HELP! Re: Is XTension Dead?

[James Sentman]

One of Michaels original ideas for the software was that you wouldn’t need the most recent machine and that it could run happily in a closet for years and once you have gotten a few scripts running it would be great if you could just forget it’s even there. For a lot of folks that is what they do.

There is a lot of excitement in the raspberry pi world recently. Some of the software available for HA on it is really quite good. I am on many of the mailing lists and groups talking about it and I admit that many days it’s quite discouraging. Sometimes it feels like they are far ahead and I’m just this one guy who has other responsibilities and cannot work 24 hours a day trying to keep up with the latest fad. But then as I look into it in more detail I have to laugh about what would happen here if I suggested that y’all needed to learn to manually edit a YAML file or some such ;) 

Though raspberry pi’s are cheap in their MSRP you almost literally cannot buy one right now even if you wanted to. And if you do find a high end one with enough memory to run those softwares it will run you a quarter of the price of a new M1 mac right now as there are just not enough of them available. I’ve been trying to pick up one of the new pi zero’s that have so much more horsepower to experiment with since they were announced and never has anyone kept them in stock for more than 90 seconds and then months with nobody getting even a single one.

The reality is more complex than that very simple equation. 

If you have a machine that can be upgraded past High Sierra you should consider doing that. High Sierra was a very problematic release and caused me more wasted time finding hacky solutions to Apple’s user space freezing issues than I spent actually developing new code. This went on for months after the release and they never fixed it until the next major release. XTension does it’s best to keep things going once you walk away, but it was never a 100% solution and it wastes CPU and memory to have an extra helper app running in the background talking to XTension waiting for it to hang up. 

I know I spoke recently about increases in requirements for some of the newer stuff that I am working on. As of this moment I have a fairly full features replacement plugin architecture completed that can replace Video Pitcher as a method to get IP video into the system. I’ve completed the jpeg refresh and mjpeg plugins based on it and am about to start work on the RTSP stream and others. I would like to make it all ONVIF but at the moment I can’t get any of even the test software for that to run in any meaningful way, so I’m concentrating on the stuff that does work. This opens the doors for some very fun stuff. Apple already does some face recognition but leaves it up to the cameras to only record when they see someone or something that they are looking for. This may change of course, but I haven’t seen it yet. The new stuff that I’ve included can do all that and much more and it absolutely requires a newer OS version. I have no intention of releasing a build in the next few years that wont still at least run on the older OS versions, but I cannot add support for them to the newer plugins. If you want to take that up with Apple thats fine by me ;) But I cannot fix what they do when they change the executable formats and re-arrange the library layouts. So newer versions of the app will continue to run but newer plugins, like the video system, will not operate on the older OS versions.

In between any other plugins or fixes I want to make I am going to finalize the video system over the next few months. Including a secondary plugin system to enable the various camera control API’s that are documented out there including ONVIF once I hammer that into submission here. The next item on my list is a fresh start on new web interfaces or dashboards or whatever people are calling them now days. The stuff that is possible now and much easier now than when I built those is amazing. Unfortunately nobody seems to do it the way that I need to so there is still a lot of low level work to do. After that I go back to working on extending the internal scripting away from AppleScript to Python. Don’t worry, as long as Apple lets AppleScript function I’ll continue to support it. But I will be offering python and possibly other scripting languages as an option. 

Once that is done then it would be possible to cross compile much, but not all, of the main app to run on a raspberry pi. Most of the display code would have to be rebuilt as I am using a lot of custom apple methods for drawing the interfaces where it is required. If I do a good enough job on the new web interfaces then it might be best to go with a headless version that you can do all of the config and work on via a web interface. This is really the direction I’m heading in I think. All of the python based plugins can be made to run on the Pi almost as they are. Some of them include binaries as well as python code and those will need to be rebuilt but I can’t think of one that should pose much of a problem. So possibly a background only version for the Mac and MAYBE even for the pi at some point in the distant future. But now I’m really just waxing poetic. Thats far enough ahead that things may change drastically in the meantime. 

I use a LOT of raspberry pi’s around here. I think I have 14 or so doing various things, all linked into XTension in one way or another. Many are security cameras or UPS monitors or remote receivers for the SDR radio plugin and several are running the apple streaming client on them so that I can beam from my phone to speakers in the room without having to deal with Bluetooth which I’ve just never seen an implementation of that worked nearly as well as the airplay stuff. So I like the Pi, I’d love to do a version for the pi, but I don’t know if thats practical right now or ever really. Especially with the new video stuff coming along I’m not sure that even the newest and highest memory pi’s have enough horse power to manage more than a few incoming streams. We shall see.

But definitely thanks for trying to depress me by suggesting nobody wants to run a Mac in a closet anymore, thats great ;) 

So definitely not dead. I’d love to sell more of it, but even in the absence of an invigorated market I have no intention of stopping working on it as I use it myself and I can’t imagine trying to port my system to anything where I didn’t also control the code base, that would be just very frustrating. I don’t know what will happen, but I certainly haven’t made any plans to end it.


> On Oct 20, 2022, at 10:23 PM, Mark Nettleingham <markfn at comcast.net> wrote:
> 
> Doug,
> 	I’m running XTension on an M1 Mac. and everything is just fine. The M1 is more responsive that any other system that I’ve deployed since 2001. However, my system may be simpler that others. I have a Vera Plus, XTB232, Barix Barionet, and am currently playing around with the Home kit interface (so far with less that useful outcome on the Home Kit).
> 
> 	I’m NOT planning on upgrading to Ventura anytime soon.
> 
> Have Fun,
> Mark
> 
>> On Oct 20, 2022, at 8:34 PM, Doug Kiekow <dougy at iphouse.com <mailto:dougy at iphouse.com>> wrote:
>> 
>> James, XTension and the list all appear alive and well here. The mac as an HA platform on the other hand is either dying, or already dead. The computational power required for HA is akin to that of a $50 raspberry pi, not a near $1000 mac. The mac suffers from bloatware and a tendency to break things with each software update, and becomes more of a desktop iphone each update. I'll continue to use it while it works, but I stopped upgrading with High Sierra, and would not even consider an M1 mac.
>> 
>> dougy

Thanks,
 James


James Sentman                       http://www.PlanetaryGear.org		http://MacHomeAutomation.com




-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.machomeautomation.com/pipermail/xtensionlist/attachments/20221021/4ff1d0a1/attachment.html>