementation is somehow broken.
Here's a version that prints the first 10 rewrites in my implementation: [example](https://kapdemo.dhsdevelopments.com/clientweb2/#c=KDIqNSkgeyBsaW1pdOKGkDEwIOKLhCBu4oaQMCDii4QgY+KGkOKNtSDii4QgZOKGkOKNtSzCrzEg4ouEIC0gKHsgbuKGkG4rMSDii4QgaW86cHJpbnRsbiDijbXDl2RbMeKNs+KNqG1hdGg6ZGVub21pbmF0b3IgY8OX4o21XSB94o2je+KNujwwIG9yIG4+bGltaXR9KSDijbogfSAxMuKNtDNyNyAxcjExIDE0cjMgMTFyNSAxcjMgNDM3cjIyMSAzcjIgMzVyMjMgMjlyMTkgMjNyNyAxM3IyOSA0NDg1cjUyNw== )

2025-03-22 07:56:03 UTC

- reply

absolutely. If we move the clock backwards to before the arrival of computer-based searching, the difficulty was in finding the information in the first place, but once you had it you could be reasonably sure that it was true (at least a true statement made by whomever made it, religious texts notwithstanding).
Today, this is completely flipped on its head. Information is easy to get hold of (in fact, it's being pushed to you) but assessing its accuracy is becoming harder and harder, even for someone who is actively trying to research a subject.
The problem is that people still think research is about finding information, leading them to think they've done 'research' when they followed a series of youtube links.

2025-03-22 07:42:05 UTC

- reply

also, I think for a lot of people the time wasted due to errors are offloaded to others. I've often had to spend more time reworking someone's "helpful" assistance than it would have taken to do the work outright.

2025-03-20 08:23:07 UTC

- reply

It's a web-based tool that helps companies work on presales question&answer documents.
Especially for large projects, you have thousands of questions from the customer that covers all their requirements, and these have to be answered, and then used as a basis for creating an SOW (statement of work).
This tool is a collaborative editing tool to help with this.

2025-03-20 08:08:15 UTC

- reply

You using rfpio ...i mean... Responsive?
It's the worst feature I've ever seen. You stop typing for less than a second and immediately something completely irrelevant shows up that takes you out of the frame of mind you're in in order to compose a sentence.

2025-03-17 14:23:33 UTC

- reply

ise is also used elsewhere, like Australia though?

2025-03-15 05:51:01 UTC

- reply

negatives are kind of difficult to prove unless you have a strict set of axioms (i.e. maths) so I'm not sure you should worry about not being able to prove it.
It's on them to show that their worldview is likely, or even possible (and not just reply with some kind of Pascal's wager reply)

2025-03-15 05:36:25 UTC

- reply

to them, they did. It's perfectly clear to them.
I think I could follow the reasoning actualy. It's completely stupid and makes no sense, but if you just take their assumptions and worldview for granted it's not too complicated.
This means that their assumptions are both stupid and wrong, but I think we already knew that.

2025-03-12 17:04:52 UTC

- reply

Given how these things costs money (and probably loses money) on each query, I thought they'd be happy to not push it on people who don't want it?