After being stuck on V1.8 of the IDE for the longest time I finally upgraded to V2. Very impressed with all kinds of things in V2.3.3 - esp the reduced compile times for ESP8266 and ESP32 code which were tediously long on V1.8.
But, there is a problem. So.... to begin....
My config is:
Windows 10 22H2 Build 19045.5011 16GB RAM 3.2GHz Intel processor.
I can confirm that the exact same problem discussed a couple of years ago in the (closed) topic:
is still alive and well. In summary if your sketchbook or sketch involved a UNC pathname, then when you open an INO file all you get is a empty template project (with just setup and loop) BUT, if you close the IDE and restart it, then that same project is loaded, all present and correct.
I note the suggested workaround of mapping the UNC share concerned as a drive letter - which yes, I have done - but drive mapping like that is a less than ideal solution in many setups because a network problem or a problem on the server can slow or even crash client systems using mapped drives - even though they may not be using the mapped drive at the time.
The last comments in the topic above in January 2023 were hoping for a fast fix to the problem of using UNC paths for sketchbook and project files. Do we know when such a fix is now expected?
That's disappointing. Some sources say that up to 70% of Arduino IDE users are on Windows. I think it's fair to say that UNC usage is pretty fundamental to how a lot of technically inclined people use Windows systems.
I assume when you say UNC you mean a shared network file system. Depending in what you are using (most likely SMB) there are other ways to use them without UNC built into the (SMB) protocol
I was not aware there was any SMB involved in Windows to Windows UNC access? Surely Samba stuff is only used on Windows when it shares files systems with Linux based systems? Or did I miss something....?
SMB is implemented on all 3 of the major systems, I ported it to a 4th under contract. I don't remember enough about UNC to be sure, but it's worth checking out.
Today, I think everyone is moving on from SMB but the concepts don't change, just the acronyms.
So - to summarise - as far as anyone knows there no progress on this problem after three years? Despite it potentially affecting 70% of the user-base. Seems like it would merit a little higher priority than that. But, what do I know.
Hi @arduinoal900. The Eclipse Theia IDE framework is an open source project. You are welcome to submit a high quality pull request to fix the bug in the framework:
You can learn about how to do that from the contributing guide here:
Once the bug is fixed in Eclipse Theia, it will eventually propagate to Arduino IDE the next time the Arduino IDE developers bump the version of the application's Eclipse Theia dependency.
I closely monitor the forum and GitHub for reports of this type of problem and carefully maintain a list of them here:
The list would be far, far longer than that if 70% of the user base were affected. That is not to dismiss the significance of the bug on the fraction of users who are affected, but I think that hyperbole is counterproductive.
Thank you - I will put in such a request. I do realise that the fix is outwith Arduino's direct control - but nonetheless I imagine Arduino has some level of importance to the technology originator.
I didn't say that 70% of the user base were affected, I said that the problem has the potential to affect 70%. Neither you nor I have the means to know the real number of people affected. The previous thread on this matter suggested the compromise workaround of using a mapped drive and there is no way to tell how many people are using that compromise. Given how a lot of school and college IT systems are setup, I don't imagine it's a tiny number.
So, I logged a refresh request on this with Eclipse Theia (via github) and the speedy response I got seems to indicate that they are waiting for coop with Arduino on progressing a fix for this issue? They have apparently had a help-wanted (whatever that is) on this for a while.
I hope this is not going to be a case where us users feel like a table tennis ball?
I thought I clearly communicated that my suggestion was for you to submit a fix for the bug:
You were fully aware that a formal bug report had already been submitted. So why did you submit a duplicate bug report? That is counterproductive.
There should only be a single GitHub issue for each distinct bug. Duplicate reports can have a very harmful effect, either by wasting the time of the maintainers who struggle to keep the issue trackers of popular open source projects in a useful state, or by fragmenting the available information if the maintainer doesn't catch the duplicate.
Yes, sorry I created confusion here. I misunderstood that the invitation was for me to pull the code and submit a fix (which sadly is probably outside my capabilities) I thought you were inviting me to log a "Me Too" submission on this issue - which is why I called it a "refresh" request - and my submission was more detailed which I hope may prove helpful.
