LisPi

Hi, I'm Lispi, Lisp (Technomancer) Wizard (to eventually be).

You might know me from LisPi (npub1rp2…mhtw)

I like Free Software, #Emacs and resilient computing a lot.

I also like anime girls, animes with cute girls doing cute things and artwork with them too. Cute stories are good too.

Some Pins:

Software and Assumed Privilege, common problems: https://mastodon.top/@lispi314/111253066257920146

Writing Privacy-preserving software & services 101: https://mastodon.top/@lispi314/110849018589421824

Current avatar: Cyn from Murder Drones, in her cute maid drone appearance.

#Kopimism #FreeSoftware #CommonLisp

Public Key

npub1326y4p8mw4l750yc8e52n0xguvcl5j0dn88ul5r2gjva2sh2ay7qjqgzqx

NIP-05 Address

lispi314_at_udongein.xyz@momostr.pink

Profile Code

nprofile1qqsg4dz2snah2ll28jvru69fhnywxv06f8kenn706p4yfxw4gt4wj0qpz3mhxue69uhhyetvv9ujuerpd46hxtnfduqs6amnwvaz7tmwdaejumr0ds7n6azk

Publishing to

relay.damus.io

nos.lol

Author Public Key

npub1326y4p8mw4l750yc8e52n0xguvcl5j0dn88ul5r2gjva2sh2ay7qjqgzqx

Show more details

Last Notes

2026-03-23 17:44:29 UTC

- reply

Nah, the American obsession (I wish it was only Americans) with living for work & work alone is the thing that smacks very incompatibly with the notion of hobbyful behavior.

2025-11-23 12:35:50 UTC

- reply

Peer stuff has major usability issues (from insufficiently polished software, not the tradeoffs inherent to the schema) and tends to get abandoned so fast it rapidly suffers from bitrot if it was ever properly finished to start with.

2025-08-14 03:16:00 UTC

- reply

https://en.wikipedia.org/wiki/Wplace
Oh. So it's like the Reddit thing but larger?
I have no idea.

2025-08-14 02:52:26 UTC

- reply

I was never curious enough to even check what it was.

2025-08-06 02:04:16 UTC

- reply

I think in some kind of Saboteur meets Hitman meets Deus Ex it might be usable for some fun things.

2025-07-22 07:20:50 UTC

I've seen people advocating for XDR for IPC going "dbus/protobuf/etc isn't necessary" but I think they haven't really looked looked at its actual capacities.
No, XDR is very much not adequate and even from a pure C point of view it is distinctly unimpressive.

2025-05-01 01:09:47 UTC

- reply

There isn't really a need for conventionally living in society to develop the philosophy required for morality of various sorts.
By conventionally I mean to cover the argumentative "any contact with one other than oneself is a form of society". Although, really, even that isn't truly required.
Further, your conceptualization of morality presupposes that it only applies to one's relation with similar beings (people), rather than something more universal or animistic.

2025-04-03 05:33:53 UTC

- reply

So coming up with some kind of distributed (operating) system would be necessary for practical results, it couldn't be accomplished with a single device at current time without severe limitations.
That complicates matters, but doesn't quite make it impossible.
Legacy support would be an issue.

2025-04-02 09:18:07 UTC

- reply

> It's expensive, but worth it imo.
Is it though? Even if one manages to find a manufacturer that is somehow not malicious, ordering a batch of CPUs tends to require astromonically expensive batch orders.

2025-04-01 02:55:34 UTC

- reply

Technically they aren't, I suppose.
Cheap parts /could/ possibly actually behave within their specs reliabily, which would be fine.
They usually don't though. (Not that there's any real correlation between such reliability and price, corposcum love to scam people with garbage.)
(And so cheap gets colloquially used for "unreliable" or "worthless" without necessarily referring to the price.)

2025-04-01 02:28:10 UTC

- reply

Aren't those synonymous?
Using cheap parts /is/ bad engineering.

2025-03-23 09:03:56 UTC

Considering remote power & timing analysis, I wonder, would it be possible for FPGA-based systems to periodically shuffle/randomize/spoof those aspects while maintaining a particular a logical configuration?

2025-03-21 00:45:07 UTC

- reply

Oh, they made them non-GNU?
Then yes, any scripts written wrong (without checking versions) will break.

2025-03-20 21:48:14 UTC

- reply

Is GCCRust ready?
I do approve of moving away from C. Those tools legitimately never should've been made in C.
Unfortunately at the time there was no such thing as Free/Libre Ada compilers.

2025-03-20 07:31:33 UTC

- reply

Uh-huh, right... how much do you want to bet they very much invent their own culprits on a frequent basis?

2025-03-20 06:58:32 UTC

- reply

Makes El Salvador sounds like a totalitarian shithole.
How far off the mark is that impression?

2025-03-20 00:02:01 UTC

- reply

Brought to you by Mozilla #Firefox using dozens of #SQLite databases and clearly not batching the transactions.
Did no one stop to think for a second of the performance implications of dozens of close-together fsyncs|dsyncs on HDDs?
Also, why in the world so many databases? It's completely unnecessary and even though they *could* be used as attached databases, that means one can only use them in non-WAL mode (yay, blocking readers) to maintain proper global transactionality (https://sqlite.org/lang_attach.html https://sqlite.org/wal.html).
Did they even read the docs before doing this nonsense?

2025-03-19 09:19:05 UTC

- reply

The worst part is that browsers have features for users to set their preferences, and CSS actively supports checking those preferences.
If the site doesn't actively override the color schemes, the browser will even ensure on its own that things follow those preferences.
So they have to actively go out of their way with an override doing it wrong.

2025-03-16 20:53:19 UTC

- reply

> putting the router at one end of the house causes problems for the other far end of the house.
The main aspect is that the router shouldn't matter (and ideally shouldn't even provide wifi functionality, it raises cost uselessly), there should be wifi access points wired to it.

2025-03-15 11:15:34 UTC

- reply

Did you do something based?

2025-03-15 07:54:29 UTC

- reply

Broadcast messages? That's... pretty old and simple-ish as a concept.
Or do you mean the SaaSS nature of that specific implementation which offers a surveillance opportunity after compromise?

2025-03-13 22:13:04 UTC

- reply

> One of the things Anders Hejlsberg said in his presentation was that they wanted good memory layout control, AOT compilation, good cross platform coverage. In particular, C#’s AOT was not good enough.
I see
> I guess that “AOT” may be a requirement to avoid requiring people to have another runtime installed (I know I would be miffed if I had to install a JVM in order to compile my front end project)
If it's packaged in one's distro I don't really see the issue, unless it requires a ridiculous amount of storage.

2025-03-13 21:16:32 UTC

- reply

Which to me sounds more like a "why not make C# faster" situation.
Kotlin being faster is one I have to assume they just didn't bother to really test. The various JVMs can get /pretty/ fast if tuned properly, and start off quite decent.

2025-03-11 23:44:08 UTC

- reply

Did he do something silly?

2025-03-08 06:23:28 UTC

Does Guttman erase even work with any modern hardware anymore?
Even with CMR HDDs are the logical addresses even really related to the physical addresses at this point?
A full disk erase probably does it (ignoring bad sector remap overprovision buffer), but anything less seems questionable to me.

2025-03-07 04:20:13 UTC

- reply

I think they might be built with the assumption of SSDs' weight, rather than HDDs.
If it's properly supported or reinforced that's not an issue but worth keeping in mind.
> Reading further this thing might violate the PCI-E specification?
> You're supposed to only draw up to 10W from a PCI-E 1x slot
> This fucker needs 20W
Isn't PCIe supposed to be 75W?

2025-03-05 21:18:48 UTC

- reply

That's what refurbished devices mostly are (with some checkups to ensure they're fine & part replacements).
None of them are within a reasonable price range /unless/ they're also too outdated to run GrapheneOS (like older Pixels).

2025-03-05 05:38:42 UTC

- reply

One could perfectly well pass /proc/$PID/fd/$file files, I think, so paths should work too.
It works with memfd too. One can test it with python3 'import os; fd = os.memfd_create("test")' and looking in the process' procfs directory for the fd file entry.
some-program <(echo "uwu")
Is the bash way of passing an fd iirc.

2025-03-05 05:25:34 UTC

Weird meatspace-avatar'd people with zero posts direct messaging me.
Sus.

2025-03-05 05:14:15 UTC

- reply

Very bad idea, do not produce fluoride gas or hydrogen fluoride for no reason.
Unless you've actually bothered to establish a sufficient containment zone.

2025-03-05 04:54:25 UTC

- reply

How does it handle fds as arguments that way anyway?
It had been my understanding that fd numbers were per process, so the only way I can see that work is if it effectively requires the caller to use exec/execve or yet another indirection step into a wrapper single-threaded program to acquire FDs just to pass them on by using exec/execve.

2025-03-05 04:29:24 UTC

- reply

Clone and fork are broken in a lot of languages and especially with multithreaded processes (even in C).
I was thinking of running the other processes from an executable binary of some sort (amounts to the same even if it immediately loads some bytecode like a JAR or whatever).
What annoys me is not the notion of calling an external binary. It's the use of the generic popen/system()-like interface rather than something that properly encodes the call semantically (even if it then just uses popen in the background) and manages flags using language-idiomatic means.

2025-03-05 04:14:18 UTC

- reply

It's unfortunate because even refurbished Pixels are ridiculously expensive.

2025-03-04 02:27:45 UTC

- reply

It's a mispelled gift.

2025-03-03 12:31:26 UTC

- reply

I think relative_path would do it, but it doesn't specify whether it touches the filesystem (the explanations seem to suggest it doesn't).

2025-03-03 11:28:56 UTC

- reply

You mean just a string path manipulation library?

2025-03-03 05:51:16 UTC

- reply

Well that or you implement monads & utils for all of them.

2025-03-02 00:59:19 UTC

- reply

That presupposes a non-generalized rejection of those.

2025-03-01 08:59:16 UTC

Is there a #Java library that does like Python3's PyFilesystem?
Offering an ever increasingly restricted view of a filesystem tree & enforcing operations only happening within?
In theory the basic JDK Filesystem class /could/ do it... but literally all enforcement is unspecified and left up to the implementation to choose, meaning it's completely useless.

2025-03-01 04:41:39 UTC

- reply

So much like any unsatisfactory social contract one can ignore it?

2025-02-28 11:53:29 UTC

- reply

So, did you read on all the cryptographic problems referenced about PGP?
Because it isn't battle-tested and unscathed. It's battle-tested and it looks like the ship is tilting dangerously to one side and has about 2 of its 18 boilers not blown yet.

2025-02-28 11:33:00 UTC

- reply

https://soatok.blog/2024/11/15/what-to-use-instead-of-pgp/
Also NNCP if one wants email-like (more UUCP-like) interaction with proper security, though there are tradeoffs on its ephemeral key design due to the high-latency use-cases it purports to support.

2025-02-28 01:58:13 UTC

- reply

In my view, the one was the user.
All licenses I've seen try to address the issue of undermining the rights of others also inherently prevent the right to self-defense against those who would undermine such rights, at least while using that license, or end-up a ridiculously verbose ever-growing stack of exceptions and conditions (as it gets patched again & again).
It's rather self-defeating.
I think it is impossible for a license, and the guiding freedom principles behind it, to truly foresee & address every possible case.
That is why I see these as: These are freedoms one has in computing, and normally within the bounds of society. If one decides to ignore the bounds of society, one has full responsibility for one's actions and take it upon oneself to attest that one's actions are in response to an unjust and/or oppressive society, and ideally in service of creating a better society.
Slaveowners would object to my designing (and sharing the designs) of weapons meant to help liberate slaves and bodily harm the slaveowners if they attempt to resist. A license and the principles behind it should not defend them.
This is the same reason why I see carveouts against weaponry or weaponizable technologies (nevermind that literally everything is weaponizable with enough imagination) as problematic.
While not all use of force necessarily requires force in response for defense, it often does. Of course that point is partly irrelevant since those are mostly all proprietary and militaries just completely ignore licenses and freedoms both in the course of their operations.

2025-02-28 00:00:46 UTC

- reply

I don't strictly want Free/Libre Software (it really should've just been called Libre Software, if it really had to be an English name) to disable corporate use, as that would run counter to the Four Freedoms in important ways and cannot really be reconciliated with them.
My anticorporatism and anticapitalism *does* influence Free/Libre Software, and I consider it is also necessary for the Freedom sought by Free/Libre Software to be actualized.
The original foundation of the movement ignores far too much of intersectionality, by its isolation from other activists and anarchists at the time (and now, since the foundation seems to have kind of fossilized), and that strictly limits the scope of its potential on its own.

2025-02-27 23:30:39 UTC

- reply

Well, I'd be on the "we must abolish FOSS" (the conflation, specifically) side, because Open Source (purposely) misses the point of Free Software. They shouldn't be conflated.
It is very much the point of OSS to be corposcum-friendly collaborateurs.

2025-02-27 21:56:13 UTC

- reply

Damn you have a lot of money. Well that or it's a hell of a lot cheaper wherever you are.

2025-02-27 21:54:04 UTC

- reply

> If it's possible for Trump's fascism to fail because of lack of popularity, it will be because of the remnants of conservatism, not because of the fascist part.
Aren't violent uprisings also the manifestation of lack of popularity?
I'd say that every failed fascist regime was the result of unpopularity at some level or another.

2025-02-27 21:52:39 UTC

- reply

Less meat is cheaper though. More money for computer parts/girls artwork merch/etc.
But yeah, trying to optimize my food expenses is why I stopped buying some types entirely and added a lot more beans & legumes.

2025-02-27 10:22:40 UTC

Any data is critical until the user tells you that it is not.
You as the programmer do not get to decide that.

2025-02-27 02:49:04 UTC

- reply

I'm somewhat more used to that since that's the default with pretty much any distributed and/or massively parallel program.
New abstraction levels are required at that point, both up and down the chain.
Conceptually lazy resolution of a value isn't all that different from actor-wise parallelism & messaging.

2025-02-27 02:31:22 UTC

- reply

Yeah it can get in the way, I prefer when languages provide the option for it but don't make it mandatory.
Enforcement could be done via a macro-backed DSL.
Clojure in particular tried to peg itself into a very specific role without being particularly permissive of others and its fallbacks are less than elegant or pleasant to use.
Some say it couldn't have been implemented as a Common Lisp library for ABCL like I say it could, but I think that Coalton is a good counterargument to that.

2025-02-27 02:23:55 UTC

- reply

Haskell or Scheme?

2025-02-25 03:39:21 UTC

- reply

A lot of Guix and Common Lisp software is configured with code that is fully subject to the usual error-handling of the language.
It is a lot more common with Common Lisp to interact with a program and reconfigure it or modify it while it is running than with other languages. This is a thing dynamic languages are kind of built for. Common Lisp was actively designed to enable that.

2025-02-24 00:42:26 UTC

- reply

I think the post got orphaned (redrafted: https://udongein.xyz/objects/f4854aa1-8fbc-45db-bae6-7f8d66a674ff).
But yeah, the rest of the world generally gets it a lot worse than USA for buying tech.
Malicious I'm not so sure, but fake (those scams do tend to get noticed and removed), brittle or generally unreliable is a common problem for cheap SSDs & flash memory.
Mind, I have it much better than Australia does.

2025-02-24 00:39:07 UTC

- reply

Technically it is already fully feasible with the use of Apparmor or SELinux in conjunction with Guix right now.
It just isn't done automatically.

2025-02-24 00:26:56 UTC

- reply

In USA with American dollars.
That doesn't exist here except for mystery SSDs from mostly-unknown manufacturers.

2025-02-24 00:22:44 UTC

- reply

More importantly, every different version of those libraries wastes memory independently, and on most users' machines memory is at a far more costly premium than storage.
Guix with some namespace management or sandboxing would thus be superior to flatpak, because it would still permit shared memor & shared copy-on-write libraries.

2025-02-23 05:43:06 UTC

- reply

Succint explanation of why I'm glad we have water sanitation.
Cholera is not nice.

2025-02-22 08:48:06 UTC

- reply

Huh, that's some unexpected and inordinate instability.

2025-02-22 08:27:05 UTC

- reply

You restart your Emacs?

2025-02-22 08:22:28 UTC

- reply

Is it?
I got more into computers from fiction that basically had anticorporatism, anticapitalism and hacktivism (and a lot of other types too) at their core. (So yeah, cyberpunk & post-cyberpunk mostly.)
The notion of getting a job was only ever to pay the bills, I didn't trust corporate work to ever make a positive mark on the world. I planned on having to make up for that in my free time.
As for luddites, a fun quote of @npub1fdr…lvhs on luddites "Person who goes beyond asking what technology does, to asking who it does it for and who it does it to." (https://pluralistic.net/2023/02/07/full-stack-luddites/#subsidiarity) (ahistorical technophobia is missing the point)

2025-02-21 03:21:47 UTC

- reply

Not in the name of the company having gone batshit?

2025-02-20 01:25:11 UTC

- reply

Oh no, they're not at current time.
I'm saying that they should be, because there is no practical way to trust the manufacturers of mostly-unchangeable CPUs not to be malicious.
So being able to functionally reprogram and remake them as needed gets around that (because micromanufacturing of CPUs at home is not happening anytime soon). But that only creates a new trust & bootstrapping problem, to which I propose the answer in OP.

2025-02-20 01:08:06 UTC

- reply

Don't have the storage for it?
If it is bandwidth that is an issue, it could be an option to just put the older video media on torrents.

2025-02-19 22:17:11 UTC

But if all (most) CPUs are FGPAs, how does one bootstrap and assume the payload is not malicious?
Programmers using discrete electronics and punched tape would be a human-auditable (if tedious) way of bootstrapping.
A minimum viable target & programs for it to bootstrap everything else would be needed.
I consider this analogous to the #Guix bootstrap seed endeavor.
#FGPA #Bootstrap #Bootstrapping #CPU #Hardware #Security #Auditability

2025-02-18 03:39:32 UTC

- reply

Aren't the three most popular Schemes Racket, Guile and Chicken, or two fairly pragmatic about including the kitchen sink (or having it as a library) and one that aims to make using C libraries a primary use-case?
On the pragmatic level, they seem pretty good. On the functional purity one can still /use them/ that way but one doesn't have to and most libraries don't.

2025-02-17 20:51:26 UTC

- reply

One would be losing a fortune in laundromats, among other inconveniences.

2025-02-16 04:41:50 UTC

- reply

I think most of the issues you're encountering though is that Ubuntu doesn't qualify in the "stable and no problems" distros since over a decade. It enshittified.
Lookup the hiring practices problems of Canonical to see where the rot starts. Bunch of ableist and plain bizarre shit.

2025-02-16 04:28:44 UTC

- reply

What even is that? They don't provide SHA256SUM & SHA256SUM.gpg files like Debian? Why would verification require storage at all?

2025-02-16 04:16:42 UTC

- reply

Weird package maintainer.
This should still be feasible in basically any package manager in some way or another.

2025-02-16 01:10:17 UTC

- reply

I am extremely interested in this, do you have a link?

2025-02-15 23:32:58 UTC

- reply

Chez is most likely superior.
Kawa is within "native Java" performance range, not hardware native.

2025-02-15 23:22:29 UTC

- reply

There's also Kawa Scheme, if something performance-sensitive is needed.

2025-02-15 23:03:15 UTC

- reply

It's still seeded just fine on Postman.i2p, and one can put in a request in the wishlist should that change.

2025-02-14 21:48:39 UTC

- reply

Anyone getting that close to one is annoying.
I'd be about as annoyed by someone measuring my torso or legs. (Well the latter might amuse me slightly, if it looks inconvenient.)

2025-02-13 20:48:24 UTC

- reply

Seems like they never even considered the "what if you're sitting right next to a fascist country" scneario.
Bunch of idiots. There's a reason Switzerland avoided invasion and it's not because they were being peaceniks.
> And I don't believe the NATO police would die for our NW Territories, if Russian icebreakers showed up on the Arctic shore to start a new lithium mine.
Indeed.
> If there are no police, you actually, I'm sorry, DO need a gun in the house.
Arguably even or especially if there is police, if that police serves fascists. Being harmless doesn't keep them in line or prevent escalation, it just means they're the only side enabled to escalate and they do.

2025-02-13 20:35:28 UTC

- reply

I'm kind of appalled by the notion Canada has no nukes.
I would hope that at least some blacksites avoided losing their stocks of countermeasures & deterrents.

2025-02-11 11:04:11 UTC

- reply

Is it even literal or some inexplicable figurative meaning?

2025-02-09 08:29:17 UTC

- reply

On the other hand, if the memory is failing I really don't think relying on it for anything is a good idea.
At least, not without replacing the memory (have fun with that :/) or the motherboard (probably easier, actually).

2025-02-09 06:12:39 UTC

- reply

A good coat will handle that.
(Mild joking aside, yeah, a bit more preparation is necessary but cold places are nice and the political portion here is good too.)

2025-02-08 23:35:02 UTC

- reply

This is a significant part of why the standard being frozen isn't all that much of a problem.
It's possible to fix the language with tools within the language.
Only a few things like package-local nicknames require non-standard extension. There are too few extensions to have formalized them into something like a Lisp equivalent of SRFI's.
I distinguish "non-standard extension" as things which cannot be portably implemented with only information from the standard. PLNs require hooking into implementation-defined internals to change how symbols are looked up. There are however libraries available to paper over the minor differences in implementing that.
Basically every implementation that still lives supports PLNs.

2025-02-08 20:53:41 UTC

- reply

Possibly a reason/excuse to mess with other high-availability setups?

2025-02-08 11:27:28 UTC

- reply

That's consumer.
Tron's remains the correct take on the meaning of users.

2025-02-08 04:02:32 UTC

- reply

Cost perhaps.
Power isn't true. Power over fiber is a thing (https://en.wikipedia.org/wiki/Power-over-fiber), it uses wavelength-specific photovoltaics which have higher efficiency than wideband receptors. The military uses those, for example.

2025-02-08 03:58:33 UTC

- reply

That begs the question though: Why not switch to optical signaling?
The crosstalk issue is eliminated and the usable frequencies are *significantly* higher than those allowed by copper.

2025-02-07 20:56:53 UTC

- reply

Streaming data in Elisp is a bit awkward as most of the common APIs are based on batch whole-file access, which with any large dataset will crash Emacs.
Scheme and Common Lisp have no such issues, of course.

2025-02-07 02:37:55 UTC

- reply

Perhaps they remained at that rather than disappear entirely because they helped some vital purpose?

2025-02-06 23:44:40 UTC

- reply

> "Can't we use the information superhighway to deliver our organically grown vegetables?" - "No."
Arguably we do, considering how much the internet is used in the logistics chain.
> "Maybe in five years, we can travel the information superhighway instead of flying." - "No."
We should. Flying should be avoided whenever it is possible.
A train or a boat is okay. The internet is usually used for booking, though not exclusively.
Virtual visits are even better.
> "I think people need the information highway more than food or shelter." - "No, they definitely need food and shelter more."
They're pretty complementary these days. Can't really get a job without using the internet anymore (other than a few exceptions, but most groceries don't even take paper applications anymore). Most shelter will also use the internet for initial acquisition.

2025-02-06 20:52:31 UTC

- reply

What is?
I think other than the hints, all of those are available through other JVM tools...

2025-02-06 20:42:25 UTC

- reply

> There's not a significant amount of lithium metal in a lithium ion battery.
Lithium hexafluorophosphate is used for the electrolyte part though.
And a contained fire isn't a problem. Uncontained fluoride fumes are a problem.
It probably doesn't respond to water like I suspected it might negatively do, but since it is potentially self-oxidizing I doubt the water will do anything. At that point since the fuel and oxidizer components are unremovable, the only way to actually stop the redox reaction before it completes is to remove the heat.
So unless one has a cryogenic refrigeration setup handy, the only option is containment in an appropriate pressure vessel as stopping it isn't happening.
(To me these difficulties in containment & fire prevention are more than argument enough that lithium-ion batteries should never have been allowed public sale.)

2025-02-06 06:43:26 UTC

- reply

> My plan is to use an open top metal container and just flood it until the battery burns out, should that undesirable possibility occur.
Isn't lithium an alkaline group metal? Those catch fire or explode in contact with water. Or do you keep liquid nitrogen handy (also a fume hazard, but purely because of oxygen displacement, a rebreather is all that's needed to make that safe)?

2025-02-06 06:27:43 UTC

- reply

The main issue, to me is that the fumes are toxic in both cases. Fluoride gas is a significant component of them.
Just venting that is, also, undesirable.
The better solution, I think, would be for the sale of high-pressure fireproof containers to become more normalized. Such that one *can* in fact set a lithium UPS battery on fire inside, seal it closed and not have any problems.
Obviously the container would be hazardous waste at that point, ideally it should have some sort of high-pressure seal latch/lock so that it cannot be opened after a high-pressure/high-temperature incident.
At that point one could call in public/commercial/etc services for removal of contained hazardous waste.

2025-02-05 10:40:44 UTC

- reply

Ah, now that's an issue I had been hoping wasn't the case.
Probably why those bags (that don't hold pressure) are used.

2025-02-05 01:28:51 UTC

- reply

Most consumers don't care because they have been shackled to server-centric proprietary malware and SaaSS where top-down control by malicious corposcum & their ilk is the norm.
And indeed, I note the distinction between user/peer and consumer (pejorative). It is entirely on purpose to maintain dependency on the server-centric schema and the abuse of users (and thus the status of much of it as software which abuses users, malware; seeing the users so reduced to mere helpless consumers) it enables.
I believe GPON being rarely symmetric is a choice rather than a technology requirement, since it has support for symmetric use as a set of standards.
I would also like a lift in upload.

2025-02-05 01:09:48 UTC

- reply

Ah yes, the usual "Australia gets fucked, always, no exceptions" scenario.
It's quite vexing, especially since asymmetric completely defeats the point of fiber-to-the-premises to start with.

2025-02-05 00:49:51 UTC

- reply

I'd honestly take 300/300 over 1000/50.
Asymmetric just sucks so much for p2p stuff.

2025-02-03 02:42:44 UTC

- reply

> Specialized equipment, like servers with redundant PSUs, are generally installed by people that "know what they're doing" and as such are not subject to that rule.
Yeah this is what was confusing me with the whole matter. Why it was fine for servers but not the rest.
> would plug them into the same circuit
If it says split circuit or split phase one obviously has to follow the instructions.
> If they're not isolated, the exposed metal bits on the unplugged end can rise to 120v if only one is plugged in.
That would be bad. One would hope such faulty equipment would fail safety inspection.
Unless you meant some ignoramus of a user deciding that it ought to work, without adding any of the necessary intermediary equipment to actually make that work.
> The daisy chaining is perfectly safe: there's a two-screw "bus bar" on each side of the device that handles all the 'passthrough' current.
I somewhat see... (will look it up later, probably). I was also thinking of how that interacts with devices that involve a DC step, like computers & such.

2025-02-03 01:11:52 UTC

- reply

> A single load using multiple plugs is generally "frowned upon" for very good reasons
Why is that, anyway? Adding fuses or breakers would handle partial failure just fine.
> so the only valid reason to have a 20A circuit feeding 15A outlets is simply to have more 15A outlets, or multiple <15A high-draw loads on the same circuit
That is what the electrical code intends, yeah.
> The 'daisy chain' configuration I was talking about is for the in-wall wiring: power from the breaker goes to one device, then it's "daisy chained" through that device to the next one.
That sounds like it could have undesirable side-effects for anything other than a UPS or power bar.
> Also, the only case where you can't use a 15A outlet on a 20A circuit is when you only have a single outlet
Why is that?

2025-02-02 23:57:53 UTC

- reply

> And for the mandatory 20A circuits in the kitchen? 15A outlets "because fuck you that's why."
I don't think that's even legal. I'm not sure who has liability in this case.
How much power does your desktop draw? (And how are you powering it now?)
Also an alternative option: Adapters from the oven/range plug to the proper circuit you need.
Of course this is somewhat inconvenient if you actually /use/ the range, but if you don't it's perfectly viable.
> My only option for a UPS is one that either does some hinky shit to charge from a separate circuit than the load is on, or stringing a big fat 10-gauge extension cord across my shoebox from the kitchen after bodging a 5-15 make-a-cord on the wall end.
I'm not quite sure I understand what you mean by this part but it sounds neat, could you expand on the last part?

2025-02-01 04:38:02 UTC

- reply

Perhaps switching to ESR might be a good idea.