Looking at recent phones they seem to have USB C connectors for charging
and data, but the underlying USB protocol is USB 2.

Any specific reason?

I vaguely recall a Samsung a good few years back having a special
connector for a special cable to allow higher speed transfers, but that
was back in the mists of time.

Cheers

Dave R

--
AMD FX-6300 in GA-990X-Gaming SLI-CF running Windows 10 x64

Am 16.08.23 um 17:26 schrieb David:

> Looking at recent phones they seem to have USB C connectors for charging
> and data, but the underlying USB protocol is USB 2.
>
> Any specific reason?

You are supposed to store your data in the cloud (where the big players
can make money with it).

Local file system access is intentionally crippled and becomes more and
more restricted with each new Android version, following the example of
iOS. Under these circumstances, it's pretty much pointless to upgrade
the local storage interface, even though most recent Qualcomm and
Mediatek SOCs do provide hardware support for USB3.1/3.2.

David <wibble@btinternet.com> wrote:

> Looking at recent phones they seem to have USB C connectors for
> charging and data, but the underlying USB protocol is USB 2.

USB2 implementers likely already paid for the Vendor ID, so they
probably don't want to continue paying for USB2 while also coming out
with USB3 devices.
https://www.usb.org/getting-vendor-id

USB3 requires licensing. It's free, but a license is required.
https://www.usb.org/document-library/usb-30-adopters-agreement

USB-C and USB Type-C are trademarked.
https://www.usb.org/usb-type-cr-cable-and-connector-specification

VanguardLH wrote:

> USB2 implementers likely already paid for the Vendor ID, so they
> probably don't want to continue paying for USB2 while also coming out
> with USB3 devices.

Does a manufacturer need separate VIDs for USB2 and USB3?

I see several companies on the VID list with just one ID who produce
both types of device.

Andy Burns <usenet@andyburns.uk> wrote:

> VanguardLH wrote:
>
>> USB2 implementers likely already paid for the Vendor ID, so they
>> probably don't want to continue paying for USB2 while also coming out
>> with USB3 devices.
>
> Does a manufacturer need separate VIDs for USB2 and USB3?
>
> I see several companies on the VID list with just one ID who produce
> both types of device.

True, but I also see vendors, like HP and Samsung, that have many VIDs.
The VID does not specify the USB speed, and that's not part of the
definition of a registered VID. A VID identifies a mfr with a product
decription. One VID could only identify a mfr and cover all their
products, or identity a mfr and describe just one of their products.

Having more than one VID allows a manufacturer differentiate between
their products. Even if the mfr had a VID for their USB2 phone and a
VID for their USB3 phone, you can't what is the USB speed based on the
VID for each phone, but the different VIDs could identify which product
was USB2 or USB3 by describing a different product. A mfr with just 1
MID would have to use it for every product and model they produce. They
might do that for, say, a family of printers in a model series, but for
printers in a different model family they'd probably want a different
VID. If Samsung or HP had just 1 VID, there'd be no way to do a VID
lookup to see just what of those brands you had which of their products
in your setup. It's up to the mfr to decide how many VIDs they want for
how granular they want for identifying numerous products of theirs,
especially when they produce totally dissimilar products (printers,
joysticks, card readers, optical writers, Bluetooth devices, etc).

If a mfr only has 1 VID but produced USB2 and USB3 phones, the same VID
would get used for both their USB devices although those devices differ
on USB speed. Tis likely a mfr would create a different model number to
reflect USB3 support, but the single VID would confuse the old USB2
models with the new USB3 models.

As for not putting USB3 into smart phones, wouldn't a USB3 capable chip
cost more than a USB2 chip? Even though legacy phones seem to us to be
very pricey, the mfrs still have to compete against each other.

You could go to, for example:

https://www.epey.co.uk/phone/usb-version/3-1-gen-1-usb-3-0/

which lists phones with USB 3.1 gen 1 ports. Without going through
everything I did trying to find equivalent phones that differed only by
USB2 versus USB3, it became apparent the USB3 phones had different
feature sets; however, USB3 phones are not as rare as the OP claimed.
That site found 2864 USB2 phones and 463 USB3 phones. Yes, there are
fewer USB3 phones, but I wouldn't consider 14% are USB3 as rare.

Interesting is that USB Type-A maxes at 10 Gbps while USB Type-C maxes
at 10 Gbps. However, USB2 is up to 450 Mbps and USB3 is up to 4.8 Gbps.
So, the USB protocols don't yet support the max speeds of the physical
connections.

Also, I'm not sure why anyone would expect toy computers to compete
against desktops regarding data transfer rate. Phones aren't designed
to be file servers. The phone makers likely want you to buy more online
storage quota than give you a free ride with higher data rates across
local connections.

VanguardLH wrote:

> If a mfr only has 1 VID but produced USB2 and USB3 phones, the same VID
> would get used for both their USB devices although those devices differ
> on USB speed. Tis likely a mfr would create a different model number to
> reflect USB3 support, but the single VID would confuse the old USB2
> models with the new USB3 models.

But that's why a combination of VID and PID is used ...

Hergen Lehmann <hlehmann.expires.12-22@snafu.de> wrote:
> Am 16.08.23 um 17:26 schrieb David:
>
> > Looking at recent phones they seem to have USB C connectors for charging
> > and data, but the underlying USB protocol is USB 2.
> >
> > Any specific reason?
>
> You are supposed to store your data in the cloud (where the big players
> can make money with it).
>
> Local file system access is intentionally crippled and becomes more and
> more restricted with each new Android version, following the example of
> iOS. Under these circumstances, it's pretty much pointless to upgrade
> the local storage interface, even though most recent Qualcomm and
> Mediatek SOCs do provide hardware support for USB3.1/3.2.

Well, for most people, the biggest amount of data is probably photos,
videos and audio and that data is still accessible (despite crippled
local file system access) and therefor high USB data speed would still
be nice.

Also Samsung, - i.e. one of the biggest vendors - has a Windows (and
Mac) backup (and other purposes) companion for its phones - Smart Switch
-, which uses USB.

That said, the USB(-C) port of my low-to-mid-range Samsung Galaxy A51
is indeed USB2.

Am 18.08.23 um 12:17 schrieb Frank Slootweg:

> Well, for most people, the biggest amount of data is probably photos,
> videos and audio

You are supposed to store such data in the cloud and use the phone's
memory only as a cache for recently used data.

> and that data is still accessible (despite crippled
> local file system access) and therefor high USB data speed would still
> be nice.

Yes, it would be nice. A MicroSD slot would also be nice, as it allows
for quick&easy transfer of the whole media archive to a new device,
without having to squeeze many Gigabytes through some bottleneck, and
without the need for the old device to be still functional.

Still, both features are absent from most modern devices. Even high-end
ones, where additional costs of a few cents are certainly not a factor.

Hergen Lehmann <hlehmann.expires.12-22@snafu.de> wrote:
> Am 18.08.23 um 12:17 schrieb Frank Slootweg:
>
> > Well, for most people, the biggest amount of data is probably photos,
> > videos and audio
>
> You are supposed to store such data in the cloud and use the phone's
> memory only as a cache for recently used data.

Yes, I know, but I'm a stubborn Dutch (cheap? :-)) geezer, who likes
to keep most things local. (I've only a (encrypted) differential backup
of my most important stuff in the cloud, so I can recover from theft,
fire, etc. with my off-site backup and the delta in the cloud.)
> > and that data is still accessible (despite crippled
> > local file system access) and therefor high USB data speed would still
> > be nice.
>
> Yes, it would be nice. A MicroSD slot would also be nice, as it allows
> for quick&easy transfer of the whole media archive to a new device,
> without having to squeeze many Gigabytes through some bottleneck, and
> without the need for the old device to be still functional.
>
> Still, both features are absent from most modern devices. Even high-end
> ones, where additional costs of a few cents are certainly not a factor.

For Samsung it's apparently not "Even high-end ones", but *especially*
high-end ones. My low-to-mid-range Samsung Galaxy A51, *does* have a
MicroSD slot, so I'm a happy camper.

In article <j3l81q7tbnqi.dlg@v.nguard.lh>, VanguardLH <V@nguard.LH>
wrote:

>
> USB3 requires licensing. It's free, but a license is required.

all versions of usb require licensing. usb 3 is the current version,
which replaced usb 2, which replaced usb 1.1.

On 8/18/2023 8:32 AM, Frank Slootweg wrote:
> Hergen Lehmann <hlehmann.expires.12-22@snafu.de> wrote:

>> You are supposed to store such data in the cloud and use the
>> phone's memory only as a cache for recently used data.

> Yes, I know, but I'm a stubborn Dutch (cheap? :-)) geezer,

I pay US$2/mo for 100G. I consider that cheap...

> who likes to keep most things local. (I've only a (encrypted)
> differential backup of my most important stuff in the cloud,

That makes sense. I also encrypt my important sensitive stuff.

But I don't for music, photos, etc. That makes them easily available on
ALL of my devices including the ones that don't have enough storage to
hold it locally.

So no matter where I am I can whip out a device and bore folks with
pictures. It backfired on me recently though. A greatgrandkid was
showing me her 1st grade class photo. So I whipped out my Fire tablet
and showed her mine. All I got was an eye roll...

> so I can recover from theft, fire, etc. with my off-site backup and
> the delta in the cloud.)

A few years back my neighbor lost all his electronics in a burglary with
nothing backed up. Makes one get religion...

>> A MicroSD slot would also be nice, as it allows for quick&easy
>> transfer of the whole media archive to a new device, without having
>> to squeeze many Gigabytes through some bottleneck, and without the
>> need for the old device to be still functional.

Or just use use a MicroSD to USB adapter. No bottleneck...

>> Still, both features are absent from most modern devices. Even
>> high-end ones, where additional costs of a few cents are certainly
>> not a factor.

My GUESS would be that folks aren't using them much anymore. I haven't
used mine in over a year, maybe 2?

> For Samsung it's apparently not "Even high-end ones", but
> *especially* high-end ones. My low-to-mid-range Samsung Galaxy A51,
> *does* have a MicroSD slot, so I'm a happy camper.

Enjoy it while it lasts. My latest toy came with ONLY a USB-C port. But
I already had a box of adapters... ;)

Andy Burns <usenet@andyburns.uk> wrote:

> VanguardLH wrote:
>
>> If a mfr only has 1 VID but produced USB2 and USB3 phones, the same VID
>> would get used for both their USB devices although those devices differ
>> on USB speed. Tis likely a mfr would create a different model number to
>> reflect USB3 support, but the single VID would confuse the old USB2
>> models with the new USB3 models.
>
> But that's why a combination of VID and PID is used ...

Okay, my supposition got shot down. The VID can be looked up without
even having the product. The PID is part of the presentation data the
client gives the host which get saved as that device's enumeration data
in the registry (HKLM\SYSTEM\CurrentControlSet\Enum, 3 USB subkeys).
VID you can get without having the product. PID requires the USB
handshaking, so you must have the product and connect it. Have you seen
online lists of VIDs that also catalog the PIDs?

VIDs are registered, a 16-bit mfr identifier, and burned into the
[EEP]ROM in the device. However, you don't need to have the physical
product to get the registered VID that is available in online lists.

PIDs are not registered, a 16-bit descriptor (product number) of
whatever the mfr chooses to burn for the enumeration data in the
[EEP]ROM. Mfrs don't have to (and cannot) register the PID descriptors.
What I've seen in the VID lists for major mfs, like Samsung, HP, NEC,
Kodak, Logitech, and so on is they have several VIDs. Why would they
bother with more than one VID if adding a PID was all that was needed to
physically identify a product? VID+PID enumeration of VID+PID (with one
VID, and many PIDs) should be enough, but apparently it isn't.

Once a mfr purchases a VID registration, they could use it with physical
PIDs (burned into the device, not registered) for all their products (up
to 65536 products). Some might do it that way: 1 VID + many PIDs.
Others don't, and register multiple VIDs. I'm not a USB product maker,
so I don't know why a mfr would register multiple VIDs with each
incurring the registration process and cost money. I could guess they
want to differentiate between different types of product (say, printer
versus camera), or between product family lines, but that's just a guess
from cursory observation of VID definitions.

The VID+PID is needed by a host to determine the driver for the USB
device. That's only needed by a slave device. The host doesn't need a
VID. The slave must identify itself to the host. The host does not
need to identify itself to the slave. Slaves (aka devices) need
VID+PID. Hosts do not.

Say I give you a VID of 0810. You can look up its registration to find
that it is for Logitech and, in addition, for their QuickCam product
line, but that doesn't give you the PID which is needed for the VID+PID
identifier needed by the OS to figure out which driver gets used for the
device. Just knowing the registered VID doesn't give you the PID. You
get the PID, and the VID, when you have the device, and after you plug
it into the USB bus, and after the USB handshaking to get the VID+PID to
enumerate the device.

https://www.ti.com/lit/pdf/slla154, Section 3.2 says the device gets
re-enumerated when connected to the USB bus. Not sure about that. I've
had USB devices that failed because the enumeration data got corrupted
or became incorrect in the registry, so I had to unplug the USB device,
delete the device's enumeration data from the registry, and re-plug the
USB device to force a new copy of the device's enumeration data into the
registry whereupon the device was recognized and functioned okay.

As for the OP asking why there are no USB3 phones, there are. He didn't
look before buying his phone. If critical to him, it should've been one
of the specifications as criteria for which phone he bought.

I have to wonder if the OP's remembrance of some special USB cable for
an old phone was when there were charge-only USB cables (just have the
5V and gnd lines) and those that were full USB cables (power and
differential signal lines). For phones that didn't transfer data, the
power-only USB cable was a way for the mfr to cheap out on the item.

AJL <noemail@none.com> wrote:
> On 8/18/2023 8:32 AM, Frank Slootweg wrote:
> > Hergen Lehmann <hlehmann.expires.12-22@snafu.de> wrote:
>
> >> You are supposed to store such data in the cloud and use the
> >> phone's memory only as a cache for recently used data.
>
> > Yes, I know, but I'm a stubborn Dutch (cheap? :-)) geezer,
>
> I pay US$2/mo for 100G. I consider that cheap...

Yes, it's quite reasonable. For now, I - easily - get by with the free
15GB.

> > who likes to keep most things local. (I've only a (encrypted)
> > differential backup of my most important stuff in the cloud,
>
> That makes sense. I also encrypt my important sensitive stuff.
>
> But I don't for music, photos, etc. That makes them easily available on
> ALL of my devices including the ones that don't have enough storage to
> hold it locally.
>
> So no matter where I am I can whip out a device and bore folks with
> pictures. It backfired on me recently though. A greatgrandkid was
> showing me her 1st grade class photo. So I whipped out my Fire tablet
> and showed her mine. All I got was an eye roll...

I've no need for that yet and 'all of my devices' is currently only 2,
but maybe that'll change one day, who knows.

And indeed some/most of our loved ones have their stuff in the cloud,
but they're (mostly) Apple users, so what do *they* know!? :-)

> > so I can recover from theft, fire, etc. with my off-site backup and
> > the delta in the cloud.)
>
> A few years back my neighbor lost all his electronics in a burglary with
> nothing backed up. Makes one get religion...

Haven't lost anything yet (knocks on head/wood), but I was - amongst
others - a professional backup specialist, so these things come
naturally.

> >> A MicroSD slot would also be nice, as it allows for quick&easy
> >> transfer of the whole media archive to a new device, without having
> >> to squeeze many Gigabytes through some bottleneck, and without the
> >> need for the old device to be still functional.
>
> Or just use use a MicroSD to USB adapter. No bottleneck...

Yes, for the transfer scenario there are other options. For having
more storage in the phone, a MicroSD card sometimes can still be useful,
but with my current phone - even with many offline maps - I haven't yet
needed it, the 128GB storage is 'only' about half full.

> >> Still, both features are absent from most modern devices. Even
> >> high-end ones, where additional costs of a few cents are certainly
> >> not a factor.
>
> My GUESS would be that folks aren't using them much anymore. I haven't
> used mine in over a year, maybe 2?

Haven't used mine at all, nearly 3 years. But hush, we don't want to
wake 'Arlen' et al, do we now!?

> > For Samsung it's apparently not "Even high-end ones", but
> > *especially* high-end ones. My low-to-mid-range Samsung Galaxy A51,
> > *does* have a MicroSD slot, so I'm a happy camper.
>
> Enjoy it while it lasts. My latest toy came with ONLY a USB-C port. But
> I already had a box of adapters... ;)

On 8/18/23 12:07 PM, Frank Slootweg wrote:
> AJL <noemail@none.com> wrote:
>>
> Yes, for the transfer scenario there are other options. For having
> more storage in the phone, a MicroSD card sometimes can still be useful,
> but with my current phone - even with many offline maps - I haven't yet
> needed it, the 128GB storage is 'only' about half full.

Likewise. I keep SOME photos on the phone, but they all get transferred
to the computer and thence to 6 or more backups. Edited ones go to the
google-cloud for sharing.

>> >> Still, both features are absent from most modern devices. Even
>> >> high-end ones, where additional costs of a few cents are certainly
>> >> not a factor.
>>
>> My GUESS would be that folks aren't using them much anymore. I haven't
>> used mine in over a year, maybe 2?
>
> Haven't used mine at all, nearly 3 years. But hush, we don't want to
> wake 'Arlen' et al, do we now!?
>
>> > For Samsung it's apparently not "Even high-end ones", but
>> > *especially* high-end ones. My low-to-mid-range Samsung Galaxy A51,
>> > *does* have a MicroSD slot, so I'm a happy camper.
>>
>> Enjoy it while it lasts. My latest toy came with ONLY a USB-C port. But
>> I already had a box of adapters... ;)

I bought some from China. They're crap and don't fit tightly on either
end. How hard could it have been to make them better? I transfer
photos wirelessly, but I'd like to be able to use the old-style 10-foot
charging cable in the car.

Just ordered some from Amazon. Easy to return if they're crap.

--
Cheers, Bev

The Real Bev wrote:
> AJL wrote:

>> My latest toy came with ONLY a USB-C port. ButI already had a box of adapters... ;)

>I bought some from China. They're crap and don't fit tightly on either
>end. How hard could it have been to make them better?

Little kids have little hands so their assembly skills suffer.

>Just ordered some from Amazon. Easy to return if they're crap.

I got mine from Amazon (via China)...

On 8/18/23 4:06 PM, AJL wrote:
> The Real Bev wrote:
>> AJL wrote:
>
>>> My latest toy came with ONLY a USB-C port. But
> I already had a box of adapters... ;)
>
>>I bought some from China. They're crap and don't fit tightly on either
>>end. How hard could it have been to make them better?
>
> Little kids have little hands so their assembly skills suffer.
>
>>Just ordered some from Amazon. Easy to return if they're crap.
>
> I got mine from Amazon (via China)...

I've never actually figured out whether or not Prime is worth paying for
-- I'm afraid that it's not, but I don't drink and instant gratification
is a good substitute. We actually got something the day we ordered it!

--
Cheers, Bev
"John Wayne toilet paper -- It's rough, it's tough,
and it don't take no crap from nobody."

On 8/18/2023 8:04 PM, The Real Bev wrote:

> I've never actually figured out whether or not Prime is worth paying
> for I'm afraid that it's not, but I don't drink and instant
> gratification is a good substitute.

Depends on your usage.

With packages being delivered almost every day at my house (ordered
mostly by you know who) it saves me money in postage alone. And when I
get specific birthday requests from the out of town greatgrandkids
Amazon wraps and ships them for me. And even here in town it's often
easier for me to order from Amazon and have it at the door a day or two
later than to have to run to the store. Then I also use their Prime
video and book loaning services.

So I like it (as do around 150 million others in the US) but YMMV...

> We actually got something the day we ordered it!

My best time was 4 hours from order to door. Amazing to me the system
that must be behind that efficient service. However I live within 5
miles of two humongous Amazon plants so that may help a bit...

On 8/18/23 9:08 PM, AJL wrote:
> On 8/18/2023 8:04 PM, The Real Bev wrote:
>
>> I've never actually figured out whether or not Prime is worth paying
>> for I'm afraid that it's not, but I don't drink and instant
>> gratification is a good substitute.
>
> Depends on your usage.
>
> With packages being delivered almost every day at my house (ordered
> mostly by you know who) it saves me money in postage alone. And when I
> get specific birthday requests from the out of town greatgrandkids
> Amazon wraps and ships them for me. And even here in town it's often
> easier for me to order from Amazon and have it at the door a day or two
> later than to have to run to the store. Then I also use their Prime
> video and book loaning services.

Book loaning? I need to check that out. Never heard of it.

I recently renewed my library card and was amazed at the services it now
offers. State park passes. Wifi hotspots. Chromebooks. E-books.
They accept dead flashlight batteries. Plus they had a book AT MY LOCAL
BRANCH that I couldn't find through the usual sources or at an
exorbitant price at Amazon. Our libraries deserve first cut of whatever
funds our cities extort from the populace.

> So I like it (as do around 150 million others in the US) but YMMV...
>
>> We actually got something the day we ordered it!
>
> My best time was 4 hours from order to door. Amazing to me the system
> that must be behind that efficient service. However I live within 5
> miles of two humongous Amazon plants so that may help a bit...

We're close too, but I'm not sure how many.

--
Cheers, Bev
"The danger to America is not Barack Obama but a citizenry capable of
entrusting a man like him with the presidency." -- Unknown

On Fri, 18 Aug 2023 20:04:23 -0700, The Real Bev
<bashley101@gmail.com> wrote:

>On 8/18/23 4:06 PM, AJL wrote:
>> The Real Bev wrote:
>>> AJL wrote:
>>
>>>> My latest toy came with ONLY a USB-C port. But
>> I already had a box of adapters... ;)
>>
>>>I bought some from China. They're crap and don't fit tightly on either
>>>end. How hard could it have been to make them better?
>>
>> Little kids have little hands so their assembly skills suffer.
>>
>>>Just ordered some from Amazon. Easy to return if they're crap.
>>
>> I got mine from Amazon (via China)...
>
>I've never actually figured out whether or not Prime is worth paying for
>-- I'm afraid that it's not, but I don't drink and instant gratification
>is a good substitute. We actually got something the day we ordered it!

Worth it? Probably not, for most people. But it's often a big
convenience for me, and I stay with it.

VanguardLH <V@nguard.lh> wrote:
> Okay, my supposition got shot down. The VID can be looked up without
> even having the product. The PID is part of the presentation data the
> client gives the host which get saved as that device's enumeration data
> in the registry (HKLM\SYSTEM\CurrentControlSet\Enum, 3 USB subkeys).
> VID you can get without having the product. PID requires the USB
> handshaking, so you must have the product and connect it. Have you seen
> online lists of VIDs that also catalog the PIDs?

Yes:
https://the-sz.com/products/usbid/
http://www.linux-usb.org/usb-ids.html

However, in practice you can't know the VID of a device in advance, because
there isn't a strong mapping from manufacturer's corporate identity to the
device hardware. For many devices the brand on the outside does not match
the chip that's inside, and it's the chip that provides the VID:PID.
Sometimes there's a flash memory which can be reprogrammed to present as the
vendor, but sometimes it just appears with the chip manufacturer's ID.

> VIDs are registered, a 16-bit mfr identifier, and burned into the
> [EEP]ROM in the device. However, you don't need to have the physical
> product to get the registered VID that is available in online lists.

You do, in practice.

> PIDs are not registered, a 16-bit descriptor (product number) of
> whatever the mfr chooses to burn for the enumeration data in the
> [EEP]ROM. Mfrs don't have to (and cannot) register the PID descriptors.
> What I've seen in the VID lists for major mfs, like Samsung, HP, NEC,
> Kodak, Logitech, and so on is they have several VIDs.

There's no 'registration' beyond paying for the VID. The lists above are
just users collecting data from devices they have.

> Once a mfr purchases a VID registration, they could use it with physical
> PIDs (burned into the device, not registered) for all their products (up
> to 65536 products). Some might do it that way: 1 VID + many PIDs.
> Others don't, and register multiple VIDs. I'm not a USB product maker,
> so I don't know why a mfr would register multiple VIDs with each
> incurring the registration process and cost money. I could guess they
> want to differentiate between different types of product (say, printer
> versus camera), or between product family lines, but that's just a guess
> from cursory observation of VID definitions.

Last time I checked (a while back) the fee was $1500. Which is pocket
change for a big manufacturing organisiation. So they may request multiple
VIDs for organisational convenience (each business unit can issue PIDs
without having to set up a company-wide registry), or for corporate
structure reasons (Samsung SSDs v Samsung phones, or Samsung North America v
Samsung Korea). It is quite likely that big vendors like Samsung may
exhaust the PID space of a single VID (think about all the variations of
products there are).

> Say I give you a VID of 0810. You can look up its registration to find
> that it is for Logitech and, in addition, for their QuickCam product
> line, but that doesn't give you the PID which is needed for the VID+PID
> identifier needed by the OS to figure out which driver gets used for the
> device. Just knowing the registered VID doesn't give you the PID. You
> get the PID, and the VID, when you have the device, and after you plug
> it into the USB bus, and after the USB handshaking to get the VID+PID to
> enumerate the device.

That's how USB works. You know nothing about the device until it's
inserted, at which point you enumerate it and decide what to do based on
the VID:PID and the device class information.

> As for the OP asking why there are no USB3 phones, there are. He didn't
> look before buying his phone. If critical to him, it should've been one
> of the specifications as criteria for which phone he bought.

Originally, I think it was the Galaxy S5 that supported USB 3 (via the
micro-USB 3.0 connector with the extra sidecar). But the connector was
awkward and it cost power to use it, and I think vendors concluded it wasn't
worth it.

Nowadays I imagine the power is much less of a concern, but they've got into
the habit of not including it. There is also an issue with interference
between USB 3 and 5GHz wifi, which might be harder to manage in a small form
factor like a phone.

While Apple and Google want you to use cloud services, I'm not sure Samsung
or Xiaomi do. So I'm not buying that as a reason. However they are using
silicon from Qualcomm, and maybe QC have been 'persuaded' not to include USB
3.

It sounds like the iPhone 15 Pro will get USB 3 speeds via the USB-C
connector, so maybe there will be resurgence from Qualcomm and Android OEMs
to compete.

Theo

Theo <theom+news@chiark.greenend.org.uk> wrote:

> VanguardLH <V@nguard.lh> wrote:
>
>> Okay, my supposition got shot down. The VID can be looked up without
>> even having the product. The PID is part of the presentation data
>> the client gives the host which get saved as that device's
>> enumeration data in the registry (HKLM\SYSTEM\CurrentControlSet\Enum,
>> 3 USB subkeys). VID you can get without having the product. PID
>> requires the USB handshaking, so you must have the product and
>> connect it. Have you seen online lists of VIDs that also catalog
>> the PIDs?
>
> Yes:
> https://the-sz.com/products/usbid/

If I enter a VID, I get several products listed by the vendor. How can
searching on, say, "USB hub" give me the vendor, or even a specific
product?

VIDs are registered. PIDs are not, and can be whatever the mfr want to
code into the [EEP]ROM for their product. The onus on keeping PIDs
unique under a VID is an onus upon the mfr, not the USB IF.

> http://www.linux-usb.org/usb-ids.html

I went instead to http://www.linux-usb.org/usb.ids to see the list.

Those product descriptors are not registered with the VID. Those are
comments added by the author of the list.

https://www.usb.org/sites/default/files/vid_only_form_070119.pdf
No entry to specify the product descriptor, just a request for VID only.

The VID is assigned to the mfr. The mfr makes up their own PID. For
example, from the list, the 0810 PID is reused by 9 *different* vendors.
Within a VID, a PID assigned by the mfr often does identify a product,
or it might identify a family of products. Some VIDs don't even have a
PID listed.

It's nice there are online lists of VID+PID, but only the VID is
registered. The PID, if any, is assigned by the mfr, and the mfr
decides how to assign PIDs. Comments added to a VID list to note PIDs
are by observation, not by official registration for lookup. The
comments on PIDs are to provide human readable description instead of
just seeing cryptic digits.

The PID, like the VID, is a just a 16-bit number, not a string. The mfr
decided on the PID value. Users added comments on what a PID might mean
by observation, and reported their discovery to the list makers.

Go to https://usb-ids.gowdy.us/read/UD/. Select a VID, like 0810. That
list says the VID belongs to Personal Communication Systems. Click on
that entry. The next page shows PIDs reported by users: both PID value,
and what the user described. Notice the "Add item" link where users can
report a new PID and add some notes to it. If you then click on a PID
entry, you'll see even more user notes. There you'll even see a Discuss
link (but I don't have an account to see if you're directed to web-based
forums, a mailing list, or what).

VIDs are registered with USB IF. The PIDs are crowdsourced data. The
PID info could be handy, but it's not part of the VID registration.

VID: registered.
PID: Mfr assigned.
VID+PID: Product numbers created by mfr to identify their products. No
descriptive strings.

> However, in practice you can't know the VID of a device in advance,
> because there isn't a strong mapping from manufacturer's corporate
> identity to the device hardware. For many devices the brand on the
> outside does not match the chip that's inside, and it's the chip that
> provides the VID:PID. Sometimes there's a flash memory which can be
> reprogrammed to present as the vendor, but sometimes it just appears
> with the chip manufacturer's ID.

About the only time I've dealt with the VID (other than my own
troubleshooting) is when someone asks for help on figuring out just what
is a device enumerated in the registry by VID+PID. So lists, like
above, can be helpful in determing whose and which product (or product
family) might be the device. While the chip reports the VID (assigned
to the vendor) and PID (assigned by the vendor), there are strings along
with the enumerated registry data. For example, in the registry, I go
to HKLM\SYSTEM\CurrentControlSet\Enum\USB, pick a VID&PID entry.
Subentries show strings that seem to have come from the device probably
as presentation data during USB handshaking. While the chip doles out
the VID&PID identifier, seems it also doles out description info, too.
That's probably where the crowdsourced PID descriptors come from.

>> PIDs are not registered, a 16-bit descriptor (product number) of
>> whatever the mfr chooses to burn for the enumeration data in the
>> [EEP]ROM. Mfrs don't have to (and cannot) register the PID descriptors.
>> What I've seen in the VID lists for major mfs, like Samsung, HP, NEC,
>> Kodak, Logitech, and so on is they have several VIDs.
>
> There's no 'registration' beyond paying for the VID. The lists above
> are just users collecting data from devices they have.

Ah, thanks for verifying my interpretation of those VID lists. The PIDs
looked to be from users reporting what they discovered.

>> Once a mfr purchases a VID registration, they could use it with physical
>> PIDs (burned into the device, not registered) for all their products (up
>> to 65536 products). Some might do it that way: 1 VID + many PIDs.
>> Others don't, and register multiple VIDs. I'm not a USB product maker,
>> so I don't know why a mfr would register multiple VIDs with each
>> incurring the registration process and cost money. I could guess they
>> want to differentiate between different types of product (say, printer
>> versus camera), or between product family lines, but that's just a guess
>> from cursory observation of VID definitions.
>
> Last time I checked (a while back) the fee was $1500.

Hmm, I thought it was $6000 USD.

> It is quite likely that big vendors like Samsung may exhaust the PID
> space of a single VID (think about all the variations of products
> there are).

VID: 16-digit number (2^16 = 65536)
PID: 16-digit number (2^16 = 65536)

What happens when there are more folks trying to register VIDs than the
64K range available? Reminds me how IPv4 (32-bit) ran out of addresses,
so we had to migrate to IPv6 (128-bit).

For each VID, there could be 65K product numbers (PIDs). I suppose,
especially over decades, that a mfr would run out of PIDs under a
particular VID, so they'd have to get another VID to give them another
65K count in product numbers.

Something I hadn't considered in this discussion is when a PID is
needed. VIN+PID is needed to specify a driver for the device. However,
if the device is generic (keyboard, mouse, other HID device, camera,
speaker, mass storage), a PID isn't needed. During USB handshaking, the
presentation data specifies the type of device, so it could specify a
generic one if a device-specific driver is not needed. A generic driver
gets used with a generic device, but I think the device type must be
identified by the device to the host.

The problem with the expense of needing a VID is its price. For big
companies, no problem. Just part of the expense of doing business.
However, open source projects might need a VID, too, and free stuff
generates no revenue to offset the cost of getting a VID. There are VID
owners that will share their VID(s), and dole out PIDs for free to
open-source authors, like:

http://wiki.openmoko.org/wiki/USB_Product_IDs
https://pid.codes/howto/
(a search shows several more)

On 2023-08-16 11:26, David wrote:
> Looking at recent phones they seem to have USB C connectors for charging
> and data, but the underlying USB protocol is USB 2.
>
> Any specific reason?
>
> I vaguely recall a Samsung a good few years back having a special
> connector for a special cable to allow higher speed transfers, but that
> was back in the mists of time.

Recently I wanted to download some photos to my computer, and as at the
moment all the usb-3 sockets in the front were in use, I connected to
one of the usb-2 connectors (at the front of the computer). There were
more connectors at the back, but I did not want to get up.

Just reading the directory took for ever. So yes, usb-3 is faster on my
phone. Full usb-3 speed, I do not know. Maybe not.

Phone is a Motorola G52

--
Cheers,
Carlos E.R.

"Carlos E. R." <robin_listas@es.invalid> wrote:

> Recently I wanted to download some photos to my computer, and as at
> the moment all the usb-3 sockets in the front were in use, I
> connected to one of the usb-2 connectors (at the front of the
> computer). There were more connectors at the back, but I did not want
> to get up.
>
> Just reading the directory took for ever. So yes, usb-3 is faster on
> my phone. Full usb-3 speed, I do not know. Maybe not.
>
> Phone is a Motorola G52

Hmm, the following sites say the Moto G52 uses USB 2:

https://www.gsmarena.com/motorola_moto_g52-11457.php
https://www.phonearena.com/phones/Motorola-moto-g52_id11899

Even Motorola says so, too:

https://www.motorola.com/we/smartphones-moto-g-52/p?skuId=445

Are the front-side USB connectors built into the desktop's system case
which likely means they go to headers on the mobo, and perhaps some of
the back-side USB connectors are not backpanel ports from the mobo, but
instead on a daughtercard? Are there any backside/backpanel USB2 ports?
If so, the front-side USB2 ports might be sharing the same USB
controller on the mobo with backside USB2 ports (if the backside USB2
ports were active instead of unused or sitting idle).

What type of files were in the directory/folder? Image files? Is your
File Explorer configured to show thumbnails of your files? That
requires interrogating (opening) the files to get at a thumbnail of
image files, and that slows down updating all those thumbnails of files.
Users will sometimes complain that File Explorer is showing an
ever-progressing green animation in the address bar that takes a long
time. Once they disable showing thumbnails, all that processing
overhead is gone. There is a thumbnail cache, but quite often it has to
get rebuilt, so you see the green progress animation in the address bar.

<Aside>

I asked about the origin of the USB controller because it can affect the
realized transfer rate. I have front-side USB3 ports in my system case
that go to USB3gen3.1 headers on the mobo, not to USB2 headers.

I deliberately bought a mobo with lots of USB ports and headers. On the
backpanel: 7 type A USB3gen3.1 ports, and 1 type C USB3gen3.1 port. At
the top front of the case, are 2 type A USB ports connected to a
USB3gen3.1 header on the mobo. With just the mobo's backpanel and
case-mount USB ports, I have 10 total.

I have 7 more front-side USB ports, but those are in an expansion module
that usurped a 5.25" bay in the mini-tower case, and connect to
USB3gen3.1 headers on the mobo. Power is from a PSU SATA connector for
full 0.9A on each USB3 port. Connect using 20pin (well, 19pin since
blank one used as a key) to USB3gen3.1 header on mobo, so 2 USB mobo
ports shared across 7 external ports, so typically I only have 1 or 2
USB devices active at the same time, but may leave plugged in other idle
devices.

In all, I have 15 USB3gen3.1 ports available. The mobo does have two
USB2 headers, but they're unused. If I ever needed more USB ports, I
could use backplates with USB ports to connect to the remaining USB3
header (2 USB) and the 2 USB2 (1 USB each) headers to add up to another
4 USB ports, but in the 4 years since I built this setup I've not needed
more USB ports. As you can see, I don't like using external powered USB
hubs.

What you have available in your setup depends on your mobo, what it has
for backpanel USB ports, what it has for USB headers on the mobo, and
how you utilized them. If yours is a pre-built, you get whatever the
manufacturer gave you. I don't think I've owned a prebuilt PC for
probably about 3 decades. I like designing my own, and getting what I
want instead of settling for what someone else will give me.

</Aside>

On 2023-08-20 16:26, VanguardLH wrote:
> "Carlos E. R." <robin_listas@es.invalid> wrote:
>
>> Recently I wanted to download some photos to my computer, and as at
>> the moment all the usb-3 sockets in the front were in use, I
>> connected to one of the usb-2 connectors (at the front of the
>> computer). There were more connectors at the back, but I did not want
>> to get up.
>>
>> Just reading the directory took for ever. So yes, usb-3 is faster on
>> my phone. Full usb-3 speed, I do not know. Maybe not.
>>
>> Phone is a Motorola G52
>
> Hmm, the following sites say the Moto G52 uses USB 2:
>
> https://www.gsmarena.com/motorola_moto_g52-11457.php
> https://www.phonearena.com/phones/Motorola-moto-g52_id11899
>
> Even Motorola says so, too:
>
> https://www.motorola.com/we/smartphones-moto-g-52/p?skuId=445

Curious. Yet I noticed the speed difference. I must measure it with a
stopwatch.

>
> Are the front-side USB connectors built into the desktop's system case
> which likely means they go to headers on the mobo, and perhaps some of
> the back-side USB connectors are not backpanel ports from the mobo, but
> instead on a daughtercard? Are there any backside/backpanel USB2 ports?
> If so, the front-side USB2 ports might be sharing the same USB
> controller on the mobo with backside USB2 ports (if the backside USB2
> ports were active instead of unused or sitting idle).

They are all sockets in the box connected to the mobo via cables, IIRC.

>
> What type of files were in the directory/folder? Image files?

the DCIM/Camera directory, so all photos.

> Is your
> File Explorer configured to show thumbnails of your files?

Normally yes, but they don't display for the phone.

> That
> requires interrogating (opening) the files to get at a thumbnail of
> image files, and that slows down updating all those thumbnails of files.
> Users will sometimes complain that File Explorer is showing an
> ever-progressing green animation in the address bar that takes a long
> time. Once they disable showing thumbnails, all that processing
> overhead is gone. There is a thumbnail cache, but quite often it has to
> get rebuilt, so you see the green progress animation in the address bar.
>
> <Aside>
>
> I asked about the origin of the USB controller because it can affect the
> realized transfer rate. I have front-side USB3 ports in my system case
> that go to USB3gen3.1 headers on the mobo, not to USB2 headers.
>
> I deliberately bought a mobo with lots of USB ports and headers. On the
> backpanel: 7 type A USB3gen3.1 ports, and 1 type C USB3gen3.1 port. At
> the top front of the case, are 2 type A USB ports connected to a
> USB3gen3.1 header on the mobo. With just the mobo's backpanel and
> case-mount USB ports, I have 10 total.
>
> I have 7 more front-side USB ports, but those are in an expansion module
> that usurped a 5.25" bay in the mini-tower case, and connect to
> USB3gen3.1 headers on the mobo. Power is from a PSU SATA connector for
> full 0.9A on each USB3 port. Connect using 20pin (well, 19pin since
> blank one used as a key) to USB3gen3.1 header on mobo, so 2 USB mobo
> ports shared across 7 external ports, so typically I only have 1 or 2
> USB devices active at the same time, but may leave plugged in other idle
> devices.
>
> In all, I have 15 USB3gen3.1 ports available. The mobo does have two
> USB2 headers, but they're unused. If I ever needed more USB ports, I
> could use backplates with USB ports to connect to the remaining USB3
> header (2 USB) and the 2 USB2 (1 USB each) headers to add up to another
> 4 USB ports, but in the 4 years since I built this setup I've not needed
> more USB ports. As you can see, I don't like using external powered USB
> hubs.
>
> What you have available in your setup depends on your mobo, what it has
> for backpanel USB ports, what it has for USB headers on the mobo, and
> how you utilized them. If yours is a pre-built, you get whatever the
> manufacturer gave you. I don't think I've owned a prebuilt PC for
> probably about 3 decades. I like designing my own, and getting what I
> want instead of settling for what someone else will give me.

No, not prebuilt, but other mobo features were more important to me :-)

I have an actual true RS232 connector, for instance :-)

>
> </Aside>

--
Cheers,
Carlos E.R.

"Carlos E. R." <robin_listas@es.invalid> wrote in message
news:kkh14jF24epU1@mid.individual.net...
> On 2023-08-20 16:26, VanguardLH wrote:
>> "Carlos E. R." <robin_listas@es.invalid> wrote:
>>
>>> Recently I wanted to download some photos to my computer, and as at
>>> the moment all the usb-3 sockets in the front were in use, I
>>> connected to one of the usb-2 connectors (at the front of the
>>> computer). There were more connectors at the back, but I did not want
>>> to get up.
>>>
>>> Just reading the directory took for ever. So yes, usb-3 is faster on
>>> my phone. Full usb-3 speed, I do not know. Maybe not.

Both my previous phone (Samsung Galaxy S2) and my current one (Samsung
Galaxy A54) take a *very* long time to display the contents of the DCIM
folder when accessed by USB cable from a Windows PC. It seems to show each
new icon/filename (depending on the Windows Explorer "View" setting for that
folder) at a rate of one every second or so - so it took ages to display the
whole set of photos so I could select the recent ones that I wanted to copy
to my PC for permanent storage. OK, I could have deleted more of the old
photos once I'd copied them to the PC, but it is nice to keep some on the
phone in case I ever want to access them when I'm away from home and my PC.

I'm not sure whether this is slow USB (eg USB1) or a weird protocol (ie the
folder on the phone doesn't appear to the PC as a normal external USB HDD).

It is sometimes quicker to select the photos and email them to myself, to be
picked up at the PC, though that loses the file timestamp on the photos - it
becomes the time when the photos were saved from the email, rather than the
time when the photo was taken. Obviously the filename that includes the
date/time remains.