Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Official 5.5 Beta Release Thread
  • Recent changes introduced an issue with the UPnP finder mechanism, which breaks UPnP based drivers. The UPnP stuff is still OK, but the driver will think the device isn't there because it tries to find a device with the previously remembered device id.
  • Recent changes made some drivers that use a socket fail on load because they were calling the version of the ctor that tries to open it but the IPV4/6 type isn't known at that point. They should use the default ctor since that info is set later when the socket is opened. The immediate one reported was the GC-100 driver.
  • Recent changes introduced some problems into invoking external programs, which showed up in a failure of the IV to invoke the external helpers that host web browser instances and web cam engines.
Dean Roddey
Explorans limites defectum
  • Take a whack by eye at adding support to the Z-Wave driver for the Zooz ZEN15 light switch and StelPro STZW402+ thermostat.
  • Add support for some more channels of the Somfy blinds controller
  • Update the CML JSON parser's FindXXXValue() methods to return the provided default if the target node is a null node, in addition to when it's just not found which is already the case.
  • When we updated the data server to use the new chunked file format, the little tool that creates the image packs was not correctly updated. It was not storing the images as chunked files but just the raw image content. So they would fail when you tried to import them.
  • Crank up the poll time on the ISY driver since it looks like it may be getting overwhelmed with async events. So poll faster and wait a shorter time on each poll, to both get asyncs quicker and be more responsive to outgoing commands.
Dean Roddey
Explorans limites defectum
  • 910 broke local serial ports. The local port factory must be the first thing in the list and some changes somehow lost getting that done, so only GC-100, JaP ports were available.
  • Do some more tweaking on the ISY driver to try to keep it from being overwhelmed by aysnc events. CML has to poll so we can only do so much. Beyond that it'll have to be redone in C++.
  • Enable the MSII motion sensor in the ISY driver
Dean Roddey
Explorans limites defectum
  • The ZEN15 Z-Wave widget I just added isn't a light, it's a power outlet. So I set it up to be the wrong type.
  • The interface editor's widget palette isn't ignoring change events from the widget list when the template itself changes. Since that event doesn't have a valid widget index, it causes an index error to be logged when you change the template.
  • Our floating point rounding methods don't handle negative numbers as advertised. For us, rounding a negative number down means it becomes less negative (-2.5 becomes -2.0) and up means more negative (it becomes -3.0.) This is somewhat different from the standard C style floor/ceil/round methods. Also closest works the same way. -2.6 to the closest goes to -3.0, and -2.4 to closests to goes to -2.0.
  • In preparation for a future mixed Linux/Windows world, stop downloading C++ drivers from the MS. Just install them on the machines. That downloading scheme was from the original CQC concept where C++ drivers were going to be all of them, so installing them was a lot of overhead on each machine. But, ultimately they are almost all CML, and only a small number are C++, so we can get rid of a lot of complexity by just installing the C++ ones on each machine. Then Linux machines will load theirs and Windows machines will load theirs, all naturally without any craziness.
  • Update the install image builder to adjust for the above. All of the drivers are now just lumped in with the CQC executables and libraries.
  • Update the installer to adjust for the above. No separate drivers directory to copy, and remove any old ones still around.
  • For Linux support we need to separate the building of the paths to loadable modules (which are platform specific) from the building of paths for things like loadable text files and resource files, which we create and so the names are not platform specific. Currently this fails on Linux because it's using the Linux specific base path to build the path to the loadable text. So just totally rework how the building of module related names is implemented.
  • Dump some really old and no longer used C++ drivers.
  • Add support to the ISY driver for IOLinc relays, at least the sensor part of it. The relay is not so useful because it's not directly controllable. They will show up as BinSensor_xxx fields.
  • Undo all the work done a couple drops back to rework the way collections deal with thread safety. It turned out to be stupid.
Dean Roddey
Explorans limites defectum
  • The list browser widget will cause an index error if it has less than a page full (and hence free space at the bottom) and you click past the last one when there is a defined click action.
  • The Radio RA2 driver is sending out motion events but not storing the motion state to the field.
  • The Radio RA2 driver complains if the low set point is >= to the high. Let them be equal at least, though that may cause some confusing in user logic if the difference is zero.
  • Now that we are no longer downloading C++ drivers (client or server) we don't get the info about whether they have associated loadable resource files. Since they are local now, just look for the files before we load and set the flags that way.
  • When importing quite small images into CQC you can get an error because the small size caused an inversion of initial and max sizes of a memory buffer used in the process.
  • Interface containers (overlays in this case) are supposed to inform all their children when they are hidden or shown, to allow them to react if needed. Mostly they don't, but web brower and web cam widgets are special cases because they are separate windows that need to be hidden/shown.
  • Add what I think is correct support to the Radio RA2 driver for the contact inputs. We'll see if it works. We assume if it as a component id of 32 or 33 then it's an input, else an output.
  • Correct something in the RA2 driver now before it's too late. Contact closures are semantically read only. The read/write contacts on the Visor should be marked as relays, since those are semantically read/write. So the field names will change for the previously added r/w contact closures.
Dean Roddey
Explorans limites defectum
  • Somewhere the deselection of widgets (via shift-click) got broken because the index of the widget to remove was being used instead of the unique id.
  • Update the RA2 driver to log the outgoing msg that was sent if we get a rejection from the Lutron controller.
  • The client side driver for IR/trigger drivers got missed when we updated everything to install C++ drivers from local installed files instead of downloading them. It's a sort of special case, and it's installer image stuff didn't get updated so it wasn't getting into the right place in the installer and hence not copied to the user systems.
  • The RA2 driver's configuration of optional standard triggers on fields isn't sticking because the client and server side drivers disagree slightly on how the yes/no values are represented.
Dean Roddey
Explorans limites defectum
Version 5.4.917 is now posted. We don't speak of 915/916 anymore. You will have to go back to 914 or before and move forward again using 917. Unless something is really wrong, this will likely become the new 'must pass through' version to move forward to the eventual open source version.

So the problem for those who weren't following along is that I had my version logic wrong. I should have been checking for anything at 914 or before going to anything beyond 914. But I had that badly wrong so, depending on versions, it might not upgrade the data or it might try to upgrade already upgraded data if you ran the installer against. This made a mess of things.

This one should get it right.
Dean Roddey
Explorans limites defectum
Here is a 5.4.919. It's primarily for Bryan to see if it fixes an issue with the WebRIVA camera extension. But if someone would like to just upgrade who is on 917, just to make sure it goes well. It should, but just to verify. I'd like to get to the point pretty quickly where we can declare the current stuff to be the 6.0 open source version and update the web site and such.
Dean Roddey
Explorans limites defectum

I upgraded to 917 from 914 and then to 919. 

Unfortunately it didn't work like 914. Maybe it shouldn't. So let me run it again in the morning and capture the console log. I'm down about 6 tequilas right now. So let me test it in the morning. My first impression is that only one create webcam widget was called but the SetURL was not called but let me work on it in the morning. 

I cant get from914 to 917 message pops up after i provide name and password, I get message saying (Unhandled exception in GUI Thread) also get below command is not implemented by object

Possibly Related Threads…
Thread Author Replies Views Last Post
  6.x Beta Release Discussions Thread gReatAutomation 30 5,686 12-21-2022, 12:53 PM
Last Post: pilotguy7ca
  Official 6.x Beta Release Thread Dean Roddey 2 1,450 04-16-2021, 05:55 AM
Last Post: Dean Roddey
  5.5 Beta Discussions Thread Dean Roddey 291 85,262 04-05-2021, 04:10 PM
Last Post: Dean Roddey
  6.X Discussions Thread gReatAutomation 1 1,234 04-01-2021, 03:23 PM
Last Post: Dean Roddey
  Official 5.4 Beta Discussion Thread Dean Roddey 441 183,704 06-15-2019, 02:33 AM
Last Post: Bugman
  Official 5.4 Beta Release Thread Dean Roddey 55 33,883 06-07-2019, 07:02 PM
Last Post: Dean Roddey
  Official 5.3 Beta Discussion Thread Dean Roddey 815 402,068 07-05-2018, 12:44 PM
Last Post: Dean Roddey
  Official 5.3 Release Thread Dean Roddey 27 21,113 07-05-2018, 12:44 PM
Last Post: Dean Roddey
  Official 5.2 Beta Discussion Thread Dean Roddey 244 159,730 10-14-2017, 07:57 PM
Last Post: Dean Roddey
  Official 5.2 Beta Release Thread Dean Roddey 13 14,953 10-09-2017, 06:49 PM
Last Post: Dean Roddey

Forum Jump:

Users browsing this thread: 1 Guest(s)