Libraries went bad Saturday October 27.

For what it's worth... Last Saturday October 27 mid afternoon U.S. central time, APPARENTLY someone updated some of the libraries and screwed them up. For several months, I have been working on a very complex project initially using an ESP32, (which did not work) and now using ESP8266 along with an Arduino Metro Mini to control a whole bunch of PWM fans inside an enclosed server rack. Up until last Saturday everything was working and I was down to the final "tweaking" adjustments.

I made a very minor change to the .ino on Saturday and recompiled. Suddenly, I am getting all sorts of weird errors that I have never seen before, using both the Arduino IDE and in Visual Micro. As best as I can determine, someone probably changed something in one of the libraries that is causing problems.

I am a retired professional Programmer/Analyst. I have seen this sort of thing happen before.

Luckily I had previously downloaded the Arduino Master library from Github in ZIP format, that was still in my downloads folder.

So I was able to work around the issue by failing back to the older libraries in that .zip file.

I am sure other folks without a backup or copies of the "non-messed up" libraries" y are probably having problems.

Just thought I would report this "for what it's worth".

You should post the wierd errors

rlloper:
I made a very minor change to the .ino on Saturday and recompiled.

Did you download new versions of the libraries before re-compiling?

This is one of the reasons why I don't favour centralized libraries. For anything critical I put copies of the libraries with my program code.

...R

What libs specifically ?
Not sure even Arduino has enough people to check every single lib based on a vague comment !

Another point is that I am pretty certain they dont DEPLOY between Friday morning and Sunday night (European time probably local Italy) to avoid issues just like this.

I have been having a similar problem, and not sure that it is related.

My Wire library has suddenly stopped cooperating with i2c devices attached to arduino UNOs and MEGAs (2560). This problem started occuring about 2 weeks ago, and as of right now appears to be limited to my lab desktop and my home laptop (the Adafruit example code for each i2c device still loads and runs on a labmate's desktop, though I am the only active arduino user in the lab right now, so he may be running an older version or something).

At first I thought it was a breakout specific hardware failure, until I tried several different breakouts on several different MEGAs and UNOs, and validated with an i2c scanner sketch that sketches compiled and loaded from my machines do not see i2c devices. Yet, the same code and hardware loaded from a different machine work just fine.

I'm posting here mostly to just follow this issue in several places and hopefully find an answer. I've posted about it in another location: Wire.begin() not working for i2c devices - Networking, Protocols, and Devices - Arduino Forum I haven't yet determined whether something has updated elsewhere and I'm suffering from the fallout, or if the damage is contained to my desktop/laptop only, and is an issue specific to me.

@OP:

  • is your problem possibly tied to failure of the i2c bus, either via software or hardware means?
  • Have you tried to isolate which library(s) are giving you trouble, or if its just all of them?
  • Do you find out about the errors via the compiler output in the IDE, or does the hardware just not cooperate after compiling and uploading your sketch?

I'm not convinced that agmotes problem is the same as rlloper. Let's continue discussion of agmotes problem on the dedicated thread, where there has already been much discussion.