sjenkins

Interesting, I migrated really early and had all my 20 or so plugins came along. Didn’t have the issue where the example got reactivated either. Wonder if you were one of the lucky ones who did it in a version with some issues; I certainly read about them & wondered why mine went so slick. It’s not as if I did anything beyond hitting the button. So you know I have been running this plugin in the new environment since then so all should be ok once over.

@tazman , Wondering if the devfile was the example one which is pushed on a new install. Did it have example in the name? Just to be clear when you say migration are you talking upgrading to the latest version or migration to eisy-ui from pg3?

Hi all, hope this plugin is working out for you. I am pushing v1.5.6 with a doorbell node as well as removing the package flag. Thanks to @shergenr for pushing me to adding this right about the time I got a G6 Entry (finally gave up on the G4 Doorbell ever coming back into stock). Test it out and make sure nothing else broke, but I did validate Ring, Person, Face. Didn't validate, Motion as I don't have that set-up on the doorbell. The Ring flag is held for 3s to make sure its not missed visually in the easy-ui , a program will catch it on switch, but you may not want to rely on status. I removed package from all the nodes as it is an alarm so currently out of scope as the api doesn't push it. Although, I am thinking that we could use the webhook feature of Alarms to get around their exclusion from the api. It would require some fiddling as we would need to set up a common nomenclature for webhooks to talk to this plugin. Let me know the interest out there. ## 1.5.6 - Doorbell camera nodedef (``unifi_camera_doorbell``): classify G4 doorbells and G6 Entry when the public API omits ``isDoorbell`` (``hasPackageCamera``, hardware ``type``, G6 + entry naming) - Doorbell ring (GV18): honor ring events when ``ring`` is not listed in ``smartDetectTypes``; pulse RING when the API sends only end-timestamp ring events; hold ring status 3 seconds before NORING so Status and programs are visible - Remove package status (GV5) and PACKAGE/NOPACKAGE commands from all camera nodedefs; ignore package smart-detect events — the public Integration API rarely delivers them as usable status - After updating: reinstall profile 1.5.6, run DISCOVER, and recreate doorbell/camera nodes if Package still appears on Status

@shergenr , Sorry, it took me some time to get to this on my list, but I pushed a new version which has a specific doorbell node. Test it out and make sure nothing else broke, but I did validate Ring, Person, Face. Didn't validate, Motion as I don't have that set-up on the doorbell. Fyi package is an alarm so currently out of scope as the api doesn't have them. Although I am thinking that we could use the webhook feature of Alarms to get around those kind of things if there is interest. Yes I do have the light issue with motion as well, so I have dialled that down on all my front facing cameras. As you say the other smart features pick up most of what you need.

@garybixler , Maybe I wasn’t clear in my comment above; that behaviour is on purpose. Zero and ON do not make sense at the same time. So as someone hits dim again and again there are two choices when you approach zero. Either turn the device off and enter zero or “bounce” off zero to a minimum number, in this case 5 was chosen. You can be ON and enter or Brighten/Dim from 1-100. This minimum ON was meant for using this virtual as a generic dimmer. If you want zero you switch the device OFF.

@garybixler , Hi Gary, you are right special characters are not allowed as node names in any UD node. The 5 is being inserted when the node is on and you try to set the dimmer to zero. You can set 1,2,3,4. If you want zero then turn the node off. This is a bit of a carry over of using this as a dimmer node. So you cannot be on and zero at the same time. Hope this helps.

I don’t think there is any reason we can’t get there. I have had a fully booked week so since receiving my G6 entry I haven’t tackled the issue yet. My thought is we will have a doorbell device with the appropriate nodes we can trigger or status off. Everything the api offers will be passed through to the node.

I have the same situation with Eisy on my IoT Vlan. Just took a look at my zone firewall settings. Without seeing yours , I really just have that vlan able to get to external and gateway. I have a UDM pro max gateway. Nice, thanks for the update. Give yourself a solve for that one !

Yes, agree we need that one. I believe its on their list.

I hear your frustration, I have been on this journey for the last few months. It truly is in the nineties here, we can debate where. Functionally the 98% is accurate, the polish makes up the other percentage. I have deleted admin from my launch bar and haven’t used it for weeks. I use easy-ui daily. As said above you can sort by any column, the number box does prefer the pulldown but if you click in the box and type it does work. Network stuff does work but needs some polishing. Have not experienced the legal character thing. There are lots of design decisions in this thing which are like new clothes replacing favourite old ones. You did the right thing by trying out what you use and listing them here. I would personally have dropped the dramatic title but understand the frustration. I think as Michel and Javi have on the roadmap we are going to take to the end of year for near replacement & then first quarter to smooth it all out. The nice thing is they are also giving us new features in parallel due to the new platform. Glass half full here.

Agree try the refresh, then the hard refresh, then wipe browser memory for that page. Sometimes a logout and login, close and open browser, finally an eisy restart. Have had final success at each step in certain situations of this in the last few months of dynamic profiles learning as I implement it with my users.

Having a thick 'Tin foil hat' when it comes to ai , even though I use it at work all the time I decided to set up llama on a pretty capable macmini. It is an M4pro and has 64G of memory, its day job is running plex. It worked just fine, but even with this setup its painfully slow on the first interaction and only really slow on further interaction. Very sensitive to which model is chosen as far as speed goes. Saying all that, it works! Got it to check out statuses and control some stuff. It even made a program when I asked it to flash the lights in a room (that creeped me a bit). This has lots of potential. Some learnings: the ```json wrap is required (yes its in the example but I didn't think it would need that) provider is llamacpp for the self serve llama model is the full id , like "ggml-org/gpt-oss-20b-GGUF:MXFP4" url is your local http://10.0.10.4:9931 in my case not much feedback from the agent when its talking to the llm, hard to tell the difference between locked up and thinking. I kept a tail on the logs to understand if errors were coming. the ct number needs to be north of 100k or it will think for a long time and fail hope this helps!

So to fix the permission error here is what I needed to do on the eisy: [admin@eisy ~]$ mkdir -p /usr/home/admin/eisy-ai/generated_plugins [admin@eisy ~]$ chmod -R 777 /usr/home/admin/eisy-ai This seems to fix the log jam for now Admin Edited for typo, or change in dir name

Got production caught up to dynamic JSON profiles as well as fixing a Ratgo depreciated label issue. Let me know of any feedback. I have seen in other plugins (Tahoma) that with the jump to the JSON profile migrated plugins from the old PG3 to the new Eisy-ui needed an Eisy reboot to remove a profile error. When installed straight from Eisy-ui it seemed to be fine. YMMV. 3.2.3 \ DONE VirtualGarage Ratgdo: ESPHome REST URLs use entity display names (e.g. /light/Light) instead of deprecated object_id paths; SSE state events accept new id format ahead of ESPHome 2026.7 3.2.2 \ DONE dynamic JSON profiles (Phase 1): updateJsonProfile instead of static profile/ zip for eisy-ui sync ; requires IoX 6.0.6+, PG3x 3.4.5+, udi_interface 3.4.5+; install.sh renames profile/ to profile.static on EISY; profile/ XML remains source for data/base_profile.json; bump profile_version in server.json when JSON base changes 3.2.1 \ DONE fix config overwrite: stop re-injecting example devfile and sample devices on every parameter loal DONE remove example customParams from server.json so Polyglot updates do not reset user configuration DONE clear "Waiting on valid configuration" notice when discovery fails

Dynamic profiles are new and the way UD is suggesting as the way forward. They help with making sure Eisy-ui is always up to date but they add complexity to the programs. So this is my second run at them. Working well at home but need some feedback. If you get an error, try an Eisy reboot, if still an issue then send me logs please. ## 0.0.29 - Fix eisy-ui dynamic profile on migrate: add empty links.ctl / links.rsp on every nodedef and linkdefs: [] in data/base_profile.json so eisy-ui updateprofile validation passes (fixes “Profiles loaded but nodedefs missing” after PG3→eisy-ui migration); bump profile_version to 2

Thanks @Javi for the explanation, understand you need to prioritise development just wanted it on the observed list To @Techman ’s point, I am on osx with Safari, I’ll try with a couple other browsers and see if there are differences and bring back. Maybe it’s a browser quirk.

I am on Staging 1.3.11_1 but this has been there for a few versions. JUST in the Program section (not Nodes, Integers, State, etc) about every 3-5 sections the screen has a blue line flash under the program heading. If you have activated search, it will clear the search. Both visually irritating, but mades search a race for time or unusable. hope this helps.

@Javi , Agree status per line could be confusing all the time, it would be useful in a debug/test mode. You are writing a program and you save. Turn on the test/debug, it gives you colour by line. Saves running back to nodes and variables as you test a recent program. This would just be the program you have selected. Turns off as you move away. I could have used this last night. I was using UD Mobile & two web tabs to check out all the sources of data which affected the program. Just a thought, this would match how you use debuggers when writing code (which is another approach to this, allow the programmer to put watches on vars, nodes , progs, on a screen as you debug).

@MReeves hi Mark, Do you have a gen2 and gen3 gateway? I am assuming that is how the hybrid on the app is working. If you do run two instances of the plugin, one pointing to the gen2 gateway and the other to the gen3 gateway. The app as it is decides which one to be based on the gateway it is talking to. If there is something about the hybrid I am missing let me know.

Agree , some kind of visual management would ease the stress of every user.

so this is back office but what it does for eisy-ui is make updates to the node list happen immediately, instead of getting stuck when you add/remove nodes or install a new version of the plugin in the same slot. It works pretty slick and is the modern way to handle profiles as we move away from the admin to eisy-ui. My only worry is you can see below you have to be decently up to date on your controller, which is not the case for everyone out there. Interested in feedback here. You can install this either from the old non-production store or it shows up as a beta trial (1-year trial) if you install from the new eisy-ui plugins(beta). 3.2.2 \ DONE dynamic JSON profiles (Phase 1): updateJsonProfile instead of static profile/ zip for eisy-ui sync; requires IoX 6.0.6+, PG3x 3.4.5+, udi_interface 3.4.5+; install.sh renames profile/ to profile.static/ on EISY; profile/ XML remains source for data/base_profile.json; bump profile_version in server.json when JSON base changes

I've been running this for a few days and seems ok. This is back office except it should mean that eisy-ui picks up on profile changes immediately. My plan is to run this across all my plugins, its the current modern way to do profiles. The OGP bit is a bug fix where some users were seeing extra nodes named OGP. Let me know how it goes. ## 0.0.28 - Dynamic JSON profiles: push profile via updateJsonProfile() instead of static profile/ zip — improves eisy-ui sync; requires IoX 6.0.6+, PG3x 3.4.5+, udi_interface 3.4.5+; install.sh renames profile/ to profile.static/ on the EISY so PG3 does not race static upload; profile/ XML remains in the repo as source for data/base_profile.json; bump profile_version in server.json when the JSON base changes - Per-device nodedef on Discover: new shade nodes pick shadertsid, shadenotiltid, shadeonlyprimid, or shadeid from TaHoma capabilities — existing nodes are unchanged until the user runs Discover and a new node is created (removed nodes rediscovered, or manual delete + Discover) - OGP skip hardened: skip all ogp:// devices and ProtocolGateway uiClass (not only ogp:Bridge controllableName) — fixes duplicate OGP nodes when the gateway returns partial records or multiple OGP entries; run Discover to remove stale OGP nodes

You may want to open a ticket with UD

Minor cleanup and a notice upon connection success. ## 1.5.5 - Temporary Polyglot success notice after connect and discovery (clears after 30 s), matching Ta$

sorry for the delay, painting day today. @Techman , I saw your DM as well , what was the other server you tried? and, of course, have you tried a reboot? I just tested and am able to install from both stores, and am able to install on top of beta -> production -> beta looks like a local issue to you.