How to determine the end-of-life (EOL) for each Sonos product?
Now that Sonos have announced that their products have an end-of-life (EOL), sometimes in as little as 5 years from purchase, can they please clarify their expected EOL projections for all products? Classifying them as “modern” or “legacy” does not help and is far too simplistic. Products can move status overnight without warning.
It would be useful to know that a Sonos product bought today for several hundred pounds is expected to be viable for, say, 5yrs. If Sonos wishes to be a software company, they will need to embrace a more mature approach to this issue, and offer an SLA within each product sale to set customer expectations.
For example, the Windows 10 support life-cycle has a five-year mainstream support phase that began on July 29, 2015, and a second five-year extended support phase that begins in 2020 and extends until October 2025. Microsoft manages this level of expectation management and costs a fraction of the cost of a home speaker system.
Page 3 / 3
There are other factors besides technical capabilities alone that will ultimately determine when a product no longer receives regular updates. Customer expectations is obvious. Obviously, the current event is not going over well, so it stands to reason that Sonos will not want to do this again any time soon. That could change though if other brands start dropping support for their older devices as well, or if the technology speeds up rather quickly for whatever reason to the point where customers practically expect the older units to no longer supported (as an extreme case). As an example, if both Amazon and Google drop support for their first gen products, that will have a big effect on customer expectations.
I am not predicting that any of this will happen, I actually think it’s highly unlikely we will see any change in support for the next few years. But I can’t guarantee anything and may channge my opinion depending on what happens in the market going forward.
Exactly. We know the facts now, we don’t know the future. Can we make an educated guess given Sonos’ obvious efforts to keep the 32 MB units running for going on 15 years? Sure. Can we accurately predict future developments and their impact on the current ‘modern’ limit of 128/64 MB RAM/storage? No, not at all.
They will probably continue the recycle program 30% discount. 64mb is peanuts compared to what the One has. Sonos could shut down (legacy) 64mb play1 very soon (2 years). But don't worry about this because as jgatie says they could keep it going for longer.
I said nothing of the sort. Please don’t put words in my mouth. And 64 MB might be peanuts, but it is double what the legacy items have
I attributed the thought that Sonos could keep the old Play1 going for much longer than the 5 year date to you. I thought that was implied in your posts, sorry if it wasn't.
There are Play5 with 64 mb ram, these are being dropped are they not? Or are these Gen2 Play5?
How do you identify a Connect with 256mb ram? Does anyone know the hardware number for this?
There are other factors besides technical capabilities alone that will ultimately determine when a product no longer receives regular updates. Customer expectations is obvious. Obviously, the current event is not going over well, so it stands to reason that Sonos will not want to do this again any time soon. That could change though if other brands start dropping support for their older devices as well, or if the technology speeds up rather quickly for whatever reason to the point where customers practically expect the older units to no longer supported (as an extreme case). As an example, if both Amazon and Google drop support for their first gen products, that will have a big effect on customer expectations.
I am not predicting that any of this will happen, I actually think it’s highly unlikely we will see any change in support for the next few years. But I can’t guarantee anything and may channge my opinion depending on what happens in the market going forward.
Exactly. We know the facts now, we don’t know the future. Can we make an educated guess given Sonos’ obvious efforts to keep the 32 MB units running for going on 15 years? Sure. Can we accurately predict future developments and their impact on the current ‘modern’ limit of 128/64 MB RAM/storage? No, not at all.
Exactly, best for all customers to protect their hard earned dollars consider when the device was released and look at the ram chart when making new purchases as there are still devices in the lineup that could be ditched due to their *cough* feeble hardware. Hold on to older hardware that isn’t legacy because it could be supported for another decade. May can’t come soon enough.
But from what I can see the Play:1 only has 64MB Storage, which is why I’m getting worried.
I understand your point about it more being to do with genuine system limitation rather than a statutory period of time, but I would argue that if Sonas do say next year that the Play:1 technically won’t work, then the onus would be on Sonos to find a very decent solution (Replacement, deep discount, or some form of technical solution that keeps them viable)
That graphic is the old version. It’s been updated:
And yes, I was mistaken. The old Play:1 has 128 MB RAM, but only 64 MB storage.
Thanks for this, appreciate the updated chart.
Does anyone know how you distinguish whether you have a play 1 which 64mb/128mb or the play 1 with 256mb/256mb? I find it very odd there are 2 types of play:1s
I’ve had a look at my products and can’t see anything on the box that indicates which one I have and it’s not indicated when I look at my sonos device online.
Surely the play1 with 64mb/128mb has to be at risk of going legacy next?
Something to keep in mind, the Play 1 should be compared to the Sonos One SL, not the other Ones as they have need for much more internal hardware to deal with voice issues.
The early Play 1s are similar to the Play 3s in memory, that is going to justify a lot of effort on Sonos’ part to keep them working.
Sonos has in the past used resources of higher memory devices to support lower memory ones, the music index offloading for example. With the worst memory devices gone the chance to move to a newer Linux kernel and have room for key improvements in the now available space and offloading things that just won’t fit should serve for many years.
When looking at memory in the newer devices compared to older ones keep in mind the cost differences, not only for the memory but in putting it into a system. Look at the Raspberry Pi as an example, they want $35 for a 1 GB version, for $45 you get 2 GB and for $55 you get 4 GB. Putting in extra memory today is far different cost-wise than it was 15 years ago.
When looking at memory in the newer devices compared to older ones keep in mind the cost differences, not only for the memory but in putting it into a system. Look at the Raspberry Pi as an example, they want $35 for a 1 GB version, for $45 you get 2 GB and for $55 you get 4 GB. Putting in extra memory today is far different cost-wise than it was 15 years ago.
This is something I’ve thought about a bit, how much memory costs over the years. I keep thinking about the memory available on thumb drives. It seems to me that in 2015 and possibly earlier, it was pretty easy to get thumdrive’s with memory over 1 GB for cheap. So that begs the question of why didn’t Sonos regularly increase the memory of their devices as the costs got cheaper?
I don’t have a background in manufacturing, so I can only guess, but I think there are a couple reasons. One is that Sonos surely bought the memory in bulk, possible a contractually negotiated rate. They could not easily upgrade the memory without taking a hit on the memory they have already purchased, or breaking a contract deal. Could be 100% wrong on that. The other reason is that you cannot just replace the memory and nothing else. It’s my understanding that you would need to do some tweaks to the overall design, maybe change some other hardware as well, and then test out the new hardware before releasing. We’ve also recently learned that it’s best to make hardware changes with an official generation change, which isn’t always ideal.
Probably other things I don’t know about, or things I’ve got wrong. I think you can say though that upgrading the memory isn’t going to be as easy as dumping the new memory in the parts bin at the assembly line.
The capacity and cost of different types of memory is important too, Sonos can’t chance iffy memory like a $10 thumb drive maker can get away with.
Redesigning a circuit board to accept new memory is expensive, testing the new design is even more expensive.
Then you have the regulatory hurdles from the FCC to UL (in the US) and probably several others.
Once you have the upgraded hardware manufactured you need to add it to your software test plan and every additional device makes that harder, more time consuming and more expensive.
It has been many years since I was involved in embedded computers and I do not miss the problems that brings versus a “plug stuff in at will” design.
The embedded design still has a key factor of reliability as a big selling point.
When looking at memory in the newer devices compared to older ones keep in mind the cost differences, not only for the memory but in putting it into a system. Look at the Raspberry Pi as an example, they want $35 for a 1 GB version, for $45 you get 2 GB and for $55 you get 4 GB. Putting in extra memory today is far different cost-wise than it was 15 years ago.
This is something I’ve thought about a bit, how much memory costs over the years. I keep thinking about the memory available on thumb drives. It seems to me that in 2015 and possibly earlier, it was pretty easy to get thumdrive’s with memory over 1 GB for cheap. So that begs the question of why didn’t Sonos regularly increase the memory of their devices as the costs got cheaper?
I don’t have a background in manufacturing, so I can only guess, but I think there are a couple reasons. One is that Sonos surely bought the memory in bulk, possible a contractually negotiated rate. They could not easily upgrade the memory without taking a hit on the memory they have already purchased, or breaking a contract deal. Could be 100% wrong on that. The other reason is that you cannot just replace the memory and nothing else. It’s my understanding that you would need to do some tweaks to the overall design, maybe change some other hardware as well, and then test out the new hardware before releasing. We’ve also recently learned that it’s best to make hardware changes with an official generation change, which isn’t always ideal.
Probably other things I don’t know about, or things I’ve got wrong. I think you can say though that upgrading the memory isn’t going to be as easy as dumping the new memory in the parts bin at the assembly line.
... And this theory makes complete sense when one considers that John MacFarlane stepped down back in 2017 when he realized that future enhancements like voice control and other things that Amazon, Google, and other companies were getting into would make it nearly impossible for Sonos to effectively complete with their 32 MB Legacy offerings. Whether or not they had external microphones was not really the nail in the coffin as much as having an entire ecosystem built around 32 MB or 64 MB architecture.
