Not at all happy about this. This is basically my thinking and what I wrote in the SU forum...
----------------------------
OK, I now understand there is a focus on trying to get extensions to work in the cloud sometime in the future. I suppose we can expect the whole app (PRO + FREE) to end up in the cloud at some point. If not, then this truly is a waste of valuable developer resources.

Pros

• Eventually when SU Pro goes to the cloud, we’re back to one code base set.
• Nothing to download (except I suspect there will need to be a ‘cached’ app somewhere so when you’re not connected to the cloud you can still do work like GApps.)
• Easy to update all at once.
• Can perhaps run on slower machines as most heavy processing can be done in the cloud (aka OnShape).
• If prefs ever get implemented, it will be nice to have them stored online for all computers.
• No license per computer and no dreaded licensing conundrum Trimble has put us through year after year.
• Makes the Google Chromebooks team happy!

Cons

• Not pipeline or ecosystem friendly. If you can’t quickly save and open the program in a different app. In fact I’m not even sure how to actually download the file. If it’s not as simple as SAVE to FOLDER then it’s a fail for working in any sort of workflow, pipeline or other ecosystem. And because there is no renderer, no animator and other missing items, it’s necessary to go to 3rd parties to get things done.
• By focussing on the cloud, valuable resources are usurped which could be used to actually add some new and needed features. Those resources are now focussed on making the proper frameworks, trying to get existing extensions to play, client-side code, and the whole cloud architecture, among other things.
• Takes a long time to put everything in place-- and meanwhile SketchUp core features keeps getting further and further behind. SU2018 is such an example. It’s more than clear all the dev resources are now focused on the cloud as this is the lamest update I can ever remember.
• Sometimes instant updates create serious problems and if not adequately tested can be very problematic for professionals on a schedule. I hope you consider the ability for customers to roll back to previous versions.

So, instead of finally getting a fillet or chamfer command, or UV mapping (after all it is 2017), or better booleans, or even a decent renderer, we are now watching the programmers spend all their time on trying to figure out load balancing on servers and how to get around the limitations of different browsers.

I sure would’ve like to have been in THAT meeting where the decision was made to allocate major development resources to move the whole codebase to the cloud. Perhaps the decision was made by programmers knowing they need to add significant cloud development capabilities to their resume?

It is more and more apparent to me there is little focus at Trimble on what users need to actually model and render buildings, products, and 3D prints in 2018. There have been very few tools added over the past decade. Yes, I know we all EXPECT the plugin developers to create them for Trimble. Still, even the magnificent Fredo can’t easily build the standard filleting tool every other poly modeler has because of the internal architecture of SU.

Yes, I know the original design was a departure from standard vertex poly modelers…and a good one at that. Even so, there are so many ways to make it better. Trimble seems very reluctant to actually add the necessary features, and instead they’ve decided to now spend significant resources towards isolating the program, and it’s users, in the cloud. I’m just wondering if the programmers are actual 3D guys, or former Google Cloud computing developers?

------------

Before anyone jumps on MoI4 saying many of the same things could be said about features, understand updating a codebase to 64bit and native on a Mac is worth a whole update in itself. SU did that a couple years back and the community cheered loudly.