4 ms·

I'm astounded that they would seriously consider closing GCP after committing so much time, money, and effort into building it. Every time there's a GCP story

I'm astounded that they would seriously consider closing GCP after committing so much time, money, and effort into building it. Every time there's a GCP story on HN commenters say "I won't use it because Google can just pull the plug." I thought this opinion was silly as GCP is the #3 cloud provider in the world, not a free service like Reader. Guess I was wrong, nothing is safe. Google wants #1 or nothing at all.

Lammy 7y ago
I never thought they would shut down Google+ either. It may have been a “failure” in the sense of not beating Facebook, but it had enough traction among hacker circles especially that I assumed it was here to stay just as the cross-Google identity system if not a standalone Facebook-alike.
asplake 7y ago
However tempting (and it is very tempting), how much they’ve already spent should never figure in this kind of decision (aka the sunk cost fallacy). What matters is the return they will get from future commitments.
- mc32 7y ago
  That fallacy sunk the Desertron after many years of cost overruns...
- jjeaff 7y ago
  That, and the strong signaling it would create if they did drop it. That signal being, never ever use a Google product for Enterprise purposes.
  - leppr 7y ago
    More precisely, always have a backup plan. Nothing wrong with running on GCP if you're using it in a platform-abstract way (either by using it as a simple hosting provider, or by going through a layer like the "serverless" framework).
Guest42 7y ago
Could they sell parts of it?
rconti 7y ago
Yikes. Seriously making me question my initiatives as well.
cdumler 7y ago
In my opinion, it's just a consequence of the culture of stack ranking. The only thing that promotes you is if you deliver something. Not supporting something well. Not having an established, loyal customer base. Or, even if you provide good support/bug fixing. It's make something go public. Doesn't matter in what state. It's why I wasn't in the least shocked about Stadia's problems.
- throw_away 7y ago
  Amazon has stack ranking and hasn't fallen into this particular trap. It all depends on what you prioritize and what you reward. At Amazon, I got the sense that it was actually harder to get a promo if you were on a secret new product team. I'd heard younger engineers talk about doing stints in less fashionable departments as the ladders seemed easier to climb. Google just seems to care about the launch. Not how happy customers are. Not how much money is pulled in. Not how well it fits in with the overall strategic vision of the company. Just launch points.
ganeshkrishnan 7y ago
GCP can seriously be #1 because their kubernetes platform and cloud platform is much better than other cloud providers I have used. I started off with their Google Cloud Engine (?) and Big Table they provided and even though it was not really pure Java, I loved it. And then they pulled the plug on my accounts because some of my android apps had "likeness or similarity to other apps" which I have absolutely no idea what it meant. Then they pulled the plugs on my colleague's account because of equally vague reason. It was time for me to give up GCP forever.

Lammy 7y ago

I never thought they would shut down Google+ either. It may have been a “failure” in the sense of not beating Facebook, but it had enough traction among hacker circles especially that I assumed it was here to stay just as the cross-Google identity system if not a standalone Facebook-alike.

asplake 7y ago

However tempting (and it is very tempting), how much they’ve already spent should never figure in this kind of decision (aka the sunk cost fallacy). What matters is the return they will get from future commitments.

mc32 7y ago
That fallacy sunk the Desertron after many years of cost overruns...
jjeaff 7y ago
That, and the strong signaling it would create if they did drop it. That signal being, never ever use a Google product for Enterprise purposes.
- leppr 7y ago
  More precisely, always have a backup plan. Nothing wrong with running on GCP if you're using it in a platform-abstract way (either by using it as a simple hosting provider, or by going through a layer like the "serverless" framework).

Guest42 7y ago

Could they sell parts of it?

rconti 7y ago

Yikes. Seriously making me question my initiatives as well.

cdumler 7y ago

In my opinion, it's just a consequence of the culture of stack ranking. The only thing that promotes you is if you deliver something. Not supporting something well. Not having an established, loyal customer base. Or, even if you provide good support/bug fixing. It's make something go public. Doesn't matter in what state. It's why I wasn't in the least shocked about Stadia's problems.

throw_away 7y ago
Amazon has stack ranking and hasn't fallen into this particular trap. It all depends on what you prioritize and what you reward. At Amazon, I got the sense that it was actually harder to get a promo if you were on a secret new product team. I'd heard younger engineers talk about doing stints in less fashionable departments as the ladders seemed easier to climb. Google just seems to care about the launch. Not how happy customers are. Not how much money is pulled in. Not how well it fits in with the overall strategic vision of the company. Just launch points.

ganeshkrishnan 7y ago

GCP can seriously be #1 because their kubernetes platform and cloud platform is much better than other cloud providers I have used. I started off with their Google Cloud Engine (?) and Big Table they provided and even though it was not really pure Java, I loved it. And then they pulled the plug on my accounts because some of my android apps had "likeness or similarity to other apps" which I have absolutely no idea what it meant. Then they pulled the plugs on my colleague's account because of equally vague reason. It was time for me to give up GCP forever.