Linux founder tells Intel to stop inventing 'magic instructions' and 'start fixing real problems'

[Yousuf Khan]

Linus Torvalds' comments came from this article: https://is.gd/6zpZRL

> His comments came in a mailing list (via Phoronix) discussing an article suggesting AVX-512 might not be part of Intel's upcoming Alder Lake architecture. If that comes to pass, it will be just fine by Torvalds.
>
> "I hope AVX512 dies a painful death, and that Intel starts fixing real problems instead of trying to create magic instructions to then create benchmarks that they can look good on. I hope Intel gets back to basics: gets their process working again, and concentrate more on regular code that isn't HPC or some other pointless special case," Torvalds said.
>
> Intel introduced AVX-512 in 2013, initially as part of its Xeon Phi x200 and Skylake-X processor lines. It has also found its way into more current CPU architectures, including Ice Lake.
>
> The instruction set is designed to bolster performance in various types of workloads, such as scientific simulations, financial analytics, artificial intelligence, data compression, and other tasks that can benefit from more robust floating point operations.
>
> Nevertheless, Torvalds views AVX-512 as an example of "special-case garbage," noting that in regards to floating point performance, "absolutely nobody cares outside of benchmarks."
>
> "I absolutely detest FP benchmarks, and I realize other people care deeply. I just think AVX-512 is exactly the wrong thing to do. It's a pet peeve of mine. It's a prime example of something Intel has done wrong, partly by just increasing the fragmentation of the market," Torvalds said.

I think he's absolutely right, and previously we didn't see how much
Intel was wasting its time making these AVX instructions because it's
gaping security flaws were not yet known. We just assumed that the more
sophisticated these floating-point instructions got, the more power they
must draw naturally. But previous generations of FP instructions stayed
well within the power envelope of the processor, whereas these AVX
instructions have been known to go well outside the standard power envelope.

Yousuf Khan

[VanguardLH]

Yousuf Khan <bbb...@spammenot.yahoo.com> wrote:

> Linus Torvalds' comments came from this article: https://is.gd/6zpZRL

Full URL:
https://www.pcgamer.com/linux-founder-tells-intel-to-stop-inventing-magic-instructions-and-start-fixing-real-problems/#referrer=https%3A%2F%2Fwww.google.com&amp_tf=From%20%251%24s&ampshare=https%3A%2F%2Fwww.pcgamer.com%2Flinux-founder-tells-intel-to-stop-inventing-magic-instructions-and-start-fixing-real-problems%2F

Linus is known for publishing his tirades on Windows, and even on Linux
variants. He lambasts everyone.

Tweaking hardware to look good in benchmarks is news to you? Video chip
makers have been doing this forever, making their hardware or firmware
look better in particular benchmarks (sometimes their own benchmarks
tweaked for their hardware) but for which the benchmarks have no
practical implementation illustrating actual performance in real use.

AVX wasn't just about improving FP instructions. The number of cores
available back then was maybe up to 4 allowing concurrent thread
processing. With more cores to parallelize the computing, AVX becomes
less necessary. The latest CPUs (although far outside the consumer
price range) have 64 cores, maybe more. Sorry, but bitching in
hindsight is the easy way to look superior. I don't see Linux bitching
back *then* when AVX showed up. His forward-looking crystal ball was
just as cloudy as everyone else's. So, how many cores were in your home
computer back in 2013 when AVX came out? AVX isn't just about upping
the bit-width of FP calculations, but also about parallelization. How
many desktops nowadays have any apps on them that can use all 4 cores?

Not all CPUs are waiting to do something for end users. Some are
involved in highly complex computing, like animated computer graphics.
You think Zootopia was composed on a home computer? So, you think Intel
(or AMD) are going to tool up for a completely separate production line
for consumer vs high-graphics design platforms? There is an economy in
production by reusing existing manufacturing processes. Do consumer
platforms utilize AVX? Rarely. Why didn't Linus bitch when Intel added
Streaming SIMD Extensions (SSE)? How about all those non-gaming users
that don't care even about the old SSE extensions? Oh my God, the CPU
has something they don't need.

I suppose next Linus will bitch about increased parallelization in
Mozilla's Firefox. The next engine, Servo, takes advantage of the
memory safety and concurrency features of the Rust programming language.
Servo will use parallelism by using more cores for the rendering engine,
layout, HTML parsing, image processing, decoding, and other tasks that
can be isolated (into separate processes or threads to run on more
cores). Servo also makes further use of GPU-assisted acceleration, so
code running on a different processor. Would the GPU be needed if there
more core CPUs (real or multi-core) to parallelize the FP instructions?

I think GPU-assisted acceleration in web browsers started back in 2010,
but was just for web browsers. I remember some other apps used the GPU
for faster FP processing, but they seemed few and far between. More
video games are using AVX (AVX 2 more than AVX 512) since it is part of
the DirectX12 API. LOTS of users play video games on their home
computers, so AVX is really not that rare for use on low-end computing
platforms. AVX used to be shunned by game devs due to complexity in
coding.

Scalar, non-AVX :

void interpolate(vector<vector<int>>& mat)
{
for(int i=2; i<mat.size()-1; i=i+2)
for(int j=0; j<mat[0].size(); j++)
{
mat[i][j] = mat[i-1][j] + 0.5f * (mat[i+1][j] - mat[i-1][j]);
}
}


Using AVX:

void interpolate_avx(vector<vector<int>>& mat)
{
for(int i=2; i<mat.size()-1; i=i+2)
for(int j=0; j<mat[0].size(); j=j+8)
{
_mm256_storeu_si256((__m256i *)&mat[i][j],
_mm256_cvtps_epi32(_mm256_add_ps(_mm256_mul_ps(_mm256_sub_ps(_mm256_cvtepi32_ps(_mm256_loadu_si256((__m256i
*)&mat[i+1][j])), _mm256_cvtepi32_ps(_mm256_loadu_si256((__m256i
*)&mat[i-1][j]))), _mm256_set1_ps(0.5f)),
_mm256_cvtepi32_ps(_mm256_loadu_si256((__m256i *)&mat[i-1][j])))));
}
}

However, when mandated to programmers to code a game for maximum
performance, the AVX code runs 6.5 times faster! Simple coding with
slower performance, or complicated coding with faster performance. The
tradeoff is more cost in coding work, debugging, and optimizing hence
more time to achieve faster performance. Considering have video games
have upped the number of moving objects, physics modeling, and moving
texture change, some video games have insane requirements compared to
games dated over a decade ago.

Video games are real use of AVX. It's not just making benchmarks look
better. Guess Linus doesn't have bleeding edge hosts (in technology and
to his pocket) on which to run the most demanding video games. Is Linus
even a gamer? Oh wait, yeah, not that big a selection for Linux.

[Brian Gregory]

On 15/07/2020 19:42, VanguardLH wrote:
> Is Linus even a gamer? Oh wait, yeah, not that big a selection for Linux.

He isn't anyway.

--
Brian Gregory (in England).

[John Doe]

Yousuf Khan <bbb...@spammenot.yahoo.com> wrote:

> Linus Torvalds' comments came from this article: https://is.gd/6zpZRL

How much did he make off of Linux?
(I will look, but seems like an amusing fact.)

[Yousuf Khan]

On 7/15/2020 2:42 PM, VanguardLH wrote:
> Not all CPUs are waiting to do something for end users. Some are
> involved in highly complex computing, like animated computer graphics.
> You think Zootopia was composed on a home computer? So, you think Intel
> (or AMD) are going to tool up for a completely separate production line
> for consumer vs high-graphics design platforms? There is an economy in
> production by reusing existing manufacturing processes. Do consumer
> platforms utilize AVX? Rarely. Why didn't Linus bitch when Intel added
> Streaming SIMD Extensions (SSE)? How about all those non-gaming users
> that don't care even about the old SSE extensions? Oh my God, the CPU
> has something they don't need.

Well, no, the SSE extensions were a big improvement over the old
stack-based FPU model. Directly accessible FP registers rather than
pushing and popping indirectly off of a stack. Even AMD's 3DNow achieved
this, requiring even less changes to the hardware (it just fixed the
existing FPU stack model), although AMD did not yet have sufficient
marketshare to push it widely onto the market. I think the point Linus
is making is that AVX takes FPU's to a state that no one asked for. When
the first version of AVX came out, and no one used it, well okay just a
mistake, then the second version came out, hoping that it would correct
the deficiencies of the first one, still kind of understandable. When
even that one wasn't used, and now we're at like version 3 or 4, none of
which are being used, then that's obviously gone too far.

Yousuf Khan

[T]

On 2020-07-15 11:42, VanguardLH wrote:
> Is Linus
> even a gamer? Oh wait, yeah, not that big a selection for Linux.

Linux is not tied with Windows for gaming. Take
a gander at:

Fedora 31 | Features, Gaming, and New Daily Driver
https://www.youtube.com/watch?v=1P8oBlOTBho

[J. P. Gilliver (John)]

On Wed, 15 Jul 2020 at 13:42:37, VanguardLH <V...@nguard.LH> wrote:
>Yousuf Khan <bbb...@spammenot.yahoo.com> wrote:
>
>> Linus Torvalds' comments came from this article: https://is.gd/6zpZRL
>
>Full URL:
>https://www.pcgamer.com/linux-founder-tells-intel-to-stop-inventing-magi
>c-instructions-and-start-fixing-real-problems/#referrer=https%3A%2F%2Fww
>w.google.com&amp_tf=From%20%251%24s&ampshare=https%3A%2F%2Fwww.pcgamer.c
>om%2Flinux-founder-tells-intel-to-stop-inventing-magic-instructions-and-
>start-fixing-real-problems%2F

I'm a little surprised at VLH for the above: surely it's rather _more_
than a Full URL: I think you could truncate it before the # sign. What
follows are "referrer" and "From", with another couple of URLs in there
(with the "://"s and subsequent "/"s turned into their hex equivalents).

>
>Linus is known for publishing his tirades on Windows, and even on Linux
>variants. He lambasts everyone.
>
>Tweaking hardware to look good in benchmarks is news to you? Video chip
>makers have been doing this forever, making their hardware or firmware

[]
Not exclusive to computing hardware of course! The last _big_ one I can
remember is Volkswagen getting _caught_ detecting when their engines
were undergoing the annual emission tests (as required by most
countries) and running accordingly, but I'm sure there are myriad
examples. (Note: not myriad _of_.)
--
J. P. Gilliver. UMRA: 1960/<1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

Everyone learns from science. It all depends how you use the knowledge. - "Gil
Grissom" (CSI).

[VanguardLH]

Already pointed out: your "none of which are being used" is wrong. It
is being used. Video games use it, and those are not rare on Windows
platforms. Any game using DirectX 12 are utilizing AVX2. Scientific,
statistical, financial, encryption, and other programs can use it. Any
program using .NET Framework can use AVX. The latest versions of
Prime95 are optimized to use AVX. While it is used to stress test, that
was not its original or current intent which was to discover prime
numbers. Is prime hunting something that home users do? Of course not,
but it illustrates AVX *is* used.

https://www.tomshardware.com/reviews/stress-test-cpu-pc-guide,5461-2.html
"By default, Prime95 automatically selects the newest instruction set
extension, such as AVX, AVX2, or even AVX-512."

Your claim AVX is not used is false.

To test, go into the BIOS settings and change the AVX offset, and then
monitor the core frequencies, like with MSI's Afterburner. Surprise, a
lot of video games use AVX. You'll see the core frequencies go down
relative to the AVX offset when running an AVX-enabled program. I don't
play many new games (I still wish the Thief series keep evolving since
stealth is so poorly done in newer games), but have read SofTR,
Darksiders 3, Monster Hunter Word, AC: Odyssey, and Overwatch use AVX.
Overclockers trying to maintain the highest but stable clock rates whine
when core frequencies drop due to AVX, and have to change the AVX offset
to up the freqs.

https://www.google.com/search?q=overclock%20avx%20offset

If the games weren't using AVX, overclockers wouldn't be stymied over
the reduction in core freqs (and possible instability from vcore
reduction).

https://www.youtube.com/watch?v=BXBSVT9lpGw

Notice the AVX mode has higher frame rates. Also, it seems the right
side (for AVX) seems sharper overall. Timemark 2:23 starts the charts.
Later the author shows AVX doesn't improve performance in all games that
implement AVX. Sometimes AVX helps, sometimes not (but it's not worse).
I'm not into game programming, so I'll let someone else expert in that
note why AVX doesn't do better than SSE.

[VanguardLH]

"J. P. Gilliver (John)" <G6...@255soft.uk> wrote:

> On Wed, 15 Jul 2020 at 13:42:37, VanguardLH <V...@nguard.LH> wrote:
>>Yousuf Khan <bbb...@spammenot.yahoo.com> wrote:
>>
>>> Linus Torvalds' comments came from this article: https://is.gd/6zpZRL
>>
>>Full URL:
>>https://www.pcgamer.com/linux-founder-tells-intel-to-stop-inventing-magi
>>c-instructions-and-start-fixing-real-problems/#referrer=https%3A%2F%2Fww
>>w.google.com&amp_tf=From%20%251%24s&ampshare=https%3A%2F%2Fwww.pcgamer.c
>>om%2Flinux-founder-tells-intel-to-stop-inventing-magic-instructions-and-
>>start-fixing-real-problems%2F
>
> I'm a little surprised at VLH for the above: surely it's rather _more_
> than a Full URL: I think you could truncate it before the # sign. What
> follows are "referrer" and "From", with another couple of URLs in there
> (with the "://"s and subsequent "/"s turned into their hex equivalents).

is.gd, the URL shortening service that the OP used, does not provide a
preview mode. With TinyURL, you can add the "preview" hostname to see
where shortened URL points.

Well, is.gd does have a preview mode, but it's clumsy. You go to:

https://is.gd/previews.php

click on the "... see preview page ...", leave the web browser open, and
then click on the shortened URL the OP provided. Their page then shows
the full original URL and, yep, it has all that crap in it. Or you can
use on of the URL lengthener sites to reveal the original URL.

I gave the full URL that the *OP* provided with the shortened version.
Complain to the OP about not truncating URLs to their minimum. If he
had, he would not have needed the URL shortening service. The full URL:

https://www.pcgamer.com/linux-founder-tells-intel-to-stop-inventing-magic-instructions-and-start-fixing-real-problems/

is perhaps longer than the typical line length viewed in NNTP clients,
but slicing up URLs that are longer than the logical (viewed) line
length by injecting newlines (slicing URLs into multiple physical lines)
is a defect of the sender's client. Physical lines can be up to 998
characters long (that's the old-time recommendation). Maybe some NNTP
clients have problems when viewing physical line lengths longer than
their viewable line length making the URL not clickable, and why I've
seen some posters enclose the long URL within angle brackets, like
<URL>, as a workaround for deficient clients.

In any case, I showed the original (full) URL of what the OP used when
they generated the shortened version (well, a short redirection URL). I
showed the original URL. I didn't edit what the OP supplied.

>> Tweaking hardware to look good in benchmarks is news to you? Video
>> chip makers have been doing this forever, making their hardware or
>> firmware
>
> Not exclusive to computing hardware of course! The last _big_ one I can
> remember is Volkswagen getting _caught_ detecting when their engines
> were undergoing the annual emission tests (as required by most
> countries) and running accordingly, but I'm sure there are myriad
> examples. (Note: not myriad _of_.)

https://www.bbc.com/news/business-34324772
Dated 10-Dec-2015

https://www.businessinsider.com.au/chart-this-is-what-happened-to-volkswagen-after-the-emissions-scandal-2015-12
Dated 15-Dec-2015
Notice the chart showing the huge drop in sales.

I wonder how a car knows a gas sniffer is poking up its ahole. Oooh,
warm that up first before sticking it in. I suppose the car's computer
could notice the car wheels weren't rotating when the engine got revved
up and the steering wheel wasn't turning.

My state dropped emissions testing (for consumer vehicles which the
owner had to pay an $8 fee before they could get tabs) a long time ago.
I had a '92 bought used in '94 and kept for 24 years that never required
emission testing. I still have a '02 bought used in '04 that has never
required emissions testing. Emissions testing in my state ended back in
Nov 1999. Six other states don't have emissions testing, either. Our
requirement ceased after the levels of CO, ozone, and other pollutants
fell below the specs for the federal Clean Air Act; however, some states
are lobbying for stricter emissions control (exceeding EPA guidelines
and becoming more green-centric), so my state might go back to vehicle
testing despite our state has a green light. Must've been in sub-EPA or
more green-centric states where VW got busted for cheating.

[T]

On 2020-07-16 11:35, T wrote:
> is not

"is now"

Stinking typos

[VanguardLH]

You didn't provide a timemark for the related content, and I wasn't
going to watch all of the 22 minute video, so I moved the slider to skim
through it. The author started talking about Steam on Linux which could
now detect the native OS platform to know which game titles to present.
Steam represents about 78% of the marketshare for computer games. I saw
something about them using a compatibility shim to run Windows games on
Linux platforms eliminating the need to run Steam and the Windows games
inside of WINE. Wonder how the benchmarks reflect the performance of a
Windows game running inside of WINE versus running the Windows game atop
Steam's shim.

https://itsfoss.com/steam-play/

Oh, so Steam Play simply provides a fork of WINE as its shim between the
native OS platform and the Windows-only game. The Windows games will
likely be impacted the same whether ran inside of WINE or Steam's
variant of WINE. I didn't even bother to address running anything
Windows inside of WINE or via any other emulation layer, like VMWare
Player for Linux running Windows as a guest OS and then running a
Windows game inside of that virtual machine. That something is doable
doesn't mean it should be.

That still means the games were *not* developed for the Linux platform.
They were written for the Windows platform. Guess I should've qualified
my statement by saying:

"Oh wait, yeah, not that big a selection of native Linux games. "

Do hardcore gamers even bother with WINE? Conversely, everything Linux
can be played on Windows, too, so the user could use a Windows platform
to play native Windows games and emulated Linux games. Is there much
draw for that scenario? You can even play Android apps on Windows by
using a shim aka emulator, like Bluestacks. There's native-on-native,
and then there are less-than-ideal workarounds.

[Rene Lamontagne]

You make joke, Yes? :-)

Rene

[J. P. Gilliver (John)]

On Thu, 16 Jul 2020 at 15:22:44, VanguardLH <V...@nguard.LH> wrote:
>"J. P. Gilliver (John)" <G6...@255soft.uk> wrote:
>
>> On Wed, 15 Jul 2020 at 13:42:37, VanguardLH <V...@nguard.LH> wrote:
>>>Yousuf Khan <bbb...@spammenot.yahoo.com> wrote:
>>>
>>>> Linus Torvalds' comments came from this article: https://is.gd/6zpZRL
>>>
>>>Full URL:
>>>https://www.pcgamer.com/linux-founder-tells-intel-to-stop-inventing-magi
>>>c-instructions-and-start-fixing-real-problems/#referrer=https%3A%2F%2Fww
>>>w.google.com&amp_tf=From%20%251%24s&ampshare=https%3A%2F%2Fwww.pcgamer.c
>>>om%2Flinux-founder-tells-intel-to-stop-inventing-magic-instructions-and-
>>>start-fixing-real-problems%2F
>>
>> I'm a little surprised at VLH for the above: surely it's rather _more_
>> than a Full URL: I think you could truncate it before the # sign. What

[]

>I gave the full URL that the *OP* provided with the shortened version.

Fairy nuff.
[]

>>> Tweaking hardware to look good in benchmarks is news to you? Video

[]

>> Not exclusive to computing hardware of course! The last _big_ one I can
>> remember is Volkswagen getting _caught_ detecting when their engines

[]

>I wonder how a car knows a gas sniffer is poking up its ahole. Oooh,
>warm that up first before sticking it in. I suppose the car's computer
>could notice the car wheels weren't rotating when the engine got revved
>up and the steering wheel wasn't turning.

Could be. If done for a long time, maybe.

>
>My state dropped emissions testing (for consumer vehicles which the
>owner had to pay an $8 fee before they could get tabs) a long time ago.

Interesting; I didn't know some states didn't have any limits.
[]

>testing despite our state has a green light. Must've been in sub-EPA or
>more green-centric states where VW got busted for cheating.

There is a world outside "the states" (-:! Given that that particular
violation was Diesel engines, and I don't think Diesel cars are that
popular in the USA, it might well have first come to light somewhere in
EU. Ironically, the Germans are particularly keen on "green" matters.
(Though are also rather against nuclear, which is again ironic, as a lot
of their coal output is a rather dirty variety.)

--
J. P. Gilliver. UMRA: 1960/<1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

The death of democracy is not likely to be an assassination from ambush.
It will be a slow extinction from apathy, indifference, and undernourishment.
-Robert Maynard Hutchins, educator (1899-1977)

[Jonathan N. Little]

VanguardLH wrote:
> You didn't provide a timemark for the related content, and I wasn't
> going to watch all of the 22 minute video, so I moved the slider to skim
> through it. The author started talking about Steam on Linux which could
> now detect the native OS platform to know which game titles to present.
> Steam represents about 78% of the marketshare for computer games. I saw
> something about them using a compatibility shim to run Windows games on
> Linux platforms eliminating the need to run Steam and the Windows games
> inside of WINE. Wonder how the benchmarks reflect the performance of a
> Windows game running inside of WINE versus running the Windows game atop
> Steam's shim.

No all Steam games on Linux are Windows games running in WINE, many are
truly ported. I've got the Borderland games and they are ports not
running in WINE. Proton and WINE just extend selection for old games
avoiding the need to port.

--
Take care,

Jonathan
-------------------
LITTLE WORKS STUDIO
http://www.LittleWorksStudio.com

[T]

Titus starts in with Lutris at about 7:25

[Paul]

Are we playing "Sodoku" yet ?

Paul

[VanguardLH]

"Jonathan N. Little" <lws...@gmail.com> wrote:

> VanguardLH wrote:
>
>> The author started talking about Steam on Linux which could now
>> detect the native OS platform to know which game titles to present.
>> Steam represents about 78% of the marketshare for computer games. I
>> saw something about them using a compatibility shim to run Windows
>> games on Linux platforms eliminating the need to run Steam and the
>> Windows games inside of WINE.
>

> No all Steam games on Linux are Windows games running in WINE, many
> are truly ported. I've got the Borderland games and they are ports
> not running in WINE. Proton and WINE just extend selection for old
> games avoiding the need to port.

And does that somehow lopside the bias of game count to make native
Linux games far exceed those native games available on Windows? Of
course not. A game that had been ported to be a native Linux app is
obviously no longer a Windows app. A ported Windows game becomes a
native Linux game. The count of Linux games, whether native or ported
to make native, are still meager compared to the count of native games
on Windows.

Steam Play (Steam for Linux) detects the platform for the game probably
via a manifest for the game specifying its native platform. If it's a
native Linux game, it just loads it in Linux. If a Windows game, it
uses its WINE variant (aka Proton) to run the Windows game in that
emulator running atop Linux.

https://www.extremetech.com/gaming/275972-steam-for-linux-now-runs-windows-only-games

The confirmed list of Windows-only games that are compatible atop
Steam's Proton variant of is small. That's just the ones that Steam has
confirmed are compatible.

https://emulation.gametechwiki.com/index.php/Proton

has a hyperlink pointing to a list of app manifests/mappings for 100+
compatible Windows games. There's also a link to user-reported game
compatibility.

Proton converts the DX 10/11 calls to Vulkan to help keep game
performance similar to the Windows game when ran native on Windows.
That is dependent on the available of proprietary Linux drivers for
video hardware, the efficiency in coding the Linux video driver, and
compatibility of the Linux driver provided by the video maker to run on
other Linux variants (most don't list just Linux but some variant as
supported).

Regardless of the workarounds, they still don't alter which are native
Linux games and which are native Windows games, and the huge disparity
in counts between them. As noted, you can run Linux stuff on Windows,
but that doesn't magically mutate them into native Windows apps. Linux
marketshare floats around 2%, so obviously the game authors are going to
target the market that has the biggest ROI. That's not Linux.

I'm not a Windows proselytizer or Linux defender. I believe in using
the platform that best suits the task. Sorry, I still don't see Linux
as the best choice for a gaming platform.

[Char Jackson]

With some newsreaders, such as my old copy of Agent 2.0, brackets aren't a
workaround for a deficient client. They are simply markers to let the
composition window know that the configured line length value should be
ignored for text between the brackets.

[Jonathan N. Little]

VanguardLH wrote:
> "Jonathan N. Little" <lws...@gmail.com> wrote:
>
>> VanguardLH wrote:

<snip>

> Steam Play (Steam for Linux) detects the platform for the game probably
> via a manifest for the game specifying its native platform. If it's a
> native Linux game, it just loads it in Linux. If a Windows game, it
> uses its WINE variant (aka Proton) to run the Windows game in that
> emulator running atop Linux.
>
> https://www.extremetech.com/gaming/275972-steam-for-linux-now-runs-windows-only-games
>

2 years old. The landscape is changing rapidly. Of the top 100 games 1/3
now ported. More will be in the future with new games.

<https://www.protondb.com/>

[VanguardLH]

He first shows installing Lutris (Open Gaming Platform) before
installing Steam Play. Is Lutris even needed to use Steam's dispatcher
to decide if a game's manifest says it is Windows-only to then run it
under Steam's Proton variant of WINE? Isn't Lutris a Linux game library
manager and launcher, and perhaps across multiple sources (Steam,
battle.net, GOG)?

The video author says Lutris has no documentation. Really? Learning is
solely by trial-and-error, or pleading for info in a user community?

Before all that, he installed rpmFusion to get all the libs that Redhat
doesn't include on which Lutris and Steam Play might be dependent.
Install this, a must. Maybe install that. Then install Steam Play. I
take it Lutris and Steam Play won't grab, download, and install any libs
they are dependent upon. Seems this could be further streamlined for a
bigger lure to users to leave Windows. Maybe the chained installs are
needed just for Fedora.

[Rene Lamontagne]

I don't know how to play 'Sodoku'
Too old to start now. :-)

Rene

[Char Jackson]

On Thu, 16 Jul 2020 19:06:14 -0500, Rene Lamontagne <rla...@shaw.ca>
wrote:

I know how to play but it's tedious, so I wrote a Sudoku solver in Excel
(using VBA). It's more fun to watch the puzzle being solved.

[T]

I can't comment on any of your questions I am not a gamer.

Folks will leave Windows when enough of they applications
they need are ported over. M$ rules the universe when
it comes to applications and some outer ring of hell
when it comes to quality and security.

[VanguardLH]

T <T...@invalid.invalid> wrote:

> Folks will leave Windows when enough of they applications they need
> are ported over. M$ rules the universe when it comes to applications
> and some outer ring of hell when it comes to quality and security.

I think "catch 'em early" works better. School have and still do train
students on Windows. Chromebooks have penetrated schools more then
Linux. Users that, by choice, switch to Linux sometime later in their
lives are doing so due to curiosity, training, job requirements, using
the best platform for a critical task, or enlarge their expertise.
That's why Linux penetration has only been about 2% of the consumer PC
market. There already is good penetration into commercial use.

When schools are predominatly training students in an OS then the market
penetration goes up. The students take with them what they learned.
Microsoft learned that long ago. So did Apple. With so many Linux
variants and only a few commercial vendors (e.g., Redhat), free is not a
sufficient reason for mass migration to Linux. Get a gradually larger
student population to take Linux expertise into their homes and
workplace. Capture the minds and hearts of future computer users. Is
Linux deployed in pre-college schools for getting students intimate with
that OS?

http://linuxfederation.com/linux-part-school-education
(Yeah, it's a blog, so no datestamp as typical of blogs.)

https://opensource.com/article/18/3/linux-forward-schools

For well-rounded computer eduction, students should really be exposed to
multiple operating systems. Learn 'em, and let 'em choose.

However, businesses and even schools need support from the OS vendor.
Free doesn't include technical support. Those institutions don't look
firstly at the cost of a license. They look for support and its cost.
Not having robust support is costly. In-house training still has costs
and adds delay to acquire expertise. Like buying a printer, you figure
the Cost of Ownership is in the rate of use of the consumables (paper,
ink), and lastly consider the cost of the printer. The cost of OS
licenses is never discussed when we plan deployment of hosts, and
supporting them. The loss of use for a critical business app or suite
due to lack of support far exceeds free versus paid OS or software.
Cobol programmers are in high demand ($75K/year average base pay),
because colleges stopped teaching it long ago, so there aren't many
Cobol programmers around after attribtion of old farts that have retired
died off. Same for Fortran nowadays ($80K/year average base salary).
Companies are willing to pay for the expertise that is hard to find.
They couldn't give a gnat's fart about the costs for Cobol compiler
licenses. Losing a critical business program due to no support costs
way more, maybe even cause the company's demise.

I have a sneaking suspicion that Microsoft is planning a migration to a
Linux/Windows hybrid kernel with a Windows GUI. After all, Windows NT,
and up, which had an NT-based kernel still carried along the familiar
desktop GUI from the 9x/DOS frankenjob GUI. First it was Linux in their
Azure cloud service. Then they began releasing apps for Android and
Linux. They rolled in a Linux compatibility layer (Windows Subsystem
for Linux, or WSL, but no Linux kernel code) to run Linux binary
executables. Rolling in subsystems into Windows isn't new. NTFS is a
file subsystem, as are FAT, exFAT, and CDFS. WSL v2 was announced May
2019 which moved to a real Linux kernel (as a subset of Hyper-V
features). In 2016, WSL only provided an Ubuntu image. the Fall
Creators Update in Oct 2017 move to SUSE images. With WSL v2 in May
2019, Linux support moved to a Hyper-V VM-based backend instead of the
system-call adaption (compatibility) layer. We've been familiar with
VMMs (Virtual Machine Managers) using virtual machines running guest
OSes on Windows (or visa versa on Linux) for a long time. Microsoft
decided to use the Hyper-V VMM. They wanted a kernel-mode model instead
of user-mode solutions. Because of the extremely high adoption of
Windows versus Linux, there has been concern that WSL could be a way for
Microsoft to "embrace, extend, and extinguish Linux".

At first, WSL was available only for Pro and Enterprise editions of
Windows 10 x64. On July 2019, they granted its used on Home editions.
I run optionalfeatures.exe (run with admin permissionsto effect
changes), scroll down, and WSL is listed. I haven't yet played with
WSL, so it's currently disabled.

https://betanews.com/2018/03/06/debian-linux-windows/

"I am of the opinion that if you want to run an operating system based
on that open source kernel, then you should just do so natively -- not
on top of Windows."

Well, that is not accurate. Hyper-V (a native hypervisor) is a VMM but
it does *NOT* run in user-mode to manage VMs. It is a kernel-mode
service. Probably because the Linux images are represented as "apps" in
Microsoft's store is why that author thinks it is an app running atop of
Windows.

"Hyper-V implements isolation of virtual machines in terms of a
partition."

That's not a portion of an HDD or SDD where sectors are allocated in a
group for use by an OS or data. That's a hypervisor partition. IBM
mainframes 30+ years ago used similar hypervisors with OS isolation
partitions. I was helping the sysadmin migrate to a new version of VSE,
MVS, or VM by installing and configuring the new version of the OS in a
different partition that was not accessible to the users. When we were
ready, and late at night when the users were gone and after announcing
the switch (because any users connected to the OS version in the old
partition would get disconnected), we swapped which was the primary OS
partition. The users came in and found a new version of the OS was
ready. If there was a problem, we could switch back to the old OS
partition.

OS partition (Hyper-V) hierarchy
https://upload.wikimedia.org/wikipedia/commons/thumb/0/06/Hyper-V.png/675px-Hyper-V.png

Yes, every hypervisor is itself an OS, but the working Windows image and
Linux image are not "running atop Windows". Users aren't using the
Hyper-V OS for their work. They're using the VM of Windows managed by
Hyper-V. Well, that's how it works for the server version. Only admins
go into the Hyper-V OS to configure it. That's a distant memory since I
haven't looked at Hyper-V for years.

[VanguardLH]

Char Jackson <no...@none.invalid> wrote:

> Rene Lamontagne <rla...@shaw.ca> wrote:

>
>> Paul wrote:
>>
>>> Are we playing "Sodoku" yet ?
>>

>> I don't know how to play 'Sodoku'
>> Too old to start now. :-)
>
> I know how to play but it's tedious, so I wrote a Sudoku solver in Excel
> (using VBA). It's more fun to watch the puzzle being solved.

Does it take coffee and bathroom breaks, too?

[T]

My experience has been different:

When I was a youngster, the colleges trained on Apple.
The minute grads hit industry, they switched to Windows.

And my current experience with small business I have
constantly tried to figure out how to get folks on
Linux. It is virtually impossible, as the apps they
need only run in Windows.

It is the apps the customer cares about. They could
care less if they were run int Flying Zucchini OS, if
it ran their apps.

I know this to be the case as my customer SELDOM know
what OS they are running.

Also, the recent computer science grads I have come
across make my head spin. They know virtually nothing
about computers or programming. Seriously, they barely
know what a mouse is. And they are in debt up to the
asses with student loans.

[T]

Did you watch the video?

[VanguardLH]

VanguardLH <V...@nguard.LH> wrote:

> {Using Windows Subsystem for Linux}

From what I've seen of the WSL videos, and because the Linux "apps" are
lightweight images of Linux, what I see is running the Linux image dumps
you to a bash shell in terminal mode (aka command line aka console
mode). You don't get a GUI desktop, like Gnome or KDE. Alas, most
Windows users don't know about shells, console-mode, or entering
commands. They'll want a GUI for, say, the WSL/Ubuntu image. The
Windows 10 WSL bash shell doesn't officially support GUI Linux desktops.
Microsoft intended WSL's bash shell for developers running Linux
terminal-mode programs although it seems you can load GUI apps via shell
commands. I suppose Microsoft also didn't want to waste resources on
developing a GUI desktop when there have been lots of others already.

While it's possible to dual-boot into native Windows or into native
Linux as the base OS, dual-booting means you only get to use one OS at a
time. Hyper-V (hypervisor) is the base OS running the working Windows
or Linux images inside a VM, but the apps within those VMs are native to
that guest OS. Windows users can get acquainted with Linux while still
using Windows, and using both concurrently (without using user-mode
VMMs, like VirtualBox or VMware Player). However, not many Windows
users are going to endear themselves to Linux if stuck in terminal mode.

I read one solution is to run an X Server (on Windows) that connects to
the Linux VM. That would grant access to a Linux application or
desktop's GUI. Then you install the desktop in the Linux VM, like
running "sudo apt install lxde" (for the LXDE desktop). Looks like you
follow with "export DISPLAY=:0" and "export LIBGL_ALWAYS_INDIRECT=1". I
only remember the DISPLAY var getting set when I used Reflection X or
Hummingbird (Xming is a free alternative) eons ago to connect to
numerous *NIX hosts on my Windows workstation. However, if I install a
desktop GUI into the Linux image running in a VM managed by Hyper-V, why
would I need an X server to see that desktop (on the same host)? To me,
the X server was to see the desktop on a different workstation. X11 is,
after all, a network protocol. After installing a desktop into the
Linux image (LXDE, Gnome, KDE), why wouldn't it show when I switch to
the view window for that VM? I would think I'd have the desktop load on
startup of the Linux image (e.g., startlxde). Or won't the Linux
desktop replace the terminal window?

Maybe the suggestion to install a GUI desktop in the Linux image, have
it load on Linux startup, but use an X server to see the Linux desktop
is to eliminate having to leave open (even if minimized) the terminal
window. Been about 10 years since I had to use any Linux variant. That
was back when I was working and before I [mostly] retired.

Although Windows users were weaned on a GUI desktop, even for OS config
tools, I suspect they'd have to learn to bounce out of the Linux desktop
back into the bash shell for some OS configs. Need to make the
transition to Ubuntu, SUSE, or whatever Linux as painless and intuitive
as possible for them to adopt Linux. How many Windows users enjoy
reverting to a command shell to enter console-mode commands?
Penetration into the user market is not led by techies in their personal
use of the OS.

https://www.youtube.com/watch?v=nKCe9UE-quA (*)

I started watching that, but my eyes demanded some sleep. My initial
reaction from watching part of the video before dropping a shortcut to
it to watch later was "Geez, no wonder Windows users don't use Linux".
Yeah, the tweaks were to get the Linux VM working well along with a GUI
Linux desktop connected to using X11 from a Windows X client, but I
remember this kind of shit when I used to use native Linux, too.

Once you get past all the WSL/Linux setup shit to get a user workable
setup, seems like that would help get more Windows users familiar with
Linux.

(*) I'm sure glad I installed the Enhancements for Youtube add-on in
Firefox to have it skip past the in-video ads. You see an
interruption for the ad but only one frame shows, and the extension
skips back to the video to continue playing it. There's a version
by the same author to use in Chrome, too. Makes enjoyable again
watching of long videos at Youtube.

[VanguardLH]

"Jonathan N. Little" <lws...@gmail.com> wrote:

> VanguardLH wrote:
>> "Jonathan N. Little" <lws...@gmail.com> wrote:
>>
>>> VanguardLH wrote:
>
> <snip>
>
>> Steam Play (Steam for Linux) detects the platform for the game probably
>> via a manifest for the game specifying its native platform. If it's a
>> native Linux game, it just loads it in Linux. If a Windows game, it
>> uses its WINE variant (aka Proton) to run the Windows game in that
>> emulator running atop Linux.
>>
>> https://www.extremetech.com/gaming/275972-steam-for-linux-now-runs-windows-only-games
>>
>
> 2 years old. The landscape is changing rapidly. Of the top 100 games 1/3
> now ported. More will be in the future with new games.
>
> <https://www.protondb.com/>

Where's the impetus to port if Steam's Proton (variant of WINE) along
with using proprietary video drivers for Linux (if available) lets
Windows-only games run on Linux?

Any benchmarks showing performance differences (FPS, CPU/core
frequencies, video quality, temperatures, etc) between a ported Windows
game (making it a native Linux game) versus using Steam Proton and
proprietary video Linux drivers?

If there's no or little performance impact, can't see game authors
spending the time and resources to port from Windows with 88%
marketshare to Linux with a 2% marketshare.

protondb.com is a database of Windows-only games that have been
user-reported as compatible by using Proton (don't know if proprietary
Linux video drivers were used, though, or if Vulkan is solely relied on
to retain video performance). Is there a toggle or view there showing
how many Proton-compatible Windows-only games have been ported to Linux
hence eliminating the need for Proton? Games played on Linux using
Proton are not ported games.

[Paul]

VanguardLH wrote:
> VanguardLH <V...@nguard.LH> wrote:
>
>> {Using Windows Subsystem for Linux}
>
> From what I've seen of the WSL videos, and because the Linux "apps" are
> lightweight images of Linux, what I see is running the Linux image dumps
> you to a bash shell in terminal mode (aka command line aka console
> mode). You don't get a GUI desktop, like Gnome or KDE. Alas, most
> Windows users don't know about shells, console-mode, or entering
> commands. They'll want a GUI for, say, the WSL/Ubuntu image. The
> Windows 10 WSL bash shell doesn't officially support GUI Linux desktops.
> Microsoft intended WSL's bash shell for developers running Linux
> terminal-mode programs although it seems you can load GUI apps via shell
> commands. I suppose Microsoft also didn't want to waste resources on
> developing a GUI desktop when there have been lots of others already.

That was solved within two days of release of WSL.

Someone put an Xserver on their Windows box, and claimed
to run Firefox in WSL and displayed it on the XServer.
(I didn't see a picture of this at the time.)

But that doesn't cover every possible application you
might want to run. It was just a bar bet that
"we can get something to run under a GUI".

Paul

[VanguardLH]

T <T...@invalid.invalid> wrote:

> Also, the recent computer science grads I have come across make my
> head spin. They know virtually nothing about computers or
> programming. Seriously, they barely know what a mouse is. And they
> are in debt up to the asses with student loans.

We'd get CSci university interns to help in Software QA. They were
trained to follow instructions, and nothing more (no intuition, no
imagination, no motivation). We had test procedures, but some were just
templates that had to get filled out when new features or changes showed
up in the software. They got the same training (classes and
instructional CDs) the rest of us got. The interns just had no grasp of
how to dig into software to test it, and how to document their testing
despite having an detailed template but which they had to fill in during
and after testing. Way too much handholding. The interns that got the
retro tests (for old functionality) where the procedure was completely
written did okay, because they didn't have anything to do but read
instructions. Yes, they were interns and had to learn, but they were
like 1st-year students instead of near-grads. No initiative, no talent
for testing, and poor writing skills.

I remember someone remarking that college isn't about training their
students for a particular job. It's to train them on how to learn. Not
evidenced by the interns that we got. I think we used interns for 6
months: the contract length. Never again thereafter. A failed
experiment trying to up the count of QA testers to shorten our testing
schedule which always got squeezed by Dev delivering late and Sales
arranging early deployment to customers. We ended up outsourcing some
retro tests (fully written on old functionality) to the Dev and Field
Support groups if some were available. Once we explained our test
scheduled and Sales wanted it shorter, we said either we don't test all
the old stuff and hope it works, or we get helpers to make their
schedule. Dev was hard to get, so we used Sales to pressure them.
Field Support was easy to get unless they were at a job site, plus they
were experienced on how customers used the product, not how Dev thinks
it should work per the Functional and Engineering Spec docs. Eventually
I wrote scripts for all the retro tests that did the setup, checked
dependency on the results of other tests, and logged results or alerted
on failed tests. The scripts became the tests with the doc template
just outlining what the scripts did.

Of course, our bad experience with interns could've been with the ones
we happened to get at that time. We paid our interns minimum wage.
They weren't allowed overtime (great for them that they could quit by
the clock while the rest of us were goal-oriented and left when we got
to a stopping point that provided a good resume point). They only
worked half days since they were still going to school. They had the
option to become employees at the end of their internship. They got
experience and a salary. We felt it unfair to exact manpower from
unpaid workers. They were [supposed to] help us, and we wanted to
reward them. We didn't care FLSA considers interns as not employees
which means interns don't have to get compensated
(https://www.dol.gov/agencies/whd/fact-sheets/71-flsa-internships). We
were, um, lenient in our report to the college for our assessment of the
interns. We still wanted them to get academic credit.

There were no later experiments using interns to better gauge the
usefulness of that workforce source. Before the contract ended, my
manager asked for reviews on their performance. I told my boss that I'd
write scripts to do the work of the interns. I didn't get overtime, but
I did accrue flex time that I could add to my PTO. I took some long
vacations or extended weekends when QA wasn't in prep or crunch mode.

[VanguardLH]

Char Jackson <no...@none.invalid> wrote:

> With some newsreaders, such as my old copy of Agent 2.0, brackets
> aren't a workaround for a deficient client. They are simply markers
> to let the composition window know that the configured line length
> value should be ignored for text between the brackets.

Ah, it's a sender's client trick to prevent line breaks. Understood.
In my NNTP client, I don't need to do that for URLs as they kept intact
in one physical line; however, it does have a Word Wrap toggle that I
can click to insert a composition marker (not in the sent copy) to keep a long string from line wrapping. I use it occasionally, like for a wide data table where line wrapping makea it unintelligible. (I used it on this line as an example.)
It keeps the long string as one long physical line. The reader's client
might enforce line splitting at their configured line length. Nothing I
can do about that. My tricks sounds similar to your bracketing trick.

However, I've seen those long strings in a long line include the angle
brackets. They might be a hint in the composition window in the
sender's client, but they were also included in the sent copy. As a
test, could you reply with a long string, like 200 characters, enclosed
in your non-wrap markers, so I could see if the submitted copy has the
non-wrap markers or not?

[Andy Burns]

VanguardLH wrote:

> I wonder how a car knows a gas sniffer is poking up its ahole. Oooh,
> warm that up first before sticking it in. I suppose the car's computer
> could notice the car wheels weren't rotating when the engine got revved
> up and the steering wheel wasn't turning.

A bit more complex than that, but basically spotting conditions of the
standardised tests and switching into an alternate ECU mode

<https://media.ccc.de/v/32c3-7331-the_exhaust_emissions_scandal_dieselgate>

Jump to 57:00 if you just want the money shot, but the whole thing is
worth a watch ...

[Andy Burns]

VanguardLH wrote:

> Where's the impetus to port if Steam's Proton (variant of WINE) along
> with using proprietary video drivers for Linux (if available) lets
> Windows-only games run on Linux?
>
> Any benchmarks showing performance differences (FPS, CPU/core
> frequencies, video quality, temperatures, etc) between a ported Windows
> game (making it a native Linux game) versus using Steam Proton and
> proprietary video Linux drivers?

Can't point you to a specific video, but I daresay Wendell has one that
covers it with a gaming-targeted distro.

<https://www.youtube.com/c/TekLinux/videos>

[Andy Burns]

VanguardLH wrote:

> From what I've seen of the WSL videos, and because the Linux "apps" are
> lightweight images of Linux, what I see is running the Linux image dumps
> you to a bash shell in terminal mode (aka command line aka console
> mode). You don't get a GUI desktop, like Gnome or KDE.

Not this year (unless you install an X11 server within windows, or an
RDP server within windows) but they're working on it ...

<https://devblogs.microsoft.com/commandline/the-windows-subsystem-for-linux-build-2020-summary/#wsl-gui>

[Andy Burns]

Andy Burns wrote:

> VanguardLH wrote:
>
>> You don't get a GUI desktop, like Gnome or KDE.
>
> Not this year (unless you install an X11 server within windows, or an
> RDP server within windows) but they're working on it ...

^^^^^^^
Linux

[Rene Lamontagne]

I think I slept through the best parts.

Rene

[Jonathan N. Little]

VanguardLH wrote:
> "Jonathan N. Little" <lws...@gmail.com> wrote:
>
>> VanguardLH wrote:
>>> "Jonathan N. Little" <lws...@gmail.com> wrote:
>>>
>>>> VanguardLH wrote:
>>
>> <snip>
>>
>>> Steam Play (Steam for Linux) detects the platform for the game probably
>>> via a manifest for the game specifying its native platform. If it's a
>>> native Linux game, it just loads it in Linux. If a Windows game, it
>>> uses its WINE variant (aka Proton) to run the Windows game in that
>>> emulator running atop Linux.
>>>
>>> https://www.extremetech.com/gaming/275972-steam-for-linux-now-runs-windows-only-games
>>>
>>
>> 2 years old. The landscape is changing rapidly. Of the top 100 games 1/3
>> now ported. More will be in the future with new games.
>>
>> <https://www.protondb.com/>
>
> Where's the impetus to port if Steam's Proton (variant of WINE) along
> with using proprietary video drivers for Linux (if available) lets
> Windows-only games run on Linux?
>
> Any benchmarks showing performance differences (FPS, CPU/core
> frequencies, video quality, temperatures, etc) between a ported Windows
> game (making it a native Linux game) versus using Steam Proton and
> proprietary video Linux drivers?

Well I can say that with the ported game Borderlands 2 and Pre-Seaquel
on this laptop with "Enhanced" Intel GPU was unplayable with Windows 10.
Now have Ubuntu 16.04 and they are quite playable, albeit not with maxed
out graphics settings.

>
> If there's no or little performance impact, can't see game authors
> spending the time and resources to port from Windows with 88%
> marketshare to Linux with a 2% marketshare.

Well the refinements to Proton only recently narrowed the performance
gap, many ported games can perform better on Linux than Windows. Also
when Windows OS becomes SAAS and folks will have to subscribe to use
Linux will be come more attractive. Linux allows more system resources
to be applied to the application at hand and to to telemetry and
advertisers... For gamers performance is paramount, so now the last
holdout nVidia is beginning to cooperate and get onboard so as Windows
bloats as Linux performs gamers will go with the performance.

>
> protondb.com is a database of Windows-only games that have been
> user-reported as compatible by using Proton (don't know if proprietary
> Linux video drivers were used, though, or if Vulkan is solely relied on
> to retain video performance). Is there a toggle or view there showing
> how many Proton-compatible Windows-only games have been ported to Linux
> hence eliminating the need for Proton? Games played on Linux using
> Proton are not ported games.
>

That has the green bar for native...that's the ported game percentage.

[J. P. Gilliver (John)]

On Thu, 16 Jul 2020 at 22:35:33, T <T...@invalid.invalid> wrote:
>On 2020-07-16 21:47, VanguardLH wrote:

[]

>> When schools are predominatly training students in an OS then the
>>market
>> penetration goes up. The students take with them what they learned.
>> Microsoft learned that long ago. So did Apple. With so many Linux

And, before that, Bell labs with UNIX; when I was at uni., it was pretty
universal on the mainframes (this was around 1980, when home computers,
inasmuch as they existed, were all incompatible, mostly running
[incompatible versions of] BASIC); I understood that Bell let academic
institutions for a peppercorn fee, for that very reason.

>> variants and only a few commercial vendors (e.g., Redhat), free is not a
>> sufficient reason for mass migration to Linux. Get a gradually larger

For the vast majority of home users (including a large proportion of
those who use their computer in their profession), Windows is seen as
"free" anyway - in that it comes with the computer. Sure, it's an
element of the price, but I've not seen it listed separately since the
days when people had their PC made to their spec. - which for the (vast
I think) majority of users, is decades ago. [The same applies to other
OSs - Apple, Chrome, etc.; the price of the OS is not shown, any more
than that of the case, mobo, HD (or SSD), memory ... in _most_ of the
places people buy computers these days.]

I'm speaking of the UK, but I _think_ it's the same in the USA. Here,
the main "High Street" (US: Main Street) or mall places where computers
are on sale are the larger supermarket branches, and specialist shops -
and of the latter, we sadly only have one now.
[]

>> For well-rounded computer eduction, students should really be
>>exposed to
>> multiple operating systems. Learn 'em, and let 'em choose.

Ideally, yes. In practice, even if hardware/licensing etc. weren't a
problem, time is. (As well as all the other pressures on a teacher.)

>> However, businesses and even schools need support from the OS
>>vendor.
>> Free doesn't include technical support. Those institutions don't look
>> firstly at the cost of a license. They look for support and its cost.

There, of course, the fact that Windows _isn't_ free makes the
difference: MS provides support to schools (in both the UK and US
meaning of "schools"), whereas "Linux" doesn't (for both cost reasons
and that it isn't a single entity).

>> Not having robust support is costly. In-house training still has costs
>> and adds delay to acquire expertise. Like buying a printer, you figure
>> the Cost of Ownership is in the rate of use of the consumables (paper,
>> ink), and lastly consider the cost of the printer. The cost of OS

You _should_, but looking at what people buy, I think few do! (Not
helped by most stores _not_ showing the cost of a set of cartridges by
each printer.) [Not to mention the recent abomination - IMO - of "ink as
a service".]
[]

>> died off. Same for Fortran nowadays ($80K/year average base salary).

[Where do I sign up? Though I imagine my skills - Fortran IV, send off a
coding form and get back a pile of punched cards and some printout,
1970s - are a bit rusty ... (-:]
[]

>> I have a sneaking suspicion that Microsoft is planning a migration
>>to a
>> Linux/Windows hybrid kernel with a Windows GUI. After all, Windows NT,

You might be right.

>> and up, which had an NT-based kernel still carried along the familiar
>> desktop GUI from the 9x/DOS frankenjob GUI. First it was Linux in their

I remember when NT gradually took over: NT4 had the 9x/XP GUI, but
NT3.51 had (more or less) the Windows 3.1 (or 3.11) GUI, and held on for
quite a while: at my employers, NT3.51 systems were not replaced by NT4
as a matter of course.

[Long chunk that's beyond me here. (I'm sure mostly correct.)]
[]

>And my current experience with small business I have
>constantly tried to figure out how to get folks on
>Linux. It is virtually impossible, as the apps they
>need only run in Windows.
>
>It is the apps the customer cares about. They could
>care less if they were run int Flying Zucchini OS, if
>it ran their apps.

Indeed. Same for a lot of home users: as long as it does browsing,
email, and (mostly via the browser) social media, and in some cases word
processing, they don't care (or in a _few_ cases even _know_) what OS
they have. And this is _not_ a put-down of such users - they can be
quite intelligent, just not _interested_. The car analogy is imperfect
but relevant.
[]

>Also, the recent computer science grads I have come
>across make my head spin. They know virtually nothing
>about computers or programming. Seriously, they barely
>know what a mouse is. And they are in debt up to the
>asses with student loans.

I feel the same (more in next post), though to be fair they probably
_do_ know more about _some_ things than you (and I) do. The fact that
_we_ may consider those things less important is of (endlessly arguable
and probably not productively so) relevance.
--
J. P. Gilliver. UMRA: 1960/<1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

"... all your hard work in the hands of twelve people too stupid to get off jury
duty." CSI, 200x

[J. P. Gilliver (John)]

On Fri, 17 Jul 2020 at 02:54:22, VanguardLH <V...@nguard.LH> wrote:
>T <T...@invalid.invalid> wrote:
>
>> Also, the recent computer science grads I have come across make my
>> head spin. They know virtually nothing about computers or
>> programming. Seriously, they barely know what a mouse is. And they
>> are in debt up to the asses with student loans.
>
>We'd get CSci university interns to help in Software QA. They were

[]

>instructional CDs) the rest of us got. The interns just had no grasp of
>how to dig into software to test it, and how to document their testing

[]

>like 1st-year students instead of near-grads. No initiative, no talent
>for testing, and poor writing skills.

I remember - I _think_ it was in the last decade, but it might have been
more - being startled when I spoke to a young computing graduate, to
find he'd never done any assembler. At that time, after my initial
double-take, I thought to myself: the field is big enough, that there'll
be plenty of room for him, and in practice he'll probably never have any
trouble finding interesting and well-paid employment.

>
>I remember someone remarking that college isn't about training their
>students for a particular job. It's to train them on how to learn. Not

That is certainly part of it, especially if they hadn't picked that up
at school (UK meaning). It's also - at _some_ levels - when the brain is
at peak ability: I remember holding two conversations at once, something
I'm not sure I could do to the same extent now. (In contrast, my now
slower brain has more _experience_. And that combines with my
"generalist" outlook.)

>evidenced by the interns that we got. I think we used interns for 6
>months: the contract length. Never again thereafter. A failed

Why did you use them in the first place - was it because of some form of
state subsidy, of about that duration?
[]

>Field Support was easy to get unless they were at a job site, plus they
>were experienced on how customers used the product, not how Dev thinks
>it should work per the Functional and Engineering Spec docs. Eventually

(-: [Users constantly amaze you in the ways they use things. (Doesn't
just apply to software, of course.) Occasionally, it's very innovative!]
[]

>were, um, lenient in our report to the college for our assessment of the
>interns. We still wanted them to get academic credit.

You were, in short, decent guys.

>
>There were no later experiments using interns to better gauge the
>usefulness of that workforce source. Before the contract ended, my
>manager asked for reviews on their performance. I told my boss that I'd
>write scripts to do the work of the interns. I didn't get overtime, but
>I did accrue flex time that I could add to my PTO. I took some long
>vacations or extended weekends when QA wasn't in prep or crunch mode.

[J. P. Gilliver (John)]

>On Thu, 16 Jul 2020 at 22:35:33, T <T...@invalid.invalid> wrote:

[]

>>It is the apps the customer cares about. They could

couldn't

>>care less if they were run int Flying Zucchini OS, if
>>it ran their apps.

[]
I know "could care less" is the US version of this expression, but it's
inaccurate. Think about it: if you could care less, that implies that
you do care a little - which is not what you mean; you actually mean
"couldn't care less".

--
J. P. Gilliver. UMRA: 1960/<1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

Politics: A strife of interests masquerading as a contest of principles.
- Oscar Wilde, quoted by Ron Bauerle 2015-7-24

[Andy Burns]

Andy Burns wrote:

> VanguardLH wrote:
>

>> You don't get a GUI desktop, like Gnome or KDE.
>
> Not this year (unless you install an X11 server within windows

Having mentioned it, I thought I'd better try it ...

I already have WSL2 and Fedora32 installed, so I installed VcXsrv on my
Win10, and then tries xeyes which runs fine, glxgears which runs so fast
that the gears almost look stationary, and thunderbird.

[nospam]

In article <rerdb6$4jb$1...@dont-email.me>, <T...@invalid.invalid> wrote:

> Also, the recent computer science grads I have come
> across make my head spin.

as well they should. technology changes rapidly to where anyone's head
will spin, including theirs.

> They know virtually nothing
> about computers or programming. Seriously, they barely
> know what a mouse is.

what does knowing what a mouse is have to do with knowledge of
computers or programming?

> And they are in debt up to the
> asses with student loans.

most college students do.

[nospam]

In article <rvx$q$fknbE...@255soft.uk>, J. P. Gilliver (John)

<G6...@255soft.uk> wrote:

> I remember - I _think_ it was in the last decade, but it might have been
> more - being startled when I spoke to a young computing graduate, to
> find he'd never done any assembler. At that time, after my initial
> double-take, I thought to myself: the field is big enough, that there'll
> be plenty of room for him, and in practice he'll probably never have any
> trouble finding interesting and well-paid employment.

there is no need for assembler anymore, except in very rare
circumstances.

[Char Jackson]

On Fri, 17 Jul 2020 03:10:26 -0500, VanguardLH <V...@nguard.LH> wrote:

>Char Jackson <no...@none.invalid> wrote:
>
>> With some newsreaders, such as my old copy of Agent 2.0, brackets
>> aren't a workaround for a deficient client. They are simply markers
>> to let the composition window know that the configured line length
>> value should be ignored for text between the brackets.
>
>Ah, it's a sender's client trick to prevent line breaks. Understood.
>In my NNTP client, I don't need to do that for URLs as they kept intact
>in one physical line; however, it does have a Word Wrap toggle that I
>can click to insert a composition marker (not in the sent copy) to keep a long string from line wrapping. I use it occasionally, like for a wide data table where line wrapping makea it unintelligible. (I used it on this line as an example.)
>It keeps the long string as one long physical line. The reader's client
>might enforce line splitting at their configured line length. Nothing I
>can do about that. My tricks sounds similar to your bracketing trick.

Agreed. When composing a post with a long URL, we use whatever 'trick' or
method that particular client allows or requires in order to override the
configured line length. For me it's angle brackets and for you it could be
the word wrap toggle.

With my client, the angle bracket method only works in specific
circumstances. For example, it seems to require the "http://" string or the
"www." string, and I'm not sure what else. With my client, the angle
brackets become part of the post, and thus travel with the post. In your
case, there's nothing added to the post and thus nothing extra travels with
the post.

Note that some clients don't require any of this, from what I hear. They
simply recognize URLs and do the right thing, not just during composition
but also during subsequent reading, even when URLs are split across lines.
When my client encounters a split URL, it doesn't recognize the entire URL
and requires me to remove the line break(s) and any quote markers. That's
easy to do, but it's not automatic.

>However, I've seen those long strings in a long line include the angle
>brackets. They might be a hint in the composition window in the
>sender's client, but they were also included in the sent copy.

That's exactly how my client does it. The angle brackets become part of the
message.

>As a
>test, could you reply with a long string, like 200 characters, enclosed
>in your non-wrap markers, so I could see if the submitted copy has the
>non-wrap markers or not?

If it's regular text with spaces and no http:// or www. string, then angle
brackets aren't going to do anything special here. They would just be
included as text in the post. To do your test, I would have to use the left
angle bracket, either of the two URL markers, (there may be others that I
don't know), followed by text without spaces, (for me, a space marks the
end of a URL), ending with a right angle bracket.

Example:
<www.To%20do%20your%20test,%20I%20would%20have%20to%20use%20the%20left%20angle%20bracket,%20either%20of%20the%20two%20URL%20markers,%20(there%20may%20be%20others%20that%20I%20don't%20know),%20followed%20by%20text%20without%20spaces,%20(for%20me,%20a%20space%20marks%20the%20end%20of%20a%20URL),%20ending%20with%20a%20right%20angle%20bracket.>

That should post as a single line, but what happens in anyone's client when
they retrieve and view it is out of my hands.

[T]

On 2020-07-17 07:55, J. P. Gilliver (John) wrote:
> I know "could care less" is the US version of this expression, but it's
> inaccurate. Think about it: if you could care less, that implies that
> you do care a little - which is not what you mean; you actually mean
> "couldn't care less".

Interesting, so the one with the double negative is the correct one. My
publik skool education really sucked.

Also interesting, in America we say "that horse is different
FROM that one". In the UK they say "that horse is different
TO that one". Or so the shows from the UK on Netflix use it.

You will find this 5:49 video very interesting. This is
Simon Wistler, my fourth favorite Brit:

The Truth About the Split Infinitive:
https://www.youtube.com/watch?v=qc3cgdjUWI8

One of my customers just hired a English grad. To make
small talk with her, I asked her for her take on the
Split Infinitive. I could tell she barely knew
what I was talking about. I quickly changed the
subject as I could tell it made her uncomfortable

[T]

He is a bit of a blow hard. But he does provide
great information at times. I got "Debloat 10"
from him. It does perk up Windows 10.

[Char Jackson]

On Fri, 17 Jul 2020 15:55:00 +0100, "J. P. Gilliver (John)"
<G6...@255soft.uk> wrote:

>>On Thu, 16 Jul 2020 at 22:35:33, T <T...@invalid.invalid> wrote:
>[]
>>>It is the apps the customer cares about. They could
>
>couldn't
>
>>>care less if they were run int Flying Zucchini OS, if
>>>it ran their apps.
>[]
>I know "could care less" is the US version of this expression, but it's
>inaccurate. Think about it: if you could care less, that implies that
>you do care a little - which is not what you mean; you actually mean
>"couldn't care less".

Please don't attribute that mangled expression to all of us over here. :)

[Paul]

It's a totally different experience if you youtube-dl it,
then scroll through the boring parts. fedora.mkv 148,284,273 bytes

This streaming idea is never going to catch on.

And nobody will ever need more than 640K.

Paul

[T]

On 2020-07-17 00:54, VanguardLH wrote:
> T <T...@invalid.invalid> wrote:

> No initiative, no talent for testing, and poor writing skills.

I have the poor writing skills myself. I am a product of America's
Publik Skool system. Fortunately I have lots of talent and initiative.
My wife taught me to write in college.
Publik Skool taught "look see" instead of "phonics". I swear
at times I have to look at my drivers license in order to
spell my name correctly.

And college was no better. The English department taught "Creative
Writing". And they frequently made fun of the
sports department. Pardon me, but the odds of me
being a successful novel writer are worse than me trying
out and being accepted by the Green Bay Packers!

Fortunately, the business department threw a required
by engineers course in business writing. ON MY GOD!
I nearly lost 15 pounds in the course. It was the
hardest course I took. And years and years later,
the most useful course I took. My final exam was a
50 page proposal for my senior project that both
the teacher and my engineering counselor graded
and divided in half. About killed me.

> I remember someone remarking that college isn't about training their
> students for a particular job. It's to train them on how to learn.

In America it is looked at differently. In engineering,
you only learn about 15% of what you need to know
from college. The rest you learn on the job. What
the engineering degree tells your employer is
that your are will and capable of starting a long term
project that involves considerable personal hardship
and stick it through to the end. They can count
on you to tackle a project and finish it. You
are not a snow flake.

I was lucky that I went to a "teaching" university,
rather than a "research" university. My university
had a lab for every engineering course, except
for the ones with two labs. We were highly sought
after by industry because we knew which end of the
soldering iron was hot and hit the ground running.
I was told several times by scouts that our
teaching university cut two years off of the training
time required to get a research grad up to speed.

The computer grad I spoke of was from a research
university (UCLA).

One of my customers had her niece with a recent
degree in architecture come work for her.
She could not read a blue print. On questioning,
she said what she learned about was a bunch of
pretty buildings. Her's was a research university
too.

Do you know if your interns came from a teaching or
a research university?

[VanguardLH]

http://www.youtube.com/watch?v=RZf2dik9DlM&t=1m4s
"We're still not at performance parity with Windows."

https://www.youtube.com/watch?v=onOiOs-z0ws&t=10m22s
"Generally, the performance is a bit lower [on Linux] than the same
hardware on Windows."

Those videos were uploaded to YT back in mid-2018. Being a little bit
slower doesn't matter on a lot of games because not everyone plays
hyper-anxiety super-fast changing video games. Reminds me of Paul's
response: "Are we playing 'Soduko' yet?"

I feel we're delving too much into Linux in a newsgroup for Windows.
Linux is great for some things, not for everything. Same for Windows.
Same for Android. That's the nature of general purpose operating
systems.

[J. P. Gilliver (John)]

On Fri, 17 Jul 2020 at 12:49:31, T <T...@invalid.invalid> wrote:
>On 2020-07-17 07:55, J. P. Gilliver (John) wrote:
>> I know "could care less" is the US version of this expression, but
>>it's inaccurate. Think about it: if you could care less, that implies
>>that you do care a little - which is not what you mean; you actually
>>mean "couldn't care less".
>
>Interesting, so the one with the double negative is the correct one. My
>publik skool education really sucked.

Well, don't take my word for it: think about it. If you disagree with my
explanation, keep using the other one! I certainly don't claim to be
correct all the time.

>
>Also interesting, in America we say "that horse is different
>FROM that one". In the UK they say "that horse is different
>TO that one". Or so the shows from the UK on Netflix use it.
>

Interesting: I always thought Brits argued about whether it was
"different to" or "different from", until a leftpondian said "different
than", then the Brits ganged up against that!
[]

>One of my customers just hired a English grad. To make
>small talk with her, I asked her for her take on the
>Split Infinitive. I could tell she barely knew
>what I was talking about. I quickly changed the
>subject as I could tell it made her uncomfortable
>

(a) Grammar isn't _that_ much taught here either [not even in "grammar
schools", I think!], and (b) the split infinitive isn't actually wrong,
except in some grammar books that came out around the turn of the (last)
century - but those writing them didn't give any explanation _why_ they
considered it wrong.
>
My view is that it's best to avoid it if you can - but not to make
extremely clumsy circumstances in order to do so.

Of course, the most famous split infinitive was nicely parodied by
Douglas Adams, in (at least the original radio and TV versions of)
THHGTTG:


Far back in the mists of ancient time, in the great and glorious days of
the former Galactic Empire, life was wild, rich and largely tax free.

Mighty starships plied their way between exotic suns, seeking adventure
and reward amongst the furthest reaches of Galactic space. In those days
spirits were brave, the stakes were high, men were real men, women were
real women, and small furry creatures from Alpha Centauri were real
small furry creatures from Alpha Centauri. And all dared to brave
unknown terrors, to do mighty deeds, to boldly split infinitives that no
man had split before- and thus was the Empire forged.

--
J. P. Gilliver. UMRA: 1960/<1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

I hope you dream a pig.

[J. P. Gilliver (John)]

Very sorry! Glad it's not universal in US. But I haven't seen it at all
used in UK.

--
J. P. Gilliver. UMRA: 1960/<1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

[VanguardLH]

"J. P. Gilliver (John)" <G6...@255soft.uk> wrote:

> VanguardLH <V...@nguard.LH> wrote:
>
>> I think we used interns for 6 months: the contract length. Never
>> again thereafter.
>

> Why did you use them in the first place - was it because of some form of
> state subsidy, of about that duration?

I wasn't in the decision loop. At a weekly status meeting, our manager
said, "Guess what?" followed by a groan. Whenever he said that, we'd
get quiet waiting for yet another edict from management.

Training (the same group that taught us) took care of classes,
instructional CDs, and documentation but QA did the progress monitoring
and tutoring to make sure the interns got up to speed in 2 weeks. We
stretched it to 3 weeks to also train them on our QA procedures. They
did okay during the training phase probably because it was similar to
school, but more accelerated, like a seminar. However, When they were
on their own to do the actual testing, and despite the retro tests were
complete (no decisions to make, and previously reviewed with feedback
from outside our group to make sure any tech could follow them), was
when they got, um, slow and "lost". We changed from weekly status
meetings to still doing those but with me going around to everyone (not
just interns) to get a daily status update to see who needed more help,
discover any snags, talk to the boss about possible resource
reallocation, or gauge the severity of peril to our testing schedule.

Maybe our expectation was too high. We had programmers that left
because they couldn't take the stress or didn't have the flair for
digging into a product to thoroughly test it. I got offered a
programming position but declined because it was too boring. We had
expection of getting and training new-hires, just as we were once, but
the interns just never became adept. Could be they knew they were going
back to just school and they'd be leaving us hence no motivation for
long-term motivation (although there was the prospect of getting hired
if they performed well). Would you keep going to the gym to stay
healthy if you knew you were getting killed in 6 months?

[VanguardLH]

T <T...@invalid.invalid> wrote:

> And college was no better.

I *hated* college. So freaking s-l-o-w. 3 months to read a book and
take a test. Really? C'mon, that's ridiculous. And the class didn't
even cover the entire textbook. Then they switched from quarters to
semesters, so even longer to be bored. I tested out of as many classes
as they permitted, and they wouldn't let me test out of more. They
claimed the classroom experience must also be included for a
well-rounded education. Yeah, sit in a chair and listen to a prof orate
a textbook, and his oration interrupted by stupid questions.

[Brian Gregory]

On PCs maybe.

I bet some embedded stuff for ultra cheap mass market stuff is still
done in assember, or something only very slightly higher level.

--
Brian Gregory (in England).

[J. P. Gilliver (John)]

On Fri, 17 Jul 2020 at 19:32:04, VanguardLH <V...@nguard.LH> wrote:
>T <T...@invalid.invalid> wrote:
>
>> And college was no better.
>
>I *hated* college. So freaking s-l-o-w. 3 months to read a book and
>take a test. Really? C'mon, that's ridiculous. And the class didn't
>even cover the entire textbook. Then they switched from quarters to
>semesters, so even longer to be bored. I tested out of as many classes
>as they permitted, and they wouldn't let me test out of more. They
>claimed the classroom experience must also be included for a

There is _something_ in that ...

>well-rounded education. Yeah, sit in a chair and listen to a prof orate
>a textbook, and his oration interrupted by stupid questions.

... but you do remind me of the definition of a lecture as "a means for
the text to pass from the notes of the lecturer into the notes of the
students without passing through the minds of either".

--
J. P. Gilliver. UMRA: 1960/<1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

It's a beta orgy, not a product. - Mayayana in alt.windows7.general, 2018-3-8

[Brian Gregory]

On 17/07/2020 20:49, T wrote:
> Also interesting, in America we say "that horse is different
> FROM that one".  In the UK they say "that horse is different
> TO that one".  Or so the shows from the UK on Netflix use it.

Utter rubbish.
I've never said different to in my life.
Sounds daft.
Just as daft as saying I'm excited for Christmas.
It's supposed to be I'm excited about Christmas.

[J. P. Gilliver (John)]

Yes. Define "need". Compact code is noticeably more efficient - so runs
faster. Yes, for a lot of things, the returns don't justify the effort -
for a lot of things that are only done once, or where speed doesn't
matter, or - these days - to _some_ extent where modern processor power
can hide the inefficiency of the code.

I suspect IrfanView, for example, is mostly coded in either assembler,
or at least quite low-level code (or just possibly using an excellent
optimiser - which are rare with ultra-high-level languages, such as
scripting interpreters).

--
J. P. Gilliver. UMRA: 1960/<1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

[J. P. Gilliver (John)]

surprised, hopefully, comprise/consist, myriad ...

much/many/less/few(er) ...

then there is the US/UK different use of small auxiliary words, above
all "up" ... but also of, with, to ...

--
J. P. Gilliver. UMRA: 1960/<1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

[nospam]

In article <hnf0un...@mid.individual.net>, Brian Gregory

<void-invalid...@email.invalid> wrote:

> >> I remember - I _think_ it was in the last decade, but it might have been
> >> more - being startled when I spoke to a young computing graduate, to
> >> find he'd never done any assembler. At that time, after my initial
> >> double-take, I thought to myself: the field is big enough, that there'll
> >> be plenty of room for him, and in practice he'll probably never have any
> >> trouble finding interesting and well-paid employment.
> >
> > there is no need for assembler anymore, except in very rare
> > circumstances.
> >
>
> On PCs maybe.

on just about everything.

it's *really* difficult for a human to write assembler that's better
than what a modern compiler can produce, plus doing so would take a
*lot* longer.

> I bet some embedded stuff for ultra cheap mass market stuff is still
> done in assember, or something only very slightly higher level.

you'd lose that bet.

[nospam]

In article <v83OOyp+...@255soft.uk>, J. P. Gilliver (John)

<G6...@255soft.uk> wrote:

> >>> I remember - I _think_ it was in the last decade, but it might have
> >>>been
> >>> more - being startled when I spoke to a young computing graduate, to
> >>> find he'd never done any assembler. At that time, after my initial
> >>> double-take, I thought to myself: the field is big enough, that there'll
> >>> be plenty of room for him, and in practice he'll probably never have any
> >>> trouble finding interesting and well-paid employment.
> >> there is no need for assembler anymore, except in very rare
> >> circumstances.
> >>
> >
> >On PCs maybe.
> >
> >I bet some embedded stuff for ultra cheap mass market stuff is still
> >done in assember, or something only very slightly higher level.
> >
> Yes. Define "need". Compact code is noticeably more efficient - so runs
> faster. Yes, for a lot of things, the returns don't justify the effort -
> for a lot of things that are only done once, or where speed doesn't
> matter, or - these days - to _some_ extent where modern processor power
> can hide the inefficiency of the code.

compact code is not necessarily more efficient. a simple example is an
unrolled loop, which is less compact yet will run faster. another is
having a large lookup table versus calculating a value each time.

however, modern processors are far from simple, which is one reason why
a compiler can do a better job of optimization than humans can.

and then there's the issue of portability. anything written in a high
level language can be recompiled for another processor, normally with
little to no problem, whereas anything in assembler would need to be
entirely rewritten from scratch.

> I suspect IrfanView, for example, is mostly coded in either assembler,
> or at least quite low-level code (or just possibly using an excellent
> optimiser - which are rare with ultra-high-level languages, such as
> scripting interpreters).

it's highly unlikely any of it is in assembler.

[Jeff Barnett]

On 7/17/2020 8:30 PM, nospam wrote:
> In article <v83OOyp+...@255soft.uk>, J. P. Gilliver (John)
> <G6...@255soft.uk> wrote:
>
>>>>> I remember - I _think_ it was in the last decade, but it might have
>>>>> been
>>>>> more - being startled when I spoke to a young computing graduate, to
>>>>> find he'd never done any assembler. At that time, after my initial
>>>>> double-take, I thought to myself: the field is big enough, that there'll
>>>>> be plenty of room for him, and in practice he'll probably never have any
>>>>> trouble finding interesting and well-paid employment.
>>>> there is no need for assembler anymore, except in very rare
>>>> circumstances.
>>>>
>>>
>>> On PCs maybe.
>>>
>>> I bet some embedded stuff for ultra cheap mass market stuff is still
>>> done in assember, or something only very slightly higher level.
>>>
>> Yes. Define "need". Compact code is noticeably more efficient - so runs
>> faster. Yes, for a lot of things, the returns don't justify the effort -
>> for a lot of things that are only done once, or where speed doesn't
>> matter, or - these days - to _some_ extent where modern processor power
>> can hide the inefficiency of the code.
>
> compact code is not necessarily more efficient. a simple example is an
> unrolled loop, which is less compact yet will run faster. another is
> having a large lookup table versus calculating a value each time.

That depends on the length of the loop body and the number of
iterations: crossing cache boundaries can extract performance hits. In
fact, unrolling some loops can punish performance by factors of tens or
even hundreds. If you carry your argument far enough and force paging
you may cause arbitrary slow downs. As to look up tables the problems
are the same. If all the values and indices fit in a small amount of
storage (i.e. they are COMPACT), well and good. Otherwise all hell can
break loose.

> however, modern processors are far from simple, which is one reason why
> a compiler can do a better job of optimization than humans can.

Almost always true and more so as the years go by.

> and then there's the issue of portability. anything written in a high
> level language can be recompiled for another processor, normally with
> little to no problem, whereas anything in assembler would need to be
> entirely rewritten from scratch.

Unless you are coding in C or one of its derivatives. There are enough
places where a C compiler can choose the meaning of an expression (and
not necessarily consistently) that you have to code in a modest subset
of the language. And find a compatible compiler for portability.

>> I suspect IrfanView, for example, is mostly coded in either assembler,
>> or at least quite low-level code (or just possibly using an excellent
>> optimiser - which are rare with ultra-high-level languages, such as
>> scripting interpreters).

Scripting interpreters are not a good example of an "ultra-high-level"
anything. Most are, in fact, among the worst documented, least reliable,
and least consistent from version to version of any software meant to be
used by non developers of that software. Of course you may argue that
they are completely consistent since documentation typically gives no
serious details of what is supposed to be done in boundary cases.

> it's highly unlikely any of it is in assembler.
>

--
Jeff Barnett

[T]

Well stated.

When I hit college, I was a vet. I had no patience
with condescending, arrogant, pompous blowhards.
They taught me differently than the did the children
that also attended the classes. I was responsible
for getting two teacher removed for not doing
their jobs.

Several teachers said they liked teaching vets because
there was no subterfuge. We were there for a reason
the EXPECTED them to do their part. No having to
herd cats.

I also hate(d) school. The first day on campus after
six years of military service and one medal, when I
stepped foot on campus, I literally shook. "What in
God name am I doing to myself?" I graduated in 4-1/4
years whilst working full time WITHOUT ONE SINGLE
SOLITARY EXTRA CREDIT. I was working as a full
engineer in my 3rd year of college.

[T]

We build little computers with parts. Then went to
CPU. All assembly code!

[T]

In my college, they mixed the upper tier and lower tier
courses. Electronics 101 had about 200+ folks in it.
Standing room only. At the end, there was only about
15 of us. We all knew each other by name

[T]

On 2020-07-17 21:06, Jeff Barnett wrote:
> Scripting interpreters are not a good example of an "ultra-high-level"
> anything. Most are, in fact, among the worst documented, least reliable,
> and least consistent from version to version of any software meant to be
> used by non developers of that software. Of course you may argue that
> they are completely consistent since documentation typically gives no
> serious details of what is supposed to be done in boundary cases.

Perl 5 has wonderful documentation (PerlDocs).

Raku's (Perl 6's) absolutely suck. They are
written as a refresher for those that already
know what they are doing. They are not completely
useless, just really close to useless. And often
have mistakes in them.

I program a lot in Raku. I love the language,
but not the docs.

[nospam]

In article <retsfc$rpo$1...@dont-email.me>, Jeff Barnett <j...@notatt.com>
wrote:

> >>> I bet some embedded stuff for ultra cheap mass market stuff is still
> >>> done in assember, or something only very slightly higher level.
> >>>
> >> Yes. Define "need". Compact code is noticeably more efficient - so runs
> >> faster. Yes, for a lot of things, the returns don't justify the effort -
> >> for a lot of things that are only done once, or where speed doesn't
> >> matter, or - these days - to _some_ extent where modern processor power
> >> can hide the inefficiency of the code.
> >
> > compact code is not necessarily more efficient. a simple example is an
> > unrolled loop, which is less compact yet will run faster. another is
> > having a large lookup table versus calculating a value each time.
>
> That depends on the length of the loop body and the number of
> iterations: crossing cache boundaries can extract performance hits. In
> fact, unrolling some loops can punish performance by factors of tens or
> even hundreds. If you carry your argument far enough and force paging
> you may cause arbitrary slow downs. As to look up tables the problems
> are the same. If all the values and indices fit in a small amount of
> storage (i.e. they are COMPACT), well and good. Otherwise all hell can
> break loose.

they were simple examples to make a point. they were not meant to be a
treatise on every possible scenario.

> > however, modern processors are far from simple, which is one reason why
> > a compiler can do a better job of optimization than humans can.
>
> Almost always true and more so as the years go by.

yep.

> > and then there's the issue of portability. anything written in a high
> > level language can be recompiled for another processor, normally with
> > little to no problem, whereas anything in assembler would need to be
> > entirely rewritten from scratch.
>
> Unless you are coding in C or one of its derivatives. There are enough
> places where a C compiler can choose the meaning of an expression (and
> not necessarily consistently) that you have to code in a modest subset
> of the language. And find a compatible compiler for portability.

c is not an issue unless the programmer makes mistakes and ignores any
compiler warnings.

> >> I suspect IrfanView, for example, is mostly coded in either assembler,
> >> or at least quite low-level code (or just possibly using an excellent
> >> optimiser - which are rare with ultra-high-level languages, such as
> >> scripting interpreters).
>
> Scripting interpreters are not a good example of an "ultra-high-level"
> anything. Most are, in fact, among the worst documented, least reliable,
> and least consistent from version to version of any software meant to be
> used by non developers of that software. Of course you may argue that
> they are completely consistent since documentation typically gives no
> serious details of what is supposed to be done in boundary cases.

irfanview isn't using a scripting interpreter.

[nospam]

In article <reu16r$dni$2...@dont-email.me>, <T...@invalid.invalid> wrote:

>
> We build little computers with parts. Then went to
> CPU. All assembly code!

that doesn't make any sense.

[VanguardLH]

One day a prof arrived just before his class, put his hat on his desk,
but got busy elsewhere. When he got back and very late for class, all
the students had left. The next day he declared, "When my hat is on my
desk, that's the same as I'm here." The next day the students came in,
put hats on their chairs, and left.

[Andy Burns]

nospam wrote:

> J. P. Gilliver wrote:
>
>> startled when I spoke to a young computing graduate, to
>> find he'd never done any assembler.
>

> there is no need for assembler anymore, except in very rare
> circumstances.

I know that *I* would have a much poorer understanding of how computers
work if I hadn't cut my teeth with assembler (and BASIC) on 8 bit micros.

[Andy Burns]

J. P. Gilliver (John) wrote:

> Char Jackson wrote:

>
>> J. P. Gilliver (John) wrote:
>>
>>> I know "could care less" is the US version of this expression
>>

>> Please don't attribute that mangled expression to all of us over here. :)
>
> Very sorry! Glad it's not universal in US. But I haven't seen it at all
> used in UK.

I must admit I did think all Americans used the "could care less"
variation. Similarly I thought all Americans used "I'm pissed" (which
in the UK means "I'm drunk") rather than "I'm pissed off" but I've
started to hear the latter more now.

[Brian Gregory]

Tiny CPUs often have tiny instruction sets that aren't well suited to
any high level language.

The first project I did on a PIC16C55 we did in assembler.
It was not really too difficult. The whole thing fitted on about 3 pages.

We did later get a "C compiler" for those chips but it was pushing it to
call the language "C". It was mostly the syntax of C but nothing but
static variables. Weird syntax to configure the chip the way you wanted.
I can't even remember how you did I/O on it; must have been another
weird extension to the syntax. It made it easier but not as much as you
might think.

[Rene Lamontagne]

Good One VanguardLH :-) :-)

Rene

[nospam]

In article <hng6qb...@mid.individual.net>, Brian Gregory

<void-invalid...@email.invalid> wrote:

> >>
> >> We build little computers with parts. Then went to
> >> CPU. All assembly code!
> >
> > that doesn't make any sense.
> >
>
> Tiny CPUs often have tiny instruction sets that aren't well suited to
> any high level language.

he didn't specify tiny cpus with tiny instruction sets or writing
software for them.

he said built computers with parts and *then* went to cpus.

computers built before there were cpus were not little. they were minis
and mainframes, which end users did not build.

> The first project I did on a PIC16C55 we did in assembler.
> It was not really too difficult. The whole thing fitted on about 3 pages.
>
> We did later get a "C compiler" for those chips but it was pushing it to
> call the language "C". It was mostly the syntax of C but nothing but
> static variables. Weird syntax to configure the chip the way you wanted.
> I can't even remember how you did I/O on it; must have been another
> weird extension to the syntax. It made it easier but not as much as you
> might think.

in other words, assembly code not needed.

[nospam]

In article <hnfrgs...@mid.individual.net>, Andy Burns

<use...@andyburns.uk> wrote:

>
> I know that *I* would have a much poorer understanding of how computers
> work if I hadn't cut my teeth with assembler (and BASIC) on 8 bit micros.

that doesn't give you a good understanding about modern processors,
which are far more complex than those 8-bit micros.

[Yousuf Khan]

On 7/16/2020 3:19 PM, VanguardLH wrote:
> Already pointed out: your "none of which are being used" is wrong. It
> is being used. Video games use it, and those are not rare on Windows
> platforms. Any game using DirectX 12 are utilizing AVX2. Scientific,
> statistical, financial, encryption, and other programs can use it. Any
> program using .NET Framework can use AVX. The latest versions of
> Prime95 are optimized to use AVX. While it is used to stress test, that
> was not its original or current intent which was to discover prime
> numbers. Is prime hunting something that home users do? Of course not,
> but it illustrates AVX *is* used.

No, there is a difference between "are utilizing" and "can use". You
used both terms in different sentences up above, probably because you're
not actually sure which one it is, and you wanted to CYA.

"Are utilizing" implies that the games have no choice in the matter, and
they are using AVX even if they don't know it. This would presumably
mean that AVX is being used within the DirectX 12 API itself, and
operates in the background regardless of direct utilization by the game
itself.

That is not how the DX12 API operates. It is a bare-metal API, allowing
the games themselves to control most low-level aspects of the visual
production. This is unlike DX11 the previous API, which was more
hands-on, controlling the low-level aspects. I could see if DX11 had
been further developed, they perhaps might have started to use AVX
within the API itself to help some aspects of performance. But they went
in the completely opposite direction with DX12, I really don't see how
AVX benefits the DX12 low-level API, as it's mainly just a series of
calls to the GPU. For DX12, the "can use" AVX is the appropriate term,
not the "are utilizing".

But even within the game itself over the level of the API, AVX
utilization is very rare. Games could use AVX even under DX11 or even
DX9 or 10 beforehand, but it just didn't use it in the API. There was no
prohibition of using AVX within the application itself. These days they
are more likely to pass off most FP calculations off to the GPU than to
try to do it within the CPU anymore, so AVX is dead in the water.

> https://www.tomshardware.com/reviews/stress-test-cpu-pc-guide,5461-2.html
> "By default, Prime95 automatically selects the newest instruction set
> extension, such as AVX, AVX2, or even AVX-512."
>
> Your claim AVX is not used is false.

On the contrary, this exactly proves my point. Torvalds was complaining
about how it's only being used in benchmarks but no real apps. Prime95
is exactly an example of a benchmark and stress testing app. Nobody is
actually using Prime95 for anything other than stress testing and
benchmarking. It's not like as if you're going to be finding any new
prime numbers with a PC anymore, those have now firmly entered the realm
of supercomputers/HPC.

Yousuf Khan

[Andy Burns]

nospam wrote:

> Andy Burns wrote:
>
>> I know that *I* would have a much poorer understanding of how computers
>> work if I hadn't cut my teeth with assembler (and BASIC) on 8 bit micros.
>
> that doesn't give you a good understanding about modern processors,
> which are far more complex than those 8-bit micros.

But it was a good foundation from where to keep up with progress to
newer processors, and higher level languages.

I just think if you chuck e.g. an rPi at a youngster today they won't
get the same understanding of it from logic gates upwards ...

[nospam]

In article <hngmdt...@mid.individual.net>, Andy Burns

<use...@andyburns.uk> wrote:

> >> I know that *I* would have a much poorer understanding of how computers
> >> work if I hadn't cut my teeth with assembler (and BASIC) on 8 bit micros.
> >
> > that doesn't give you a good understanding about modern processors,
> > which are far more complex than those 8-bit micros.
>
> But it was a good foundation from where to keep up with progress to
> newer processors, and higher level languages.

true, although it's not as relevant as you might think.

> I just think if you chuck e.g. an rPi at a youngster today they won't
> get the same understanding of it from logic gates upwards ...

that depends on what they do with it, although an arduino would be a
better choice.

[Andy Burns]

nospam wrote:

> Andy Burns wrote:
>

>> I just think if you chuck e.g. an rPi at a youngster today they won't
>> get the same understanding of it from logic gates upwards ...
>

> an arduino would be a better choice.

No disagreement there ...

[T]

On 2020-07-18 05:04, Brian Gregory wrote:
> On 18/07/2020 07:09, nospam wrote:
>> In article <reu16r$dni$2...@dont-email.me>, <T...@invalid.invalid> wrote:
>>
>>>
>>> We build little computers with parts.  Then went to
>>> CPU.  All assembly code!
>>
>> that doesn't make any sense.
>>
>
> Tiny CPUs often have tiny instruction sets that aren't well suited to
> any high level language.

That was not the purpose. It was to teach you how they worked.

> The first project I did on a PIC16C55 we did in assembler.
> It was not really too difficult. The whole thing fitted on about 3 pages.
>
> We did later get a "C compiler" for those chips but it was pushing it to
> call the language "C". It was mostly the syntax of C but nothing but
> static variables. Weird syntax to configure the chip the way you wanted.
> I can't even remember how you did I/O on it; must have been another
> weird extension to the syntax. It made it easier but not as much as you
> might think.

We never got much past an 8080. And some Arithmetic Log Unit
which I can't remember.

[T]

I second that!

My goodness th condescending, arrogant, pompous blow hards
we all had to suffer. But not all of them fortunately.
the best teaches were the part times that had jobs in
industry with some kind of work arrangement with their
employers. Now that is where the rubber met the road!

[T]

A good video on this is Superman and Hot Fuzz
discussing British Slang:

Henry Cavill and Simon Pegg Teach You English Slang:

https://www.youtube.com/watch?v=5eIj1nDUs9g

So in the UK would "I'm pissed off" mean you are
sober?

:-)

[T]

Oh and our GI's picked up the word "Crap", as in
"I am going to the crapper" from your guys:
Thomas Crapper & Co. Crapper's trade marks
were all over your crappers.

And GI's picked up "Piss", as in "I am going to
go take a piss", from the French word for the
crapper: pissoir.

Apparently our British Allies didn't quite
educates us properly on the latter.

[VanguardLH]

Yousuf Khan <bbb...@spammenot.yahoo.com> wrote:

> Prime95 is exactly an example of a benchmark and stress testing app.
> Nobody is actually using Prime95 for anything other than stress
> testing and benchmarking.

Yep, elide over the intent of the authors of Prime95, because that would
be another example of several shown where AVX is used.

[Paul]

If they're not careful, they're going to scare the kids
away from overclocking :-/

https://www.mersenneforum.org/showthread.php?t=24445

Paul

[T]

On 2020-07-17 07:55, J. P. Gilliver (John) wrote:
>> On Thu, 16 Jul 2020 at 22:35:33, T <T...@invalid.invalid> wrote:
> []
>>> It is the apps the customer cares about.  They could
>
> couldn't
>
>>> care less if they were run int Flying Zucchini OS, if
>>> it ran their apps.
> []
> I know "could care less" is the US version of this expression, but it's
> inaccurate. Think about it: if you could care less, that implies that
> you do care a little - which is not what you mean; you actually mean
> "couldn't care less".

My American trophy wife -- the ultimate word smith -- has
a "from and to" rule:

"I take something FROM you"

"I give something TO you"

[VanguardLH]

Andy Burns <use...@andyburns.uk> wrote:

> VanguardLH wrote:
>
>> I wonder how a car knows a gas sniffer is poking up its ahole. Oooh,
>> warm that up first before sticking it in. I suppose the car's computer
>> could notice the car wheels weren't rotating when the engine got revved
>> up and the steering wheel wasn't turning.
>
> A bit more complex than that, but basically spotting conditions of the
> standardised tests and switching into an alternate ECU mode
>
> <https://media.ccc.de/v/32c3-7331-the_exhaust_emissions_scandal_dieselgate>
>
> Jump to 57:00 if you just want the money shot, but the whole thing is
> worth a watch ...

That was interesting. I jumped to the timemark, but decided it was too
much out of context, so I watched it all. Interesting how the
alternative model was used for most of the driving (underdoses the NH3
to convert NO) but switched to the main model during the conditions of
NEDC testing. Man, they sure are tricky. Also interesting is that VW
isn't the only one, but BMW was caught a decade before but managed to
silently alter their model. Apparently lots of brands do this trickery.
It's almost a requirement for the car to be road-worthy, but still pass
the emission testing for when the car is tested under nothing like road
use of the vehicle.

[J. P. Gilliver (John)]

Not sure how that relates to the "care less" discussion. If it relates
to the debate whether to us different from, different to, or different
than, I'm not sure how it would help there either.
--
J. P. Gilliver. UMRA: 1960/<1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

At the end of the day, I wasn't asking to kill the pandas, I was simply asking
for an audit in terms of conservation resources, and I stand by every word.
-Chris Packham, quoted in Radio Times, 29 May - 4 June 2010

[Brian Gregory]

Well not needed yes but that compiler was a bit of a POS and we had to
pay quite a lot for it. For that chip it was arguably not worth it for us.

However for some of the slightly bigger chips it was definitely worth it.

[Andy Burns]

VanguardLH wrote:

> Interesting how the alternative model was used for most of the
> driving (underdoses the NH3 to convert NO) but switched to the main
> model during the conditions of NEDC testing. Man, they sure are
> tricky.

If you're going to set a test for vehicles, better make it a good test
because companies will spend a *lot* of money to try and game it ... I
believe WLTP has now replaced NEDC.

[T]

On 2020-07-19 07:09, J. P. Gilliver (John) wrote:
> On Sun, 19 Jul 2020 at 01:28:52, T <T...@invalid.invalid> wrote:
>> On 2020-07-17 07:55, J. P. Gilliver (John) wrote:
>>>> On Thu, 16 Jul 2020 at 22:35:33, T <T...@invalid.invalid> wrote:
>>> []
>>>>> It is the apps the customer cares about.  They could
>>>  couldn't
>>>
>>>>> care less if they were run int Flying Zucchini OS, if
>>>>> it ran their apps.
>>> []
>>> I know "could care less" is the US version of this expression, but
>>> it's  inaccurate. Think about it: if you could care less, that
>>> implies that  you do care a little - which is not what you mean; you
>>> actually mean  "couldn't care less".
>>
>> My American trophy wife -- the ultimate word smith -- has
>> a "from and to" rule:
>>
>> "I take something FROM you"
>>
>> "I give something TO you"
>>
>>
> Not sure how that relates to the "care less" discussion. If it relates
> to the debate whether to us different from, different to, or different
> than, I'm not sure how it would help there either.

It does not. It was just an interesting observation
between the US and the UK