I wonder this as well but outside the HN community. I don’t think many folks I know still attribute Google with the flakiness that this community does.
Yeah, but it matters disproportionately to early adopters. If a brand becomes toxic with early adopters it never gets in front of anyone else. Except through acquisition, I suppose.
Cloud compute is an even starker example: Google lost the second-mover cloud space to Microsoft and they lost it because the relevant decision-makers knew Google's reputation.
Microsoft beat Google to second in cloud because it already had a strangle hold on a sizable portion of the enterprise computing market, not because executives at important companies (large enterprises) knew Google has a reputation for killing obscure consumer products. Microsoft already had relationships with many large and medium businesses/organizations, often spanning multiple decades, a business software ecosystem, and an army of salespeople, account managers, and support people, making the transition to cloud a natural next step.
Google, on the other hand, has had to build these relationships from scratch with it’s only advantage being advanced software and infrastructure which reduce costs.
AWS doesn't have cheap Windows licenses, MSSQL, Office, and Active Directory. They're still beating Azure. Because they were first, of course -- but the point is that true competition is all about playing to your advantages and mitigating your disadvantages. Google had enormous advantages and squandered them. Maybe Microsoft's advantages were genuinely stronger, but from where I'm standing it looks like Google didn't even try.
I've seen the AWS vs Azure vs GCP drama play out a few times, once on a team with significant Microsoft legacy and twice without. In my judgement, the MS legacy, even at its most potent, was a smaller consideration than Google's unmitigated weakness, every manager's worst nightmare: the concrete risk that Google would cancel something and the manager would get blamed for not seeing it coming. Maybe I've just been blessed to work at non-dysfunctional companies where non-technical management asks the opinion of technical management on technical matters, but technical management is highly cognizant of this risk and has been for a decade.
My recommendation would have been to go on offense. Make AWS's constant over-promising and under-delivering a meme. Make Microsoft's we-have-altered-the-deal-pray-we-dont-alter-it-further enterprise pricing a meme (compare to: google and gmail are still free). Build counterveiling fears in that technical manager's mind, no lying required, because Amazon and MS earned those criticisms and they earned them hard.
>Google has a reputation for killing obscure consumer products
It has a reputation for killing cloud products like classic VPNs in GCP too, a significant part of our cloud infrastructure. Distrust is earned in this case.
I’d be interested in hearing your argument for that being the reason why Azure is beating GCP in marketshare. The conventional wisdom that Azure can do Windows for cheaper than everybody else and has an easier time getting enterprise customers they already have a relationship with onboard always made sense to me.
I just don't see the value in Stadia over buying a console. The only use case seems to be for someone too poor to afford a console, but that someone has enough money to buy a full price video game on Stadia.
Wow, I didn't realize that Stadia games are running on Debian and Vulcan. https://stadia.dev/about/ That is a hard sell that you have to (likely) port your game to a completely different OS and graphics API for a presumably small customer base.
There's a large number of games that use premade engines (eg. Unity, Unreal). As long as the engine supports Linux and Vulkan (I believe both Unity and Unreal do), there's no real issue in supporting that platform.
For the day to day user products I can see that not impacting sales much. But is their unwillingness to commit hurting GCP adoption? The people who would participate in adopting GCP (except in corporate environments) are people who might also frequent here.
Steve Yegge, a former Googler & GCP Customer begs to differ. From my experience GCP does deprecate a lot of libraries, SDKs, and APIs. None of which are covered in the SLA policy.
You might want to work somewhere else so you hear some differing opinions, pretty much nobody I know wants to implement on top of google products including GCP because of the perceived churn.
What were you using it for? I'll be upfront in saying that my experience with GCP was awful, but I can appreciate that others found it useful and would be curious to learn more.
That's because they haven't quite done this to any of their big consumer products or their core features.
Anecdotally I know some people who express irritation at the rotating plethora of messaging apps. I also know some mildly inconvenienced by the whole Google Play Music/Youtube Music nonsense, although I never got the impression that GPM was a market leader in the streaming space.
Cloud compute is an even starker example: Google lost the second-mover cloud space to Microsoft and they lost it because the relevant decision-makers knew Google's reputation.