Make the source code publicly available under a permissive open source license so the community can run the software.
Scrape/generate a read-only archive of the site and host the static files for the next ~2 years before finally dropping the domain.
Less spray & pray in the first place? It's already clear that their product strategy is fruitless, can you name a Google product from the last 5 years that stuck, loved by masses and deserves to be kept operational by their own measures?
I would try switching to spending the money on acquisitions.
Even if they can’t do that every time, it should at least be one of the options under consideration.
I feel like there’s often an argument that “keeping it running costs more than you think, because it would need updating” (or “it depends on legacy infrastructure that’s going away”), but I’m skeptical of that in the general case. Other than possibly security issues, there’s often little to get in the way of servers running happily with little or no maintenance for years.
In this particular case, I don’t know, although the lack of any positive examples to point to suggests that it’s not an option that’s typically given any weight in decision-making.
Whose department should take the budget hit for maintaining a no longer viable project?
Speaking as a user, I don’t care. Isn’t Google swimming in profit?
(I realize of course that you don’t achieve the kind of success in the first place that brings those colossal profits without making hard-nosed decisions along the way -- arguably)
Whose reputation should be dinged for running a non viable product?
Again, speaking as a user, I’d put value in keeping the “non-viable” product around, if it were one I happened to use.
I’m not expecting my answer to change anything, but it’s a general answer to a general question. What could they do differently? Answer: eat the loss.
I have very little interest in what happens behind the product so long as it both works and continues to work. Google experiments like crazy, and appears to kill off products just as quickly. The fact that they occasionally get a “hit” and earn hundreds of millions does not offset my fear that anything of theirs I depend on will be cancelled without warning.
Honestly, at this point Google products should come with some kind of versioning and labelling for commitments.
It would be far easier to choose Google services (and less irritating as a consumer) if services were labelled alpha, beta, production, and LTS, and if each of those labels had a specific duration (e.g. alpha services can vanish immediately without warning; beta have a limited lifecycle where they need to become profitable or at least self-sustaining, production are generally stable and can be relied on for a year, and LTS will be supported for a specific period of time, backed by enterprise support agreements).
Depends what you mean by bad; it hurts trust. For instance the Appengine to Appengine 2 transition is a particularly painful one for a consumer who built an entire business on it and I was part of a migration to address this.
I had heard Google is unreliable in this aspect before but after this particular experience I will unlikely use or suggest Google for anything serious.
Depends on where you stand... If you're one of the people who comes to rely on the product, then your answer would probably be "yes".
The cost of the experimentation is people wondering what product they'll discontinue next and whether it'll be one that _they_ rely on. If you find yourself asking this often enough, you'll probably start moving out of the Google ecosystem. How is this good for Google?
Yes, I understand that products need to pay for themselves, but Google keeps drawing people into services that they then shut down. It's frustrating for users and damaging to Google. What's the cost of the damage to their reputation?
I do, for two reasons. One, Google's culture and managerial system is pretty poor, in such a way that it rewards starting things and not finishing them. That hurts consumers. Two, it feels contrary to the design of public corporations. Should Google be allowed by its owners to spend money trying out hundreds of new ideas? Or should it give profits back to shareholders who can then go do these things themselves while Google just tries to be come a really efficient, and insanely profitable search provider?
I'd advocate for the latter as a healthier way to run society with fewer tech giants of lovecraftian size slipping their tentacles into every cranny that one can stuff a million dollars into.
It is a Google's graveyard. They will sunset this one (stop accepting new "bodies") after 2 to 5 years, and open a new, "improved" one, with worse UX and lost features.
Wow - first time I've seen that. Some are a little over-eager... Hangouts for example basically continues as Google Meet.
What I wonder is how many startups were effectively killed by these dead products, either by destroying a business model and chances of getting investors or through acqui-hiring the best talent?
I didn't say there was zero value, my statement also came with the assumption that upon starting this service, google never made it clear it would be temporary. I may be wrong about that.
Also, not everyone can afford to re-host their assets, not everyone will be able to, some might not even realize it'll be lost.
The broken trust comes from the idea that google likely promoted this as something that will be a new ongoing community that's now being shutdown unceremoniously.
The lack of trust towards Google's products come not just from the services they're providing they shut down, but they try to build up communities then abandon them with no recourse.
It wasn't just the assets they hosted, it was the people they were trying to bring together that are now scattered again because they trusted google.
"google never made it clear it would be temporary"
Why would you ever assume any service is permanent?
"The broken trust comes from the idea that google likely promoted this as something that will be a new ongoing community"
Anyone who "trusted" that a free service would work forever is being unrealistic.
"they try to build up communities then abandon them with no recourse"
They're hosting open source content. You can fork that. Anyone can fork it. If your content were hosted on Poly, you could use Takeout to download all of it in one go.
"that are now scattered again because they trusted google"
If there's any blame here, I would think it would be that there's not a group of people (like us at HN) who work closely with sites like Poly to back them up before they are shut down. Once we have a track record, we can ask Google to say not just, "We're shutting Poly down," but "and it's moving to Hacker News Back Things Up .com."
If someone wants to bridge that gap, moving this content to the Internet Archive or something, that sounds worthwhile.
Didn't really know about this platform, but to me it looks like it could perfectly be kept in some kind of read-only mode without costing all that much for Google.
Google Poly looks like a fun experiment that unfortunately didn't pan out to their expectations. Google APIs/GCP have paying users that businesses depend on. It wouldn't be fair to have them in the same room.
I was about to ask if TiltBrush relied on this. Thanks for confirming. It's hard to buy anything made by Google that hasn't been around for close to a decade, unless you don't care whether or not that product stays around.
I mean, it's a language and open source. If the community really hates the direction Google takes it, they can fork it and support whatever features they want.
I mean, that's what you get when the internal metrics are skewed towards shipping new products, paying no attention to maintaining or improving existing ones.
Yet another Google service which I only know about by reading the news that it is shutting down. This one looks pretty great. Would be nice to see announcements when those things are launching, not when they are shutting down ;)
Every time Google is shutting down a product there are the same kind of reactions like “I will not trust google running anything more than a year”, “killedbygoogle.com”.
I’m wondering, have you ever heard about this product? I for one have no clue what this is. How many more products does google have that you never heard about and that are not shut down?
Counterpoint: since Google has firmly established a reputation for killing their own products, people who would otherwise find value in certain Google products (beyond search ads gmail and android) avoid them and never hear about them to begin with.
I wonder if they are creating a self-fulfilling prophecy: google launches a product, people avoid it assuming it’ll get killed off, google says the user metrics don’t justify the investment, and they kill it.
I think HN tends to overplay google's product discontinuations.
If you ask someone who doesn't read HN if they've ever been frustrated with Google shutting down a product, my guess is 95%+ of people wouldn't be able to remember a single product that was shut down that they used.
One solution for Google might be to brand the products without the "Google" name prefix, until Google is absolutely sure it will be around for a decade+.
True enough, but this is among the largest gatherings of technology decision makers (at varying levels, presumably some very high level) out there and this comes up every thread about Google.
Hosting some files to serve as a counterpoint would not be totally unjustifiable from a marketing perspective.
If you ask someone who doesn't read HN if they've ever been frustrated with Google shutting down a product, my guess is 95%+ of people wouldn't be able to remember a single product that was shut down that they used.
That is true, but it ignores the fact that at least 50% of Google's product offerings are aimed at the sort of people who read HN.
Maybe. But even in gaming circles, people often cite "Google may discontinue it" as one of the main reasons for avoiding Stadia, which is a Google gaming service. I believe their reputation goes well beyond the HN tech circle.
You might well be right with your 95 per cent guess, but it
has proved frustrating and even embarrassing for me over the years to have recruited friends, family, colleagues and associates to some Google products/services only for the rug to be pulled later from under those who jumped aboard on my recommendation.
This has happened with Picassa, with Google+ (which had great integrations with other G services, and features I liked - such as one that let you group people into, say separately addressable units such as friends, family, colleagues, associates etc) and also the original Google chat (still around, but functionally degraded and headed for the chop, I fear).
I've loved some of these because of their features/usability. To me Google+ having few users and not having content pushed in my face was a bonus. Text chatting happily while in a Gmail tab (usually open when I'm working) but with minimal distraction and not having to context-switch - suits me. And now a lovely library of free 3D assets is for the chop.
Once upon a time I earned a little kudos for having my finger on the pulse and spotting decent emerging tech/trends. I was laughed at by many when, in 1994, I switched from a senior role in print news to one in the nascent online space - but many people took an interest because of my recommendation or at my urging.
I am no seer, I am rarely an evangelist, but I hate it that many people I know have invested time/effort in Google stuff on my recommendation. I feel as though I have let them down.
I know Google Poly since I'm in game dev (but as a hobby these days, so all the art has to come from my non-3d-artist self or places like Google Poly)
The fact that they won't just freeze uploads and leave it up is just stupid.
People always bring up "What about maintenance costs in terms of dev time!", but at the core this is literally static assets. No fancy API, no crazy isomorphic app or something needing fat servers.
Just 3D modeling files.
Literally they could generate a static page for each asset that uses the preview image instead of the 3d viewer, stick it in their cloud storage solution, and have it exist, untouched, in perpetuity.
The cost wouldn't even register on a microscopic scale for a place like Google
The reason google kills everything is because they do everything via monorepo. So if they don't have enough devs to literally -keep a service up to date with the currently evolving stack- they have to kill the project. This is why things get killed so easily.
Makes sense to me from a "why canyt you just leave it on" standpoint. But it still makes me very leeery of trusting any project of theirs long term (besides search/gmail).
This. Spent almost 3 years at Google and my view is that it is an intended side benefit of how they do development.
If something that relies on say, one of their log services, and that log service is getting end-of-lifed then all the dependent services that aren't really critical will also get shutdown. It's a forcing mechanism to keep the amount of cruft down. If that service that relies on the old service is that important they'll put the resources on the project long enough to port it to the new service.
This makes sense for engineers but is detrimental from a user point of view. Google effectively taking what would be a few days of engineering time to update some API calls and instead amplifying it and passing it on to users.
When considering whether to learn/migrate to a Google service, everyone should consider this ratio at which Google values your time vs. that of their engineers.
That's the distinction I made above, this isn't like a traditional "app" where the core of it's functionality is dynamic
It doesn't need to stay up to date with an evolving stack
Static files in Google's equivalent of an S3 bucket with public downloads enabled, that's it.
By design these types of blob storage services "just work" as a consumer indefinitely (otherwise there wouldn't be much of a point to them)
-
I get the whole "who does it get billed to, who's going to take a day to write this script that makes the preview page" deal, but like, this is literally the one time where just a little bit of initiative can win some brownie points with minimal investment, can't they do it?
The right thing to do by Google would be to make the product available as a stand-alone application, so users can still use it even if it doesn't run on Google's infrastructure.
More work, yes. But certainly much better than what they are doing now.
Not to mention that its been built on top of the Google internal development stack and getting it working outside of that environment would require a substantial re-write.
How is it vendor lock-in if the internal code base that was never built in a way to work in the outside world can't just be taken and dropped on Github?
> How would they do that? Who would pay the server costs?
The user would run the server, or another company, or a group of enthousiasts. It doesn't need to be expensive, especially if the user doesn't require many other users on the system.
That doesn't really make sense. poly.google.com looks to be essentially a repository/library of 3D renderings people want to share. There is no "standalone app" here, it's just a sharing site.
> The right thing to do by Google would be to make the product available as a stand-alone application, so users can still use it even if it doesn't run on Google's infrastructure.
Did any of your companies ever do that? How did it go?
Companies discontinue products all the time due to acquisitions, bankruptcy, restructuring or simply re-orienting investment and I've never seen them held to this standard of having to host it somewhere external. I'm not even really sure how that would work.
It's definitely a funny one. I hadn't heard of it but it appears to be an Instagram alike for 3D models and VR. If Google was trying to establish themselves as a first mover platform in the VR space, it's odd that they would can such a product before VR itself has finished taking off. It signals that they actually have no interests in VR going forward but it's Google so you never know, they will probably have 4 of these services at some point in the future and you'll never know when or where one starts and stops. For an advertising company, Google is remarkably bad at advertising their own products.
Google has already shown that they're willing to try: they've made Tango (AR, discontinued), Google Glass (AR, enterprise-only), ARCore (AR, active), Cardboard (VR, discontinued), and Daydream (VR, discontinued).
These resemble portfolio hedging bets on potential future paths of technology playforms that could affect their ad biz. Google seems to buy a spread of small bets with several-year expiry and high convexity (i.e. big payout if equilibrium shifts in that direction). I suspect a number of these come out of discussions anticipating what could be next after mobile.
I’ve never been able to understand their VR approach. They actually brought ~$500 mobile VR daydream headsets to market (with their HW partners) with cameras doing inside-out positional tracking but no tracked controllers. Everyone I talked to in the VR scene at the time (at conferences etc) that had tried some HW knew 3D input was critical for UX and the near term future and this was a dead end. But Google puts a ton of effort into launching an obviously not-yet-viable product, then after doing the most technically difficult part doesn’t follow through on the last step and lets FB eat their lunch.
I could understand putting in less effort, or more effort, but failing the way they did was mind boggling and just seems like lack of a strategy on the management level.
Given what I know about Google culture, which has also been brought up in other threads here, this isn't that surprising. I think because of the extent to which their promotion process relies on demonstrating technical achievement, it's hard for them to get enough people to work on things if they aren't technically challenging.
Yep, I hadn't heard of this one. Several of their shutdowns have effectively been product announcements and shutdowns rolled into one for me.
They announce a shutdown, I don't know what it is so I check it out. I get excited for a brief moment until I remember it's getting shutdown.
I've not only heard of it, I use it frequently. It's by far the easiest way to access permissively licensed (Creative Commons attribution mostly) content for making things in WebGL. Given the fact that Poly is essentially a very basic CRUD app with simple files attached there's can't really be a good reason to shut it down - it's simple, mature, and can't take much more than a few hours of maintenance a month for a single developer.
Either Google has a real problem finding developers capable of maintaining things, or their decisions to keep things running are based purely on cash with no regard for either users or their own reputation as a company that can keep services running.
They have a real problem finding developers that will want to work on it. You don't get promoted at Google for maintaining a product that isn't growing.
It's strange that getting promoted is such a universal goal. Many or most people who start on the promotion ladder get stuck in the miserable middle. There's surprisingly little reflection on whether or not the extra money is worth the extra hassle, or at what point one should avoid further promotion.
I'm sure that's what Google developers tell themselves to feel better about being unable to keep something running for more than a couple of years, but the fact is Google could easily recruit developers to maintain things. They pay well, they make things people have heard of, and not everyone is insanely ambitious that they actively refuse to work on things that won't further their career.
This is an HR problem. Two HR problems really, as exemplified by both your sentences.
> They have a real problem finding developers that will want to work on it.
At most companies, you don't get to choose projects you want to work on. You get assigned a project whether you like it or not. Almost every other company functions like this, Google's insistence on being different is shooting them in the foot.
> You don't get promoted at Google for maintaining a product that isn't growing.
Maybe Google should be selecting for people who aren't interested in chasing promotions. There are plenty of 'workman' developers in the world who are okay with spending the rest of their careers maintaining legacy systems. This probably ties into good ol' Silicon Valley age discrimination: they spend so much time chasing young hotshots with chips on their shoulders, while most of the workman developers are in their 50s.
You're looking at it from the wrong angle. It's not that an engineer says no but rather, other projects that are growing and have more resources/weight behind them come calling and looking for engineers that want to transfer into their project. A stalled/non-growing project can't block a transfer like that.
If Google really wanted to solve this problem they could by making the promotion criteria different. Their actions say they don't really see it as a problem, though.
I've read a guess: As known, Google uses monorepo and own libraries, so they need to maintain all applications to keep up to date to work with libraries. That's why Google kills minor but worth services.
Duo is pretty big. It's pre-installed on every single android phone, and on Samsung phones it is the default video calling app too.
Also, you cannot remove it. You can uninstall the icon, but the underlying code is so deeply integrated into the OS that people can still call you and you will get a Duo video call even after uninstalling it.
That's interesting. I have a Fairphone 3 with a stock Android 10 and it does not come with the Duo pre-installed. The icon for video chats in Contacts app is greyed-out. After installing the Google Duo, it becomes active.
Is it possible to remove/deactivate it via adb or other tools?
Yes, it is a standard Play Store / Google Services on Fairphone. However, it comes with a very small amount of applications pre-installed. I was really surprised when I turned it on for the first time.
That's actually (at least recently) quite "popular" product I would say, because they included it in most Android smartphones to offer video calls. I am tempted to say Google Duo is spread on millions of devices, however it is kind of an app, which is rarely used at all by most of users. Also, there are thousands of competitors, where even Google Meet does feels better than Duo.
Yes, but what about the people who did? Here's the thing, Google is a big company. When a big company offers a service or product you expect a long shelf life so you can justify the time/money investment to use it. If you stumble across an open source or just some other random project by a single dev, you have an instinctive expectation that the project could die at any time. Thus, you invest your effort accordingly unless you're a hobbyist that enjoys the exploration (which is not most people).
Google has the focus of a 12 year adhd kid who just ate a whole birthday cake, unless when it comes to data collection. Google doesn't produce quality products anymore. They found their niche and pretend to do other things to try continuing that "do no evil" lie of a mission statement they used to have.
Here's the lesson, you can't trust google with anything. You base your personal or business infrastructure on Google you can expect one of two things:complete privacy invasion or they're going to destroy whatever you're using because they never took it seriously in the first place.
> Google has the focus of a 12 year adhd kid who just ate a whole birthday cake, unless when it comes to data collection.
A decade ago I was an intern at Google. One of my mentors said something that has stuck with me: "Google found a hose that money pours out of and it's name is online advertising. All we do now is desperately try to find another hose."
Google's strategy for a long time was: Hire every clever person you can; give them some creative freedom; see if any of them come up with a trillion dollar idea.
Only now they've 'grown up'. 20% time is dead. There is no second hose. They've mostly given up on the idea of a clever person finding a new money-hose. They just focus on the one they have now.
> They've mostly given up on the idea of a clever person finding a new money-hose. They just focus on the one they have now.
The financial numbers suggest they have not given up and have succeeded incrementally towards other sources since. How else do you correct the above statement?
There's no posted evidence for the idea that Google only makes money on advertising, either. Can we have a discussion where we assume that everyone isn't an idiot who needs a link for every statement? Or do you seriously think advertising is the only way Google makes money? If so, sibling comment posted some nice links that one would do well to read.
They're publicly traded. Their revenue sources are open for the public to read. The other guy saying they're diversified linked to their quarterly to show 80% ad revenue. That's pretty much "they're an ad company".
A decade ago when the OP was an intern it was 100%! Non-ad revenue source have grown by infinity percent!
I made a very specific claim and I'm not sure what everyone's problem is. Google has found other sources of revenue and is actively trying to grow those sources. There are other hoses. I have never claimed it's not an ad company or that ads have somehow gone away. Because that would be dumb and factually wrong.
Being "well diversified" is subjective. But in freelance/contracting, the rule of thumb is to never have a customer account for more than 25% of your yearly income. You diversify your client base so when one walks away, it's never too big of a hit.
Ads make up 80% of their revenue. That's not diversified. That's an ad company that dabbles in other things. If Shell or BP's revenue was 80% oil and 20% solar, wouldn't you still call them oil companies that dabble in solar?
If it constitutes significantly less than 20% of their revenue, it's not a "hose that pours money out". So no, Google has not yet found a second hose, even if they've found other revenue streams.
That $36B doesn't come from a single source. Google Cloud is their single largest non-ad revenue stream, at $3.4B in Q3 (or 7.4% of total revenue). The rest of their non-advertising revenue ($5.5B, or 11.8% of total revenue) is lumped into the "other" category, which implies that they have a lot of smaller revenue streams that add up to a larger figure—not a single hose. So I guess you could maybe call Google Cloud a "hose" at ~$14B per year, but it certainly doesn't print money like their advertising business, and is a relatively small source of revenue for such a large company.
Everything is relative. Back then, for Google's size, their ad business was basically printing money. Today, given their size, Google Cloud and the rest of their non-ad businesses are barely significant—over 80% of their total revenue still comes from ads, and that is only very slowly shifting. Maybe in another 13 years, Google Cloud (or another venture) will catch up to the ad business (though I'm skeptical), but today, that's nowhere near being true, and they still largely depend on that single source of revenue. So are they more diversified than they used to be? Sure. But to say that they have found a second hose that's anywhere near as profitable as advertising is misleading at best, or blatantly false at worst, at least as it stands today.
I'd like to see one for operating income, though, because I know AWS has a higher operating income than Amazon's Retail arm, even though Retail has a much higher revenue.
> There is no second hose. They've mostly given up on the idea of a clever person finding a new money-hose.
I think they are quite shortsighted. The idea of running Google as a conglomerate, turning it into Alphabet, was good. But somehow, they have not been able to get rid of those layers of middle management that kill innovation. I've heard that Calico and other ventures are really political, and lot's of people that moved there hoping to do great work are quite frustrated.
Which is a shame, because they do have know how and technology to achieve great things. Just out of the protein folding results of this week you could spin off dozens of highly successful startups.
>20% time is dead. There is no second hose. They've mostly given up on the idea of a clever person finding a new money-hose. They just focus on the one they have now.
Isn't that how they've always operated? Gmail was launched in 2004 and that sure wasn't a money-hose. But it did serve as yet another platform to support their money-hose since they could scan your emails and serve you ads.
It's really only a recent change to actually monetise Gmail beyond ads. They sell it as part of G Suite and as part of Google One. Not sure if Google One brings in the dough but surely G Suite makes enough to justify its existence.
>When a big company offers a service or product you expect a long shelf life so you can justify the time/money investment to use it.
I bought a Chromebook in 2014. It still works, but support was discontinued earlier this year. That sounds okay to me. I wonder if Poly was popular? I've never heard about it 'til today.
> If you stumble across an open source or just some other random project by a single dev, you have an instinctive expectation that the project could die at any time.
Ironically at this point, I'd place more trust on an open-source project because at the very least, if the maintainer abandons it, I could fork it and potentially self-host.
Alternatively, it's time to adjust expectations and understand that if a product or service doesn't hit the growth curve Google expects, they sunset it.
A lot of businesses would be more successful if they learned how to do this across their own products and features. Instead engineers drown in KTLO, adding to cognitive load not only for employees but customers too.
Yes, I know it and browse it frequently - a great resource, if you've an interest in or need for this type of 3D asset. (I used to run a 3D animation software company.)
It's about branding and trust. Before any product is a popular one that prints money it's a small one looking for early adopters.
When those first adopters are deciding whether or not use the new product, part of that equation is what happens if this doesn't catch on? Should I trust them to jump on board now, wait, or go with someone else?
For instance if someone was choosing betwen two similar and new products from Microsoft and Google one factor weighing in Microsoft's favor is they will care about support, and an exit plan even if the product doesn't take off.
Looking at the frequency of the art uploaded and the effort in it I bet the artists and those who used that platform would indeed care about it being killed by Google.
The problem is that if Google holds this attitude, that means I as a user must absolutely stay away from anything that isn't promoted extremely heavily by Google.
And if I and everyone else thinks this way, as Google is conditioning them to, then no one will ever adopt Google's non heavily promoted products, which will make their failure a self fulfilling prophecy.
I loved Google poly! It was a great place to find models for A-frame scenes. A-frame is VR basically built in HTML. I really liked it as a teaching tool especially for kids. Like learn HTML elements by finding a bear on Google poly and putting it in your scene.
It was a little frustrating in some places since a lot of it was oriented around their VR drawing tool (Tilt Brush) which my system didn't support.
Honestly the fact that I haven't heard of it and those comments are probably related. My trust in google to maintain things is so underwater that I really don't feel like dedicating the attention to even learn their new product offerings. For me it's all in one ear and out the other.
There's a phenomenon I've witnessed in dev teams that I always thought was a separate issue but I'm starting to question that conclusion.
Like with Covid, like with SLAs, it's not enough to think of ratios of success or failure. You have to also look at frequency and duration.
A flaky test might fail a build 1 time in 200, but as your team gets bigger, build frequency rises, your tests grow, and eventually you're getting failed builds frequently enough that people start to see them as a regular occurrence, and that negatively affects their opinions about the whole experience. I've seen people bash the system when failures happen weekly, I've seen others 'turn' after a couple of statistical clusters and then fall to confirmation bias long afterward.
Google has so many irons in the fire that I think we've reached that same threshold for a lot of people. Shutting down the worst 1% of your projects a year sounds like a completely reasonable business plan. Until you have 1000 projects, and now you're shutting one down every five weeks on average. People will talk.
And if there's no transparency in that process, how do I know that my favorite tool isn't next, or on the list for next year? Odds are low, but not zero.
I had never heard of this thing before. I know there is this running joke of Google killing their stuff all the time, but do they just suck at launching, or at least promoting, services?
200 comments
[ 0.27 ms ] story [ 249 ms ] threadI would try switching to spending the money on acquisitions.
Even if they can’t do that every time, it should at least be one of the options under consideration.
I feel like there’s often an argument that “keeping it running costs more than you think, because it would need updating” (or “it depends on legacy infrastructure that’s going away”), but I’m skeptical of that in the general case. Other than possibly security issues, there’s often little to get in the way of servers running happily with little or no maintenance for years.
Do you know whether it was considered?
> Other than possibly security issues, there’s often little to get in the way of servers running happily with little or no maintenance for years.
Security costs alone can be substantial. Whose department should take the budget hit for maintaining a no longer viable project?
Whose reputation should be dinged for running a non viable product?
In this particular case, I don’t know, although the lack of any positive examples to point to suggests that it’s not an option that’s typically given any weight in decision-making.
Whose department should take the budget hit for maintaining a no longer viable project?
Speaking as a user, I don’t care. Isn’t Google swimming in profit?
(I realize of course that you don’t achieve the kind of success in the first place that brings those colossal profits without making hard-nosed decisions along the way -- arguably)
Whose reputation should be dinged for running a non viable product?
Again, speaking as a user, I’d put value in keeping the “non-viable” product around, if it were one I happened to use.
I’m not expecting my answer to change anything, but it’s a general answer to a general question. What could they do differently? Answer: eat the loss.
The question is whether the products started in 2021 will last until 2022...
It would be far easier to choose Google services (and less irritating as a consumer) if services were labelled alpha, beta, production, and LTS, and if each of those labels had a specific duration (e.g. alpha services can vanish immediately without warning; beta have a limited lifecycle where they need to become profitable or at least self-sustaining, production are generally stable and can be relied on for a year, and LTS will be supported for a specific period of time, backed by enterprise support agreements).
But that's just crazy talk!
I had heard Google is unreliable in this aspect before but after this particular experience I will unlikely use or suggest Google for anything serious.
The cost of the experimentation is people wondering what product they'll discontinue next and whether it'll be one that _they_ rely on. If you find yourself asking this often enough, you'll probably start moving out of the Google ecosystem. How is this good for Google?
Yes, I understand that products need to pay for themselves, but Google keeps drawing people into services that they then shut down. It's frustrating for users and damaging to Google. What's the cost of the damage to their reputation?
However, Google share price and market saturation tell me the whatever damage has been done is worth no follow up from them.
I'd advocate for the latter as a healthier way to run society with fewer tech giants of lovecraftian size slipping their tentacles into every cranny that one can stuff a million dollars into.
> Your assets will be available at poly.google.com until June 30, 2021
I doubt they will in future.
Does Google have any more space in their graveyard?
Someday, the Google graveyard will look like this: http://www.worldwartours.be/files/img_9456_1.jpg
What I wonder is how many startups were effectively killed by these dead products, either by destroying a business model and chances of getting investors or through acqui-hiring the best talent?
Well, I guess that was a mistake for whoever did.
That almost reads like a hint of self awareness for their continued killing of things they release.
There is value. Yes, there's also cost to having to re-host your assets somewhere else.
Also, not everyone can afford to re-host their assets, not everyone will be able to, some might not even realize it'll be lost.
The broken trust comes from the idea that google likely promoted this as something that will be a new ongoing community that's now being shutdown unceremoniously.
The lack of trust towards Google's products come not just from the services they're providing they shut down, but they try to build up communities then abandon them with no recourse.
It wasn't just the assets they hosted, it was the people they were trying to bring together that are now scattered again because they trusted google.
Why would you ever assume any service is permanent?
"The broken trust comes from the idea that google likely promoted this as something that will be a new ongoing community"
Anyone who "trusted" that a free service would work forever is being unrealistic.
"they try to build up communities then abandon them with no recourse"
They're hosting open source content. You can fork that. Anyone can fork it. If your content were hosted on Poly, you could use Takeout to download all of it in one go.
"that are now scattered again because they trusted google"
If there's any blame here, I would think it would be that there's not a group of people (like us at HN) who work closely with sites like Poly to back them up before they are shut down. Once we have a track record, we can ask Google to say not just, "We're shutting Poly down," but "and it's moving to Hacker News Back Things Up .com."
If someone wants to bridge that gap, moving this content to the Internet Archive or something, that sounds worthwhile.
There was Poly.
Titlbrush could export / import to Poly. That was added after the launch of TiltBrush, yes?
Sure, if you bought TiltBrush after Poly integration was added, I could see how that's taking away a feature you used, and that's a bummer.
(Go ahead and downvote me a ton)
PHP, Python, JavaScript, Ruby, Go, Java, Swift, Obj-C, Rust…
Welcome to capitalism, friend.
At least you're paying for GCP.
I’m wondering, have you ever heard about this product? I for one have no clue what this is. How many more products does google have that you never heard about and that are not shut down?
I wonder if they are creating a self-fulfilling prophecy: google launches a product, people avoid it assuming it’ll get killed off, google says the user metrics don’t justify the investment, and they kill it.
If you ask someone who doesn't read HN if they've ever been frustrated with Google shutting down a product, my guess is 95%+ of people wouldn't be able to remember a single product that was shut down that they used.
One solution for Google might be to brand the products without the "Google" name prefix, until Google is absolutely sure it will be around for a decade+.
They ended Music, and bait and switch their photo policy.
Traumatic events for me. I've left both services, won't get another pixel, and attempted to switch to Firefox.
Hosting some files to serve as a counterpoint would not be totally unjustifiable from a marketing perspective.
That is true, but it ignores the fact that at least 50% of Google's product offerings are aimed at the sort of people who read HN.
This has happened with Picassa, with Google+ (which had great integrations with other G services, and features I liked - such as one that let you group people into, say separately addressable units such as friends, family, colleagues, associates etc) and also the original Google chat (still around, but functionally degraded and headed for the chop, I fear).
I've loved some of these because of their features/usability. To me Google+ having few users and not having content pushed in my face was a bonus. Text chatting happily while in a Gmail tab (usually open when I'm working) but with minimal distraction and not having to context-switch - suits me. And now a lovely library of free 3D assets is for the chop.
Once upon a time I earned a little kudos for having my finger on the pulse and spotting decent emerging tech/trends. I was laughed at by many when, in 1994, I switched from a senior role in print news to one in the nascent online space - but many people took an interest because of my recommendation or at my urging.
I am no seer, I am rarely an evangelist, but I hate it that many people I know have invested time/effort in Google stuff on my recommendation. I feel as though I have let them down.
The fact that they won't just freeze uploads and leave it up is just stupid.
People always bring up "What about maintenance costs in terms of dev time!", but at the core this is literally static assets. No fancy API, no crazy isomorphic app or something needing fat servers.
Just 3D modeling files.
Literally they could generate a static page for each asset that uses the preview image instead of the 3d viewer, stick it in their cloud storage solution, and have it exist, untouched, in perpetuity.
The cost wouldn't even register on a microscopic scale for a place like Google
The reason google kills everything is because they do everything via monorepo. So if they don't have enough devs to literally -keep a service up to date with the currently evolving stack- they have to kill the project. This is why things get killed so easily.
Makes sense to me from a "why canyt you just leave it on" standpoint. But it still makes me very leeery of trusting any project of theirs long term (besides search/gmail).
If something that relies on say, one of their log services, and that log service is getting end-of-lifed then all the dependent services that aren't really critical will also get shutdown. It's a forcing mechanism to keep the amount of cruft down. If that service that relies on the old service is that important they'll put the resources on the project long enough to port it to the new service.
When considering whether to learn/migrate to a Google service, everyone should consider this ratio at which Google values your time vs. that of their engineers.
It doesn't need to stay up to date with an evolving stack
Static files in Google's equivalent of an S3 bucket with public downloads enabled, that's it.
By design these types of blob storage services "just work" as a consumer indefinitely (otherwise there wouldn't be much of a point to them)
-
I get the whole "who does it get billed to, who's going to take a day to write this script that makes the preview page" deal, but like, this is literally the one time where just a little bit of initiative can win some brownie points with minimal investment, can't they do it?
More work, yes. But certainly much better than what they are doing now.
How would they do that? Who would pay the server costs?
Edit: Out of curiosity, is there anyone reading this comment that would be willing to pay for the infrastructure costs to keep Poly alive?
The user would run the server, or another company, or a group of enthousiasts. It doesn't need to be expensive, especially if the user doesn't require many other users on the system.
Did any of your companies ever do that? How did it go?
A commenter here shared an example of a similar product shut down by Microsoft: https://en.wikipedia.org/wiki/Remix_3D
Disclaimer: Googler, but in no way connect to the Poly team.
Edit: Added Remix 3D example.
I could understand putting in less effort, or more effort, but failing the way they did was mind boggling and just seems like lack of a strategy on the management level.
Either Google has a real problem finding developers capable of maintaining things, or their decisions to keep things running are based purely on cash with no regard for either users or their own reputation as a company that can keep services running.
> They have a real problem finding developers that will want to work on it.
At most companies, you don't get to choose projects you want to work on. You get assigned a project whether you like it or not. Almost every other company functions like this, Google's insistence on being different is shooting them in the foot.
> You don't get promoted at Google for maintaining a product that isn't growing.
Maybe Google should be selecting for people who aren't interested in chasing promotions. There are plenty of 'workman' developers in the world who are okay with spending the rest of their careers maintaining legacy systems. This probably ties into good ol' Silicon Valley age discrimination: they spend so much time chasing young hotshots with chips on their shoulders, while most of the workman developers are in their 50s.
If Google really wanted to solve this problem they could by making the promotion criteria different. Their actions say they don't really see it as a problem, though.
Also, you cannot remove it. You can uninstall the icon, but the underlying code is so deeply integrated into the OS that people can still call you and you will get a Duo video call even after uninstalling it.
Quite a menace.
Is it possible to remove/deactivate it via adb or other tools?
I haven't tried, I don't mess with my phone too much because I don't want to trip the Knox fuse.
I'm surprised an Android 10 phone does not have Duo pre-installed. Does it have the standard Play store and associated google apps?
Duo has 1 billion+ installs in play store, which basically means they pre-installed it on every phone.
https://play.google.com/store/apps/details?id=com.google.and...
Google has the focus of a 12 year adhd kid who just ate a whole birthday cake, unless when it comes to data collection. Google doesn't produce quality products anymore. They found their niche and pretend to do other things to try continuing that "do no evil" lie of a mission statement they used to have.
Here's the lesson, you can't trust google with anything. You base your personal or business infrastructure on Google you can expect one of two things:complete privacy invasion or they're going to destroy whatever you're using because they never took it seriously in the first place.
A decade ago I was an intern at Google. One of my mentors said something that has stuck with me: "Google found a hose that money pours out of and it's name is online advertising. All we do now is desperately try to find another hose."
Google's strategy for a long time was: Hire every clever person you can; give them some creative freedom; see if any of them come up with a trillion dollar idea.
Only now they've 'grown up'. 20% time is dead. There is no second hose. They've mostly given up on the idea of a clever person finding a new money-hose. They just focus on the one they have now.
https://abc.xyz/investor/static/pdf/20201030_alphabet_10Q.pd...
Q3 2019, ads were 84% of revenue Q3 2020, ads were 81% of revenue
It's not a massive change, but it's changing and ads are at the lowest percentage of overall revenue they've ever been.
"More diversified than it's ever been" can be factually true here, but for the purposes of this discussion was highly inaccurate.
> They've mostly given up on the idea of a clever person finding a new money-hose. They just focus on the one they have now.
The financial numbers suggest they have not given up and have succeeded incrementally towards other sources since. How else do you correct the above statement?
Growing from $1B to $10B is really difficult, unless you have virtually no competition.
I made a very specific claim and I'm not sure what everyone's problem is. Google has found other sources of revenue and is actively trying to grow those sources. There are other hoses. I have never claimed it's not an ad company or that ads have somehow gone away. Because that would be dumb and factually wrong.
Ads make up 80% of their revenue. That's not diversified. That's an ad company that dabbles in other things. If Shell or BP's revenue was 80% oil and 20% solar, wouldn't you still call them oil companies that dabble in solar?
My statement is that yes, there is a second hose.
I've never said they're not primarily an ad company.
What's the threshold for something being a "hose"? A $36B business is a pretty big business!
https://en.wikipedia.org/wiki/List_of_largest_companies_in_t...
I'd like to see one for operating income, though, because I know AWS has a higher operating income than Amazon's Retail arm, even though Retail has a much higher revenue.
I think they are quite shortsighted. The idea of running Google as a conglomerate, turning it into Alphabet, was good. But somehow, they have not been able to get rid of those layers of middle management that kill innovation. I've heard that Calico and other ventures are really political, and lot's of people that moved there hoping to do great work are quite frustrated.
Which is a shame, because they do have know how and technology to achieve great things. Just out of the protein folding results of this week you could spin off dozens of highly successful startups.
Isn't that how they've always operated? Gmail was launched in 2004 and that sure wasn't a money-hose. But it did serve as yet another platform to support their money-hose since they could scan your emails and serve you ads.
It's really only a recent change to actually monetise Gmail beyond ads. They sell it as part of G Suite and as part of Google One. Not sure if Google One brings in the dough but surely G Suite makes enough to justify its existence.
I bought a Chromebook in 2014. It still works, but support was discontinued earlier this year. That sounds okay to me. I wonder if Poly was popular? I've never heard about it 'til today.
Ironically at this point, I'd place more trust on an open-source project because at the very least, if the maintainer abandons it, I could fork it and potentially self-host.
A lot of businesses would be more successful if they learned how to do this across their own products and features. Instead engineers drown in KTLO, adding to cognitive load not only for employees but customers too.
When those first adopters are deciding whether or not use the new product, part of that equation is what happens if this doesn't catch on? Should I trust them to jump on board now, wait, or go with someone else?
For instance if someone was choosing betwen two similar and new products from Microsoft and Google one factor weighing in Microsoft's favor is they will care about support, and an exit plan even if the product doesn't take off.
And if I and everyone else thinks this way, as Google is conditioning them to, then no one will ever adopt Google's non heavily promoted products, which will make their failure a self fulfilling prophecy.
It was a little frustrating in some places since a lot of it was oriented around their VR drawing tool (Tilt Brush) which my system didn't support.
Like with Covid, like with SLAs, it's not enough to think of ratios of success or failure. You have to also look at frequency and duration.
A flaky test might fail a build 1 time in 200, but as your team gets bigger, build frequency rises, your tests grow, and eventually you're getting failed builds frequently enough that people start to see them as a regular occurrence, and that negatively affects their opinions about the whole experience. I've seen people bash the system when failures happen weekly, I've seen others 'turn' after a couple of statistical clusters and then fall to confirmation bias long afterward.
Google has so many irons in the fire that I think we've reached that same threshold for a lot of people. Shutting down the worst 1% of your projects a year sounds like a completely reasonable business plan. Until you have 1000 projects, and now you're shutting one down every five weeks on average. People will talk.
And if there's no transparency in that process, how do I know that my favorite tool isn't next, or on the list for next year? Odds are low, but not zero.