ppitkin

Still don't see how going via some cloud based gateway can ever be faster than the direct connection. I control my system via the admin console and mobilinc and both respond almost instantaneously. I even run 12 cameras over the mobilinc at the same time with no issues. I really haven't seen anyone come up with a real benefit of the portal where it adds real value. Credential management is a non issue as usernames and passwords are easy to change if necessary. Dynamic DNS solves any issues with having to use an IP address. Where I need secure connections I just use VPN and so I don't need to bother running everything with HTTPS. What I don't understand is why Alexa needs to run via the portal. Can someone explain?

Hi, I'm still very confused as to what the portal is meant to do and why I should possibly need it. I'm quite happy with doing my external access through port forwarding on my router and using VPN connections. Why will the Alexa integration only work with the portal?

ISY is configured to use gmail. I don't think the ISY ever gets to send the mail to gmail as it has a problem with more than 32767 characters. If I use another application I can send mails with more than 32767 characters via gmail smtp

So I have done some more testing.... I have gone back to basics: Plain text email, no parameters, no html or xml. I have determined the following: 1) a new line eats up 4 characters of space (not 1 or 2 as would have been expected for LF/CR or just LF) 2) The limit appears to be the 32767 integer limit 3) The limit includes the email subject text as well as the body text. 4) I would expect (but didn't test) that the "to" and "From" email adresses are also included in that limit. So my guess would be that the ISY is building a charater string for the email information and the lenght of the character string is limited to 16 bits. Michel, could you check if this is what is happening and look for a possible workaround if ift is. For my HTLM email with paramter substitution I quess there must be some sort of processing on the parameter replacement that causes it to hit the limit even though the number of charaters in the mail is actually much lower - but that is just a guess

Hi Michel, SMTP delay is already set to the mayimum of 10 seconds (10,000 milliseconds) - It doesn't appear possible to set it higher than that in the admin console. I was playing around yesterday trying to find what the character limit was. It seems to change depending on the information in the file. By this I mean if the characters are pure text (and no substitution is performed by the ISY, or if they are a tag (e.g ${sys.node.xx xx xx 1.st} ) where substitution will be performed. Naturally its difficult to determine the exact size of the message that the ISY will be trying to send out after substitution but I have looked at the HTML received in the mail and it appears to be around the 15221 characters where the limit is reached. In addition I have looked to see if this is an issue with the gmail SMTP interface and I am able to send out mails to gmail of a much larger size than this without problems. Not sure is any of that helps in any way. peter.

My problem is the sending of emails when the size of the mail reaches a certain limit - I don't see that this would be influenced by the DNS or static IP address. Messages go out fine if they are under ca 19950 characters.

Hi Michel, I'm already using a gmail account for SMTP'ing my messages - I experienced a lot of delays in the early days when using the UDI default server so I switched. Below are the entries from the error log for the last few days. Unfortunately you need to understand the codes to make any sense of it so it doesn't help me much. What I have just seen is that my daily updates haven't been sent since Sunday and there are "Sun 2015/09/27 23:59:03 System -50001 -19 " entries at the time that it should be sent. The reason is after testing I forgot to reduce the number of characters back down below the threshold mentioned in my original post (probably forgot to press the SAVE button before exiting ). I have just edited the file and sent a test mail and it worked perfectly. I don't believe it's an HTML error or any sort as I can remove or add charaters anywhere in the file and it will either work or not work depending on the total number of characters. Fri 2015/09/25 03:31:12 System -140007 CLI-HAMW Fri 2015/09/25 03:31:12 System -140007 CLI-HAMW Fri 2015/09/25 03:31:12 System -140007 CLI-HAMW Fri 2015/09/25 03:36:12 System -140007 CLI-HAMW Fri 2015/09/25 03:36:12 System -140007 CLI-HAMW Fri 2015/09/25 03:36:12 System -140007 CLI-HAMW Fri 2015/09/25 08:39:42 System -170001 [uDSockets] HTTP:30 error:6 Fri 2015/09/25 08:39:52 System -170001 [uDSockets] HTTP:32 error:6 Fri 2015/09/25 16:56:19 System -140005 CLI-HAMW Fri 2015/09/25 16:56:24 System -140005 CLI-HAMW Fri 2015/09/25 16:56:29 System -140005 CLI-HAMW Fri 2015/09/25 17:06:19 System -140005 CLI-HAMW Fri 2015/09/25 17:06:24 System -140005 CLI-HAMW Fri 2015/09/25 17:06:29 System -140005 CLI-HAMW Fri 2015/09/25 17:21:20 System -140005 CLI-HAMW Sat 2015/09/26 03:00:37 System -170001 [uDSockets] HTTP:30 error:6 Sat 2015/09/26 03:00:47 System -170001 [uDSockets] HTTP:32 error:6 Sun 2015/09/27 06:41:57 System -50001 -19 Sun 2015/09/27 06:45:19 System -50001 -19 Sun 2015/09/27 06:47:41 System -50001 -19 Sun 2015/09/27 07:10:25 System -50001 -19 Sun 2015/09/27 07:22:15 System -50001 -19 Sun 2015/09/27 07:30:30 System -50001 -19 Sun 2015/09/27 07:41:14 System -5012 32 Sun 2015/09/27 12:39:47 System -170001 [uDSockets] HTTP:31 error:6 Sun 2015/09/27 17:34:14 System -170001 [uDSockets] RSub:31 error:6 Sun 2015/09/27 23:59:03 System -50001 -19 Mon 2015/09/28 03:00:43 System -10 n/a Mon 2015/09/28 03:31:13 System -140007 CLI-HAMW Mon 2015/09/28 03:31:13 System -140007 CLI-HAMW Mon 2015/09/28 03:31:13 System -140007 CLI-HAMW Mon 2015/09/28 03:36:13 System -140007 CLI-HAMW Mon 2015/09/28 03:36:13 System -140007 CLI-HAMW Mon 2015/09/28 03:36:13 System -140007 CLI-HAMW Mon 2015/09/28 04:40:00 System -10 n/a Mon 2015/09/28 19:31:25 System -140005 CLI-HAMW Mon 2015/09/28 20:11:16 System -140008 CLI-HAMW:500 Mon 2015/09/28 23:59:02 System -50001 -19 Tue 2015/09/29 00:19:15 System -5012 35 Tue 2015/09/29 23:59:04 System -50001 -19 Wed 2015/09/30 03:00:59 System -170001 [uDSockets] HTTP:32 error:6 Wed 2015/09/30 03:01:09 System -170001 [uDSockets] HTTP:31 error:6 Wed 2015/09/30 03:01:14 System -170001 [uDSockets] HTTP:31 error:6 Wed 2015/09/30 03:18:24 System -170001 [uDSockets] RSub:29 error:6 Wed 2015/09/30 03:18:29 System -170001 [uDSockets] RSub:29 error:6 Wed 2015/09/30 03:18:34 System -170001 [uDSockets] RSub:29 error:6 Wed 2015/09/30 03:18:39 System -5012 38 Wed 2015/09/30 23:59:02 System -50001 -19 Thu 2015/10/01 14:26:19 System -140000 api.aerisapi.com Thu 2015/10/01 18:09:59 System -5012 39 Thu 2015/10/01 18:10:00 System -5006 uuid:39 Thu 2015/10/01 18:10:00 System -5012 40 Thu 2015/10/01 20:16:21 System -140005 CLI-HAMW Thu 2015/10/01 20:16:27 System -140005 CLI-HAMW Thu 2015/10/01 20:16:32 System -140005 CLI-HAMW Thu 2015/10/01 20:21:21 System -140005 CLI-HAMW Thu 2015/10/01 20:21:26 System -140005 CLI-HAMW Thu 2015/10/01 20:21:32 System -140005 CLI-HAMW Thu 2015/10/01 23:59:02 System -50001 -19 Fri 2015/10/02 03:31:15 System -140007 CLI-HAMW Fri 2015/10/02 03:31:15 System -140007 CLI-HAMW Fri 2015/10/02 03:31:15 System -140007 CLI-HAMW Fri 2015/10/02 03:36:15 System -140007 CLI-HAMW Fri 2015/10/02 03:36:15 System -140007 CLI-HAMW Fri 2015/10/02 03:36:15 System -140007 CLI-HAMW

Michel, so sorry for the delay in replying. I've had a look at the field that you mentioned and it is NOT checked. Peter

I've come accross a strange problem today whilst trying to update my HTLM formatted status email that I have sent to me on a regular basis. The issue (as far as I can tell) appears to be related to the number of characters in the email. My status mail sends me all sorts of information relating to my chalets current status. (see attached image(s) for an example of what it looks like). As it has lots of references to variables and device statuses there are a large number of characters in the file. What I have noticed is that if the file contains 19934 charaters (or less) it is sent out the mail without any problems. If it has ca. 19979 (or more) characters it fails to be sent. There appears to be no error message being placed in the error log. Does anyone else have this problem? Peter.

Io_guy, Your middleman suggestion sounds like a good compromise for my setup. I have two GEMs running wired Ethernet to a single dash box. The dash box then sends out the data to SEG. For the last few weeks SEG hasn't been working correctly and Sam doesn't appear to be monitoring the Forum which is a real shame as I really like the SEG user interface for monitoring. The DashBox is improving but still needs some work. For my setup I don't want to run either the serial or the wifi connection between the GEMs and the Dashbox so the node server either has to accept the output from the dash box (and allow forwarding onto SEG) or accept the output from multiple GEMs and forward it to the DashBox. It's really great that your investing time on this. I was waiting a while until UDI has their framework finished and then I plan to start building various node servers for all my devices e.g Russound, Sonos, Onkyo receivers, media players etc. This will then start putting the ISY at the heart of my automation system. Once the Echo is available outside of the U.S. and I can integrated it daily with the ISY the options are going to be endless. Now all I need is to come up with a High Availability solution for the ISY and the PLM!

Io_guy, If you are going to develope something please make it also compatible with the Dashbox. I use the network output from the Gems to connect to the dash box so it wouldn't be available to a node server. Make sure it can also handle multiple Gems. Thx.

io_guy, If you get a node server created for a GEM put me down for 2. It would be fantastic to get it showing up as a proper node in ISY.

I have audible notification (voice) being output via my sonos speakers. ISY sends commands to an API (hosted on a Raspi) which then controls the speakers. It's possible to even define which of the speakers it uses - anything from all the speakers to just a single one. (I have speakers in every room). With the ability to now use parameters in the network calls it is fairly easy to combine this with NF technologies such that the system knows which room you are currently in and selects the speaker in that room as the output device for any voice alerts. It does mean that you always have to have your phone on your person (or some other device that the NF system can detect e.g there are some systems that use a key fob) so it's not ideal but the best that can be easily achieved today. Peter.

OK, happy to accept that. Unfortunately I can't test as my Insteon kit is located in my Holiday Chalet and not where I live and so its impossible for me to physically push a switch. I have to wait until a guest is staying there and then see what happens - not ideal! I now have to change the If conditions on roughly 70 auto OFF programs - guess that's my evening gone again. BTW I do see (in the admin console) the program being triggered and it having a TRUE state which should have meant it ran through the THEN path. I also see that the time it ran and the time that it ended corresponds to the first wait period. What I can't tell is what physical action was taken at the switch other than through the other program I have for each switch that records when the Switch was turned ON (again using Control rather than Status) and this seems to be showing that it ran at nearly the same time that the AUTO OFF program was started. This may be pure coincidence or it may be showing that the logic is working. What I also see is that when the Auto OFF program has a status of FALSE its end time is almost the same as the start time - meaning that the ELSE path was followed. There just happens to be some guests staying in the chalet since yesterday and that's why I managed to determine something was wrong in that I was not getting the notifications being sent. I hate having to debug in a 'live' environment.

Thanks for all the helpful feedback. Control was used as I only wanted it to respond to a physical action at the switch. This then allows me to turn it on and off in the program without it causing the if condition to re-trigger. Oberkc, I think that "Control is not switched off" will cause the if statement to be processed for any physical action on the switch. It should be TRUE for anything apart from turning the switch off (normally, not Fast OFF). Turning the switch OFF will therefore stop the Wait statement and exit via the ELSE i.e. do nothing. Any other physical action should cause it to start the wait loop from the beginning again. At least that is my understanding having read a large number of the posts and examples in the Forum. Apostolakisl, I think that the "Control is not switched off" is the same as what you have done (but without the status part). It also however covers Fade Down and Fast OFF. As mentioned above I didn't use status as it will cause the IF to be recalculated if you change the status of the device within the program (e.g. where I flash the lights on and off twice). I think your program only works correctly as you are setting the device to the OFF state as the last statement in the program. I may however be totally wrong and have not understood correctly how the ISY logic handles "not Switched Off" - if so I'll take back all my comments above I'll try changing to a "repeat 1" and see what that does. I noticed in the wiki that there was mention of a bug fix in one of the releases for the "Repeat 0" command - but it didn't say what the bug was or what fix was put in place. The wiki also doesn't appear to detail what Repeat 0 does - to me it implies don't repeat the following code i.e. run it only once.