On 29 Dec 2023 at 17:29:21 GMT, "The Natural Philosopher"
<tnp@invalid.invalid> wrote:

> On 29/12/2023 17:20, Tim Lamb wrote:
>> In message <kv86taFca4nU3@mid.individual.net>, Andy Burns
>> <usenet@andyburns.uk> writes
>>> Theo wrote:
>>>
>>>> You can buy 200Gbit NICs these days, which push the limits of PCIe a bit
>>>> more. Still relatively affordable used:
>>>> https://www.ebay.com/itm/176061504502
>>>
>>> 800Gbe is in development
>>>
>>>> The usual bottleneck though is the efficiency of the software
>>>> processing the
>>>> traffic - need tweaking to ensure you make best use of CPU to be able to
>>>> handle the packets, and probably a decent server CPU to have enough
>>>> horsepower.
>>>
>>> Heading towards servers with DPUs, which are to NICs what GPUs are to
>>> dumb graphics cards, packet offloading on steroids ...
>>
>> Something interesting over Christmas. S.I.L. brought an early i pad to
>> try on our wi-fi. Apparently worked fine here but not in London on a
>> fibre connection.
>>>
> iPads don't have fibre connections

It would have to be fed through a router. And the type of connection from the
router to the wider Internet would be irrelevant.

--
Labour - a bunch of rich people convincing poor people to vote for rich people by telling poor people that "other" rich people are the reason they are poor.

Peter Thompson

In message <kv8u7rFhpttU1@mid.individual.net>, Tim Streater
<tim@streater.me.uk> writes
>On 29 Dec 2023 at 17:29:21 GMT, "The Natural Philosopher"
><tnp@invalid.invalid> wrote:
>
>> On 29/12/2023 17:20, Tim Lamb wrote:
>>> In message <kv86taFca4nU3@mid.individual.net>, Andy Burns
>>> <usenet@andyburns.uk> writes
>>>> Theo wrote:
>>>>
>>>>> You can buy 200Gbit NICs these days, which push the limits of PCIe a bit
>>>>> more. Still relatively affordable used:
>>>>> https://www.ebay.com/itm/176061504502
>>>>
>>>> 800Gbe is in development
>>>>
>>>>> The usual bottleneck though is the efficiency of the software
>>>>> processing the
>>>>> traffic - need tweaking to ensure you make best use of CPU to be able to
>>>>> handle the packets, and probably a decent server CPU to have enough
>>>>> horsepower.
>>>>
>>>> Heading towards servers with DPUs, which are to NICs what GPUs are to
>>>> dumb graphics cards, packet offloading on steroids ...
>>>
>>> Something interesting over Christmas. S.I.L. brought an early i pad to
>>> try on our wi-fi. Apparently worked fine here but not in London on a
>>> fibre connection.
>>>>
>> iPads don't have fibre connections
>
>It would have to be fed through a router. And the type of connection from the
>router to the wider Internet would be irrelevant.

And the wi-fi data rate? Or does the router determine that from the
i-pad.
>

--
Tim Lamb

On 29/12/2023 19:56, Tim Lamb wrote:
> In message <umn13v$tpnh$2@dont-email.me>, The Natural Philosopher
> <tnp@invalid.invalid> writes
>> On 29/12/2023 17:39, Tim Lamb wrote:
>>> In message <ummvlh$tav6$7@dont-email.me>, The Natural Philosopher
>>> <tnp@invalid.invalid> writes
>>>> On 29/12/2023 17:20, Tim Lamb wrote:
>>>>> In message <kv86taFca4nU3@mid.individual.net>, Andy Burns
>>>>> <usenet@andyburns.uk> writes
>>>>>> Theo wrote:
>>>>>>
>>>>>>> You can buy 200Gbit NICs these days, which push the limits of
>>>>>>> PCIe  a bit
>>>>>>> more.  Still relatively affordable used:
>>>>>>> https://www.ebay.com/itm/176061504502
>>>>>>
>>>>>> 800Gbe is in development
>>>>>>
>>>>>>> The usual bottleneck though is the efficiency of the software
>>>>>>> processing the
>>>>>>> traffic - need tweaking to ensure you make best use of CPU to be
>>>>>>> able to
>>>>>>> handle the packets, and probably a decent server CPU to have enough
>>>>>>> horsepower.
>>>>>>
>>>>>> Heading towards servers with DPUs, which are to NICs what GPUs are
>>>>>> to  dumb graphics cards, packet offloading on steroids ...
>>>>>  Something interesting over Christmas. S.I.L. brought an early i
>>>>> pad  to  try on our wi-fi. Apparently worked fine here but not in
>>>>> London  on a  fibre connection.
>>>>>>
>>>>>
>>>> I pads don't have fibre connections
>>>  wi-fi fed from fibre.
>>>
>> What has that got do do with the fibre?
>
> Something to do with the router source speed?
>
> Different router/wi-fi obviously but I assume there is a data rate
> issue. Radio was not part of my engineering studies.

No. All the pad 'sees' is the wifi.

If it doesn't work at all, then the wifi isnt working, or at most the
DHCP isnt working to issue the pad with a valid IP address etc.
Apple is renowned for ignoring any standards but its own.

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

On 29/12/2023 20:21, Andy Burns wrote:
> The Natural Philosopher wrote:
>> On 29/12/2023 17:59, Andy Burns wrote:
>>> The Natural Philosopher wrote:
>>>
>>>> Tim Lamb wrote:
>>>>
>>>>> S.I.L. brought an early i pad to try on our wi-fi. Apparently
>>>>> worked fine here but not in London on a fibre connection.
>>>>
>>>> I pads don't have fibre connections
>>>
>>> But a few years ago they could be fussy about certain DHCP servers.
>>>
>>>
>> No.
>> Its a completely different part of the router
>
> I'm not suggesting the ipads don't like the router because the router
> has fibre, just that for unrelated reasons, old ipads don't like some
> router's dhcp.
>
That isn't quite what you said. I think you have clipped the part which
implied that it was the router being fussy about ITS DHCP server, up at
the ISP.

It would not surprise me in the least that the router was set to
something other than the normal protocol for today - WPA2/PSK

And Apple barfed on it.

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

On 29/12/2023 23:28, Tim Lamb wrote:
> In message <kv8u7rFhpttU1@mid.individual.net>, Tim Streater
> <tim@streater.me.uk> writes
>> On 29 Dec 2023 at 17:29:21 GMT, "The Natural Philosopher"
>> <tnp@invalid.invalid> wrote:
>>
>>> On 29/12/2023 17:20, Tim Lamb wrote:
>>>> In message <kv86taFca4nU3@mid.individual.net>, Andy Burns
>>>> <usenet@andyburns.uk> writes
>>>>> Theo wrote:
>>>>>
>>>>>> You can buy 200Gbit NICs these days, which push the limits of PCIe
>>>>>> a bit
>>>>>> more.  Still relatively affordable used:
>>>>>> https://www.ebay.com/itm/176061504502
>>>>>
>>>>> 800Gbe is in development
>>>>>
>>>>>> The usual bottleneck though is the efficiency of the software
>>>>>> processing the
>>>>>> traffic - need tweaking to ensure you make best use of CPU to be
>>>>>> able to
>>>>>> handle the packets, and probably a decent server CPU to have enough
>>>>>> horsepower.
>>>>>
>>>>> Heading towards servers with DPUs, which are to NICs what GPUs are to
>>>>> dumb graphics cards, packet offloading on steroids ...
>>>>
>>>> Something interesting over Christmas. S.I.L. brought an early i pad to
>>>> try on our wi-fi. Apparently worked fine here but not in London on a
>>>> fibre connection.
>>>>>
>>> iPads don't have fibre connections
>>
>> It would have to be fed through a router. And the type of connection
>> from the
>> router to the wider Internet would be irrelevant.
>
> And the wi-fi data rate? Or does the router determine that from the i-pad.

That is negotiated to be whatever the link and the kit at each end will
stand .

Like old fashioned modems

>>
>

--
"I am inclined to tell the truth and dislike people who lie consistently.
This makes me unfit for the company of people of a Left persuasion, and
all women"

On 29 Dec 2023 at 23:28:45 GMT, "Tim Lamb" <tim@marfordfarm.demon.co.uk>
wrote:

> In message <kv8u7rFhpttU1@mid.individual.net>, Tim Streater
> <tim@streater.me.uk> writes
>> On 29 Dec 2023 at 17:29:21 GMT, "The Natural Philosopher"
>> <tnp@invalid.invalid> wrote:
>>
>>> On 29/12/2023 17:20, Tim Lamb wrote:
>>>> In message <kv86taFca4nU3@mid.individual.net>, Andy Burns
>>>> <usenet@andyburns.uk> writes
>>>>> Theo wrote:
>>>>>
>>>>>> You can buy 200Gbit NICs these days, which push the limits of PCIe a bit
>>>>>> more. Still relatively affordable used:
>>>>>> https://www.ebay.com/itm/176061504502
>>>>>
>>>>> 800Gbe is in development
>>>>>
>>>>>> The usual bottleneck though is the efficiency of the software
>>>>>> processing the
>>>>>> traffic - need tweaking to ensure you make best use of CPU to be able to
>>>>>> handle the packets, and probably a decent server CPU to have enough
>>>>>> horsepower.
>>>>>
>>>>> Heading towards servers with DPUs, which are to NICs what GPUs are to
>>>>> dumb graphics cards, packet offloading on steroids ...
>>>>
>>>> Something interesting over Christmas. S.I.L. brought an early i pad to
>>>> try on our wi-fi. Apparently worked fine here but not in London on a
>>>> fibre connection.
>>>>>
>>> iPads don't have fibre connections
>>
>> It would have to be fed through a router. And the type of connection from the
>> router to the wider Internet would be irrelevant.
>
> And the wi-fi data rate? Or does the router determine that from the
> i-pad.

Completely irrelevant. The flow rate from end to end is managed by the TCP/IP
protocols. Each bit of wire (or not, in the case of wi-fi) has its own speed.
That this is so by and large DOESN'T MATTER. Other things being equal, and if
you have no other traffic, and you have an end to end connection that is made
up of sections running at diffeent transfer rates, the end to end transfer
rate will ramp up until the slowest link is saturated. At which point it
starts dropping packets, at which point the end-to-end protocol, TCP/IP, will
cause the sender to slow down.

--
I was brought up to believe that you should never give offence if you can avoid it; the new culture tells us you should always take offence if you can. There are now experts in the art of taking offence, indeed whole academic subjects, such as 'gender studies', devoted to it.

Roger Scruton

On 30 Dec 2023 at 09:26:19 GMT, "The Natural Philosopher"
<tnp@invalid.invalid> wrote:

> If it doesn't work at all, then the wifi isnt working, or at most the
> DHCP isnt working to issue the pad with a valid IP address etc.
> Apple is renowned for ignoring any standards but its own.

Rubbish. The early iPad must have had a bug in its DHCP software, subsequently
fixed in later ones. Bit that's iPads. No mac I've ever used has ever has a
DHCP problem.

And I'll think you'll find it's MS which has frm in ignoring standards. As we
saw in their whole attitude to the Internet and the web, with Internet
Explorer. And see my sig.

--
"The idea that Bill Gates has appeared like a knight in shining armour to lead all customers out of a mire of technological chaos neatly ignores the fact that it was he who, by peddling second-rate technology, led them into it in the first place." - Douglas Adams

In article <umonnr$17qcv$5@dont-email.me>, The Natural Philosopher
<tnp@invalid.invalid> wrote:
> On 29/12/2023 19:56, Tim Lamb wrote:
> > In message <umn13v$tpnh$2@dont-email.me>, The Natural Philosopher
> > <tnp@invalid.invalid> writes
> >> On 29/12/2023 17:39, Tim Lamb wrote:
> >>> In message <ummvlh$tav6$7@dont-email.me>, The Natural Philosopher
> >>> <tnp@invalid.invalid> writes
> >>>> On 29/12/2023 17:20, Tim Lamb wrote:
> >>>>> In message <kv86taFca4nU3@mid.individual.net>, Andy Burns
> >>>>> <usenet@andyburns.uk> writes
> >>>>>> Theo wrote:
> >>>>>>
> >>>>>>> You can buy 200Gbit NICs these days, which push the limits of
> >>>>>>> PCIe a bit more. Still relatively affordable used:
> >>>>>>> https://www.ebay.com/itm/176061504502
> >>>>>>
> >>>>>> 800Gbe is in development
> >>>>>>
> >>>>>>> The usual bottleneck though is the efficiency of the software
> >>>>>>> processing the traffic - need tweaking to ensure you make best
> >>>>>>> use of CPU to be able to handle the packets, and probably a
> >>>>>>> decent server CPU to have enough horsepower.
> >>>>>>
> >>>>>> Heading towards servers with DPUs, which are to NICs what GPUs are
> >>>>>> to dumb graphics cards, packet offloading on steroids ...
> >>>>> Something interesting over Christmas. S.I.L. brought an early i
> >>>>> pad to try on our wi-fi. Apparently worked fine here but not in
> >>>>> London on a fibre connection.
> >>>>>>
> >>>>>
> >>>> I pads don't have fibre connections
> >>> wi-fi fed from fibre.
> >>>
> >> What has that got do do with the fibre?
> >
> > Something to do with the router source speed?
> >
> > Different router/wi-fi obviously but I assume there is a data rate
> > issue. Radio was not part of my engineering studies.

> No. All the pad 'sees' is the wifi.

> If it doesn't work at all, then the wifi isnt working, or at most the
> DHCP isnt working to issue the pad with a valid IP address etc. Apple is
> renowned for ignoring any standards but its own.

No different from Microsoft in that respect.

--
from KT24 in Surrey, England - sent from my RISC OS 4té²
"I'd rather die of exhaustion than die of boredom" Thomas Carlyle

In message <umoo29$17qcv$6@dont-email.me>, The Natural Philosopher
<tnp@invalid.invalid> writes
>On 29/12/2023 20:21, Andy Burns wrote:
>> The Natural Philosopher wrote:
>>> On 29/12/2023 17:59, Andy Burns wrote:
>>>> The Natural Philosopher wrote:
>>>>
>>>>> Tim Lamb wrote:
>>>>>
>>>>>> S.I.L. brought an early i pad to try on our wi-fi. Apparently
>>>>>>worked fine here but not in London on a fibre connection.
>>>>>
>>>>> I pads don't have fibre connections
>>>>
>>>> But a few years ago they could be fussy about certain DHCP servers.
>>>>
>>>>
>>> No.
>>> Its a completely different part of the router
>> I'm not suggesting the ipads don't like the router because the
>>router has fibre, just that for unrelated reasons, old ipads don't
>>like some router's dhcp.
>>
>That isn't quite what you said. I think you have clipped the part which
>implied that it was the router being fussy about ITS DHCP server, up at
>the ISP.
>
>It would not surprise me in the least that the router was set to
>something other than the normal protocol for today - WPA2/PSK

Relatively recent install as it followed a house move/renovation. Router
age/version unknown but I will pass on your comments.
>
>And Apple barfed on it.

--
Tim Lamb

In message <kva6mgFolviU1@mid.individual.net>, Tim Streater
<tim@streater.me.uk> writes
>On 29 Dec 2023 at 23:28:45 GMT, "Tim Lamb" <tim@marfordfarm.demon.co.uk>
>wrote:
>
>> In message <kv8u7rFhpttU1@mid.individual.net>, Tim Streater
>> <tim@streater.me.uk> writes
>>> On 29 Dec 2023 at 17:29:21 GMT, "The Natural Philosopher"
>>> <tnp@invalid.invalid> wrote:
>>>
>>>> On 29/12/2023 17:20, Tim Lamb wrote:
>>>>> In message <kv86taFca4nU3@mid.individual.net>, Andy Burns
>>>>> <usenet@andyburns.uk> writes
>>>>>> Theo wrote:
>>>>>>
>>>>>>> You can buy 200Gbit NICs these days, which push the limits of PCIe a bit
>>>>>>> more. Still relatively affordable used:
>>>>>>> https://www.ebay.com/itm/176061504502
>>>>>>
>>>>>> 800Gbe is in development
>>>>>>
>>>>>>> The usual bottleneck though is the efficiency of the software
>>>>>>> processing the
>>>>>>> traffic - need tweaking to ensure you make best use of CPU to be able to
>>>>>>> handle the packets, and probably a decent server CPU to have enough
>>>>>>> horsepower.
>>>>>>
>>>>>> Heading towards servers with DPUs, which are to NICs what GPUs are to
>>>>>> dumb graphics cards, packet offloading on steroids ...
>>>>>
>>>>> Something interesting over Christmas. S.I.L. brought an early i pad to
>>>>> try on our wi-fi. Apparently worked fine here but not in London on a
>>>>> fibre connection.
>>>>>>
>>>> iPads don't have fibre connections
>>>
>>> It would have to be fed through a router. And the type of connection
>>>from the
>>> router to the wider Internet would be irrelevant.
>>
>> And the wi-fi data rate? Or does the router determine that from the
>> i-pad.
>
>Completely irrelevant. The flow rate from end to end is managed by the TCP/IP
>protocols. Each bit of wire (or not, in the case of wi-fi) has its own speed.
>That this is so by and large DOESN'T MATTER. Other things being equal, and if
>you have no other traffic, and you have an end to end connection that is made
>up of sections running at diffeent transfer rates, the end to end transfer
>rate will ramp up until the slowest link is saturated. At which point it
>starts dropping packets, at which point the end-to-end protocol, TCP/IP, will
>cause the sender to slow down.

Ok Tim. I think I have got the process now. Detached house in London so
all sorts of electromagnetics floating about. The router is ok with
their other gadgets so ditch the old pad.
>

--
Tim Lamb

On 12/30/2023 5:56 AM, Tim Lamb wrote:
> In message <kva6mgFolviU1@mid.individual.net>, Tim Streater <tim@streater.me.uk> writes
>> On 29 Dec 2023 at 23:28:45 GMT, "Tim Lamb" <tim@marfordfarm.demon.co.uk>
>> wrote:
>>
>>> In message <kv8u7rFhpttU1@mid.individual.net>, Tim Streater
>>> <tim@streater.me.uk> writes
>>>> On 29 Dec 2023 at 17:29:21 GMT, "The Natural Philosopher"
>>>> <tnp@invalid.invalid> wrote:
>>>>
>>>>> On 29/12/2023 17:20, Tim Lamb wrote:
>>>>>> In message <kv86taFca4nU3@mid.individual.net>, Andy Burns
>>>>>> <usenet@andyburns.uk> writes
>>>>>>> Theo wrote:
>>>>>>>
>>>>>>>> You can buy 200Gbit NICs these days, which push the limits of PCIe a bit
>>>>>>>> more.  Still relatively affordable used:
>>>>>>>> https://www.ebay.com/itm/176061504502
>>>>>>>
>>>>>>> 800Gbe is in development
>>>>>>>
>>>>>>>> The usual bottleneck though is the efficiency of the software
>>>>>>>> processing the
>>>>>>>> traffic - need tweaking to ensure you make best use of CPU to be able to
>>>>>>>> handle the packets, and probably a decent server CPU to have enough
>>>>>>>> horsepower.
>>>>>>>
>>>>>>> Heading towards servers with DPUs, which are to NICs what GPUs are to
>>>>>>> dumb graphics cards, packet offloading on steroids ...
>>>>>>
>>>>>> Something interesting over Christmas. S.I.L. brought an early i pad to
>>>>>> try on our wi-fi. Apparently worked fine here but not in London on a
>>>>>> fibre connection.
>>>>>>>
>>>>> iPads don't have fibre connections
>>>>
>>>> It would have to be fed through a router. And the type of connection from the
>>>> router to the wider Internet would be irrelevant.
>>>
>>> And the wi-fi data rate? Or does the router determine that from the
>>> i-pad.
>>
>> Completely irrelevant. The flow rate from end to end is managed by the TCP/IP
>> protocols. Each bit of wire (or not, in the case of wi-fi) has its own speed.
>> That this is so by and large DOESN'T MATTER. Other things being equal, and if
>> you have no other traffic, and you have an end to end connection that is made
>> up of sections running at diffeent transfer rates, the end to end transfer
>> rate will ramp up until the slowest link is saturated. At which point it
>> starts dropping packets, at which point the end-to-end protocol, TCP/IP, will
>> cause the sender to slow down.
>
> Ok Tim. I think I have got the process now. Detached house in London so all sorts of electromagnetics floating about. The router is ok with their other gadgets so ditch the old pad.
>>
>

With an Applodian ecosystem, you feed the machine Apple packets, not Generic Brand packets :-)
Just as you charge the battery with Apple electrons, not Generic unpasturized electrons.

Remember, when Apple puts up a File Sharing dialog, it breaks the state machine for
authentication, and by the time the user has typed their password, both computers
involved in the transaction, have completely forgotten the user was trying
to file share. That's what I used to watch in horror, on my G4 Quad Nostril :-)
It's just the Apple way of doing things. If your pay depends on incompetence,
well, there you go. Interface to foreign equipment ? fuck-it-up-Charley-please.

Translating my rant into English, you need to find a different way to feed it.
Maybe a separate Apple router fed from the master router.

Another example of "taming electronics".

Acer monitor. HDMI input does not work with the computer.
OK, take DP to HDMI active adapter, plug into Acer monitor,
the active adapter seems to make the "right flavor" of HDMI,
and using the DP port on the video card and the adapter, I can
use the bloody monitor. See how easy this is ? A full junk room
of intermediaries and emissaries bearing gifts, does the job.

Computer --- DP port ------> DP-to-HDMI-active-adapter ------> HDMI port on Acer monitor OK

Computer --- HDMI port --------------------------------------> HDMI port on Acer monitor Fail (black screen, no OSD error)

Paul

On 30/12/2023 09:37, Tim Streater wrote:
> Other things being equal, and if
> you have no other traffic, and you have an end to end connection that is made
> up of sections running at diffeent transfer rates, the end to end transfer
> rate will ramp up until the slowest link is saturated. At which point it
> starts dropping packets,

No, what will happen is that ACKS will be very delayed, and no further
packets will be sent until the ACKS from previous packets have been
received.

Dropped packets only happen when one end keeps stuffing packets into a
link that cant buffer whats its sent.

--
If I had all the money I've spent on drink...
...I'd spend it on drink.

Sir Henry (at Rawlinson's End)

On 30/12/2023 09:41, Tim Streater wrote:
> On 30 Dec 2023 at 09:26:19 GMT, "The Natural Philosopher"
> <tnp@invalid.invalid> wrote:
>
>> If it doesn't work at all, then the wifi isnt working, or at most the
>> DHCP isnt working to issue the pad with a valid IP address etc.
>> Apple is renowned for ignoring any standards but its own.
>
> Rubbish. The early iPad must have had a bug in its DHCP software, subsequently
> fixed in later ones. Bit that's iPads. No mac I've ever used has ever has a
> DHCP problem.
>
Maybe. Maybe not,. There are many variables in Wifi and DHCP -
frequency, protocols, security protocol.

> And I'll think you'll find it's MS which has frm in ignoring standards.

Er no. Apple is the browser that STILL cannot deal with Webm video protocols
Apple is the platform that invented firewire, used by no one else.

I could go on and on.

As we
> saw in their whole attitude to the Internet and the web, with Internet
> Explorer. And see my sig.
>

MS was big enough to set standards, Apple is not.

--
If I had all the money I've spent on drink...
...I'd spend it on drink.

Sir Henry (at Rawlinson's End)

On 30/12/2023 10:46, Tim Lamb wrote:
> In message <umoo29$17qcv$6@dont-email.me>, The Natural Philosopher
> <tnp@invalid.invalid> writes
>> On 29/12/2023 20:21, Andy Burns wrote:
>>> The Natural Philosopher wrote:
>>>> On 29/12/2023 17:59, Andy Burns wrote:
>>>>> The Natural Philosopher wrote:
>>>>>
>>>>>> Tim Lamb wrote:
>>>>>>
>>>>>>> S.I.L. brought an early i pad to try on our wi-fi. Apparently
>>>>>>> worked fine here but not in London on a fibre connection.
>>>>>>
>>>>>> I pads don't have fibre connections
>>>>>
>>>>> But a few years ago they could be fussy about certain DHCP servers.
>>>>>
>>>>>
>>>> No.
>>>> Its a completely different part of the router
>>>  I'm not suggesting the ipads don't like the router because the
>>> router  has fibre, just that for unrelated reasons, old ipads don't
>>> like some  router's dhcp.
>>>
>> That isn't quite what you said. I think you have clipped the part
>> which implied that it was the router being fussy about ITS DHCP
>> server, up at the ISP.
>>
>> It would not surprise me in the least that the router was set to
>> something other than the normal protocol for today - WPA2/PSK
>
> Relatively recent install as it followed a house move/renovation. Router
> age/version unknown but I will pass on your comments.

I have found the following Wifi setup to be optimal for all my guests

Mode should be mixed - 11b + 11g +11n
Security should be WPA2/PSK

These are things that need to be set up on the router

>>
>> And Apple barfed on it.
>

--
“It is hard to imagine a more stupid decision or more dangerous way of
making decisions than by putting those decisions in the hands of people
who pay no price for being wrong.”

Thomas Sowell

On 12/30/2023 10:30 AM, charles wrote:
> In article <umonnr$17qcv$5@dont-email.me>, The Natural Philosopher
> <tnp@invalid.invalid> wrote:
>
>> If it doesn't work at all, then the wifi isnt working, or at most the
>> DHCP isnt working to issue the pad with a valid IP address etc. Apple is
>> renowned for ignoring any standards but its own.
>
> No different from Microsoft in that respect.

Apple and Microsoft have different sins they like to commit.
They're not exactly the same.

The Microsoft ecosystem is open, and there are a lot more
variables involved. To reduce workload, Microsoft gradually
replaces custom drivers with class drivers. Their next project
will involve printers. How it will end up, that's a good
question, as an anemic printer control panel is never going
to work for an art printer. Declaring that the only media you
print on, is sheets of typing paper, will never work.

Paul

On 30/12/2023 12:23, The Natural Philosopher wrote:
> On 30/12/2023 09:41, Tim Streater wrote:
>> On 30 Dec 2023 at 09:26:19 GMT, "The Natural Philosopher"
>> <tnp@invalid.invalid> wrote:
>>
>>> If it doesn't work at all, then the wifi isnt working, or at most the
>>> DHCP isnt working to issue the pad with a valid IP address etc.
>>> Apple is renowned for ignoring any standards but its own.
>>
>> Rubbish. The early iPad must have had a bug in its DHCP software,
>> subsequently
>> fixed in later ones. Bit that's iPads. No mac I've ever used has ever
>> has a
>> DHCP problem.
>>
> Maybe. Maybe not,. There are many variables in Wifi and DHCP -
> frequency, protocols, security protocol.
>
>> And I'll think you'll find it's MS which has frm in ignoring standards.
>
> Er no. Apple is the browser that STILL cannot deal with Webm video
> protocols
> Apple is the platform that invented firewire, used by no one else.
>
> I could go on and on.
>
> As we
>> saw in their whole attitude to the Internet and the web, with Internet
>> Explorer. And see my sig.
>>
>
> MS was big enough to set standards, Apple is not.

Apple users would disagree.

On 30 Dec 2023 at 12:19:37 GMT, "The Natural Philosopher"
<tnp@invalid.invalid> wrote:

> On 30/12/2023 09:37, Tim Streater wrote:
>> Other things being equal, and if
>> you have no other traffic, and you have an end to end connection that is made
>> up of sections running at diffeent transfer rates, the end to end transfer
>> rate will ramp up until the slowest link is saturated. At which point it
>> starts dropping packets,
>
> No, what will happen is that ACKS will be very delayed, and no further
> packets will be sent until the ACKS from previous packets have been
> received.

My explanation was good enough for government work. The salient point is that
there is end-to-end flow control so the speed and type of intermediate links
doesn't have to be known by any equipment other than what the links are
connected to.

> Dropped packets only happen when one end keeps stuffing packets into a
> link that cant buffer whats its sent.

--
The reason you think government is the solution is because you think freedom is the problem. But the truth is that government ensures that the most evil, ruthless people end up in control, because the state is a single point of failure, and a high-value target of corruption.

Alan Lovejoy

On 30 Dec 2023 at 12:23:22 GMT, "The Natural Philosopher"
<tnp@invalid.invalid> wrote:

> On 30/12/2023 09:41, Tim Streater wrote:
>> On 30 Dec 2023 at 09:26:19 GMT, "The Natural Philosopher"
>> <tnp@invalid.invalid> wrote:
>>
>>> If it doesn't work at all, then the wifi isnt working, or at most the
>>> DHCP isnt working to issue the pad with a valid IP address etc.
>>> Apple is renowned for ignoring any standards but its own.
>>
>> Rubbish. The early iPad must have had a bug in its DHCP software, subsequently
>> fixed in later ones. Bit that's iPads. No mac I've ever used has ever has a
>> DHCP problem.
>>
> Maybe. Maybe not,. There are many variables in Wifi and DHCP -
> frequency, protocols, security protocol.
>
>> And I'll think you'll find it's MS which has frm in ignoring standards.
>
> Er no. Apple is the browser that STILL cannot deal with Webm video protocols

And which Safari version is that?

> Apple is the platform that invented firewire, used by no one else.

And they were the first to sell machines with USB, sold by no one else at the
time

Canon certainly used it on at least the video camera I still have of theirs
(from 2006 or so). That's how you downloaded video for editing.

> I could go on and on.
>
> As we
>> saw in their whole attitude to the Internet and the web, with Internet
>> Explorer. And see my sig.
>
> MS was big enough to set standards, Apple is not.

--
What power have you got?
Where did you get it from?
In whose interests do you use it?
To whom are you accountable?
How do we get rid of you?

Tony Benn

On 30/12/2023 12:19, The Natural Philosopher wrote:
> On 30/12/2023 09:37, Tim Streater wrote:
>> Other things being equal, and if
>> you have no other traffic, and you have an end to end connection that
>> is made
>> up of sections running at diffeent transfer rates, the end to end
>> transfer
>> rate will ramp up until the slowest link is saturated. At which point it
>> starts dropping packets,
>
> No, what will happen is that ACKS will be very delayed, and no further
> packets will be sent until the ACKS from previous packets have been
> received.

Not strictly true.

Many packets can be in train depending on the initial transaction; where
the receiver may only send occasional ACKs. If a dropped/missing packet
occurs then an ACK is sent for the last good packet received preceding
that, so the sender can resend a dropped packet. The packets can be sent
out of order.

> Dropped packets only happen when one end keeps stuffing packets into a
> link that cant buffer whats its sent.

Yes, but the overall link will(should) negotiate and minimise any
dropped packet.

The overall idea is to ensure continuous data through the slowest link.

Andy Burns <usenet@andyburns.uk> wrote:
> Theo wrote:
>
> > You can buy 200Gbit NICs these days, which push the limits of PCIe a bit
> > more. Still relatively affordable used:
> > https://www.ebay.com/itm/176061504502
>
> 800Gbe is in development

For a network card? There are 400Gbps NICs available but you need PCIe Gen5
x16 to feed them.

Most of the current 400GbE is used between switches.

> > The usual bottleneck though is the efficiency of the software processing the
> > traffic - need tweaking to ensure you make best use of CPU to be able to
> > handle the packets, and probably a decent server CPU to have enough
> > horsepower.
>
> Heading towards servers with DPUs, which are to NICs what GPUs are to
> dumb graphics cards, packet offloading on steroids ...

Yes, the datacentre folks have a variety of things including a 'bump in the
wire' FPGA that is on both PCIe and intercepts the networking (eg Microsoft
Catapult in Azure), and smart NICs which are NICs with a programmable
dataplane.

Theo

Theo wrote:

> Andy Burns <usenet@andyburns.uk> wrote:
>
>> 800Gbe is in development
>
> For a network card?

Not seen NICs, just switches (32x800Gbps !!) and transceivers

> There are 400Gbps NICs available but you need PCIe Gen5
> x16 to feed them.

On Sat, 30 Dec 2023 22:17:01 +1100, Paul <nospam@needed.invalid> wrote:

> On 12/30/2023 5:56 AM, Tim Lamb wrote:
>> In message <kva6mgFolviU1@mid.individual.net>, Tim Streater
>> <tim@streater.me.uk> writes
>>> On 29 Dec 2023 at 23:28:45 GMT, "Tim Lamb"
>>> <tim@marfordfarm.demon.co.uk>
>>> wrote:
>>>
>>>> In message <kv8u7rFhpttU1@mid.individual.net>, Tim Streater
>>>> <tim@streater.me.uk> writes
>>>>> On 29 Dec 2023 at 17:29:21 GMT, "The Natural Philosopher"
>>>>> <tnp@invalid.invalid> wrote:
>>>>>
>>>>>> On 29/12/2023 17:20, Tim Lamb wrote:
>>>>>>> In message <kv86taFca4nU3@mid.individual.net>, Andy Burns
>>>>>>> <usenet@andyburns.uk> writes
>>>>>>>> Theo wrote:
>>>>>>>>
>>>>>>>>> You can buy 200Gbit NICs these days, which push the limits of
>>>>>>>>> PCIe a bit
>>>>>>>>> more. Still relatively affordable used:
>>>>>>>>> https://www.ebay.com/itm/176061504502
>>>>>>>>
>>>>>>>> 800Gbe is in development
>>>>>>>>
>>>>>>>>> The usual bottleneck though is the efficiency of the software
>>>>>>>>> processing the
>>>>>>>>> traffic - need tweaking to ensure you make best use of CPU to be
>>>>>>>>> able to
>>>>>>>>> handle the packets, and probably a decent server CPU to have
>>>>>>>>> enough
>>>>>>>>> horsepower.
>>>>>>>>
>>>>>>>> Heading towards servers with DPUs, which are to NICs what GPUs
>>>>>>>> are to
>>>>>>>> dumb graphics cards, packet offloading on steroids ...
>>>>>>>
>>>>>>> Something interesting over Christmas. S.I.L. brought an early i
>>>>>>> pad to
>>>>>>> try on our wi-fi. Apparently worked fine here but not in London on
>>>>>>> a
>>>>>>> fibre connection.
>>>>>>>>
>>>>>> iPads don't have fibre connections
>>>>>
>>>>> It would have to be fed through a router. And the type of connection
>>>>> from the
>>>>> router to the wider Internet would be irrelevant.
>>>>
>>>> And the wi-fi data rate? Or does the router determine that from the
>>>> i-pad.
>>>
>>> Completely irrelevant. The flow rate from end to end is managed by the
>>> TCP/IP
>>> protocols. Each bit of wire (or not, in the case of wi-fi) has its own
>>> speed.
>>> That this is so by and large DOESN'T MATTER. Other things being equal,
>>> and if
>>> you have no other traffic, and you have an end to end connection that
>>> is made
>>> up of sections running at diffeent transfer rates, the end to end
>>> transfer
>>> rate will ramp up until the slowest link is saturated. At which point
>>> it
>>> starts dropping packets, at which point the end-to-end protocol,
>>> TCP/IP, will
>>> cause the sender to slow down.
>>
>> Ok Tim. I think I have got the process now. Detached house in London so
>> all sorts of electromagnetics floating about. The router is ok with
>> their other gadgets so ditch the old pad.
>>>
>>
>
> With an Applodian ecosystem, you feed the machine Apple packets, not
> Generic Brand packets :-)
> Just as you charge the battery with Apple electrons, not Generic
> unpasturized electrons.

> Remember, when Apple puts up a File Sharing dialog, it breaks the state
> machine for
> authentication, and by the time the user has typed their password, both
> computers
> involved in the transaction, have completely forgotten the user was
> trying
> to file share. That's what I used to watch in horror, on my G4 Quad
> Nostril :-)
> It's just the Apple way of doing things.

Airdrop is nothing like that.

> If your pay depends on incompetence,
> well, there you go. Interface to foreign equipment ?
> fuck-it-up-Charley-please.
>
> Translating my rant into English, you need to find a different way to
> feed it.
> Maybe a separate Apple router fed from the master router.
>
> Another example of "taming electronics".
>
> Acer monitor. HDMI input does not work with the computer.
> OK, take DP to HDMI active adapter, plug into Acer monitor,
> the active adapter seems to make the "right flavor" of HDMI,
> and using the DP port on the video card and the adapter, I can
> use the bloody monitor. See how easy this is ? A full junk room
> of intermediaries and emissaries bearing gifts, does the job.
>
> Computer --- DP port ------> DP-to-HDMI-active-adapter ------>
> HDMI port on Acer monitor OK
>
> Computer --- HDMI port -------------------------------------->
> HDMI port on Acer monitor Fail (black screen, no OSD error)
>
> Paul

On Sun, 31 Dec 2023 05:11:00 +1100, cantankerous trolling geezer Rodent
Speed, the auto-contradicting senile sociopath, blabbered, again:

<FLUSH the abnormal trolling senile cretin's latest trollshit unread>

On Sat, 30 Dec 2023 23:23:22 +1100, The Natural Philosopher
<tnp@invalid.invalid> wrote:

> On 30/12/2023 09:41, Tim Streater wrote:
>> On 30 Dec 2023 at 09:26:19 GMT, "The Natural Philosopher"
>> <tnp@invalid.invalid> wrote:
>>
>>> If it doesn't work at all, then the wifi isnt working, or at most the
>>> DHCP isnt working to issue the pad with a valid IP address etc.
>>> Apple is renowned for ignoring any standards but its own.
>> Rubbish. The early iPad must have had a bug in its DHCP software,
>> subsequently
>> fixed in later ones. Bit that's iPads. No mac I've ever used has ever
>> has a
>> DHCP problem.
>>
> Maybe. Maybe not,. There are many variables in Wifi and DHCP -
> frequency, protocols, security protocol.
>
>> And I'll think you'll find it's MS which has frm in ignoring standards.
>
> Er no. Apple is the browser that STILL cannot deal with Webm video
> protocols
> Apple is the platform that invented firewire, used by no one else.
>
> I could go on and on.
>
> As we
>> saw in their whole attitude to the Internet and the web, with Internet
>> Explorer. And see my sig.
>>
>
> MS was big enough to set standards, Apple is not.

That's bullshit with firewire, the lightning connector, airdrop, imessage
etc etc etc

On Sun, 31 Dec 2023 05:30:31 +1100, cantankerous trolling geezer Rodent
Speed, the auto-contradicting senile sociopath, blabbered, again:

<FLUSH the abnormal trolling senile cretin's latest trollshit unread>

--
"Anonymous" to trolling senile Rodent Speed:
"You can fuck off as you know less than pig shit you sad
little ignorant cunt."
MID: <62dcaae57b421e2b3b10e97d9c0ddf08@haph.org>