Warning from Printer Manager - Not found

I normally have my mobile phone plugged in to a USB port of my BBxM to charge while I’m at the keyboard. Plugging and unplugging it has been no problem until I upgraded the ROM to the 2010-01-05 build, whereupon I get a single-tasking error box titled “Warning from Printer Manager” with the error message “Not found” immediately every time I unplug the phone. The phone offers 5 modes for the USB connection, but I always leave it on the default, “No data transfer”. I’m not aware of any update to the phone’s firmware that coincides with the start of this issue, but unfortunately I can’t see anything in its log that tells me when any updates were applied.

But here’s the kicker: I reverted to the 2019-12-15 build of RO, and the problem is still here.

Can anyone suggest any cause, or anything I can do to diagnose the cause?

Try a different lead known to work with other chargers.

You may be over-complicating things!

Hmm. The phone is the only device I have with a USB Type-C connector. The USB cable that came with it is the only one I have with a USB Type-C plug.

PoundLand or PoundWorld tomorrow, then, if they have them!

Or Amazon or Ebay!

I had a micro lead which went faulty and “crowbarred” (presumably) my 2 port charger so the LED indicator went out and charging stopped on both outputs.

Luckily I have many such leads, and simply replacing it cured that problem.

Whether your case is similar requires a replacement lead substituting to check!

Try a different lead known to work with other chargers.

Yes, there are some “charging only” leads that don’t have the data connections (annoying as hell when you pick one up to transfer data, there’s usually no indication it isn’t fully wired).

However this is a “fix”, not a solution.

so the LED indicator went out and charging stopped on both outputs.

It’s rare that a lead goes short circuit. Usually they go open circuit, which would mean nothing useful happening.

A short circuit on the USB port? I rather imagine the Beagle would have a nervous breakdown if that happened.

I reverted to the 2019-12-15 build of RO

Do you have an older build (or try the stable – it’s about a year old isn’t it?)?

or anything I can do to diagnose the cause?

Clearly something is being seen that is tripping up Printers (or maybe Printers via DeviceFS?).

Open a TaskWindow. Enter the command *USBDevices (IIRC!). Now plug in the phone, count to ten, and do it again. Anything different?

I’ve noticed that more recent builds of Android (my S7 and S9) seem to present multiple devices. The first is whatever the phone is set up to be (MTP, PTP, whatever), and the second is a debug device used by background stuff and accessible with adb. Maybe, if the second is being seen by RISC OS, it is upsetting things?

The phone is the only device I have with a USB Type-C connector.

Really, that should be about as relevant as whether a device has a mini USB or micro USB connection. In other words, not at all. USB 3 type C will happily fall back to plain ordinary USB with older machines.
But feel free to plug it in the other way up so the data falls out the right way up. :-p

but I always leave it on the default,

Strange. My S9 defaults to MTP (but it’ll present a dummy blank device until you authorise the connection on the phone).

I’m not aware of any update to the phone’s firmware that coincides with the start of this issue,

I trust you have set updates to prompt and not just install it.
It’s always best to ignore a major update for a few days, then hit Google to see if there are any issues to be aware of with the update. Mobile phones don’t have much in the way of rollback capability…neither do the app stores…it’s as if nobody is capable of imagining anything going wrong. Hmm.

but unfortunately I can’t see anything in its log that tells me when any updates were applied.

Most likely place to look is settings, about phone, software version, kernel version. Mine says 14th Feb 2019 (I’ve not updated to Android 9 because it breaks the video editor that I use a lot).
Good luck getting any information on what actually changed. It’s like the never ending selection of app store offerings that list the changes as “bug fixes and improvements”. Of course we’ll never see proper change logs because one of the worst offenders here is Google…

Anyway, look at the USB devices and look at DeviceFS and see if anything changes when the phone is connected.

A short circuit on the USB port?

Just looked at the schematic. USB power goes via a TPS2045 which is supposed to have short circuit protection. One should just hope the power chip catches the problem before it sinks enough power to crash the board.

To explain what I mean, my netbook has USB ports that can detect short circuits and over current conditions. Unfortunately it also has a power/charging system that is more sensitive. So if there’s a short on the USB (and since the port on the left is broken, plugging anything in will short the power to the plug shielding) then the USB hub never gets a look in, the power system immediately kills the entire device to the point where I have to disconnect DC and remove the battery pack in order to reset it!

The Pi schematics are useless and don’t mention anything about the USB ports. Is there protection on the 5V output or do we rely upon the polyfuse? If there is protection, is it per port?

I think we can agree, however, that short = bad.

I was only having a punt – if I were right that’d probably be a first!

But worth checking!

“not found” from printer manager… I’m just reminded of when !Printers dropped a system variable which was not uncommmonly used by applications (I forget the actual one). Just wondering if this is in any way related? That being said, in those cases, the “not found” tended to come from programs which expected the system variable to be present, not from printer manager itself…

Of course, it would be useful if Printer Manager were to say what it hadn’t found.

But feel free to plug it in the other way up so the data falls out the right way up. :-p

I know you jest but a few weeks ago we were looking at the USB C pinout and spec. and spotted that ‘the otherway round’ is not exactly the same and you can detect which way round a cable is plugged in! I said someone will produce a device that only works properly with the cable in one orientation!

Never thought to connect my phone to the BB -xM. Doing so, USBinfo shows it being seen available for audio, camera or radio link depending on the phones settings. It’s a Pixel 3a but presents itself as a Nexus device.

Charges on the 5 Jan ROM (presume Dave meant 2020 in initial post).

(presume Dave meant 2020 in initial post)

Indeed I did (hit the adjacent key), thanks.

Further experimentation shows that removing PDF from the printers (put there by PrintPDF) solved the problem. I have no idea why. I rarely need to print to PDF, so there’s no need for it to be there normally – I can just add it when I need it.

As for the cable: it has a permanent set in it, it only sits comfortably one way up, so there’s no doubt that it was always the same way up. Turning the reversible end the other way up didn’t change the symptoms. Unplugging either end was also identical. It is a charge and sync cable, as supplied with the phone (OnePlus 3T) by the manufacturer, from whom I bought it direct. (Also a Dash charger.)

Rick, I’m not sure what you were expecting me to find from USBDevices, but it does turn up as “Android Android”. I do have a USB printer permanently plugged in, but it’s never normally powered up, and it doesn’t appear as a USB device unless it is powered.

(Rick – the phone tells me the last upgrade, and all the version info, including the software’s dates; but it says nothing about when the update was applied. I’m sure that the last upgrade was applied more recently than the software’s dates.)

Rick, I’m not sure what you were expecting me to find from USBDevices, but it does turn up as “Android Android”.

There you go. In your original post you said:

The phone offers 5 modes for the USB connection, but I always leave it on the default, “No data transfer”.

Which would imply that you just want to charge the battery and that the phone shouldn’t “appear” as anything to the host.

Yet, something shows up.

Can you look at the USB device information (off the top of my head, *USBDevInfo followed by the device number (use *USBDevices for that)). It should tell you vendor ID, product ID, etc. Copy-paste the output of the command here so we can try to figure out what RISC OS is seeing.

and it doesn’t appear as a USB device unless it is powered.

That’s normal for devices that aren’t bus powered, my keyboard (as in music) is the same.

the phone tells me the last upgrade, and all the version info, including the software’s dates;

Yes, it’ll provide you with a tonne of info. Not much of it terribly useful. It certainly won’t tell you what the update actually adds/removes/fixes. Neither will any of the subsequent updates you might look at.
My phone had an update, something like 120MiB. For “an improved user interface and other bug fixes”. That was actually surprisingly descriptive, usually it says less.

I think it comes to a pretty pass when I need to download an app to test my phone for various known vulnerabilities because the manufacturer’s update blurb doesn’t mention any of this stuff.

For what it is worth, 120MiB, looks and feels exactly the same. <shrug>

I’m sure that the last upgrade was applied more recently than the software’s dates.

That’s normal.
Any competent company would freeze all development, build a firmware, and have it tested prior to deployment.

I recall back in the early days (like Android 2 era), one of the major companies rolled out an update that went horrifically wrong. Update was installed, phone rebooted to finish up installation.
And rebooted.
And rebooted.
And…

The devices were unrecoverable.

That’s why these days any useful device will have a special hidden bootloader (usually powering up while holding one of the volume keys). This doesn’t do a lot, but it does allow you to at least try to apply updates from a file on SD card.

Plus, I would imagine reputable manufacturers do a reasonable amount of testing so update issues will likely be quirks (of varying levels of annoyance) rather than anything as major as bricking the device. It isn’t worth the publicity…

I would imagine reputable manufacturers do a reasonable amount of testing

https://seclists.org/cert/2020/12

Patch to the patch to the patch for RDP.

Ah, yes, you did say reputable manufacturers

Further experimentation shows that removing PDF from the printers (put there by PrintPDF) solved the problem. I have no idea why. I rarely need to print to PDF, so there’s no need for it to be there normally – I can just add it when I need it.

Does the active Scrap directory contain a PrintPDF folder? The default PrintPDF configuration has the print to file set to <Wimp$ScrapDir>.PrintPDF.Printout/ps (IIRC), and ISTR Printers getting unhappy if the folder isn’t there.

Running PrintPDF will create this folder if it is missing.

Does the active Scrap directory contain a PrintPDF folder?

Aha! No, it doesn’t. There are two folders in ScrapDirs: ID60f06a0a, dated last March, which contains 21 folders (including PrintPDF) and 5 files; and ScrapDir, dated 2010 Jan 6, which contains 10 folders (not including PrintPDF) and 1 file. ScrapDir is the active scrap directory.

So that explains the symptoms – thank you very much – and of course will give me a proper solution, rather than a work-round.

It would be interesting to know why an OS upgrade changed the active ScrapDir.

The phone offers 5 modes for the USB connection, but I always leave it on the default, “No data transfer”.

Which would imply that you just want to charge the battery and that the phone shouldn’t “appear” as anything to the host.

No, it implies that the device can appear in 5 modes, selectable in software. It would be hard to make a device not appear to USB at all in one mode, and it would be nonsensical (for example, the charge current would be limited to 100mA).

Anyway, we know now what’s happened, so following up this line is pointless.