seacordean

My reply crossed with yours, Mr. Bill. So, just to complete the point, I think the big difference between the solutions is step-wise debugging. Thanks for the info though. it is a partial solution.

Thanks bpwwer - I got more info from PyCharm - it looks like Polisy does not use ARM, so that is not an issue - however, as you intuited, they do not support and do not have plans to support FreeBSD. So, that is a dead-end. And thanks for the note on Slack - I was still experimenting, so was not sure I was going all-in to join the developer community, but now it seems like that is a good idea. As for PG3 vs PG2 development, I am not sure if that changes the SDLC approach, but you seem to be implying it might. I chose PG2 because deployment is inferred by Polyglot through the noderservers directory, so I was able to edit and reload (either through stop and start or through delete and add. depending on if I am changing code or profile) and I do not know how to deploy to a store in order to get PG3 to load the Node server (and that also seems a bit clumsy during dev/test cycles). So, PG3 may be my next challenge to conquer. And it seems the right place to continue these discussions would be in Slack. So, I guess this thread can be closed at this point.

Thanks for the info and quick response MrBill. Does this fully solve the issue? I want to use the editor interactively to debug, etc. (because otherwise, how can I more easily debug the parts of the code that need access to a running Poly instance?). So, it would effectively have to be executing the code on the remote live and using the debug and log information that is generated remote. PyCharm seems to do that by installing an IDE component on the RPi - if I could only get it to work through the SSH. Unless you are saying your combo does the same? Also, WinSCP seems to be Windows only and I spend more than half my time on a Mac, so a multi-platform solution would be much better.

I am posting here, as the Developer section seems to be geared to only cover specifically published nodes and not more general development questions, so I hope this is the right place. I have been developing automations on my ISY for a long time. I just started to delve into node servers, and more specifically in Polyglot on a Polisy. I have successfully developed and locally deployed my first Node server in PG2. My basic question (which may be an obvious one, but I have searched forums to no avail) is how i can easily use a good Python IDE (ideally from both a Mac and Windows) to develop on the Polisy. While I was successful in getting a Node server working I did a lot of it in the text editor that comes with the Polisy and that was grueling for all the obvious reasons. At some point, I had a moment of clarity and sanity, and got PyCharm. I did this development in a hybrid, as I did not see a way to access the live Polisy environment, so I used the IDE for all non-Polyglot related functions that surround the specific automations. I then discovered that PyCharm has an ability to work across SSH in the Pro version. So, I got that and tried to make it work. However, while the test connection worked via their SSH connection configurator, the actual move to the next step fails. So, I am not able to get the real connection to happen. I pinged PyCharm and they asked if Polisy is ARM, as they do not work with ARM. The specs on UD website just say it is an AMD chip. Given AMD is just now getting into ARM, my guess is Polisy does not use ARM, but good to confirm. If it is not ARM, this leaves me with no answer as to why I cannot get PyCharm to work. So, has anyone gotten PyCharm to work that can give me the magical answer as to why I cannot get past the configuration step? Alternatively, is someone using a different IDE that works remotely? I assume Polisy itself does not have a built-in IDE, as it is not a desktop edition of RPi, but if it does, that could work too. Thanks.

Thanks bpwwer - I think your explanation makes sense, but I may have been unclear on one point - while I do still have a couple of nodes in PG2, since they are not available in PG3, most of my nodes are PG3 for the ISY. I am not concerned about the PG2 nodes for now, as they are not critical for my programs. But 2 of the PG3 nodes are critical. So, I would like to install the PG3 nodes on IoP that were on ISY. When I restored the backup to IoP, it made it look like the nodes were installed on IoP, as they show up in the node server menu and in the device list. However, the devices have no connections, probably for the reason you gave - the internals are still addressed to the original ISY. This leaves me in a bind - I think you are saying I need to remove the nodes from the ISY and that will then free up the slots on both the ISY and the IoP. I cannot remove the nodes from the IoP, as they are unmanaged there, with no delete option. I would then need to reinstall the node servers on both ISY and IoP individually - which means I need to redo all configurations manually on both. I probably do not need to redo them on the original ISY for my situation, as once IoP is stable, I likely do not need it. That said, right now, it is my hot backup of a known working configuration of my IoP migration experiment fails, so would really prefer to leave it intact and just switch the PLM when needed. Another big downside of this approach is that all of the devices installed from the node servers that are in scenes or programs will disappear or go yellow respectively, as the device reference will be gone. This means I also need to go in and reestablish the devices in scenes and the program conditions, statuses and actions that touch these devices. The good news is that I have tried to abstract devices to one folder in my program stack and then use programs to access them - however, scenes are just broken and entries will disappear by this, so I need to document every scene that has a node device in it. This is all doable, but quite painful, manual and painstaking. And then I have a similar challenge with the Z-wave devices that also disappear from scenes and programs when migrating. Further, my automation functionality is disrupted until I can make all the changes, as nodes and z-wave are embedded in logic along with Insteon devices. I guess I need to leave a concentrated amount of time to fight through it, if this really the only answer, so I can make the transition in one fell swoop. If there is any other method that is easier, would love to know before I go through with this. And maybe this will be the last major transition until Wifi (and/or real Zigbee) gets turned on on the Polisy to talk to all the new devices supposedly coming out, so will get a break for a bit once I get through this..... Again, thanks for keeping this group informed and volunteering your insights. This community is one of the many things that make this the best smart automation around.

Hello team - fantastic threads - takes some doing and tech savvy, but I was mostly able to get my older Polisy to install PG3 (which seems to have brought IoP along for the ride) and was able to restore a backup of my isy to IoP. I got Zwave to show on the menu too (it seems to needs 2 reboots of isy, which I do not see in the threads). However, I am having issues with PG3 node servers - it looks a lot like abartello's issue - except I had already moved from PG2 to PG3 on my isy, before attempting a migration. The results of that is that all of my nodes show as unmanaged under the IoP isy in PG3 Dashboard. They all look normal in isy Dashboard. Since PG3 is supposed to be able to handle 2 isys (or more), I tried selecting IoP and then restoring backup of PG3 from my isy to IoP. Nothing changes from that. Even stranger is that all the nodes show on my IoP, but they cannot connect to anything. So, slots would start at 10 (I have 9 nodes filled on the isy). How can I get PG3 to recognize 2 different ISYs? Can I migrate PG3 to PG3? If not, how can I get the ghost nodes on my IoP to go away to even install nodes fresh to IoP, even if there is ? Do I have to delete them from isy? But then isn't it as if it cannot handle 2 isys in PG3? Any help greatly appreciated.

It seems like this resolved for most people a couple of weeks ago, however, I have the same symptom, even though likely a different cause. I needed to swap an ISY last night and restored everything, including local Polisy, onto the new ISY successfully. I also have no issues logging into ISY Portal and connecting via cloud admin. I can get into Polyglot Cloud site, and it shows the correct ISY ID in the ISY pulldown, but if I click Dashboard in Polyglot Cloud, it goes into a perpetual spinning "Loading", as folks described above. If I go to the Store and try to install something, it brings up the "Install NodeServer" box, but under that is says "undefined" and "Select Available Slot" has no choices. Further, every number of seconds or so, i get a bunch of Red error boxes on the right of the screen that say "ID specified but ISY not found" for the old ISY that no longer shows up in the ISY list in Polyglot Cloud and is also removed from ISY Portal. I am not sure if there is an issue in Polyglot Cloud with ghost entries for the old ISY. But even then, I also cannot access the new one at all. I even tried removing the new ISY from ISYPortal and re-adding it, but did not help. Any ideas for how to fix it? thx

I see what you are saying. If I look at the behavior of the program in the log from a prior successful run, I do see that Evan's Light Status returns as 0% after the command to set it to 0. This Status does not appear in the log in the run under discussion where the light did not turn off (the light was definitely not off as was suggested might explain the behavior, as confirmed by visually). The issue is that there was no return at all. I thought one of the benefits of Insteon was validation of commands having the desired effect. Since the command did not have the desired effect, shouldn't the system have done something? The only difference I could see between all the times the program ran successfully and this time was the one circuit breaker being open on the other lights in the program. Since that should not have had an effect per your earlier post, I guess this will be very difficult to debug other than to chalk it up to some signal interference that had it not shut off through the 2 tries the program made (given the consistency I have had with the system until now, this seems an unlikely scenario). I could try the event monitor (I looked at it now, but it only has the most recent items and does not go as far back as needed here). However, the switch seems to work now when triggered from the web, so I am not sure I can recreate. I think I will need to see if the next normal run of the program works, which will chalk it up as an anomaly. The debugger in me just hates leaving something without a properly attributed root cause and, even better, action to resolve for the future. Thanks

For a switch that was on the open circuit, the entry was: Evan's Light Off 0 Fri 2011/04/08 11:00:01 PM Program Log Zach's Room Fri 2011/04/08 11:00:11 PM System -2 which is strange since only the second entry where it errored should make sense since the switch had no power. For the switch that should have worked, the log entry was: Evan's Light Off 0 Fri 2011/04/08 11:00:01 PM Program Log which should mean it was shut off (I think) - and yet it did not shut off and there was the same execution and log an hour later, again with no result. This is a Switchlinc Dimmer with incandescent bulbs and I believe it is controlling 360 watts total. Thanks for the continued interest and support.

Thank you for the very quick response. Given that the device errors should not have caused any issues for the 3rd command, what else could cause an Insteon switch to fail to shut off from an off command in a program when the program indicates in a Notification that it ran? The Notification is the 4th (and last) command in the program. I have a number of programs that run through the night and commands from these seem to have run normally that night. The only command that did not have the expected result was this one line in this one program. If this is too difficult to narrow in on, I will have a chance to see it in action again at the end of the week to see if it was a one off issue, but prefer to understand root causes and try to be proactive, especially if there is a risk of recurrence. Thanks.

I have a program that runs at a certain time and shuts off lights for 3 rooms in the Then clause. It happens that last night, 2 of the switches were inaccessible due to a circuit breaker trip. The 3rd light is on a different circuit and was fine. The program has the commands to turn off the 2 switches that were out as the first 2 lines of code and the 3rd light as the 3rd line. The issue and question is that the 3rd light did not turn off at the program run time (the first 2 lights were already out due to the circuit so did not need to be turned off by the program). The log and email notification set on the program indicated the program ran at the correct time. Could this be because the first 2 commands to turn off somehow errored since the switches were inaccessible? Does a program behave differently when some of the commands are in error? Thanks.

Is there a way to set up programs to adjust for DST (i.e., end/To 1 hour later when in DST)? I thought of using date ranges to determine DST, but a. it means programming for every year (since date ranges are specific to a year) and b. DST is a different date every year based on how Sundays fall on the calendar. thx