Re: [mh] MH5.1 + The Future

Thanks Richard, those are excellent points, and indeed I feel the same way. I have usercode that is dated 2007, and they still ‘just work'

A suggestion for those that have older versions, but want to look at the new features for value.  MH has the ability to run multiple instances, I run 3 on my main server. One for the bulk of the logic, one for ‘web services — those code modules that might create pauses’, and another for non-critical MH stuff that I don’t need in the main loop. I run each of these instances on a separate IP address on my linux server. I use the xAP infrastructure to keep the states in sync. ($item->xAP_enable()). 

This would allow someone to start up a new v5+ instance of MH on a separate IP or port, configure a few generic_items with the same name as objects in another MH instance, and then test  out if the new features (web interface, homekit, alexa, google home, etc…) are useful and worth upgrading to.


> On Nov 2, 2018, at 8:03 AM, Richard F via misterhouse-users <[email protected]> wrote:
> 
> I'm also a long-time user of MH (2008), but lurk on the list picking up
> the odd hint here or there.
> 
> I have sent in some code snippets (Squeezeplayer integration recently,
> which I continue to tweak), but a fair amount of my code is very custom
> and so not easy to describe or particularly useful if you don't have my
> hardware.  Longer term, I would like to move to the MH Git branch, but
> as John, it's too important to change without a lot of testing (no
> heating, lighting... and important integrations that make all the
> difference).  With the experience I can reliably maintain/improve my
> environment without breakage - so yes maturity is in several ways a
> reason for the limited traffic on this list.  For ref. I achieve around
> "three nines" availability, and a lot of the "unavailability" is a
> result of enforced OS updates, and I value availability highest next to
> functionality.  Did I mention that MH never crashes or loses it's config ?
> 
> MH has a lot of strengths that are too easily ignored today.  No.1 of
> course is Open Source and portability - Perl may be old, but it does
> gradually improve from release to release and probably has the widest
> range of library support out there due to longevity, and I doubt will
> die before me.  A lot of new stuff comes and goes with fashion, tied to
> vendor's aspirations (including Google who are actually one of the most
> fickle - Wikipedia lists nearly 50 closed projects!  Google Plus anyone?). 
> 
> With an infrastructure that is 100% open source, nobody can entirely
> obsolete it, though takeovers of Github and now Redhat threaten even
> this) - having somewhat of a "prepper" attitude with your core system
> isn't so bad an idea - I keep spare chips etc that may go obsolete. 
> Your home infrastructure needs stability - tie yourself to (a limited
> number of) vendors whims, worse "cloud" whims, and your investment in
> time and money getting it "just right" with 3 9's reliability, may
> disappear!  IA5 may be a bit dated, it does work acceptably on a 2018
> Android and gets the job done, I'm sure IA7 would be better, but it's
> testament to good initial design that MH/IA5 keeps on working though 3
> major Linux migrations including a switch from 32 to 64 bit with barely
> a tweak.  Actually later Perl versions showed pre-existing bugs in my
> code, easily fixed.
> 
> So, a massive vote for MH, let's all continue improving it!
> 
> Richard
> 
> On 28/10/2018 00:04, john wrote:
>> On another note, I’d be interested in the community’s perspective on
>> the future of MH. It seems like there’s been a reduction in messages
>> on the mailing lists, and not a big list of contributors submitting
>> pull requests. I don’t know if this is due to the stability and
>> maturity of MH, or if users are moving away to different directions.
>> MH is definitely one of the oldest DIY HA products out there, but it
>> seems like on-premise solutions are being more difficult to find.
>> 
>> 
> 
> 
> 
> ________________________________________________________
> To unsubscribe from this list, go to: https://lists.sourceforge.net/lists/listinfo/misterhouse-users
> 



________________________________________________________
To unsubscribe from this list, go to: https://lists.sourceforge.net/lists/listinfo/misterhouse-users