WTF Bites

Bulb

Czech (and the latter are quite prone to jump into random online discussion claiming that).

Czech definitely is not very phonetic. Well, much more than English for sure, but not compared to other Slavic languages. Assimilation (where a consonant changes whether it is voiced or non-voiced depending on the next one or sometimes the preceding one) is really common, to the point that prefixes s- and z- and prepositions s, z, se and ze have their pronunciation totally mixed up and it's one of the hardest parts of the grammar for students. And a bunch of other combinations that get tweaked to be easier to pronounce, but keep the spelling based on etymology. Plus we tend to keep original spelling of loan words, so there is pleeeeenty of exceptions.

I would say that German, Italian or Spanish are no less phonetic than Czech.

Bulb

@Gąska said in WTF Bites:

Half of Cyrillic alphabet is stroke-for-stroke identical to Latin alphabet, including having the same lowercase and uppercase, and yet each of them gets unique code point. If Cyrillic can do it, there's no reason why Turkish can't do it. And Turkish has more reason to do it than Cyrillic.

It's the backward compatidebility . It's been explained a couple of times already.

Gąska

@Bulb backward compatibility is one thing. But this is pseudo-compatibility that doesn't actually increase compatibility with anything in any way whatsoever and only makes everyone's lives more miserable for no reason. I've explained it a couple of times already too.

People shouting "BACKWARD COMPATIBILITY" and ending the discussion immediately without actually checking what needs to be compatible with what is exactly how we've got into this mess in the first place. As I said before, there is absolutely no reason, even compatibility-wise, why Turkish i and Latin i must have the same code point.

Bulb

@Gąska The requirements of Unicode regarding backward compatibility are stated quite clearly.

For Cyrillic, since the encodings used in Russia have ASCII in the lower half and complete Cyrillic including all the letters that look exactly like the Latin ones, Unicode needed a complete Cyrillic block, otherwise distinct codes in the original encoding could not be always kept distinct in Unicode per the requirement.

For Turkish, the encoding used in Turkey only had the ı and İ. And also just simply ASCII in the lower half. They could have chosen to use separate codepoints for the i and I when transforming from that encoding, but because the lower half is still ASCII, a bunch of ASCII-not-actually-Turkish texts would get transcoded with the Turkish i. Given choice of two bad options, they chose the one they chose, because it was more obvious.

Gąska

@Bulb said in WTF Bites:

Given choice of two bad options, they chose the one they chose, because it was more obvious.

And that's exactly the problem. Instead of going for "more obvious", they should've actually checked which one is ~~better~~less bad.

blek

I just got this notification:

Screenshot from 2021-01-23 02-51-53.png

It's a desktop...

HardwareGeek

@blek I used to get battery status indications when I had my UPS connected, before it bit the dust. (Anybody know where to buy a UPS that actually has them? Fry's only has expensive rack-mounted ones, and I'm not sure they actually have even those. The normal, non-rack-mounted ones on their website are not available in-store, for delivery, or for shipping.)

cvi

@blek Obligatory ry: No battery falls into the category of low battery. So points for being technically correct. (After all, if you unplug the desktop, it will shut down due to lack of battery power.)

Tsaukpaetra

@HardwareGeek said in WTF Bites:

Anybody know where to buy a UPS that actually has them?

https://www.amazon.com/APC-Battery-Protector-BackUPS-BX1500M/dp/B06VY6FXMM

Amazon.com

Amazon.com: Vertiv Liebert PSA5 UPS-500VA/300W 120V, Standby Power, 8 Outlet, Network Protection, 3 Year Warranty, Shutdown software, Uninterruptible Power Supply, Battery Backup, Surge Protection (PSA5-500MT120) : Electronics

Amazon.com

All of these desktop-style units support status signalling via USB, and should be somewhat commonly available. With the exception of the first one, which requires a stupid adapter cable...

Zecc

@error said in WTF Bites:

Warning, the following zero files will be overwritten!

git nevers addresses me by name.

sloosecannon

@HardwareGeek said in WTF Bites:

@blek I used to get battery status indications when I had my UPS connected, before it bit the dust. (Anybody know where to buy a UPS that actually has them? Fry's only has expensive rack-mounted ones, and I'm not sure they actually have even those. The normal, non-rack-mounted ones on their website are not available in-store, for delivery, or for shipping.)

MicroCenter has em, if you're near one

izzion

@HardwareGeek said in WTF Bites:

@blek I used to get battery status indications when I had my UPS connected, before it bit the dust. (Anybody know where to buy a UPS that actually has them? Fry's only has expensive rack-mounted ones, and I'm not sure they actually have even those. The normal, non-rack-mounted ones on their website are not available in-store, for delivery, or for shipping.)

Last couple times I've needed one, Best Buy stocked several decent desktop/home models.

HardwareGeek

@sloosecannon said in WTF Bites:

@HardwareGeek said in WTF Bites:

@blek I used to get battery status indications when I had my UPS connected, before it bit the dust. (Anybody know where to buy a UPS that actually has them? Fry's only has expensive rack-mounted ones, and I'm not sure they actually have even those. The normal, non-rack-mounted ones on their website are not available in-store, for delivery, or for shipping.)

MicroCenter has em, if you're near one

Thanks, but the nearest one is probably 2-3 hours away. However, 46 of the 49 they have in their online catalog are available to be shipped.

sloosecannon

@HardwareGeek said in WTF Bites:

@sloosecannon said in WTF Bites:

@HardwareGeek said in WTF Bites:

@blek I used to get battery status indications when I had my UPS connected, before it bit the dust. (Anybody know where to buy a UPS that actually has them? Fry's only has expensive rack-mounted ones, and I'm not sure they actually have even those. The normal, non-rack-mounted ones on their website are not available in-store, for delivery, or for shipping.)

MicroCenter has em, if you're near one

Thanks, but the nearest one is probably 2-3 hours away. However, 46 of the 49 they have in their online catalog are available to be shipped.

holy crap, that's a first. MicroCenter actually having stock in the online store? Damn

HardwareGeek

@sloosecannon Well, it says they're available. Whether they actually have them, . But it let me place the order. Bought some cable management stuff, too, so I can get a multitude of wires under my desk instead of on the top.

sloosecannon

@HardwareGeek said in WTF Bites:

@sloosecannon Well, it says they're available. Whether they actually have them, . But it let me place the order. Bought some cable management stuff, too, so I can get a multitude of wires under my desk instead of on the top.

I mean, they've probably got em. Unless it's like 1 in stock. Probably sitting in the warehouse in Hilliard, OH somewhere........
I'd be willing to bet they've stepped up online ops since the whole pandemic thing happened...

LaoC

@Gąska said in WTF Bites:

@LaoC said in WTF Bites:

Except it being the exact same character of course.

It has a different uppercase form.

You mean lowercase? Because that is an uppercase I.

It may be "the same", but it absolutely isn't exactly the same.

All the ISO-8859 charsets have ASCII in their lower half. An ASCII I is an ASCII I, there is no difference. Text being in the Turkish language that prescribes peculiar case rules is a different thing from text being encoded in ISO-8859-9. You could perfectly well write English, declare it ISO-8859-9, and trust your i to be the lowercase of I.

If it were, there would be no problem in treating it the same. But there is.

… iff the text is in Turkish. Which it may or may not.

Half of Cyrillic alphabet is stroke-for-stroke identical to Latin alphabet, including having the same lowercase and uppercase,

[AEOPCX] as far as I can see - any others? That's a tad short of half.

and yet each of them gets unique code point. If Cyrillic can do it, there's no reason why Turkish can't do it. And Turkish has more reason to do it than Cyrillic.

If they had done that when the ISO-8851 family was designed, there would be no problem; Turkish users would have used their special forms and they would be mapped 1:1 to different Unicode code points with clear and locale-independent case rules. The ISO-8859-5 users did just that - that code page has all the Latin plus all the Cyrillic characters separated, even where they look identical, so you could be unambiguous wrt your intended script.

El_Heffe

@TimeBandit said in WTF Bites:

@Atazhaia said in WTF Bites:

And while I love having a MacBook as a personal laptop, having one for my work is... less optimal. I very much prefer Linux there.

Good news !

Joey Sneddon / Jan 20, 2021

Ubuntu is 'Completely Usable' on Apple Silicon

Ubuntu Linux is up and running on Apple Silicon thanks to developers working at Coreillium. They now hope to upstream their work on Linux for M1.

Is Ubuntu usable on Apple's new M1?

Yes.

That's weird, it wasn't usable before.

hungrier

@HardwareGeek said in WTF Bites:

@blek I used to get battery status indications when I had my UPS connected, before it bit the dust. (Anybody know where to buy a UPS that actually has them? Fry's only has expensive rack-mounted ones, and I'm not sure they actually have even those. The normal, non-rack-mounted ones on their website are not available in-store, for delivery, or for shipping.)

I got a normal, consumer grade one off Amazon a few months back without any issue, and looking at the listing now it's available, on Prime same as ever. Is this one of the rare occasions where the US Amazon store doesn't have something that's available somewhere else?

e: In case you want the particular one I got for whatever reason, it's this one

dkf

@remi said in WTF Bites:

"congers"

Like this one?

dkf

@El_Heffe said in WTF Bites:

That's weird, it wasn't usable before.

One of my colleagues uses Ubuntu on his main work system (which happens to be a Dell laptop). He finds it usable. He also has Windows on it for games apparently.

Tsaukpaetra

@hungrier said in WTF Bites:

e: In case you want the particular one I got for whatever reason, it's this one

Got that one for the small networking cabinet, it's one of the few that's slim while still keeping the outlets upright (as opposed to coming out the sides)

Zerosquare

@blek said in WTF Bites:

I just got this notification:

It's a desktop...

Maybe the CMOS battery is low?

( actually, the state of that battery is software-readable, but I've never heard of an OS actually warning you about it.)

Tsaukpaetra

@Zerosquare said in WTF Bites:

@blek said in WTF Bites:

I just got this notification:

It's a desktop...

Maybe the CMOS battery is low?

( actually, the state of that battery is software-readable, but I've never heard of an OS actually warning you about it.)

The BIOS normally should, but since configuration is nominally saved in flash as a virtual CMOS this isn't much of a problem anymore...

aitap

Here's another " is me" bite:

I finally got around to finding a copy of Harry Potter and Deathly Weapons and deciding to watch it, but quickly found out that I cannot play it on my machine. For almost all of my machines still have Debian Stretch (that is, oldstable) installed, and there is no libaov there, or anything else to decode AV-1, apparently.

Feeling too to upgrade to Buster (that is, stable) right now, I decide to try building the blasted piece of code from source. One cmake .. && make -j$(nproc) later, it even gives me an aomdec binary, which reads the mkv file and outputs a raw video stream that ffmpeg understands. Success, right? I just need to encode it into something I can play and supply the audio track.

clickety-click ./aomdec hpdw.mkv - | ffmpeg -i - -i hpdw.mkv -c:a copy -map 0:v:0 -map 1:a:0 hpdw-encoded.mkv, and IT'S ALIIIIIIVE!!!! I start watching the file as it's being written by ffmpeg, get confused about the stutters that seem to happen more and more frequently, then realise that aomdec ate all my RAM and most of the swap, too.

Okay, maybe aomdec is too simplistic for what I'm asking of it, or does some kind of buffering the wrong way when connected to ffmpeg via a pipe, and I didn't even expect it to work in the first place. I actually have a Buster machine elsewhere, let's send the file there and transcode it using a new enough ffmpeg, then everything will just work, right?

clickety click rsync clickety ssh clickety ffmpeg -i hpdw.mkv -c:a copy out.mkv tap tap

Wait. Did the remote machine just hang? No, it's responding again. What do you mean, Killed? What do you mean, Out of memory: Kill process (ffmpeg) or sacrifice child? Apparently, something causes libaom to allocate gobs and gobs of memory after around 8 minutes worth of Harry Potter and Deathly Weapons, no matter what.

And this is why I'm currently transcoding the film in 5 minute-sized chunks to join them later.

BernieTheBernie

@Luhmann said in WTF Bites:

@error said in WTF Bites:

Java

Like the coffee?

You know what could happen to you while you suffer from severe hypocaffeineaemia?
Dilbert can teach you:

(https://dilbert.com/strip/1998-6-7)

topspin

@aitap can’t you just download a vlc binary?

aitap

@topspin There are downsides to central package management, Debian-way. I'd have to upgrade to Buster or install snap to get a new enough VLC. At least I don't see any statically linked binaries for Linux on their website.

Gąska

@LaoC said in WTF Bites:

@Gąska said in WTF Bites:

@LaoC said in WTF Bites:

Except it being the exact same character of course.

It has a different uppercase form.

You mean lowercase? Because that is an uppercase I.

Whatever. The point is, the uppercase-lowercase mapping is different.

It may be "the same", but it absolutely isn't exactly the same.

All the ISO-8859 charsets have ASCII in their lower half. An ASCII I is an ASCII I, there is no difference. Text being in the Turkish language that prescribes peculiar case rules is a different thing from text being encoded in ISO-8859-9.

Are uppercase-lowercase pairs part of the alphabet or part of the specific language? Is there any other language that uses Turkish alphabet? If there is, does letter i correspond to I or İ?

If it's part of the alphabet, I don't see why it should be locale-dependent. Alphabets aren't locale-dependent. They're simply used or not used in a given locale.

You could perfectly well write English, declare it ISO-8859-9, and trust your i to be the lowercase of I.

Citation needed. I'm 99% sure that back in those days, uppercase-lowercase mappings were tied to specific encodings, and in ISO-8859-9 the uppercase of i was always İ. By which I mean, if you took ISO-8859-9 text and told the text editor to make it all caps, then every i would be replaced with İ, not I.

But maybe I'm wrong. Hence, citation needed.

If it were, there would be no problem in treating it the same. But there is.

… iff the text is in Turkish. Which it may or may not.

My point exactly. You need to know whether a given i is Turkish or not to know what you should do with it (programmatically). Whenever two symbols need to be distinguished from one another, they should be encoded differently so the difference isn't lost in transmission. But there's no way to do that. It's not even just about the code point. Unicode doesn't allow encoding this difference at all. Mixed English-Turkish text will always, always be broken.

Half of Cyrillic alphabet is stroke-for-stroke identical to Latin alphabet, including having the same lowercase and uppercase,

[AEOPCX] as far as I can see - any others? That's a tad short of half.

A, B, E, K, M, H, O, P, C, T, Y and X.

and yet each of them gets unique code point. If Cyrillic can do it, there's no reason why Turkish can't do it. And Turkish has more reason to do it than Cyrillic.

If they had done that when the ISO-8851 family was designed, there would be no problem

But if they had done it with Unicode regardless of how ISO-8851 does it, there wouldn't be any problem either.

Tsaukpaetra

@aitap said in WTF Bites:

to get a new enough VLC.

Ah, yeah, apparently that was added in 3.0, and you'd have to be on stretch or higher for that...

loopback0

@Tsaukpaetra said in WTF Bites:

@aitap said in WTF Bites:

to get a new enough VLC.

Ah, yeah, apparently that was added in 3.0, and you'd have to be on stretch or higher for that...

@aitap said in WTF Bites:

For almost all of my machines still have Debian Stretch (that is, oldstable) installed

topspin

Seriously YouTube, I do not speak Korean.

Tsaukpaetra

@topspin said in WTF Bites:

Seriously YouTube, I do not speak Korean.

It's been serving me spanish ads for half a year now, and I don't really mind.

Gąska

@Tsaukpaetra Korean is far less common in Germany than Spanish is in USA.

Zerosquare

All that money spent and data sucked to build profiles to target users, and they can't even get such a simple thing right

aitap

@Gąska said in WTF Bites:

A, B, E, K, M, H, O, P, C, T, Y and X.

Sorry for the , but Y != У. I do agree that the rest of the capitals are stroke-for-stroke identical between Latin and Cyrillic and that lowercase y is stroke-for-stroke identical to у.

PleegWat

@Gąska said in WTF Bites:

Are uppercase-lowercase pairs part of the alphabet or part of the specific language? Is there any other language that uses Turkish alphabet? If there is, does letter i correspond to I or İ?

If the uppercase letters are used to write Latin, they don't have a lowercase variant at all.

dcon

@aitap said in WTF Bites:

Feeling too to upgrade to Buster (that is, stable) right now

After all the other stuff, seems this would be easier...

dcon

@topspin said in WTF Bites:

Seriously YouTube, I do not speak Korean.

: Default language is not English, currentLanguage = SetLanguage(rand())

LaoC

@Gąska said in WTF Bites:

@LaoC said in WTF Bites:

@Gąska said in WTF Bites:

@LaoC said in WTF Bites:

Except it being the exact same character of course.

It has a different uppercase form.

You mean lowercase? Because that is an uppercase I.

Whatever. The point is, the uppercase-lowercase mapping is different.

It may be "the same", but it absolutely isn't exactly the same.

All the ISO-8859 charsets have ASCII in their lower half. An ASCII I is an ASCII I, there is no difference. Text being in the Turkish language that prescribes peculiar case rules is a different thing from text being encoded in ISO-8859-9.

Are uppercase-lowercase pairs part of the alphabet or part of the specific language?

If by "alphabet" you mean the alphabet used by a particular language, then the two things are basically synonymous. If instead you mean "encoding", it can only be the language. Unicode defines a standard casemap attribute that is part of the encoding, but also a list of special cases where rules changes by locale.
There may even be multiple correct-as-per-language-rules capitalizations for a single letter

Is there any other language that uses Turkish alphabet? If there is, does letter i correspond to I or İ?

If it's part of the alphabet, I don't see why it should be locale-dependent. Alphabets aren't locale-dependent. They're simply used or not used in a given locale.

You could perfectly well write English, declare it ISO-8859-9, and trust your i to be the lowercase of I.

Citation needed.

I thought it was common knowledge that all the 8859-n charsets are identical in the 7-bit range. But here you go, ECMA 128 page 11, "language coverage"

I'm 99% sure that back in those days, uppercase-lowercase mappings were tied to specific encodings, and in ISO-8859-9 the uppercase of i was always İ. By which I mean, if you took ISO-8859-9 text and told the text editor to make it all caps, then every i would be replaced with İ, not I.

No doubt software existed that made such assumptions. That doesn't mean it was right to do so.

If it were, there would be no problem in treating it the same. But there is.

… iff the text is in Turkish. Which it may or may not.

My point exactly. You need to know whether a given i is Turkish or not to know what you should do with it (programmatically). Whenever two symbols need to be distinguished from one another, they should be encoded differently so the difference isn't lost in transmission.

Ideally yes. That wasn't the case in Latin-5 though, so they could either have broken compatibility or this rule.

But there's no way to do that. It's not even just about the code point. Unicode doesn't allow encoding this difference at all. Mixed English-Turkish text will always, always be broken.

No. You can use COMBINING DOT ABOVE to get i to capitalize to İ even under non-Turkish rules.

Half of Cyrillic alphabet is stroke-for-stroke identical to Latin alphabet, including having the same lowercase and uppercase,

[AEOPCX] as far as I can see - any others? That's a tad short of half.

A, B, E, K, M, H, O, P, C, T, Y and X.

The lowercase of B, K, M, H and T looks like a small version of the uppercase and different from the Latin equivalent, and У is basically a bigger version of the Latin lowercase and different from Y.

If they had done that when the ISO-8851 family was designed, there would be no problem

But if they had done it with Unicode regardless of how ISO-8851 does it, there wouldn't be any problem either.

Yes there would.

Gąska

@LaoC said in WTF Bites:

I'm 99% sure that back in those days, uppercase-lowercase mappings were tied to specific encodings, and in ISO-8859-9 the uppercase of i was always İ. By which I mean, if you took ISO-8859-9 text and told the text editor to make it all caps, then every i would be replaced with İ, not I.

No doubt software existed that made such assumptions. That doesn't mean it was right to do so.

But was it WRONG to do so? If yes, citation needed.

If it were, there would be no problem in treating it the same. But there is.

… iff the text is in Turkish. Which it may or may not.

My point exactly. You need to know whether a given i is Turkish or not to know what you should do with it (programmatically). Whenever two symbols need to be distinguished from one another, they should be encoded differently so the difference isn't lost in transmission.

Ideally yes. That wasn't the case in Latin-5 though, so they could either have broken compatibility or this rule.

And their cargo cult of "compatibility" above all (because there were no practical problems that the change would have caused; nothing in real life would have broken if they made the change) is exactly the problem.

But there's no way to do that. It's not even just about the code point. Unicode doesn't allow encoding this difference at all. Mixed English-Turkish text will always, always be broken.

No. You can use COMBINING DOT ABOVE to get i to capitalize to İ even under non-Turkish rules.

Okay and how to make I "lowercase" into ı unconditionally? Or into i even in Turkish locale, for that matter?

Half of Cyrillic alphabet is stroke-for-stroke identical to Latin alphabet, including having the same lowercase and uppercase,

[AEOPCX] as far as I can see - any others? That's a tad short of half.

A, B, E, K, M, H, O, P, C, T, Y and X.

The lowercase of B, K, M, H and T looks like a small version of the uppercase and different from the Latin equivalent, and У is basically a bigger version of the Latin lowercase and different from Y.

Okay fair enough. Forgot about that.

If they had done that when the ISO-8851 family was designed, there would be no problem

But if they had done it with Unicode regardless of how ISO-8851 does it, there wouldn't be any problem either.

Yes there would.

What problem? Show me an exact scenario where some program would do the wrong thing because of this change when now it does the right thing.

Tsaukpaetra

WTF are we still talking about Unicode? Goddam people threads are free!

Gąska

@Tsaukpaetra it's such a petty topic that neither of us wants to look like we care, so we cannot make a dedicated topic, but neither of us wants to quit because it would make it look like we were wrong.

TimeBandit

Iv Choe

Apple Daily - Celebrity Buzz

Stay updated with Apple Daily, your go-to source for the latest in celebrities, technology, global news, and entertainment.

loopback0

@Gąska said in WTF Bites:

@Tsaukpaetra it's such a petty topic that neither of us wants to look like we care, so we cannot make a dedicated topic, but neither of us wants to quit because it would make it look like we were wrong.

Can we declare you both wrong and all move on?

Applied Mediocrity

@TimeBandit The fix is better than the

Authorities fixed the issue by installing a pirated version of Flash at 4:30 a.m. the following day.

cvi

@Gąska said in WTF Bites:

but neither of us wants to quit because ~~it would make it look like we were wrong.~~quitting instead of repeating ourselves ad nauseam suspiciously looks like a smart move.

My POV.