On Wed, May 31, 2023 at 5:05 PM home user <mattisonw@comcast.net> wrote:
(replying to Tim and George)

Thank-you Tim and George.

On 5/29/23 7:01 AM, George N. White III wrote:

> Early USB-3 was problematic.  Have you used USB memory sticks before?   There have been cheap USB sticks that advertise a much higher
> capacity than they actually provide.

I've been using USB sticks for several years, and USB-3.0 sticks for a few years.  For this, I'm using 32 GB Samsung USB-3.1 sticks.  Samsung
is one of the good brands, right?

The crooks copy name-brand packaging.  Bogus USB drives can be introduced anywhere in the supply chain, so the problem
is usually discovered only when it refuses to hold advertised capacity.    Some reports say they silently discard data past the
physical capacity and fill reads with garbage, so casually transferring a large file will appear to work until you actually check the
contents.
 

On 5/28/23 10:21 PM, Tim via users wrote:

> It could just be /that/ USB stick.  Did you have another to try?

After posts from Tim and George, I did more "experiments":

0. At the start, stick_1 has "mt86plus_6.20_64.iso" on it.  stick_2 has "Fedora-Workstation-Live-x86_64-37-1.7.iso" on it, but in pieces.

1. I used Fedora Media Writer to put "mt86plus_6.20_64.iso" on stick_2; no hint of trouble.  I tested it in both USB-3 ports; it worked fine.

2. I used Fedora Media Writer to put Fedora-Workstation-Live-x86_64-37-1.7.iso" on stick_1; no hint of trouble.  I tested it in the left USB-3 port; it failed.  I tested it in the right port; it succeeded.

3. I used Fedora Media Writer to put "mt86plus_6.20_64.iso" on stick_1; no hint of trouble.  I tested it in both USB-3 ports; it worked fine.

4. I used Fedora Media Writer to put Fedora-Workstation-Live-x86_64-37-1.7.iso" on stick_2; no hint of trouble.  I tested it in the left USB-3 port; it failed.  I tested it in the right port; it succeeded.

It seems that the problem is not /that/ stick.  memtest consistently working in both ports and on both sticks suggests to me that neither the ports nor the sticks are at fault.  But Fedora Live behaves inconsistently.

USB3 uses frequencies higher than USB2, so other devices can be affected by poor shielding at the ports, and
kinked cables or excessively long leads connecting port to system board cause deterioration of the signals.  Are
both ports soldered neatly to the system board with short leads?  Consider adding a USBC card to an older desktop.

All that entropy talk in the user's list recently: might some of that entropy be infecting Fedora Live?
(just kidding)

On 5/29/23 7:01 AM, George N. White III wrote:

> ... You can mount the iso Linux
> and use cmp to check for files that differ between .iso and USB.

My Downloads directory (from which the copy to the sticks comes) has one .iso file; the stick has a few folders and several files.  How would I do that cmp?

With Fedora Live apparently working adequately if the stick is in the right port, the original goal is satistied.


--
George N. White III