I am taking a break from working with PicoCalc at the moment because my one and only Picocalc is doing an I2C endurance test that I do not want to interrupt.
Lately the word “Fragile” in connection with PicoCalc has been on my mind very often, this is not a reference to the 1971 Yes album but more to the PicoCalc environment.
One fragile component of the PicoCalc is know to many in the community: the LCD screen that breaks while installing before it can even be used. My experience with this component is that when handled with care (knock on wood) is that there is a minimal risk of damaging the LCD while swapping pico modules. I taped down the screen, I am using the PicoCalc without the backcover with the mainboard held down using rubber washers (like the ones found in a water tap). This has been a great help because I have been swapping the Pico modules many times every day over the last months. At this moment I have 6 Pico modules (PicoH, PicoHW, Pico2H, Pico2HW, PicoPlus2, PicoPlus2W) that I use to test the changes I make in the PicoMite firmware). A brand new 7th PicoH is current running an endurance test since yesterday, it has been working for just over 20 hours without problems, I am waiting for a shoe to drop, meaning for an I2C error to come up. More about this later.
Without blaming anybody, in particular not the authors Geoff and Peter, the PicoMite firmware is a fragile component in the PicoCalc environment. The PicoMite features crammed into the modest Pico hardware is amazing, that so much can be done with the limited resources available is near to a miracle. It must be appreciated what the PicoMite authors have achieved but it must be handled with care. It is very easy to “damage” this excellent product when adapting to the PicoCalc and there is a need to differentiate between the original PicoMite and the PicoCalc Picomite when experiencing problems. Basic 
Basic problems are unlikely caused by the PicoCalc adaption, however user interaction (keyboard, LCD, I2C) may very well be a fault in the PicoCalc adaption (partially my work). When experiencing problems it is not helpful broadcast your findings on all available channels, most success can be obtained by writing a clear description of the problem, with a clear description of all the components involved, the expected result and what came out. In my 50 year work history many problems were already solved by doing this because the effort needed would show me what went wrong. If you still do not succeed then publish your story on this forum where we can decide how to follow up to help you.
The next fragile component is electronics. There are a few interesting channels on youtube where the authors show how they find faults and how they fix the fault, if possible.
I can recommend watching Parts-People Dell Laptop Experts, NorthridgeFix,
for the german members SMARTRONIK and if you need some humor then StezStix Fix?. All these channels have something that I do not have, more about that later.
The Pico boards and the processor can also be declared fragile, when mishandled - this can be mechanical (bending, damaged components) and electrical (overvoltage, overcurrent, etc) - these components may show errors or stop working completely. Sometimes these error are not visible or difficult to identify.
At this moment I have a box (Pirate-brand Plastic Loot Box
by Pimoroni, highly recommended) with a number of faulty pico modules that I collected over the last few weeks.
4 x PicoH giving “Illegal address - resetting”
1 x PicoH that does no see the SD card
1 x PicoH that stopped working on I2C (*)
1 x Pico2H that also stopped working on I2c (*)
1 x Pi Debug Probe with failed UART
I suspect that all of these have some problem with one or more GPIO pins, I do not know why but I am keeping these boards for later experiments. It would be interesting to replace the RP2040/RP2350 chip on the faulty boards, the cost of these chips is about €|$ 1 each. But more difficult are some of the common tools used in the youtube channels above. One being the multimeter, quite affordable today, I am using a ZOYI ZT-703S OSCILLOSCOPE MULTIMETER available for about $100. The other tools they use are Microscope (~$2000) and hot air station (~$300), this investment I can not justify just because of wanting to replace a few RP2 processors 
.
(*) The I2C problem is weird, both boards stopped working about the same time while connected to a Waveshare RP2040 used as I2C sniffer. What is interesting that I was trying to find the cause for the keyboard to become unresponsive. I took a brand new unused PicoH that now has been running for 22+ hours without faults.
There are many fragile components in the PicoCalc with PicoMite universe, please keep this in mind. Software can be fixed or a workaround can be found, electronic parts can be replaced. The most difficult of this is the turnaround time, software takes time to analyze, repair and test, these tests can be very involved because regression testing is needed to ensure that the fix has not caused a new problem somewhere else. With electronic components availability and delivery times can be critical, delivery time for a replacement LCD display (weeks) will most likely exceed that of a (few) pico boards (days) while the cost of the components may not be significant.
IMHO the PicoCalc should be delivered with three labels attached: “Fragile”, “Handle with Care” and “Don’t Panic”.
Note 1: The term “PicoMite” is used to indicate all variants of MMBasic, PicoMite, WebMite, created and maintained by Geoff Graham and Peter Mather.
Note 2: The I2C endurance test has now been running for 24 hours without errors.
Note 3: The original title was “Fragile” but this forum software is using AI to make me look incompetent to express the truth in a minimum number of characters.
. I have some apps reading temperature, humidity, xyz magnetometers, xyz accelerometers etc at 400Khz and r/w to FRAM memory too, not a single glitch.
