I’ve put dataMine 0.967 onto the app store for MCV approval. This is a quick release version, primarily to fix a bug that some users had seen with 0.966, however it also has a few changes to the GUI…
Version 0.967
[ul][li]Reworking code to support remote server[/li]
[li]Increased number of channels supported in a single graph from 4 to 8[/li]
[li]Removed margin in GUI (wasted space in mobile)[/li]
[li]Added online status indicator[/li]
[li]Added graph information window[/li]
[li]Added code to reinitialise alpha-lookup table[/li]
[li]Fixed bug with alpha-lookup table[/li]
[li]Added warning if Luup restarts[/li][/ul]
Most of the changes are reasonably small - the big point to note is that I’ve increased the number of possible channels from 4 to 8.
I guess this will not be available for a day or two, however as usual, any comments appreciated…
Hi Chris,
I had initially installed datamine 0.965 and noticed last week that on my list of Apps that it now listed version 0.966 so had presumably auto updated. Today on my list of Apps 0.967 is listed but I still can’t plot more than 4 variables so I am guessing the auto update is broken and in fact I am still on 0.965
I’m still searching through the forum to confirm this suspicion (I don’t have any asterisks after any of the plug-in’s) but wanted to check if you had a recommended way of manually updating the plug-in without losing the ability to display previously logged data.
Note that I’ve also just put a new DataMine Lua file onto one of the threads to fix a problem that may occur with the UUID feature (it may depend on how the USB is configured - I’m not really sure). I’d definitely grab that one you’ve got updated [url=http://forum.micasaverde.com/index.php/topic,13512.msg101325.html#msg101325]http://forum.micasaverde.com/index.php/topic,13512.msg101325.html#msg101325[/url].
All dataMine data and configuration is stored on the USB - other than the mount information. So, in theory, updating won’t impact your data - this is of course only true if you don’t trash your USB. This also goes for uninstalling etc - the data should be ok. If in doubt, put the USB drive into your PC and copy all the files somewhere before doing anything (in fact, I recommend doing this as a backup periodically anyway).
First the program updated itself automatically to 0.967 - no sign of this mystical star (what’s that all about by the way?).
I have three temperature variables with offsets. I now find the offsets are applied to the wrong data.
The config file shows the offsets and version number as where they belong and all as they should be
The application of the offsets appears to be rotated by one, as seen as they are ordered in the config file (even though it’s an unordered object).
So it appears the problem occurs after the config file is read and made use of.
Just by the way I tried the plugin reload by using the URL method - readers should note; if you do this more than once, the request will be cached and not work on subsequent tries. You can avoid this by adding &retry=1 to the url and then incrementing the number each time:
I’m not 100% sure since most of the plugins I have installed are ones I’ve written, and they all have a *… The star is after the version number in the apps list, and I think it indicates that it won’t be automatically updated for some reason - possibly because you’ve done something to manually change files…
[quote=“a-lurker, post:4, topic:174191”]I have three temperature variables with offsets. I now find the offsets are applied to the wrong data.
The config file shows the offsets and version number as where they belong and all as they should be
The application of the offsets appears to be rotated by one, as seen as they are ordered in the config file (even though it’s an unordered object).[/quote]
Well, I was sure you must be wrong, but, your not! :-[ I’ll take a look at this - at first glance the code looks right, but the pointer must be incremented earlier than I think. I’m sure this won’t be a problem to fix.
[quote=“a-lurker, post:4, topic:174191”]I note the following line in the log file; I have seen it once after an install - I can’t verify that it always occurs - needs more testing:
Note the double slash in the file name - that’s a concern as the device description is not found.[/quote]
I don’t think this is anything that I have control over. I don’t know if it was logged during/around the time of the software update. I’ll keep an eye out for this, but I don’t think there’s anything I can do as I don’t specify any paths when this file is entered in the apps interface.
@a-lurker - this is sorted. It was caused by differences in the way JS and Lua handle arrays/tables. I’d assumed the data coming back from Vera was in the same order as I sent it out, but that wasn’t necessarily the case… I’ve now re-correlated the response so it should be fine in the next version…
I thought this: “Reworking code to support remote server” was to reslve this… was it?[/quote]
No - sorry, this isn’t quite what I was doing…
Unfortunately, it’s not possible to completely bypass this using a web browser. The problem isn’t getting the data back - the problem is with the web client. MIOS won’t serve this up, and it’s not easy (or secure) to just put it somewhere else and have it access the data.
What I’ve been looking at is to write an Android app that can use cp.mios.com. Actually, I can “compile” the existing GUI into an app - it looks the same as if it was running on the browser, but it means you an access the data from outside the local network. In theory, it’s also possible to do it for other phones - I haven’t investigated this for a PC yet…
It’s still work in progress - I did the proof of concept and it works - I just need to resolve some timing issues and maybe tweek the interface. It is still a little way off being ready…
So, sorry - not quite what your were hoping for I’m afraid.
Now that I’m on 0.967, I’ve noticed something weird (might just be me) but having just added all new items to the monitoring listing, when they’re rendered on the graph the first marker looks to be a full 24 hours earlier than the last one, even though there is no way a value was captured at that time.
I was wondering if this was done to give you a starting point/value or something (so done by design) , as when starting it makes for a very long line from that mystery first entry to the first real entry ?
[quote=“parkerc, post:9, topic:174191”]Now that I’m on 0.967, I’ve noticed something weird (might just be me) but having just added all new items to the monitoring listing, when they’re rendered on the graph the first marker looks to be a full 24 hours earlier than the last one, even though there is no way a value was captured at that time.
I was wondering if this was done to give you a starting point/value or something (so done by design) , as when starting it makes for a very long line from that mystery first entry to the first real entry ?[/quote]
Yes - exactly, this is be design. The graph is drawn for a particular period, and it just puts the previous value at the beginning, and continues the last value to the end of the graph…
Chris
Best Home Automation shopping experience. Shop at Ezlo!