SONOS Legacy issue - You are deleting posts of enraged customers, and closing comments!

I am puzzled as to how anyone could look at the forum at the moment and conclude that Sonos are deleting critical posts.

Looks like a lot of fire fighting going on …. this mess can’t be covered up.

As a software developer myself, I completely understand the desire to improve product offerings. But when “Legacy Products” are built into the fabric of people’s homes, you have an obligation to maintain support for those loyal customers.

I have untold Sonos devices, not all of which you have shown. I have spent £££’s buying them and having them installed. Don’t let me down, or you will lose my support and endorsement - forever.

As a software developer, if you have any experience with embedded systems you know the struggle Sonos has been having with maintaining the antique Linux kernel that is all they can fit into 32 MB along with other essential bits. Sonos had to start paring out code that folks used, client access to SonosNet, the Dock support as well as being unable to do something like increase SMB to a current version to keep the antiques running.

The legacy split isn’t ideal but it will keep the older gear running for at least a few more years.

Offload the controlling technologies to the controller devices (iPad, iPhone, Android, PC). Use only the comms basics on the Sonos hardware. Future proof the hardware. Surely they can get that into 32MB. Or bring out another device that harmonises new with old. There has to be another way. I ain’t upgrading wired in components.

Offload the controlling technologies to the controller devices (iPad, iPhone, Android, PC). Use only the comms basics on the Sonos hardware. Future proof the hardware. Surely they can get that into 32MB. Or bring out another device that harmonises new with old. There has to be another way. I ain’t upgrading wired in components.

Other systems, such as lighting control and thermostats introduce a “hub”. This would be a major conceptual change for the SONOS architecture because the current products form a co-equal mesh. Unfortunately, the legacy products are becoming less and less “equal”.

With modern systems we have become both empowered and lazy. I’ve been fussing with micro computer systems since inception. In the earliest days a few kilobytes of RAM was nirvana. Now, a few tens of megabytes is a terrible constraint. I can remember my first “print driver”. I had 128 bytes available to do what I needed to do. Fast forward a bit, I recently pulled out a Raspberry Pi that has been sitting in a bag since 2014. In Raspberry Pi terms this is old, and under powered (512MB, 700MHz). Regardless, after only a few hours of familiarization I had the system up and running with a (slow) GUI and a nice selection of on board programming languages. A few evenings later I had the Pi linked up with Arduino (also living in that bag since 2014), a motion detector, audio sensor, WiFi, running the Apache web server to present the results. I am independently monitoring an automated pet feeder. Obviously I could take on the pet feeder control too, but I wanted a completely independent monitoring system. This is the power of the modern approach. Even on my ancient (1K, 1 MHz) micro board I could perform the monitoring, but without the fancy web server or access to an online time server to keep everything synchronized. (and the development environment is dramatically different)  Constrained as it is, my ancient Pi has more resources than the legacy SONOS kit.

I too got myself a Raspberry PI in late 2018. Played with it for a month and was blown away with its capabilities. I was interested in circuits and Sonic PI. Maybe for £100 Sonos could create a distributed network of PI’s, to overcome their Legacy issue. The thing I cannot understand, they are only controlling an Amp connected to a speaker after all. That technology is not changing.

As you know, you really blew it. But now I have paid good money for a “Port” and have a “bricked” ZP-90”. Now that you have backtracked on this program, I need you to replace the ZP90 and give me my money back,.

In my Computer Science classes, way back about ‘75 or so, we had an InterData 7/32 (IIRC), with shared access via VT-100 terminals and the old TI thermal line printers.  We’d test our ideas interactively using the BASIC language, in a huge 4K of memory(!), then translate the code to the required PL/1 language, punch it out on the cards, and submit the deck to run overnight (usually would fail the first run).  Saved a ton of time by getting it working first, vs going through multiple iterations of those damned card decks.

Heh, I wonder how “enraged” customers of those old InterData 7/32’s and IBM 360’s are now that they’re museum pieces.  They were WAY more expensive than the old Sonos devices, lol.