On 06/08/2023 09:42, Jan Panteltje wrote:
> On a sunny day (Sun, 6 Aug 2023 08:51:31 +0100) it happened The Natural
> Philosopher <tnp@invalid.invalid> wrote in <uanje4$25fvd$1@dont-email.me>:
>
>> On 06/08/2023 05:01, 23k.304 wrote:
>>> On 8/5/23 4:01 AM, The Natural Philosopher wrote:
>>>> On 05/08/2023 05:20, 23k.304 wrote:
>>>>> On 8/4/23 4:58 AM, The Natural Philosopher wrote:
>>>>>> On 04/08/2023 07:03, 23k.304 wrote:
>>>>>>> Still have the old IBM Technical Reference Manual.
>>>>>>>    All that stuff was meticulously documented.
>>>>>>
>>>>>> I gave that away sometime in the late 80s.
>>>>>
>>>>>    I consider it to be a "valuable artifact" :-)
>>>>>
>>>>>    At minimum, it showed how serious and deep they'd
>>>>>    go with the early PCs. That's when programming was
>>>>>    still painfully "real". Mostly you could not buy
>>>>>    what you needed - so you had to make it yourself.
>>>>>
>>>>>    And I *did* find many uses for that info.
>>>> Oh, so did I. I was assembly programming 8086s then - writing BIOS
>>>> extensions or complete BIOSES for them,
>>>>
>>>>> https://www.youtube.com/watch?v=WUVZbBBHrI4
>>>
>>>   Shoulda gone with the Frenchies, they LOVE nuke !  :-)
>>>
>>>   Never went as far as a full BIOS. Leveraging what IS,
>>>   what's commercial, what's widespread and "standard"
>>>   was generally my approach. In the time it takes to
>>>   write a BIOS, esp solo, I could do ten times the stuff
>>>   with my approach.
>>>
>>>   But to each ...
>>>
>> I was being paid to get custom designed hardware to perform.
>> This minicomputer company wanted an 8088 board to control the other
>> boards in their super fault tolerant computer. So there it was, a bare
>> 8088 board with PROm and some RAM and they wanted to be able to load
>> code from an 8" floppy, and drive a serial monitor, oh and while you are
>> at it, write us a full C library in ROM and an FORTH interpreter -
>> here's the source code in a 150 page scan of some magazine article.
>>
>> That's 6 months of my life I wont get back. They were going to write a
>> multitasking basic operating system on top of that but my contract
>> expired before they got that far.
>>
>>
>>>
>>>> Or interfacing with special hardware people had built that needed
>>>> drivers written.
>>>> But that line of business dried up, so I ended up doing yet another
>>>> career shift into running IT businesses.
>>>
>>>   Once you COULD buy lots of neat stuff off the shelf,
>>>   then yea, a focus shift made sense. Never tried/wanted
>>>   to RUN businesses though - massive pain in the ass and
>>>   biz WAY too often winds up in the shitter.
>>>
>> No option. No one wanted a 43 year old self taught programmer. They
>> wanted bushy tailed youngsters who wrote C++ and had been schooled in
>> all sorts of theoretical shit that enabled them to justify writing
>> shitty bug filled code.
>>
>> When no one is offering jobs, it sensible to invent your own. It
>> started as technical support, but as it became apparent that the other
>> people in the business hadn't a clue how to actually run a business -
>> accounts, operations, technical support and so on, I ended up running
>> more than half the employees.
>>
>> I needed a job, and that was the only way I could guarantee the business
>> didn't fall to pieces with incompetence.
>>
>> Well it fell to pieces with ego, so a couple of us started another one,
>> and I ended up running most of that too.
>> When I had enough cash accumulated, and my co directors wife started
>> trashing the business, we sold it and I gave up work altogether.
>>
>> I'd frankly had enough. So I retired at 50.
>
>
> I had a TV repair shop for years during the late seventies and early eighties.
> Worked for the national TV network here before that
> and before that designed electronics for the among other things army and navy.
> After the repair shop did many things, often via agencies.
> Electronics is everywhere so gained knowledge of a lot of fields
> from missiles to medical (worked in a university hospital too) and space and physics (accelerator).
>
> Curiosity is likely what drives me.
>
>
>
>
>
>>
>>>   Was still programming in solder into maybe half a
>>>   dozen years ago - stuff you couldn't, or wouldn't,
>>>   buy off the shelf. However that stuff was for
>>>   microcontrollers, not real CPUs.
>>
>> I still solder stuff together. And I will be for this current Pi project
>> too. I need an internal mains power supply for the server - to hell with
>> USB wall warts - and I need a special bit of kit that turns on after up
>> to an hours delay, boots a pico and runs it until the PICO has done its
>> stuff and issues a shutdown signal on a GPIO pin, and puts the whole
>> thing to sleep at a couple of µA or so for up to an hour, to get a
>> decent battery life.
>>
>> Oddly enough the best solution to that looks to be not some specialised
>> complex chip, but an old school simple 4000 series CMOS Schmitt
>> trigger, some Rs and Cs, and three boring old school transistors. 1970s
>> technology in fact.
>>
>> In quiescent state, the only current draw will be leakage current on the
>> transistors charge current on the timing capacitor and the quiescent
>> current of the CMOS Schmitt, which at 4.5v is less than a µA...the best
>> I could do with clever chips like a CMOS 555 timer was around 16µA.
>>
>> If it works I might get some PCBS made up for that. Lots of people might
>> want a PICO for use as a remote sensor on batteries on an 'occasional'
>> basis.
>>
>> So old school it is. I am hoping for at least a years battery life on
>> the oil tank sensor. Even if that means only monitoring the level once
>> an hour or less.
>
> I have done a fluid level sensor with just a 4040 counter and an echo system
> 555 timer, sending a beep down to the surface and counting ticks before the echo returns
> Was in the eighties.
> Distance sensors are now less than 2 Euro or on ebay..
> https://www.ebay.co.uk/b/ultrasonic-distance-sensor/bn_7024750003
> done a lot of experiments with those ebay things too, for example wind speed measurements:
> http://panteltje.nl/pub/wind_speed_by_differential_2_ebay_distance_meters_IMG_4891.JPG
> http://panteltje.nl/pub/44kHz_radar_time_of_flight_test_in_wind_tunnel_IMG_4105.JPG
>
> But am from origin electronics hardware, programming came later..
> I do not always see the need for any Raspberry stuff in equipment, I use Microchip 18F14K22 chips
> that I program in asm.
> Those have ADCs, PWM generators, a simple DAC, counters, etc etc and sleep mode with only nano amps.
> For an other e dollars or so you can add an USB interface.
> For the 'network' its a bit more complicated, I need an extra chip
> As the ebay modules are so cheap who cares .
> https://www.ebay.co.uk/itm/285358235060
>
> PICs can drive a nice OLED or LCD display no problem:
> http://panteltje.nl/pub/gamma_spectrometer_plus_probe_plus_geiger_counter_2_IMG_4185.JPG
> https://panteltje.nl/panteltje/pic/scope_pic/
>
> Only external code I use is the 32 bit PIC asm math routines written by somebody else.
>
> No emulators no debuggers used ever.
> serial port and a scope is all I need.
> No million dollar scope, 10 MHz dual trace is enough.
>
> ASM is easy,
> C is OK.
> The rest I do not want to know about ;-)
> Seems these day you can ask AI to write the code...
> I wonder....
> :-)
>
>
We seem to be similar in many respects. I started out in electronics.
But as a designer, really the job died in the 1970s. After that is was
all large scale integrated chips, and there was no challenge to it. So I
taught myself software.

C to me was just a faster way to write assembler.
'if..{} else {} is quicker than MOV, CMP JNE, JMP etc etc

And the way C used the stack to allocate temporary variable space to
remove the need for complex memory management or else global memory
assignment was pure genius. I think that first appeared in ALGOL. It
allowed recursive functions to be simply coded.

On 06/08/2023 10:04, Jan Panteltje wrote:
> Bit more warming, mass migration, society as we know it may fall apart.
>
Most societal collapses seem to occur because one resource or another
that is so fundamental to a societies existence, runs out or just stops
for natural reasons.
It might be water, a supply of fish, or anything. Today its energy. Our
society is utterly dependent on energy derived from coal oil and gas.
And the so called 'renewables' are no replacement for it.

Coal enabled the UK population to rise from 5-10 million to 50 million,
oil and gas has taken us to 70-80 million. No one knows, there are so
many illegal immigrants these days . Renewable energy will take us back
to 5 million - or probably less, as we do not have a pre-existent system
set up to achieve anything at all with say horses or oxen,

Nuclear might take us to 100 million. But no one seems to want that. I
think they prefer to die.

> We may need to move / spread our species to other planets to protect civilization as we know it..

I don't think we 'need' to do anything. In a Godless world we find
Nature is supremely indifferent to our success or failure as a species.
What persists in this world is whatever lucks out and finds a *way* to
persist, for a time. As circumstances change, species change.

As a general species, we have become as unbelievably arrogant as we have
become unbelievably stupid. Or perhaps we always were, but civilization
now has been built by the clever few, us engineers, and the rest of the
population is simply incapable of understanding it or keeping it going,
and worse, doesn't even understand that their very lives depend on it.

We have built a society where even the most ill adapted and incompetent
can survive and prosper, and in that mood of compassion, have
inadvertently built a world full of people who are simply too stupid to
keep it going. And given them enough political power to destroy it.
Which they are doing.

Ah well. Darwin in action. We tried.

I won't live to see the worst of the collapse.

Its a shame. Id like to see all these unbelievably arrogant and stupid
people dying in screaming agony as the realisation dawns that everything
they were told or believed in was all a load of utter shit, and that no
matter how much they scream about their 'rights', and what the
government *ought* to be doing, there is simply no one left to give them
a single thought.

Life is rough, tough and desperately unjust. Life's a bitch, and then
you die.

It's a shame that these lessons are no longer taught.

--
When plunder becomes a way of life for a group of men in a society, over
the course of time they create for themselves a legal system that
authorizes it and a moral code that glorifies it.

Frédéric Bastiat

On a sunny day (Sun, 6 Aug 2023 10:34:22 +0100) it happened The Natural
Philosopher <tnp@invalid.invalid> wrote in <uanpeu$26dmj$1@dont-email.me>:

>On 06/08/2023 10:04, Jan Panteltje wrote:
>> Bit more warming, mass migration, society as we know it may fall apart.
>>
>Most societal collapses seem to occur because one resource or another
>that is so fundamental to a societies existence, runs out or just stops
>for natural reasons.
>It might be water, a supply of fish, or anything. Today its energy. Our
>society is utterly dependent on energy derived from coal oil and gas.
>And the so called 'renewables' are no replacement for it.
>
>Coal enabled the UK population to rise from 5-10 million to 50 million,
>oil and gas has taken us to 70-80 million. No one knows, there are so
>many illegal immigrants these days . Renewable energy will take us back
>to 5 million - or probably less, as we do not have a pre-existent system
>set up to achieve anything at all with say horses or oxen,
>
>Nuclear might take us to 100 million. But no one seems to want that. I
>think they prefer to die.
>
>> We may need to move / spread our species to other planets to protect civilization as we know it..
>
>I don't think we 'need' to do anything. In a Godless world we find
>Nature is supremely indifferent to our success or failure as a species.
>What persists in this world is whatever lucks out and finds a *way* to
>persist, for a time. As circumstances change, species change.
>
>As a general species, we have become as unbelievably arrogant as we have
>become unbelievably stupid. Or perhaps we always were, but civilization
>now has been built by the clever few, us engineers, and the rest of the
>population is simply incapable of understanding it or keeping it going,
>and worse, doesn't even understand that their very lives depend on it.
>
>We have built a society where even the most ill adapted and incompetent
>can survive and prosper, and in that mood of compassion, have
>inadvertently built a world full of people who are simply too stupid to
>keep it going. And given them enough political power to destroy it.
>Which they are doing.
>
>Ah well. Darwin in action. We tried.
>
>I won't live to see the worst of the collapse.
>
>Its a shame. Id like to see all these unbelievably arrogant and stupid
>people dying in screaming agony as the realisation dawns that everything
>they were told or believed in was all a load of utter shit, and that no
>matter how much they scream about their 'rights', and what the
>government *ought* to be doing, there is simply no one left to give them
>a single thought.
>
>Life is rough, tough and desperately unjust. Life's a bitch, and then
>you die.
>
>It's a shame that these lessons are no longer taught.

Yes, when I see those 'save the world' activist glue themselves to airport runways
screwing up people's holidays...
I would like to sent those to a re-education camp.
Teach them some technology, as technology is the only thing that is able to keep us
alive in the numbers we are now - when climate changes.
Most of them do not even know how to connect a light bulb.
Brainwashed by green fanatics.
OTOH world wars, Darwin, one ant heap against the other, what's new.
May the best one win!
Maybe I am lucky, born just after WW2 and we have lived in peace here for so long now.
Fear for nuclear... shutting down nuclear power plants.. fear for CO2, fear sells!
And so does snake oil, closing farms because cows produce CO2.
Here they let wolves multiply as that species 'must be protected', while the wolves attack the sheep and farmers...
You are a criminal if you shoot one!
We need hose farmers and farms for food!
World upside down, next poisonous snakes will be protected as more of those will appear as climate heats up?
Mosquitos?

So, some space ships and some colonies on some other planets or asteroids,
why not.
A thousand years ago nobody could fly as birds were considered lighter than air and we heavier...
I personally think life is self-assembling,
you have quarks and protons, neutrons, electrons, atoms , molecules , forming living cells, those form us, we make cellphones...
ever more complex.
Would be interesting to go 'out there' and see what else nature has come up with!!

:-) bit of luck and SETI decodes a nice alien show :-)

On Sun, 6 Aug 2023 10:11:43 +0100
The Natural Philosopher <tnp@invalid.invalid> wrote:

> And the way C used the stack to allocate temporary variable space to
> remove the need for complex memory management or else global memory
> assignment was pure genius. I think that first appeared in ALGOL. It
> allowed recursive functions to be simply coded.

Yes block structure arrived in programming with Algol.

The early documentation on procedure calls in Algol I was mind
numbingly strange because it did not refer to a stack - it was written in
terms of textually putting the subroutine code where the call was with
modifications to the variable names to achieve the same effect. We
covered in college, when it got to recursion most of the minds in the
lecture hall went into meltdown. I was lucky, I already knew about stacks
and spent most of the lecture marvelling at the baroque explanations.

Algol was OK but at college it was BCPL that I really liked working
in - we didn't have C because we did have Martin Richards.

I learned C one the job (from a photocopy of the reference section
of K&R first edition) writing a table driven remote debugging tool for folks
writing games for the early cartridge based consoles - the other part of
the job was making the fake cartridges with RAM for the game code, an
overlay EPROM for startup and a serial port to run it from.

One cool part of that job was the people - my boss was Nick Toop
who designed the Acorn Atom (I learned a lot from him!) and the owner of the
company was David Levy who only just got away with his cash challenge for
anyone with a chess program that could beat him - he set a five year time
limit, it took less than six!

We had a wonderfully eerie window display, it was a beautifully made
prototype of a chess machine which appeared to be no more than a very
thick board with normal pieces on it and a marked out space to store the
pieces off the board ... You played it by moving the pieces and the
computer played by sliding the pieces around the board (the jiggling to do
a knights move was fun to watch). There was nothing visible moving the
pieces. You could even put the pieces on the board randomly, hit the "setup"
button and watch them move silently into place all by themselves.

Inside the thick board was an stepper driven X-Y mechanism holding
an electromagnet, inside each piece was a bit of steel and a tuned L-C
passive resonator, I'm sure you can work out the details.

It didn't sell, they cut costs going into production and replaced
the carved wooden case and pieces with moulded plastic and the brass gears
with nylon ones - result it looked tacky and squeaked. The prototype looked
like an expensive chess set and was silent. IMHO they should have gone with
keeping the quality high, raising the price to luxury levels and providing
software updates for a price.

--
Steve O'Hara-Smith
Odds and Ends at http://www.sohara.org/
Host: Beautiful Theory meet Inconvenient Fact
Obit: Beautiful Theory died today of factual inconsistency

On Sun, 06 Aug 2023 12:33:58 GMT
Jan Panteltje <alien@comet.invalid> wrote:

> Yes, when I see those 'save the world' activist glue themselves to
> airport runways screwing up people's holidays...

These people are idiots - they have no idea how much needs to be
done before we can stop burning oil, coal and gas - which we have no
choice about we have to do that well before it runs out otherwise we'll
never survive the energy wars.

Nobody is going to permit nuclear until far too late (it probably
already is nuclear plants are not quick things to build) which leaves
renewables as the only game in town - we *have* to make them work! It is
physically possible to make them work given that we now have reliable flow
battery designs, the rest is simply a matter of engineering much of it
civil!

If we're really lucky something better might turn up while we're at
it but we can't count on that or a sudden embracing of nuclear power - any
engineer worth their salt knows you have to work with what you've got, it's
no use wishing for what you can't have.

--
Steve O'Hara-Smith
Odds and Ends at http://www.sohara.org/
Host: Beautiful Theory meet Inconvenient Fact
Obit: Beautiful Theory died today of factual inconsistency

On Sun, 6 Aug 2023 10:34:22 +0100
The Natural Philosopher <tnp@invalid.invalid> wrote:

> Coal enabled the UK population to rise from 5-10 million to 50 million,

Pity about what it did to their lungs but yes.

> oil and gas has taken us to 70-80 million. No one knows, there are so
> many illegal immigrants these days . Renewable energy will take us back
> to 5 million - or probably less, as we do not have a pre-existent system
> set up to achieve anything at all with say horses or oxen,

Renewable energy requires two things in order to handle the load,
it needs a LOT of wind, tide, solar ... plants building and it needs a lot
of mid to long term efficient energy storage, enough to run everything out
of storage for at least a week and preferably ten days (globally the
requirements are best measured in terawatt hours and the numbers aren't
small).

The plants can be built quickly despite the NIMBY brigade - quickly
enough I don't know but it seems likely. I really like the floating solar
plants that are starting to happen - the panels work a lot better when
they're water cooled and you could even put a water source heat pump under
them.

The storage is the big one - there's nowhere near enough gravity
based storage options around, there's nothing like enough lithium either
and don't even think about lead/acid. If those were all the options I'd be
agreeing with you wholeheartedly but ...

There is plenty of vanadium, iron and chlorine for the current crop
of grid scale flow batteries - they tick a lot of boxes, long live
electrodes, zero (yes zero) self discharge (unless the tanks leak), cheap
and safe. We'd need between two hundred and a thousand litres of
electrolyte (depends on the chemistry) per person to hold ten days of two
hundred watts average each.

The task is huge but it is not too huge - it can be expected to
take decades and it is already happening as fast as production can go. When
flow battery electrode assemblies hit the DIY market things will really
be moving. I expect that within a decade followed by IBLCs getting hard to
obtain.

> Nuclear might take us to 100 million. But no one seems to want that. I
> think they prefer to die.

There is precisely no chance of getting enough nuclear plants built
before we all die and what do we do when that fuel runs out ?

The sun is good for the next five billion years, the core heat
might give out sooner but not much and the moon's orbital momentum won't
even notice if we stop the tides globally.

--
Steve O'Hara-Smith
Odds and Ends at http://www.sohara.org/
Host: Beautiful Theory meet Inconvenient Fact
Obit: Beautiful Theory died today of factual inconsistency

On Sun, 6 Aug 2023 08:21:04 +0100, The Natural Philosopher wrote:

> On 06/08/2023 02:45, Computer Nerd Kev wrote:
>> In comp.os.linux.misc The Natural Philosopher <tnp@invalid.invalid>
>> wrote:
> When you are looking at second hand SD cards plus postage, the cost
> benefit analyis told me I should pay the money for whatever was the
> current best price on new stuff , with plenty of life left, from the
> same shop that was shipping me the Pi. To leverage the postage. and that
> turns out to be 16GB. Why, I don't have a clue. I could ask an old
> college friend who used to work designing NAND flash in California, but
> I am not sure I give a rat's arse why.
>
> Again, you asses a situation and what you want, and work out the least
> cost and effort to arrive at that target. If I could get 2GB cards at 5
> for £10 I'd use those, But I cant.
>
> I was interested because I don't remember ever having an SD card for my
> ancient camera that was that small (256MB).

I use 8GB Samsung in my cameras and flight recorders. If you buy branded
cards from brands that are known the own their own chip fab, then you're
less likely to end up with some chunk of junk with memory to hold just a
fraction of its advertised capacity.

--

Martin | martin at
Gregorie | gregorie dot org

On 06/08/2023 14:51, Ahem A Rivet's Shot wrote:
> On Sun, 06 Aug 2023 12:33:58 GMT
> Jan Panteltje <alien@comet.invalid> wrote:
>
>> Yes, when I see those 'save the world' activist glue themselves to
>> airport runways screwing up people's holidays...
>
> These people are idiots - they have no idea how much needs to be
> done before we can stop burning oil, coal and gas - which we have no
> choice about we have to do that well before it runs out otherwise we'll
> never survive the energy wars.
>
> Nobody is going to permit nuclear until far too late (it probably
> already is nuclear plants are not quick things to build) which leaves
> renewables as the only game in town - we *have* to make them work! It is
> physically possible to make them work given that we now have reliable flow
> battery designs, the rest is simply a matter of engineering much of it
> civil!

No, it is not possible to make them work. Not sustainably anyway. I have
spent years studying this and that is the simple conclusion

>
> If we're really lucky something better might turn up while we're at
> it but we can't count on that or a sudden embracing of nuclear power - any
> engineer worth their salt knows you have to work with what you've got, it's
> no use wishing for what you can't have.
>
There will be either a sudden embracing of nuclear power, after the
lights go out in the white house, or the collapse of civilisation as we
know it.

There are no other viable scenarios

--
To ban Christmas, simply give turkeys the vote.

On 05/08/2023 22:14, Martin Gregorie wrote:
> On Sat, 5 Aug 2023 15:23:21 +0100, Ahem A Rivet's Shot wrote:
>
>> On Sat, 5 Aug 2023 14:36:08 +0100 Pancho <Pancho.Jones@proton.me> wrote:
>>
>>> I was actually surprised at how fast the new Ryzen was. Maybe it is
>>> time to buy a new PC.
>>
>> I haven't done that in a long time. I used to buy parts and make
>> them but these days I buy "refurbished"* ex-corporate machines which are
>> cheap and far better made than the typical consumer PC.
>>
>> * Read cleaned, OS reinstalled and stripped of asset tags.
>
> It may well be worth looking at getting a new PC.
>
> Since around 2005 my house server was a Dual Athlon PC in a noname case
> and not new when I got it in 2005 and has had at least two new disks (bog
> standard WD 500MB blue) since then.

I think it all depends, what you want. If you don't already have a PC
less than 10 years old, then second-hand deals are good.

In most respects I found a 10-year-old 2500K, OK to use. Up until a
couple of years ago it was my desktop, then its graphics card broke, and
it got relegated to HTPC, it wasn't so good at that. I've just replaced
it with an Orange Pi 5, with caveats, the Orange Pi 5 is better as a TV
computer. The Orange Pi 5 was cheap. I've also bought a couple of low
power new NUC types for about £200, for relatives, e.g. Intel N5150,
they are generally happy with them.

Mainly I've been interested in small low wattage PCs, it is only
recently that they have been quicker than my old 2500K. The new big
desktops were quick, but not that much quicker, maybe twice as fast, for
a reasonable price. Now, the Ryzen 5 7600 is looking to be 3-4 times as
fast, single thread, more than 3-4 multi thread. Which might be
noticeable, for me, I don't know. Without having one, it is hard to know.

On Sun, 06 Aug 2023 12:33:58 GMT, Jan Panteltje wrote:

> Teach them some technology, as technology is the only thing that is able
> to keep us alive in the numbers we are now - when climate changes.
> Most of them do not even know how to connect a light bulb.
> Brainwashed by green fanatics.
>
Take Kathmandu in Nepal: (I've been there twice - both by road in the
'70s). Then Swayambhu was a 20 minute walk through paddy fields from the
centre of Kathmandhu and to took as bus to Braktapur, at the east end of
the Kathmandhu Valley. Now Kathmandu occupies virtually the while damn
valley.

Some serious population reduction would be a good idea. Worldwide, but
maybe climate change will solve that problem Real Soon Now

Teach people some limits too: the world was a much nicer place in the '70s
with the population on the 2 Billion region and the number of megacities
could still be counted on one hand.

On Sun, 6 Aug 2023 14:51:10 +0100, Ahem A Rivet's Shot wrote:

> Nobody is going to permit nuclear until far too late (it probably
> already is nuclear plants are not quick things to build) which leaves
> renewables as the only game in town - we *have* to make them work!
>
Besides, the non-thinking multitude won't allow them to be built.

> It is physically possible to make them work given that we now have
> reliable flow battery designs, the rest is simply a matter of
> engineering much of it civil!
>
Yes: I've known about flow batteries for at least 20 years and they seem
like an ideal power source, so why aren't they in widespread use already?
> If we're really lucky something better might turn up while we're
at
> it but we can't count on that or a sudden embracing of nuclear power -
> any engineer worth their salt knows you have to work with what you've
> got, it's no use wishing for what you can't have.

Yep, and I'm not holding my breath for fusion to arrive any time soon.

--

Martin | martin at
Gregorie | gregorie dot org

On Sun, 6 Aug 2023 15:06:37 -0000 (UTC)
Martin Gregorie <martin@mydomain.invalid> wrote:

> Yes: I've known about flow batteries for at least 20 years and they seem
> like an ideal power source, so why aren't they in widespread use already?

It is only quite recently that there have been electrodes that don't
either degrade quickly, poison the main reaction, have too many unwanted
side reactions, need regular difficult maintenance or are otherwise unusable
commercially.

The principle is easy but the details are fiddly.

Currently there are only two good electrode designs both subject to
commercial patent and involving fairly complex chemistry and construction.
Both are in commercial use.

One works with a vanadium based electrolyte and is rated to run
for twenty years before it degrades too much and another that works with
ferric chloride as the electrolyte that apparently does not degrade at all,
IIRC the manufacturers provide a 30 year guarantee of no performance loss.
The vanadium electrolyte has a much higher energy density (around 200Wh/l
IIRC) while ferric chloride is only good for about 50Wh/l.

Both companies are actively selling into the grid storage market
with products that start at four forty foot containers and go up to as many
acres as is wanted. Their current market is the minutes to hours storage
sector which can be very profitable for the battery owners - lithium
apparently works best in the seconds to minutes sector but gets
unreasonably expensive for longer term storage.

There are a bunch of startups in early stages of funding and product
development going for the domestic market with the vanadium chemistry.
Their products look pretty but seem too small to me.

--
Steve O'Hara-Smith
Odds and Ends at http://www.sohara.org/
Host: Beautiful Theory meet Inconvenient Fact
Obit: Beautiful Theory died today of factual inconsistency

On 2023-08-06, The Natural Philosopher <tnp@invalid.invalid> wrote:

> Coal enabled the UK population to rise from 5-10 million to 50 million,
> oil and gas has taken us to 70-80 million. No one knows, there are so
> many illegal immigrants these days . Renewable energy will take us back
> to 5 million - or probably less, as we do not have a pre-existent system
> set up to achieve anything at all with say horses or oxen,
>
> Nuclear might take us to 100 million. But no one seems to want that. I
> think they prefer to die.

They believe that nook-yoo-lur is EEEvil, and can't conceive that
some of the alternatives are far worse.

Thomas Malthus got a lot of it right over 200 years ago - but back
then there were still few enough of us that people could get away
with believing that the planet's space and resources were infinite.

What's really going to destroy us is the belief by many - especially
our ruling classes - that our population can and must increase
indefinitely. People cannot allow themselves to realize that there
are hard limits - for example, the entire planet being converted into
people, crawling over each other like a swarm of bees. If our population
continues to double every 40 years, we'll reach that point in 1800 years.
A more realistic limit - one person for every square meter of dry land
on the planet - will be reached in only 600 years. But anyone with any
sense will realize that the crash will happen much sooner than that.

> Its a shame. Id like to see all these unbelievably arrogant and stupid
> people dying in screaming agony as the realisation dawns that everything
> they were told or believed in was all a load of utter shit, and that no
> matter how much they scream about their 'rights', and what the
> government *ought* to be doing, there is simply no one left to give them
> a single thought.
>
> Life is rough, tough and desperately unjust. Life's a bitch, and then
> you die.
>
> It's a shame that these lessons are no longer taught.

Oh, we don't want to bruise their precious egos.

--
/~\ Charlie Gibbs | You can't save the earth
\ / <cgibbs@kltpzyxm.invalid> | unless you're willing to
X I'm really at ac.dekanfrus | make other people sacrifice.
/ \ if you read it the right way. | -- Dogbert the green consultant

On 8/6/23 3:51 AM, The Natural Philosopher wrote:
> On 06/08/2023 05:01, 23k.304 wrote:
>> On 8/5/23 4:01 AM, The Natural Philosopher wrote:
>>> On 05/08/2023 05:20, 23k.304 wrote:
>>>> On 8/4/23 4:58 AM, The Natural Philosopher wrote:
>>>>> On 04/08/2023 07:03, 23k.304 wrote:
>>>>>> Still have the old IBM Technical Reference Manual.
>>>>>>    All that stuff was meticulously documented.
>>>>>
>>>>> I gave that away sometime in the late 80s.
>>>>
>>>>    I consider it to be a "valuable artifact" :-)
>>>>
>>>>    At minimum, it showed how serious and deep they'd
>>>>    go with the early PCs. That's when programming was
>>>>    still painfully "real". Mostly you could not buy
>>>>    what you needed - so you had to make it yourself.
>>>>
>>>>    And I *did* find many uses for that info.
>>> Oh, so did I. I was assembly programming 8086s then - writing BIOS
>>> extensions or complete BIOSES for them,
>>>
>>>> https://www.youtube.com/watch?v=WUVZbBBHrI4
>>
>>    Shoulda gone with the Frenchies, they LOVE nuke !  :-)
>>
>>    Never went as far as a full BIOS. Leveraging what IS,
>>    what's commercial, what's widespread and "standard"
>>    was generally my approach. In the time it takes to
>>    write a BIOS, esp solo, I could do ten times the stuff
>>    with my approach.
>>
>>    But to each ...
>>
> I was being paid to get custom designed hardware to perform.
> This minicomputer company wanted an 8088 board to control the other
> boards in their super fault tolerant computer. So there it was, a bare
> 8088 board with PROm and some RAM and they wanted to be able to load
> code from an 8" floppy, and drive a serial monitor, oh and while you are
> at it, write us a full C library in ROM and an FORTH interpreter -
> here's  the source code in a 150 page scan of some magazine article.
>
> That's 6 months of my life I wont get back. They were going to write a
> multitasking basic operating system on top of that but my contract
> expired before they got that far.

Your skills - or at least your PATIENCE - exceed mine :-)

Full 'C' Library in ROM ??? WTF *FOR* ???

Let's say I wouldn't have done it that way - more
a 'MCP' that manages all the other junk once it
starts.

"Super-Fault-Tolerant" has always interested and
impressed me. It's an art unto itself. There are
only a few NEEDS - nuclear-weapons systems, nuke
reactors and, coolest, space probes. The way NASA
can essentially re-wire probes, work around even
primary memory and maybe sub-processor failure and
such, over a billion-mile remote connection is
just spectacular.

>>> Or interfacing with special hardware people had built that needed
>>> drivers written.
>>> But that line of business dried up, so I ended up doing yet another
>>> career shift into running IT businesses.
>>
>>    Once you COULD buy lots of neat stuff off the shelf,
>>    then yea, a focus shift made sense. Never tried/wanted
>>    to RUN businesses though - massive pain in the ass and
>>    biz WAY too often winds up in the shitter.
>>
> No option. No one wanted a 43 year old self taught programmer. They
> wanted bushy tailed youngsters who wrote C++ and had been schooled in
> all sorts of theoretical shit that enabled them to justify writing
> shitty bug filled code.

Oh, you've NOTICED that trend eh ? :-)

Software Design by Management Buzzwords ...

I'm surprised The Infrastructure hasn't totally
imploded WITHOUT any Chinese help by now.

> When no one is offering jobs, it sensible  to invent your own. It
> started as technical support, but as it became apparent that the other
> people in the business hadn't a clue how to actually run a business -
> accounts, operations, technical support and so on, I ended up running
> more than half the employees.
>
> I needed a job, and that was the only way I could guarantee the business
> didn't fall to pieces with incompetence.

I found some "smaller outfits" that needed custom apps
and sometimes hardware but without the need to re-invent
the 8088. It worked out SO much better :-)

Oh yea, why did your former employer pick an 8088 ???
The 68K's came out about the same time and were SO
much better !

> Well it fell to pieces with ego, so a couple of us started another one,
> and I ended up running most of that too.
> When I had enough cash accumulated, and my co directors wife started
> trashing the business, we sold it and I gave up work altogether.
>
> I'd frankly had enough. So I retired at 50.

Waited till 65. PART of it was just needing Something
To Do, the other part was "PWE". Alas internal management
politics drove away BOTH the people I was training to
take over ... no time to do it again ... so they're
kinda fucked. They'll be lucky to even find a decent
office computer geek to install printer drivers and
such. The bigger stuff they'll have to outsource to
commercial interests that will promise much and then
stop answering the phone.

>>    Was still programming in solder into maybe half a
>>    dozen years ago - stuff you couldn't, or wouldn't,
>>    buy off the shelf. However that stuff was for
>>    microcontrollers, not real CPUs.
>
> I still solder stuff together. And I will be for this current Pi project
> too. I need an internal mains power supply for the server - to hell with
> USB wall warts -

Be careful, that many volts/amps right on the PCB/hat ...

Not insanely difficult though ... it all depends on
how power-efficient you HAVE to be. Full-wave rec
with cap and cheapo harmonic suppression and low-
dropout regulator might be good enough. However if
you're using A/D then a really really stable voltage
becomes highly desired even before any on-board
regulators. Switching supplies are much more efficient,
but you've gotta pay more $$$ so you don't get all
kinds of weird harmonics that WILL, mysteriously, find
its way into your inputs. Put switchers on a scope under
varying loads and check.

> and I need a special bit of kit that turns on after up
> to an hours delay, boots a pico and runs it until the PICO has done its
> stuff and issues a shutdown signal on a GPIO pin, and puts the whole
> thing to sleep at a couple of  µA or so for up to an hour, to get a
> decent battery life.

I did something similar, solar-powered, for "the field".
The target system was a Mega however - which DOES support
significant "sleep" capability. Seeed li-po charger (NOT
the other brands !) LDO reg to keep from over-volting that
and ultimately a 5W PV panel. Time chip ("ChronoDot") set
the pace normally except for occasional interrupts from
certain measurement devices. It took readings every 15
minutes, which decidedly stretched the battery life.

Alas even an RP Pico requires a LOT more power. You PAY
for all that "added capability". You'd very literally
have to TURN IT OFF ENTIRELY between operations.

I was always impressed with those cheepo solar calculators
you can buy from the office-supply store ... just a tiny
strip of amorphous PV makes them go even in dim light.

> Oddly enough the best solution to that looks to be not some specialised
> complex chip, but an old school simple 4000 series  CMOS Schmitt
> trigger, some Rs and Cs, and three boring old  school transistors. 1970s
> technology in fact.

Oldies but goodies - and CHEAP too.

> In quiescent state, the only current draw will be leakage current on the
> transistors charge current on the timing  capacitor and the quiescent
> current of the CMOS Schmitt, which at 4.5v is less than a µA...the best
> I could do with clever chips like a CMOS 555 timer was around 16µA.

Getting an EXACT voltage output from a little "hummer" supply
like that might be tricky. As said, it depends on what you're doing.

Oh, if you're powering direct from the mains, why CARE what
the quiescent current is ? That's more for battery/solar apps.
You could use a vacuum-tube based power supply :-)

> If it works I might get some PCBS made up for that. Lots of people might
> want a PICO for use as a remote sensor on batteries on an 'occasional'
> basis.

Used to etch my own PCBs. Messy, but that was
before outfits that could whip-up boards in
arbitrary quantities basically overnight.
Buy your own 3-D milling machine and you don't
even have to wait overnight.

On 8/6/23 9:06 AM, Ahem A Rivet's Shot wrote:
> On Sun, 6 Aug 2023 10:11:43 +0100
> The Natural Philosopher <tnp@invalid.invalid> wrote:
>
>> And the way C used the stack to allocate temporary variable space to
>> remove the need for complex memory management or else global memory
>> assignment was pure genius. I think that first appeared in ALGOL. It
>> allowed recursive functions to be simply coded.
>
> Yes block structure arrived in programming with Algol.

ALGOL was originally a "demo", "proof of concept", not
intended for the real world. I seem to remember that the
initial version didn't even have I/O. I've got a
compiler for the "next version" but haven't had the time
to mess around with it much. However it's STILL kind
of complex, a lot of work to emulate what's done so
neatly these days.

> The early documentation on procedure calls in Algol I was mind
> numbingly strange because it did not refer to a stack - it was written in
> terms of textually putting the subroutine code where the call was with
> modifications to the variable names to achieve the same effect. We
> covered in college, when it got to recursion most of the minds in the
> lecture hall went into meltdown. I was lucky, I already knew about stacks
> and spent most of the lecture marvelling at the baroque explanations.

Well, sort of "new concept" ... they had to cast it
in more conventional terms. A couple of images would
have made it MUCH more clear. Guess those were too
expensive :-) Nothing more useless than an attempted
TEXT DESCRIPTION of something best suited for a VISUAL
description. (Hmm ... some Linux man pages still bring
back that nightmare).

Alas, back then, there WERE serious limits to recursion
and stacks because there was so little MEMORY. You COULD
shove the stack onto a mass-storage device .. but SLOW !!!

> Algol was OK but at college it was BCPL that I really liked working
> in - we didn't have C because we did have Martin Richards.
>
> I learned C one the job (from a photocopy of the reference section
> of K&R first edition) writing a table driven remote debugging tool for folks
> writing games for the early cartridge based consoles - the other part of
> the job was making the fake cartridges with RAM for the game code, an
> overlay EPROM for startup and a serial port to run it from.

'C' was (and is) Da Boss !

> One cool part of that job was the people - my boss was Nick Toop
> who designed the Acorn Atom (I learned a lot from him!) and the owner of the
> company was David Levy who only just got away with his cash challenge for
> anyone with a chess program that could beat him - he set a five year time
> limit, it took less than six!

Early back in the IBM/PC era there was a company that
was selling "un-copyable" floppy disks. Turned out the
trick was a tiny laser hole in the medium. Phillipe
Khan ("Borland") offered a BOUNTY for anyone who could
find a work-around. I seem to remember it was just a
couple of days. The disk-makers went out of biz.

Having someone involved in making the Atom would have
made it something of a dream job.

> We had a wonderfully eerie window display, it was a beautifully made
> prototype of a chess machine which appeared to be no more than a very
> thick board with normal pieces on it and a marked out space to store the
> pieces off the board ... You played it by moving the pieces and the
> computer played by sliding the pieces around the board (the jiggling to do
> a knights move was fun to watch). There was nothing visible moving the
> pieces. You could even put the pieces on the board randomly, hit the "setup"
> button and watch them move silently into place all by themselves.

Cool. Of course the quality of the CHESS must have
also figured-in ...

> Inside the thick board was an stepper driven X-Y mechanism holding
> an electromagnet, inside each piece was a bit of steel and a tuned L-C
> passive resonator, I'm sure you can work out the details.

Yep. Would have been very visually impressive !

> It didn't sell, they cut costs going into production and replaced
> the carved wooden case and pieces with moulded plastic and the brass gears
> with nylon ones - result it looked tacky and squeaked. The prototype looked
> like an expensive chess set and was silent. IMHO they should have gone with
> keeping the quality high, raising the price to luxury levels and providing
> software updates for a price.

Bean-counters ruin all .......

Serious chess people tend to run in "higher" circles,
meaning more $$$. If you'd had a "beautiful" product
they WOULD have paid for it. Lower volume but higher
margins.

On 8/4/23 4:51 AM, The Natural Philosopher wrote:
> On 04/08/2023 06:08, 23k.304 wrote:
>> On 8/3/23 2:44 AM, Jan Panteltje wrote:
>>> On a sunny day (Wed, 2 Aug 2023 17:47:56 +0100) it happened The Natural
>>> Philosopher <tnp@invalid.invalid> wrote in
>>> <uae1bt$5slp$1@dont-email.me>:
>>>
>>>>
>>>>
>>>>>     ANYway, /var/log CAN be moved to a RAMdisk if you want.
>>>>>     Not 100% sure WHY you'd want to, but it CAN. If a few
>>>>>     very early logs get 'lost' as you re-direct /var/log
>>>>>     then that MIGHT not be all so important. If you want
>>>>>     it all on RAMdisk then you don't CARE if it all
>>>>>     vanishes on reboot. I very rarely look in /var/log
>>>>>     anyhow so ....
>>>>
>>>> As I said, whoever you are, the less writing to an SD drive there is,
>>>> the less chance there is of file system corruption when the power
>>>> goes out.
>>>
>>> To avoid restarts on power failure use a UPS (I do) or just a simple
>>> battery circuit
>>> was it not called 'battery hat' or something? with the Pi.
>>
>>
>>    I remember the "battery hat" - still sold :
>>
>>    https://www.waveshare.com/li-ion-battery-hat.htm
>>
>>    But MOST Pi's, because of the high power consumption,
>>    still run off the mains - so a UPS is probably the
>>    simplest option. The "battery hat" however might
>>    serve to deal with very short interruptions.
>>
>>    However I don't think the most stress to SD cards
>>    is on boot - but during regular USE ... the usual
>>    data churning and loading system apps from the
>>    card. If "ping" is used, well, where does it COME
>>    from on a PI ? The SD card. Each application has
>>    to be examined to see what routines are used and
>>    put them into a RAMdisk or whatever. Remember,
>>    even reading an SD card involves re-writing the
>>    thing, that's how the tech works.
>>
>>
> I am not sure it actually does. If you turn off 'last read' with 'noatime'
> Which you surely would if you cared enough

A right move ... but doesn't entirely solve the
SD-wear problem. System apps you're using a lot
really MUST be moved to RAMdisk. A certain amount
of the OS does wind up on its own RAMdisk, but
it might not be what YOU are using a lot.

Hey, Raspbian might be 2gb image on the SD card
but yer Pi, depending on model, might not have
much more RAM (or even LESS if it's a Pi-1/2).
As such the WHOLE OS cannot be put into RAM.
All the parts that can't, well, they get read
over and over from the SD.

SD's are GREAT - but on PI's and such you've
gotta THINK about it.

>>    Used to do ASM on micro-controllers, but over the past
>>    decade the better 'C' compilers are actually smarter,
>>    can do the same in even fewer bytes/cycles than most
>>    human-writ code. Not by a HUGE margin, but, on tiny
>>    devices, maybe enough.
>
> Oh yes. The days of trying to write C that didn't turn into bloat on an
> 8 bit 6809...are long gone

The compiler-writers have become VERY VERY good.
Stuff I commonly wrote in ASM on u-controllers
is actually smaller/faster now using 'C'. That's
just IMPRESSIVE.

> I occasionally looked at .a files on *86, and they wrote better
> assembler than I could.

The old 'Dilbert' narrow-tie people WERE damned good.
Modern hardware with a lot more resources kind of
made them obsolete - any 10-year-old could write
adequate code then.

> So much so that I gave up looking.

I went to 'C' and PASCAL early on. ASM became only
for a handful of very weird things neither did well
at the time.

Shit though ... I remember a job where I had to
work the (newfangled) 8087 math co-proc from
BASIC - that code being copied/translated from
older FORTRAN statistical routines. What a huge
pain in the ass. DATA statements galore ! The
'87 was NOT like normal CPUs, very odd register
sets and procedures to make it work and get
data back FROM the damned thing. Still, made
the code run about 25x faster ...

On Mon, 7 Aug 2023 00:53:05 -0400
"23k.304" <23k304@bfxw9.net> wrote:

> ALGOL was originally a "demo", "proof of concept", not
> intended for the real world. I seem to remember that the

Indeed it was - and Algol 68 was a computer scientist wet dream.

<no mention of stacks>
> Well, sort of "new concept" ... they had to cast it
> in more conventional terms. A couple of images would

There was another reason - there were a lot of machines that didn't
have stacks so the design did not force an implementation. That's a trick
many designers never seem to learn.

> have made it MUCH more clear. Guess those were too
> expensive :-) Nothing more useless than an attempted
> TEXT DESCRIPTION of something best suited for a VISUAL
> description.

Ah the ancient lament - "Oh for a muse of fire ..." - Shakespeare
pining for a 21st century special effects department!

> Early back in the IBM/PC era there was a company that
> was selling "un-copyable" floppy disks. Turned out the
> trick was a tiny laser hole in the medium. Phillipe

Sounds not too dissimilar to some of the later DVD copy prevention
tricks, trigger the error handling and ...

> Khan ("Borland") offered a BOUNTY for anyone who could
> find a work-around. I seem to remember it was just a
> couple of days. The disk-makers went out of biz.

Figures, there was a guy I was at college with in 1978 who would
have made short work of that - he broke copy protection for a living!

> Yep. Would have been very visually impressive !

It sat in the window playing demo games - every so often someone
would break step, stop dead and stare for a while.

> Serious chess people tend to run in "higher" circles,
> meaning more $$$. If you'd had a "beautiful" product
> they WOULD have paid for it. Lower volume but higher
> margins.

Precisely - and for them the record and replay feature would have
been far more interesting than the machine opponent.

--
Steve O'Hara-Smith
Odds and Ends at http://www.sohara.org/
Host: Beautiful Theory meet Inconvenient Fact
Obit: Beautiful Theory died today of factual inconsistency

On 06/08/2023 14:41, Ahem A Rivet's Shot wrote:
> On Sun, 6 Aug 2023 10:34:22 +0100
> The Natural Philosopher <tnp@invalid.invalid> wrote:

> The task is huge but it is not too huge - it can be expected to
> take decades and it is already happening as fast as production can go. When
> flow battery electrode assemblies hit the DIY market things will really
> be moving. I expect that within a decade followed by IBLCs getting hard to
> obtain.

I wish I shared your optimism.

Remember I ran the accounts as well as the technical sections of my
companies.
There is not enough space for renewable ebergy AND humans AND farming
AND the as yet to be identified storage. The whole heath robinsn
arrangement would uterly destroy the environment, bankrupt the nations
attempting it and still not create enough energy to replace itself in
twenty years.
Today we burn fossil to build renewables. That will not always be the case.
Simply put the implementation cost and the EROEI (enetrgy return over
energy invested) is so poor for renewables that it only takes one nation
to say 'sod this for a game of soldiers' and go 100% nuclear, and that
nation will rule the renewable peasants in their grass shacks forever.

I assumed that was Russia's plan, until the moron invaded Ukraine. Fund
the anti nuclear and anti-fracking and anti oil brigades, then sell
Germany gas on the QT to run the country off while the windmills
convinced people they were actually green, and then absorb or control a
Germany utterly dependent on Russian gas. And thereby essentially
control the EU.

Nope. Despite the costs, nuclear is sadly the only economically viable
option if we want to keep civilisation, which it seems many peole do not,

And its a durn sight easier to cost reduce a technology whose main cost
in in *bureaucracy*, than it is to build a safe 200GWh storage system
that wont go off with the force of an atomic bomb if someone drops a
wrench across the terminals...

And 200GWh is only just enough to keep the UK say, going for about 2
sunless flat calm winters days maybe less, if we have abandoned all
fossil duel and are 100% electric .

It is equivalent in destructive power to 172 kilotons of TNT or about
ten Hiroshima bombs worth.

And that is just the UK.

I'd rather live on the ROOF of a nuclear power station than within ten
miles of THAT.

No 'flow battery' even could come close to that sort of storage. That is
serious water up a hill. That is all that even has the potential, if you
have the water and the hill.

The only other possible storage is heat banks of molten salt, but unless
you actually drive them from a nuclear reactor, (see the Natrium
concept) they are woefully inefficient, and if you have the reactors,
you don't really need much storage.

Nope. Its all green people's wet dreams. A marvellous way to separate
the plebs from their cash and put it in the trousers of the virtue
signalling renewable energy companies, and leave nothing left to invest
in something that *actually works*.

Remember that the batteries on the grid are not there to replace the
energy of the stores of coal, oil ,gas or uranium that power the power
stations, they are there to simply replace the energy of the spinning
mass of all those alternators, that stabilise the grid in the 0-100
second range, giving time to get the pumped storage wound up before the
gas turbines can kick in.

Because if they don't, and the frequency falls, the renewable energy
disconnects itself and you get a grid scale blackout.

--
WOKE is an acronym... Without Originality, Knowledge or Education.

On 06/08/2023 15:45, Pancho wrote:
> On 05/08/2023 22:14, Martin Gregorie wrote:
>> On Sat, 5 Aug 2023 15:23:21 +0100, Ahem A Rivet's Shot wrote:
>>
>>> On Sat, 5 Aug 2023 14:36:08 +0100 Pancho <Pancho.Jones@proton.me> wrote:
>>>
>>>> I was actually surprised at how fast the new Ryzen was. Maybe it is
>>>> time to buy a new PC.
>>>
>>>     I haven't done that in a long time. I used to buy parts and make
>>> them but these days I buy "refurbished"* ex-corporate machines which are
>>> cheap and far better made than the typical consumer PC.
>>>
>>> * Read cleaned, OS reinstalled and stripped of asset tags.
>>
>> It may well be worth looking at getting a new PC.
>>
>> Since around 2005 my house server was a Dual Athlon PC in a noname case
>> and not new when I got it in 2005 and has had at least two new disks (bog
>> standard WD 500MB blue) since then.
>
> I think it all depends, what you want. If you don't already have a PC
> less than 10 years old, then second-hand deals are good.
>
> In most respects I found a 10-year-old 2500K, OK to use. Up until a
> couple of years ago it was my desktop, then its graphics card broke, and
> it got relegated to HTPC, it wasn't so good at that. I've just replaced
> it with an Orange Pi 5, with caveats, the Orange Pi 5 is better as a TV
> computer. The Orange Pi 5 was cheap. I've also bought a couple of low
> power new NUC types for about £200, for relatives, e.g. Intel N5150,
> they are generally happy with them.
>
> Mainly I've been interested in small low wattage PCs, it is only
> recently that they have been quicker than my old 2500K. The new big
> desktops were quick, but not that much quicker, maybe twice as fast, for
> a reasonable price. Now, the Ryzen 5 7600 is looking to be 3-4 times as
> fast, single thread, more than 3-4 multi thread. Which might be
> noticeable, for me, I don't know. Without having one, it is hard to know.
>
Its been a few years now, since I felt my machines were too slow for the
job they were doing.

I accept that people doing video editing or shooting up 3D aliens in
unlikely scenarios may need more, but I don't.

I am more replacing now to lower power consumption, although even that
is not 100% wasted as the computers heat this room in winter.

--
You can get much farther with a kind word and a gun than you can with a
kind word alone.

Al Capone

On 07/08/2023 02:55, Charlie Gibbs wrote:
> What's really going to destroy us is the belief by many - especially
> our ruling classes - that our population can and must increase
> indefinitely. People cannot allow themselves to realize that there
> are hard limits - for example, the entire planet being converted into
> people, crawling over each other like a swarm of bees. If our population
> continues to double every 40 years, we'll reach that point in 1800 years.
> A more realistic limit - one person for every square meter of dry land
> on the planet - will be reached in only 600 years. But anyone with any
> sense will realize that the crash will happen much sooner than that.

Cf 'Stand on Zanzibar; et al.

There is a massive dichotomy, that the ruling elite cannot bridge.
On the one hand the indigenous European populations is in fact
shrinking, because people simply don't want so many children and modern
medicine ensures they don't need to have them either. On the other,
masses of essentially less or uncivilised people are flooding in and
breeding like rabbits, as their cultural background does not include a
phenomenally low infant mortality rate, nor thin gs like old age
pensions. A big family is a support system

Now one view is that a falling population would bring far greater wealth
per individual, since the GDP of a nation is not particularly linked to
human endeavour, but more to what resources - land, sea, minerals etc -
it has legally defensible access to. Plus in a hi tech age, the quality
of its education to turn out not an unending stream of entitled self
absorbed ArtStudents™ , but a smaller stream of intenbselyt well
qualified and able people who are able to create wealth by dint of
technical innovation, or good risk assessment in the financial areas.

Such a society would inevitably be extremely productive high earning and
individually wealth, but the GDP would not markedly increase. And that
is the 'n' word in the woodpile, because the mechanics of the massive
debt incurred by all the western nations is conditional for being even
able to be *serviced*, let alone repaid, by a continuous process of
expanding population and inflations.
Consider the effect on house prices of a falling population. You
wouldn't be able to give away the cheapest houses. they would remain
empty. An asset price crash of epic proportions would set in as all the
things that are rare by dint of having too many people wanting them
would lose their value.

The banks would all fail. And that is the elephant in the room, both in
the USA and in the Eurozone, there is and has been for 20 years a
looming debt crisis that threatens to crash the entire system of debt.

And that is why, gentlemen, the immigrants are flooding in, to keep
house prices up, wage prices down, and burden an even greater number of
people with debt and erode the wealth of private capital through inflation.

The banks lent us money, and successive governments instead of investing
in assets - railways, nuclear power stations, ports, gas storage, grid
upgrades - instead 'invested in people' which is a euphemism for
creating a consumer boom with low personal credit rates and large
payouts, so that they could essentially buy the votes of the less well
heeled. And increase the number of people who are not actually creating
wealth, but living off directly or indirectly, taxpayer funding.
Now over 50% of GDP in this country circulates through the government in
terms of taxes and benefits, public sector wages and contracts to privet
companies that are entered into by the apparatus of state, or state
industries. Right down to the cleaning company that cleans the council
offices or the hospital wards...

The problem is that a falling population makes this house of debt cards
unsustainable: the western banking system would collapse. As it nearly
did with the sub prime scandal. And with bankrupt governments, 50% of
the workforce now loses its jobs. Social security collapses, pension
funds crash...we would be left holding what physical assets we owned,
and earn whatever we could barter for our skills.

And that is why, gentlemen, the banks are running us into the ground,
because if we drop population the whole system of debt will collapse and
the consequences - at least to bankers - are simply unthinkable. They
are fighting for their very existence, and they play very very dirty
indeed, and have no moral scruples whatsoever.

I don't have an answer, either. A collapsed economy and financial system
or a third world level of poverty in western countries. Not much of a
choice is it?

--
"The great thing about Glasgow is that if there's a nuclear attack it'll
look exactly the same afterwards."

Billy Connolly

On 07/08/2023 05:21, 23k.304 wrote:
> Getting an EXACT voltage output from a little "hummer" supply
>   like that might be tricky. As said, it depends on what you're doing.
>
>   Oh, if you're powering direct from the mains, why CARE what
>   the quiescent current is ? That's more for battery/solar apps.
>   You could use a vacuum-tube based power supply  😄

There are two halves to what I am building. A mains powered server in a
box controlling stuff and accepting inputs from remote sensors and
running a web server as a user interface for user level information and
control, and the sensors themselves which may, and in one case must be,
battery powered. Namely thermometers, whose response time will be in
tens of minutes at best, and a (battery powered) oil level sensor whose
response can be in an hour or two.

So the mains powered server, already has relays switching mains in it,
and I have bought a custom 'block' switched mode PSU that should run it
ok. The Pi Zero W has its own SMPS internally to generate the 3V3 and
1V8 rails it needs, and will allegedly run from 1.8V to 5.5V input wise,
and the PICO has similar, although it seems to only need 3V3.

So its (the PICO) fairly ideal for battery power, as long as it spends
plenty of time *off*...

The idea is the thermometers won't *need* batteries if there is a handy
USB power source available, but the oil sensor is not so lucky. It must
be battery powered.

--
No Apple devices were knowingly used in the preparation of this post.

On 07/08/2023 06:31, 23k.304 wrote:
> On 8/4/23 4:51 AM, The Natural Philosopher wrote:
>> On 04/08/2023 06:08, 23k.304 wrote:
>>> On 8/3/23 2:44 AM, Jan Panteltje wrote:
>>>> On a sunny day (Wed, 2 Aug 2023 17:47:56 +0100) it happened The Natural
>>>> Philosopher <tnp@invalid.invalid> wrote in
>>>> <uae1bt$5slp$1@dont-email.me>:
>>>>
>>>>>
>>>>>
>>>>>>     ANYway, /var/log CAN be moved to a RAMdisk if you want.
>>>>>>     Not 100% sure WHY you'd want to, but it CAN. If a few
>>>>>>     very early logs get 'lost' as you re-direct /var/log
>>>>>>     then that MIGHT not be all so important. If you want
>>>>>>     it all on RAMdisk then you don't CARE if it all
>>>>>>     vanishes on reboot. I very rarely look in /var/log
>>>>>>     anyhow so ....
>>>>>
>>>>> As I said, whoever you are, the less writing to an SD drive there is,
>>>>> the less chance there is of file system corruption when the power
>>>>> goes out.
>>>>
>>>> To avoid restarts on power failure use a UPS (I do) or just a simple
>>>> battery circuit
>>>> was it not called 'battery hat' or something? with the Pi.
>>>
>>>
>>>    I remember the "battery hat" - still sold :
>>>
>>>    https://www.waveshare.com/li-ion-battery-hat.htm
>>>
>>>    But MOST Pi's, because of the high power consumption,
>>>    still run off the mains - so a UPS is probably the
>>>    simplest option. The "battery hat" however might
>>>    serve to deal with very short interruptions.
>>>
>>>    However I don't think the most stress to SD cards
>>>    is on boot - but during regular USE ... the usual
>>>    data churning and loading system apps from the
>>>    card. If "ping" is used, well, where does it COME
>>>    from on a PI ? The SD card. Each application has
>>>    to be examined to see what routines are used and
>>>    put them into a RAMdisk or whatever. Remember,
>>>    even reading an SD card involves re-writing the
>>>    thing, that's how the tech works.
>>>
>>>
>> I am not sure it actually does. If you turn off 'last read' with
>> 'noatime'
>> Which you surely would if you cared enough
>
>   A right move ... but doesn't entirely solve the
>   SD-wear problem. System apps you're using a lot
>   really MUST be moved to RAMdisk. A certain amount
>   of the OS does wind up on its own RAMdisk, but
>   it might not be what YOU are using a lot.
>

No system apps are in use in the sense that they are being loaded from disk.

The are all memory resident and thats that.

Three daemons, maybe four

Apache web server plus sensor watchdog that wakes up every second, looks
at the state of various RAMDISK based files and reads the SD based
config files, probably from disk cache 99.99% of the time, and turns
stuff on and off, plus an xinetd daemon that accepts input from remote
thermometers and writes to ram disk and possibly a similar one for the
oil sensor.

And that's it.

I guess the usual cron jobs will run, vut thst so on any Pi running Raspios.

>   Hey, Raspbian might be 2gb image on the SD card
>   but yer Pi, depending on model, might not have
>   much more RAM (or even LESS if it's a Pi-1/2).
>   As such the WHOLE OS cannot be put into RAM.
>   All the parts that can't, well, they get read
>   over and over from the SD.

No, they don't. Because huge swathes of the disk image contain programs
that are *never run at all*. My pi came with for example, gcc and
friends. Thousands of
development header files and libraries that will never be invoked once
the code is stable. Once running my server is currently only using 60MB
of RAM, with 200+MB in use as disk cache. and over 100MB that isn't used
at all!

I am running apparently 110 separate processes. But these are all kernel
processes for the most part, and use very little RAM or CPU. The load
average is simply 'nothing' . In short this little Zero is completely
overspecified for the job it does. It wont wear the card out.

Nothing that is in the disk cache need ever be read from the disk. And
indeed if it is written to, the OS will only periodically flush its
buffers. Those disk caches ARE an effective ram disk protecting the SD
card from many reads and writes.
And doesn't mind being read, it's writes that screw it, anyway.

--
Climate is what you expect but weather is what you get.
Mark Twain

On 07/08/2023 09:46, The Natural Philosopher wrote:

>
> Remember I ran the accounts as well as the technical sections of my
> companies.

Didn't we all? At least, those of us who had companies.

> There is not enough space for renewable ebergy AND humans AND farming
> AND the as yet to be identified storage.

The North Sea gives enough room for the needs of the UK.

It might be uneconomic, it might not, but that space could provide
enough power for the UK.

On 07/08/2023 09:49, The Natural Philosopher wrote:
> On 06/08/2023 15:45, Pancho wrote:
>> On 05/08/2023 22:14, Martin Gregorie wrote:
>>> On Sat, 5 Aug 2023 15:23:21 +0100, Ahem A Rivet's Shot wrote:
>>>
>>>> On Sat, 5 Aug 2023 14:36:08 +0100 Pancho <Pancho.Jones@proton.me>
>>>> wrote:
>>>>
>>>>> I was actually surprised at how fast the new Ryzen was. Maybe it is
>>>>> time to buy a new PC.
>>>>
>>>>     I haven't done that in a long time. I used to buy parts and make
>>>> them but these days I buy "refurbished"* ex-corporate machines which
>>>> are
>>>> cheap and far better made than the typical consumer PC.
>>>>
>>>> * Read cleaned, OS reinstalled and stripped of asset tags.
>>>
>>> It may well be worth looking at getting a new PC.
>>>
>>> Since around 2005 my house server was a Dual Athlon PC in a noname case
>>> and not new when I got it in 2005 and has had at least two new disks
>>> (bog
>>> standard WD 500MB blue) since then.
>>
>> I think it all depends, what you want. If you don't already have a PC
>> less than 10 years old, then second-hand deals are good.
>>
>> In most respects I found a 10-year-old 2500K, OK to use. Up until a
>> couple of years ago it was my desktop, then its graphics card broke,
>> and it got relegated to HTPC, it wasn't so good at that. I've just
>> replaced it with an Orange Pi 5, with caveats, the Orange Pi 5 is
>> better as a TV computer. The Orange Pi 5 was cheap. I've also bought a
>> couple of low power new NUC types for about £200, for relatives, e.g.
>> Intel N5150, they are generally happy with them.
>>
>> Mainly I've been interested in small low wattage PCs, it is only
>> recently that they have been quicker than my old 2500K. The new big
>> desktops were quick, but not that much quicker, maybe twice as fast,
>> for a reasonable price. Now, the Ryzen 5 7600 is looking to be 3-4
>> times as fast, single thread, more than 3-4 multi thread. Which might
>> be noticeable, for me, I don't know. Without having one, it is hard to
>> know.
>>
> Its been a few years now, since I felt my machines were too slow for the
> job they were doing.
>
> I accept that people doing video editing or shooting up 3D aliens in
> unlikely scenarios may need more, but I don't.
>
> I am more replacing now to lower power consumption, although even that
> is not 100% wasted as the computers heat this room in winter.
>

It was my old monitor that heated the room, it used vastly more power
than even my old computers. I did mean to buy a new TV, this summer, but
haven't, so the old one will heat me another winter.

If you live alone in a big house, buy a fleece onesie.

<https://www.amazon.co.uk/Keanu-KM217-C-ML-Snuggle-Fleece-Charcoal/dp/B09L7CPMQ2>

They are stunningly warm, and comfortable.

On Mon, 7 Aug 2023 09:46:40 +0100
The Natural Philosopher <tnp@invalid.invalid> wrote:

> On 06/08/2023 14:41, Ahem A Rivet's Shot wrote:
> > On Sun, 6 Aug 2023 10:34:22 +0100
> > The Natural Philosopher <tnp@invalid.invalid> wrote:
>
> > The task is huge but it is not too huge - it can be expected to
> > take decades and it is already happening as fast as production can go.
> > When flow battery electrode assemblies hit the DIY market things will
> > really be moving. I expect that within a decade followed by IBLCs
> > getting hard to obtain.
>
> I wish I shared your optimism.

You have it backwards - I wish I shared *your* optimism.

You see you're right widespread adoption of nuclear power would
make things far easier if it was done well and in a timely fashion.

My problem is that I don't see any chance of the NIMBY brigade
being bypassed until long after it is far too late to build more than a tiny
fraction of the necessary power stations.

Then of course it will be fast track, crash priority get them built
as fast as possible, oh please keep the costs down the budget is tight this
year you know there's an election coming up. Why not buy a load of those new
dirt cheap plants from <insert fourth world country here> that have just
hit the market.

What could possibly go wrong ?

> Remember I ran the accounts as well as the technical sections of my
> companies.

I am reminded of an old story about army quartermasters. When you
walk into the quartermasters office with a request he opens a thick book of
regulations. One sort of quartermaster is looking for the rule that says
you can't have it the other is looking for the way through the maze of
rules that lets you have it.

Guess which one has the easier job of searching. I've been watching
this for a long time too - until fairly recently it was not looking good at
all. Cheap reliable flow batteries really do change the picture of what is
possible - you should research the technology, it's pretty cool stuff.

Phasing out oil etc. with what's available off the shelf today is
certainly possible but feasibility is dubious the numbers are huge and the
logistics daunting. A decade ago it was almost certainly not possible and
certainly not even remotely feasible. This is the right direction.

Necessity is the mother of invention - and this is one mother of a
necessity. When you were calculating the available space for everything did
you consider off shore floating solar power stations covering large areas
of open ocean ?

It can only get easier, can it get easy enough soon enough ? Who
knows but I'm certain nuclear won't be given the chance until far too late
and I really don't like the results of civil engineering being done far too
quickly. A few bridges needing bracing and crumbling buildings getting
condemned is one thing ...

--
Steve O'Hara-Smith
Odds and Ends at http://www.sohara.org/
Host: Beautiful Theory meet Inconvenient Fact
Obit: Beautiful Theory died today of factual inconsistency