Posts by DuncanLarge 1026 publicly visible posts • joined 10 Apr 2017
> That's exactly what went wrong at Fukushima - cost cutting corners
Are you confusing Fukashima for Chernobyl?
Fukashima was a state of the art 70's reactor design shortly to be retired for is newer brethren on the same site.
Poor design however failed to protect the cooling system from extreme flooding.
Funny that when we run for the hill because of "scary radiation" the local flora and fauna dont seem to have a problem with such low radiation levels.
Funny that...
I'm often confused why we are allowed to fly in planes. You get more radiation from a flight than you would if you lived for a year at Fukashima or Chernobyl, excluding living in the reactor itself obviously.
> When I gave up I was competing with 24 EiB of network space or 28,000,000,000,000,000,000 bytes.
Such a waste. The guy who created it only considered the fact that not much power is used during the farming as the only metric.
What about the power etc used during the mining, manufacturing and transportation of the drives?
> CTRL ALT DEL just instantly reboot in those days
The three fingered salute would instantly reboot a DOS machine, or if logging into an NT machine all the way up to win 10 it was the way to access the login fields, getting past the lockscreen.
I seem to remember that part of its function for an NT login was to have the login scree (or something behind it) restart each time a user logged in to help prevent user details being hijacked by ensuring that the correct login prompt was running and not a fake one. Not sure if thats true.
On a Linux system the salute usually initiates a reboot (when in a console).
I use both, ctrl-alt-del and right clicking on the task bar.
What I no longer do is use the start menu. Bloody thing barley works these days. I mean when you install a program only for it to not appear in the start menu thus needing you to search, or even NAVIGATE to programfiles well you get used to using the search box.
Even if that damn thing is also broken, where it will search the ****ing internet for notepad.exe because its still starting up in the background and hasnt realised that there is stuff installed locally yet.
> What is God's name are they doing specifying that children can't use the software?
Data collection laws require parental consent, thus if you dont get consent or the developer cant be arsed to implement a method for getting it you have one choice: ban the kiddies.
Of curse, the kids will use it, but you banned them. Just like facebook does, you thus cover your ass and if anyone asks why a kid is using audacity or facebook you can look all surprised.
> be hard to overtake
Why do we need to overtake anything? This isnt a popularity contest. We need alternatives, tahts all and thats what the forks will provide.
Those users who learn of Audacity's issues and wish to escape can do so and the fork they find the best will be what they use.
We dont need to win a race, just have different racetracks.
> Putting a printer on a Windows Server is SO 1990s.
Obviously you have no idea how printing works then.
Our cloud printing solution that prints, to the cloud, has a service running on a server which requires the print spooler.
Otherwise no users can print to the networked printers. This service will eventually be rolled out to our other sites.
Your "solution" only works for a small installation where one person prints at a time and doesnt authenticate on the printer. Thats quite 90's, we had to run like that recently when the printers failed to authenticate staff. It was a mess, with everyone's jobs getting mixed up.
> Without Speed there is no reason to uses computers instead of pen and paper
I think you vastly over estimate the speed of human computation using such methods.
Try reading up on the creation of Colossus and of the Bombe during WW2 and you may find out that humans are shit at doing computation fast, even a 486 (running appropriate software) will knock the socks off a pen and paper.
> I am curious how material that was re-mastered for CD at 16 bit a good while ago is going to be 'converted to 24 bit' without remastering from the original. Stuff that has already been re-mastered for SACD is available, but the rest?
It will sound exactly the same. The bits only affect where the noise floor is.
However if they have methods to process the noise down to 24 bit level then yes, that would be an improvement. You wont hear it however, the CD @ 16 bits has the noise already way below where anyone will hear it. If you add dither to it, then it goes even lower!
If you hear noise on a CD, then the original had that noise. Or something added it at AD conversion time. A 24 bit remaster wont sound better because its 24 bit, it may sound better because they were able to move or eliminate the noise. But if you take that 24 bit sample and convert it to 16, you wont hear the difference, unless the CD is a recording of pins dropping and you have to turn up the volume loads.
> Converting an analog medium into a digital representation and then back to analog again is, to put it mildly, less than ideal.
> If you really want to hear it as it was live you need a good quality analog recording & quality playback equipment.
> Digital is convenient but it's always going to be a compromise.
No, this is plain wrong. CD quality audio or higher will record and reproduce the original waveform perfectly.
I said PERFECTLY.
The nyquist shannon theorem perfectly captures and reproduces a band limited signal (thats very important).
All the signals we use are band limited, to human hearing, thus 44.1kHz @ 16 bit will reproduce the original waveform perfectly.
Analogue recording methods cant do that. They simply cant, adding noise to the recording itself. The dynamic range of a cassette is only 6 bits deep, if you have a very good cassette. 6 bits is a lot lower than 16. Reel to reel can go deeper, but nowhere as deep as CD. Luckily for reel to reel, you dont need to go much deeper, CD really has a lot of unused dynamic range but more is better as it lowers the noise floor to below normal listening levels.
All of the superiority of CD audio is entirely dependent on the analogue to digital conversion itself and the amplifiers etc. Again, all analogue stuff. The analogue stages, on the input or output can distort the source or add noise to the output. The CD will PERFECTLY reproduce the crap that is put into it via a crappy input and of course there is plenty of crappiness on the output, from cheap amplifiers to cheap transducers not to mention unsheilded cables that may pick up any of the EMI shit we all live in these days.
Digital audio reproduces the original waveform, band limited to the human hearing range, an a minimum of 44,100kHz and with a bit depth of 16 bits has a noise floor lower than anything ever made in the analogue domain. We could even reduce the number of bits, it just raises the noise floor, thats all it does. We could use 6 or 8 bits and we will sound just as good as any nicely recorded tape.
The character, warmth etc that everyone goes on about is nothing more than we noticing the imperfections added to the audio, during recording and playback. That why people like analogue, because it adds that imperfection. The great thing is, we can add that imperfection with the CD just by choosing devices that are not perfect, but we have a perfect original, every time.
Of course, I am talking about lossless audio here. I have not even considered any lossy coded, no matter what the bitrate or where it becomes "transparent". There are plenty of codecs that will, while being lossy, produce a waveform that is perfect "to the ear".
> precisely bugger all to help the musicians
A lot of those CD's are out of print, the musiscians may not get anything anyway.
The same argument can be said for second hand books, which obviously fails as second hand books have been a thing for a long long time.
Instead of riding on the royalties of a couple of works, artists are supposed to be encouraged to create new works, constantly.
Well, they are.
They really are.
Its just most players/devices a lazy with the error correction. My blu-ray writer I recently discovered works miracles on heavily scratched CD's, even on a 1984 disc that HAS A HOLE IN IT!!
Sure, CD error correction is nothing when compared to the error correction used in DVD, that thing is complex! But they really can take a beating.
Its just the players that are supposed ti fix the issues are not doing a good job.
That only happens in the edge cases.
Unless it was a UK made disc in the 80's, which all came from a factory that was found to have a severe defect in he line. Those will rot.
I have only had/seen 1 disc that has rotted, a cheap unknown brand CD-R, just one.
Environmental factors play a big part. A hotter humid environment will attack the disc faster.
I'm in the UK so don't have that issue. Well, not most of the year anyway.
My oldest disc has damage from an impact. I scanned it recently, the C1 and C2 errors were well within the Red Book spec, till the damaged area obviously. I scanned a brand new disc and it was generally lower than the 1984 disc but not by much. Also the damaged area in the 1984 disc (its a hole in the reflective material) was fully repaired during playback by my bly-ray writer during ripping. I could only just tell that the repair had happened (you can slightly hear it). Normal CD players apply much less error correction and skip at that point.
CD's when paired with decent full error correction are pretty resilient. even my Pink Floyd The Wall disc, that I found on the pavement in the 90's, used as a hokey puck by some kids, scratches all over the place, ripped flawlessly in that writer. I hadn't even tried resurfacing the disc!
You just need the right device.
> Spielberg factor
Does he avoid putting out the older versions like Lucas then?
I have no issue with edits and recuts etc until you get Lucas saying that the latest version is the ONLY version and anything you have thats older should be trashed and he would live to come and trash it for you.
> 44,1 kHz used to impact low-pass filtering
Many if not most players use oversampling to go way beyond 48kHz to solve that very issue.
My 90's CD player is a 6x oversampling one for example, upsampling the 44.1kHz to 264.6 kHz, which is then used with shaped dither to push the quantisation noise above human hearing, dropping he noise floor thus increasing the dynamic range from 96dB to around 116 dB.
You can get away with a very gentle filter with that. You just need to have removed most frequency content that the speakers will have trouble with.
Funny.
We got CD, ignoring issues with the actual recording, issues with lazy mastering and issues with the listeners equipment, CD give us perfect audio, lossless (albeit only stereo). Some improvements made here and there, oversampling being the main one and boom, perfect high res audio. barring the issues stated, which are beyond the scope of any playback medium to control.
Then we got MP3, which threw away loads of audio to compress the audio from something like a CD to something that the fledgling home internet could stand to distribute. Also there was the move to solid sate devices but I'm not considering that considering my first MP3 player held only 32MB which a CD simply laughed at.
We got other codecs etc, then we got faster internet and capacious solid state devices that could happily hold an uncompressed CD track, if not a flac track.
We have gone from lossless perfect audio (no, there are NO stair-steps, NO phase issues, there is only aliasing which was resolved by the 90's with oversampling), prefect because it perfectly reproduces any human hear able frequency (yes, you in the back, that DOES include multiple frequencies mixed together and the harmonics, if a HUMAN can hear it then its perfectly reproducible by CD technology) with a dynamic range that if used fully would make the listener deaf! We even extend that dynamic range further, just because we can, for just a little better noise management.
We then went to lossy audio.
Now we are back with lossless audio. Funny that, we were already there. Ok, people wanted it in their pocket and the tech we had took a while to outperform a CD, but now instead of just sitting back and enjoying a ripped CD or non-ripped one, we have those who think the CD is lossy, that it is not perfect. So off they go wasting bandwidth and storage space buying 192kHz sample rate files with no understanding that EVERY player will oversample to something as high if not higher than that, on the fly, at playback. My 90's CD player has 6x oversampling, that means it will up-sample the 44.1kHz audio to 264.6kHz.
Why? Well what they dont know is that there is NO audio supplied in the 192kHz file above 22kHz or so, which is stored exactly the same as on a CD, just with more samples, which are surplus. The reason why we oversample, on the fly at playback is because we can then move the quantisation noise above the 22kHz limit. We need the higher sampling rate to hold that JUNK audio above 22kHz, otherwise it will appear under 22kHz as aliasing/distortion. EVERYTHIG above 22kHz CAN NOT be heard and CANT be reproduced by speakers/hadphones etc. To prevent that JUNK noise from being a problem and distorting the audio because the dumb speaker will TRY to reproduce the waveform, we filter it out!
Oversampling is thus the method used since 90's CD players to push artificial quantisation noise above the human hearing range, we do this by ADDING in shaped noise (dither) that the player generates. The higher sampling rate thus allows for a filter design which is very simple and effective. The older CD players tried to filter HARD at 44.1kHz but they are never perfect and making them so was expensive. By oversampling, or even just jumping to 48kHz like with DVD audio, we can have a much better and cheaper filter. Its all filtered out above 22kHz or so and what was there was SHIT that you dont need.
All of this is done during playback, on the fly. Its merely maths and you only need a 44.1kHz sample rate to do it. Selling people "hi-res" audio is nothing more than selling people an ALREADY OVERSAMPLED file. The player does not need assistance, it can do it on the fly. There is no effing reason why the hell anyone would want to store such a file (for playback, recording and editing have other benefits here). For playback the final sellable file need never, ever to be at a sample rate greater than 44.1kHz or 48kHz. The player will create the 192kHz version on the fly during playback, no storage or bandwidth needed.
Hi-res is snake oil. A reason to have fast broadband, a reason to pay again, a reason to get a player with the same kind of "wank features" that used to be put all over CD players.
It a wank feature.
Unfortunately, to help force you to purchase the wank feature the mastering of the CD version is left to the idiots whilst the "hi-res" version is given the care that should have been there in the first place! The CD version thus sounds shit because they made it sound like shit so you get the hi-res version and bingo, the sucker now thinks he/she is hearing "hi-res" and that CD's sounded like shit all this time.
So, everyone. Lets be clear. Anyone thinking the hi-res version sounds better is only correct because:
1. They THINK they hear the difference.
2. The recording is a newer better one
3. The CD was mastered shitty
There are no other reasons. Its all down to the quality of the source material and mastering.
So buy the hi-res ones if you know thats the better recording/mastering etc. Then do the smart thing and downconvert it all back to 44.1kHz and save the space. It wont sound any different, in any way.
Yesterday I tried to reboot my system and systemd decided to take 3 mins to do so. First of all it was waiting on a job for 1:30, well I thought ok systemd, wait then kill it, then once the 1:30 timer was over it added another 1:30 to it.
I knwo what caused the issue, a filesystem that wouldnt unmount, due to a kernel bug that I triggered when I was testing something but I remember the days when if the system was told to reboot, it rebooted.
> For newer kit, this makes sense as there is not that much i386 gear that is online. Most of the still running kit is for the CNC machine (and the like) that probably never was online on even on a network.
Everything you just said contradicts the article. Tell me, if these non-networked CNC machines are the only 32 bit systems, why do Debian see them as a huge number of systems that run recent Debian versions? Did you read the article? Debian said themselves that there is a massive number of 32 bit installs but they want to drop 32 bit because they hare finding it hard to test the builds.
What you say about a non networked machine continuing to run any OS it needs is correct, basically common sense (in the IT form at least). If a German car garage can still run a C64 to help test and tune Ferrari engines I'm sure there are plenty of Windows 3.1 and OS/2 machines doing stuff.
I know of an American county who were using an Acorn machine to run the counties school heating systems over wireless link. The UK rail network also used similar machines to display the live train information on the screens on the platforms. I think they finally migrated when they got to the point they found it hard to find reliable sources of error free floppies.
Read it again mate, it's not about dropping support for the 386 processor but for x86 entirely!
Targeting 386 is the lowest common denominator for any 32bit system. They are talking about dropping support for ALL 32bit CPU's (for booting, 32bit code execution can still be supported).
Btw mate. I still run old and brand spanking new 8-bit systems ;)
> I think I typed the above nearly verbatim 25 years ago,
And you learned nothing.
You do know that these new 64 bit instructions you speak of tend to operate on larger operands? I.e ADDing two 64 bit numbers together on a 64 bit machine requires less cycles because the OPERANDS are bigger. On a 32bit machine you will need to load words from 4 separate memory locations and then perform 64 bit math using 32 bit registers. Thus much 32bit code will keep to 32 bit math, for speed.
So when that code is compiled for 64bit, it will be loading the same values in 64 bit operands, into 64 bit registers. This practically halves the time taken to perform the calculation, thus is faster.
But now all your math is using 64bit operands, see the problem? Unless you specifically use 32bit math instructions you will be loading TWICE as much data from ram and saving TWICE as much data to ram. This does not affect all instructions, and 32 bit systems were doing 128bit math too but when your basic operations now use 64bits of data:
IT MEANS YOUR SYSTEM USES MORE RAM FOR THE SAME JOB AS USED BY A 32BIT SYSTEM.
Sorry for shouting but, my god, when adults don't get basic math over 25 years it just, I'm lost for words.
> Running really old machines is actively bad for the planet, beyond a certain point.
No, it's the opposite. Dumping functional old machines is way worse.
Think about it. I have a Ryzen system as my main machine. It does way more per watt than my 586 laptop. But I use the laptop to run stuff that does not need performance, like word processing, coding (depending on language and compilation time), reading etc. So, is it better for the planet if I send that machine to landfill and use the Ryzen instead? Or I have the Ryzen off while I play with regular expressions?
You're basically saying, use the efficient Ryzen to do barely anything while bury the 586 to leach out it's poisons into the water table. Yep, so good for the planet.
> Would you really choose to do something as major as installing a new OS on a system that old, or would you use it as an incentive to replace it with something newer? I'd certainly go for the latter.
In a corporate IT environment, where the company operations rest on it, I will agree with you.
But, in most other environments, I'd consider the environment before throwing away a perfectly usable machine. My car is 10 years old, it has rust, but it will run for the next 10 years without much going wrong.
People still use Victorian, or even much older houses. Crazy.
I get new shoes when the old ones start actually falling apart.
I have books published before 1920, one is a science book that has an article on the possibility of the existence of Unicorns and the discovery of the Gulf Stream.
Computers should last for decades.
I have been doing the same.
I found out about signal before it was named signal, knowing that there was a very low chance I'd ever see anyone else use it I have been using it as my default SMS app for the last 8 years. Simply because I prefer having signal read my sms's rather than the default app. Not for privacy reasons, but back then there were many security vulnerabilities in android surrounding SMS and MMS and the default app was the usual target.
> can't set different alert tone for personal messages and group messages
Do people still do this? I thought that was a thing people played with when camera phones with midi ringtones came out. Seriously, I have never been able to figure out how to set different notifications for specific apps on most of my recent phones, they just have ONE sound, and options for enabling or disabling the permission for that app.
> What's better? Signal or telegram?
Signal, Telegram rolled their own crypto, Signal uses standard, protocols that are actually being researched for vulnerabilities and further development by the crypto community. Also as Telegram is not open source, nobody can start doing that on telegrams use of crypto, we know HOW they do it, they described it openly and even then the old addage of "don't roll your own crypto" rang the warning bells.
If you don't care about licensing and security and only about numbers then Telegram wins.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
