Prusa goes Subscription  

RE:

The actual reasons are multiple, one of  the core one ( pun intended) is the fact that in recent years cloud computing price stopped dropping ( they flattened) and since previous year the price stared to rise due to many factors:

- old hardware in datacenter need to be replaced, new hardware is much more expensive ('tanks' to AI ram and disks went into the moon)

- due to the oil price changes the cost of hardware production and transport increased, also this also influences the cost of electricity and cooling

Add to it other factors:

- people started to use more data and are to lazy to delete files and instead of 1gb they prefer to have more storage, especially when multi color prints and print volume or print resolution increases the generated gcode files

- you wanted video streaming from the camera, this is quite expensive especially  due to traffic

- due to things like GDPR some data needs to be processed in the different regions such as clients from Europe are in european datacenters, which are more expensive.

The free tier was limited, but now they added additional tiers you can actually pay to get more.

RE: Prusa goes Subscription

I'm not into putting stuff someone else's cloud, so this is a hard pass for me.

But beyond my personal preferences, I also dislike what this move broadcasts. The enshittification of Prusa has been going on for a couple of years now. First there was the stripping of features from PrusaLink, diminishing the value of the local experience. Then the privacy invading app was introduced, tying you to their service. Also, the inclusion of Printables into PrusaSlicer without either an opt out or any sensible privacy controls. And now this. It basically tells us that Prusa is now a company which on the "you'll own nothing and be happy" bandwagon.

Why not release a cloud software which can be hosted homelab style? Why doesn't Prusa respect a customer's privacy and autonomy? Prusa is fast becoming very uninteresting. True open source is becoming much more interesting again, even if it is more laborious to get going.

RE: Prusa goes Subscription

It's a difficult call to make, I guess: Does Prusa want to deliberately differentiate from Bambu by strengthening their local, non-cloud solution? Or do they see the need to follow Bambu's lead because that's what the majority of the customers (apparently) want?

For the time being it seems that they are mostly taking the latter route. PrusaLink has not received much love at all -- even nuisances that could be addressed with trivial effort, like the non-changeable bulky password that the printer generates, are still in place. And now EasyPrint sees new releases while PrusaSlicer is in eternal refactoring mode.

RE: Prusa goes Subscription

A lot of changes they do are aimed at ease of use because users want it, such as NFC antenna on the back of the printer is just to quickly and conveniently connect the printer to the wifi and nothing more.

Lot of self hosted is difficult and people just don't bother doing it at home. Making a service platform also solves a lot of issues in legal, and allows to add features without releasing a new standalone product.

Yet Prusa is not fully following the route of some companies that just lock everything to the cloud ( vide IoT scrap), and allows to use tye printer without he cloud - it will be less convenient but useable.

Regarding releasing software opensource for their platform - it would be hard to host it locally due to complexity, but would be insta cloned by competition...

RE: Prusa goes Subscription

It already exists, but not on the Prusa ecosystem. If you're in the Klipper ecosystem most of this is available under open source license.

RE: Prusa goes Subscription

I think im missing the point here but does the subscription in any way or form reduces the functionality you had until now? i never understand the approach of "what if .. intthe future". Well thats for the future. Is the same silly comment i hear about Bambu (i dont own one). What if in the future the... sky turns purple and pizzas are all vegan with soy cheese. I criticized prusa a few times about the approach to the XL, genaral for their pricing/features on their filaments, printers (ie SLS1,enclosure )etc. But here we have a company that is offering a subscribed model for added services. You still have your Prusaslicer and you still have your basic Prusa connect. Did they brake a promise? Did they charged something that before it was free? If not then whats the actual problem? I dont like subscriptions and if prusa goes down that rabit hole with its ecosystem ill just stop upgrading and go for another company. But now this is not the issue. The OP gives a feeling that not only Prusa but any company should give online storage and an managed ecosystem that costs resources for free. Come on. Lets be realistic and stop wanting everything for free. There is a cost in manpower, time and computing resources in handling an online service. If you dont like it stay on your local network. Maybe they will do what Microsoft did with office. Maybe not. 

RE: Prusa goes Subscription

Posted by: @tsamisa

I think im missing the point here but does the subscription in any way or form reduces the functionality you had until now? [...]

Those questions have been asked by others in this thread before, and I have already stated why I am concerned about the introduction of subscriptions. No, nothing has been removed from the prior functionality -- but less has been added than otherwise would have, presumably. I'd much rather see Prusa focus on PrusaSlicer 3.0 than on a hosted EasyPrint offering, for example. 'nuff said, I'm repeating myself.

RE:

I think im missing the point here but does the subscription in any way or form reduces the functionality you had until now?

Over time there certainly have been things which have deteriorated. PrusaLink got defeatured. More and more things are tied to the Prusa phone app, which is a way of locking them to data harvesting. And the Prusa app is into that, if you check the report. (why else use the ADSERVICES permissions?)

As for the "Poor Prusa Needs Money for Services" angle. I'm totally not buying it. The cost of hosting is negligible. What they're charging you is easily 10x, if not a lot more, of what it costs to run the actual service. And the Enterprise solution, give me a break. That just screams local hosting to me.

As for no promises broken. That's technically true, but  but the whole move to servitisation is not in line with the you own it yourself spirit to me. But hey, servitisation is one of the key pillars of the 4th industrial revolution that some people are pushing for, so no surprises there 🙂 Prusa's moves in the last couple of years are right in line with: "You'll own nothing and you'll be happy.", even if they say otherwise.

RE: Prusa goes Subscription

I'd much rather see Prusa focus on PrusaSlicer 3.0 than on a hosted EasyPrint offering, for example

In this we agree and also in every other aspect that Prusa neglected. But personally i like to complain, and i do, about the things that they don't deliver as they promised and not of their new practices.

Over time there certainly have been things which have deteriorated. PrusaLink got defeatured.

Not exactly. I only use Prusalink and my complain is not that is de-featured is that it never actually progressed. File for example handling is awfull. I can delete multiple files from Prusaconnect but not prusalink. But again i complain about the Prusalinks neglect and not Prusaconnects improvement.

As for the service cost im not so sure. Is always about the R&D and administration rather than just hosting. And there is some complexity into developing a cloud based multiple harware device control which has to be secure and robust. And also even if disk space is not so expensive , charging for it is a way to control unnecessary "greed".

Servitisation for me is something i dont personally like since even if it is its own thing it affect the standalone options like the way we mentioned before. The standalone solution become stagnant to gently push to cloud services. But for me since this is a hardware device that when bought it was supplied with the software needed to work , it doesnt matter if it offers services. It matters if they cannot deliver the things they promised... like let say multinozzle printing without being "experimental" or improving some shortcomings in their hardware like the lack of proper cooling on the XL. If now at some point i see that this.. promises .. are fulfilled in their cloud counterparts and not in the original standalone versions then yes the service approach causes for me an issue and promises were broken.

In the end of the day if Prusa goes to cloud/based services then i think they are the ones that will loose the most. Even now a portion of their customer base went to other brands which offer equally good hardware, more user friendly and out of the box experience with a fraction of the price. And the opensource, data secure, always updatable premium you pay is going to be unjustifiable (which for me it already is). If a Prusa becomes a BL why buy a Prusa which is more expensive?

RE:

Posted by: @tsamisa

I'd much rather see Prusa focus on PrusaSlicer 3.0 than on a hosted EasyPrint offering, for example

In this we agree and also in every other aspect that Prusa neglected. But personally i like to complain, and i do, about the things that they don't deliver as they promised and not of their new practices.

That's a fair point. It's just that the two are connected, since Prusa has finite resources. So more attention on one customer segment or product means less attention elsewhere.

It seems pretty clear that Prusa is reducing their original focus on the "dedicated hobbyist" segment, in favor of "casual user" and "professional and print farms". Does that make them evil? No, of course not. Maybe it's what they have to do because the serious hobbyist segment is shrinking relative to the others. 

But does it entice me to consider Prusa for future printer purchases? No. Many other companies target the same segments and arguably have cheaper and/or better printers. What used to differentiate Prusa, namely that they cater to an audience which matches my own preferences, is gradually fading. 

I am not actually complaining about the change of focus, but expressing my regret -- from a selfish perspective, I guess.

RE: Prusa goes Subscription

If you want a printer in every home (because you sell printers), you have to deal with the fact that not everyone has a computer, and that means cloud slicing for phones etc.

Bambu's solution was pre-sliced print profiles for download which requires the model vendor to do the work; Prusa's solution is to slice in the cloud which requires the user to do the work.

RE: Prusa goes Subscription

I may eventually "subscribe".  I like the idea of having extra storage.  

RE:

Let's do some maths, shall we?

If 50k users take the 20GB plan, the cost breakdown is about as follows, based on cost at Hertzner  on a monthly basis:

S3 bucket with 10TB storage and 30TB traffic: €114
5* AX162-R (48C96T/256GB) for slicing and other compute: €1435
5 FTE with an average workplace cost of 100k per annum: €41.666

cost per user per month: €0,86

So that is why I'm totally not buying the angle of "it costs too much".

RE: Prusa goes Subscription

You forgot about:

- underlying databases to mange users, and files

- infra to control the printers

- traffic for streaming from cameras

- few years of r&d of the prusa connect, which ends in few millions of dollars you got as credit to develop a platform

So the subscriptions are paying the dept already.

Also you want to earn money, its not a charity work...

Get over it, people. Don't buy it if you don't want to, but stop complaining. Noone forces you to pay for that service.

RE: Prusa goes Subscription

I made a small error: The S3 bucket would cost 100x, so €1140/mo. But that is still not a lot on the per cost per user basis: €1.09

RE:

Posted by: @_kaszpir_

You forgot about:

- underlying databases to mange users, and files

- infra to control the printers

- traffic for streaming from cameras

- few years of r&d of the prusa connect, which ends in few millions of dollars you got as credit to develop a platform

So the subscriptions are paying the dept already.

Also you want to earn money, its not a charity work...

Get over it, people. Don't buy it if you don't want to, but stop complaining. Noone forces you to pay for that service.

The traffic and storage is amalgamated in the pricing I mentioned and a good DevOps team is in that 5FTE. Are you arguing that the 5 FTE is not enough? Given the scope of the project, 5 FTE seems realistic to me. For Czech wage levels the €100k/annum workplace cost is already on the upper end. Why are you saying it is not?

What infra cost are you talking about? The servers and storage and their traffic are already included in that cost estimate.

I think the R&D for Prusa Connect costing "millions" is fairly overblown. A lot of the technologies already existed in some form or another. It is mostly about piecing them together.

So I'm not buying your story. It has huge holes in it.

RE:

I just want to say that you cannot count only the maintenance cost such as 5 ops + infra, you need to pay for much more things underneath, and often you need to pay for the research and development that was done in previous years.

I suspect that they got something like about 20 people working on PrusaConnect (at least not only in the engineering), the infra itself is probably in the range of 2FTE or more. I did R&D for the platforms and annual cost of the team and infra together easily hits few million dollars if you count it.

Example:
FTE 8000 EUR/month (the money the employer that needs to pay with taxes, not the money the employee gets...)
FTE count 20 ( 3 ops, 2 development teams (14), manager, legal/security etc can be a fraction, 2 people counted as infra cost circa 16k in the cloud for that project size is pretty common)
this gives you 160k EUR/month
if you get 50K paying customers then it would mean customer needs to pay about 3.20 EUR to sustain that team and infra.
Now add the years before you had that platform up and running...

RE: Prusa goes Subscription

The scope of the project is not that big. It's basically an API with quite a limited set of calls and a presentation layer attached to it that gets shown on relevant devices, either in a browser as a website or in application form and a backing of a cluster with relevant compute services. Online slicing has been solved ages ago, so that is not where the complexity lies. The rest is not quite COTS plug and play, but it is not far off either. Most of it is CRUD, which is relatively simple to implement. The hardest part is routing and balancing the traffic of the services. for which design patterns exist. It's not rocket science. 20 FTE for R&D and more than a year of development on this service seems overkill to me. 2 Ops/Sec and 3-4 Devs for a year ought to do the trick, if they're any good. But I will grant you that I overlooked the overhead for management, etc.

RE: Prusa goes Subscription

Sorry, I don't buy into those calculations. I kind of get the argument that running costs need to be recouped via a subscription. But development costs for the platform should be covered by the printer's purchase price -- just like the development costs for printer hardware, firmware, and PrusaSlicer.

Otherwise, you are really on that slippery slope to "you own nothing", and can start charging a monthly subscription for using the printer as well.

RE:

The bottom line is they have to make a profit, and I am sure they are.  It is harder to make a profit in the Czech Republic than in China because labor and material costs are higher.  The pricing is not as bad as it could be, but higher than I would like.  I guess I am getting older and more frugal.