It is a little unfair, I agree. And I think it's all too easy to be snarky about other peoples hard work (haters gotta hate), especially when it doesn't work out.
At the same time, Google seems to want it both ways. They have an idea and use their size to attract users and, perhaps even unintentionally, effectively block others from implementing the idea or scare others away (a size thing). Then, if they decide they don't like it, which can feel arbitrary, they pull out. Now the idea is tainted so others are less likely to implement it and users are left wondering what to do. They really seem like a bull in a china shop sometimes.
I like the innovation and the experimentation, but it seems like they could be more graceful about it for both users and other entities trying the same things. Perhaps they should try these experiments in a more stealth way? Pretend not to be Google? Of course, people might super-hate on them for that, huh?
Totally agree on wanting it both ways. When you associate a product with a brand, you are attempting to overcome objections by saying "This brand supports this product and is putting its reputation behind it."
If Google wants to experiment, create projects that don't have the Google brand (they just use G+ login like any other startup) and then reveal it was Google all along.
Let the projects succeed on their own merits, with their own marketing. If you tie in the Google brand you are setting expectations that this will be a long-lasting effort.
To me, this is evidence of Google's rather awkward relationship with consumers (i.e. they toss tons of branded endeavors at us, see what sticks), and not a knock against fearlessly trying things or whatever. The best brands are coy and careful with their identity... google sorta slops it around. this is a list of failed consumer facing products more than it is a list of "bad ideas".
I don't really see how it is awkward or hurts relationships with consumers (especially when most all products and services are free) -- it's a form of market research and product testing. Who needs focus groups when you have the world?
I think focus groups are inherently embarrassing. Not commenting on their utility, but the act of showcasing something to "normal people" for their "feedback" is, at its essence, pathetic.
While Google's 'everything in beta' mentality is endearing in a 'democracy'/'transparent fun-loving engineers' sort of way... it is also evidence to the fact that our information behemoth is rudderless.
Google Desktop Search - even from a cursory glance at the comments in this thread and the absence of its mention - is by far the most under-appreciated Google product ever.
That sucker was good. Great even.
It would find things across browsers, through documents and files and even collate
results from web searches that most browsers - sometimes even Chrome to this day - missed or never registered in the history, in the first place.
I would have gladly paid some $50-$75 a year to have it work across all of my devices.
And I don't typically pay for most offerings. It was that good.
Google desktop search was great. At the high of the desktop search engine race, it competed with OSX Spotlight and Windows desktop search (and various others like Yahoo X11, Copernic desktop search). Google worked on 3 major OS and run on a local web server. It looked like a local Google search and indexed various file formats incl. iFilter support.
As basically all companies stopped further developing desktop search products, I needed a proper replacement. I developed an enterprise/desktop search engine last year that basically reimplements Google desktop search including advanced search syntax and support for various file formats.
I developed an enterprise/desktop search engine last year that basically reimplements Google desktop search including advanced search syntax and support for various file formats.
If this were Reddit, this is the point where I would post the "SHUT UP AND TAKE MY MONEY" image.
Yeah, me too. It seems like desktop search as a product sector just kind of fell out of favor one day -- one moment there were a bunch of products scrapping it out, the next they were all gone. I have no idea why, I found it desktop search incredibly useful and would also happily have paid for a good product. But no really good ones exist anymore.
The even more mind-boggling thing is that for Google especially, desktop search would seem to be a huge opportunity. Imagine how precisely you could target ads to users if they let you scan their entire hard drive! Jeez louise.
Hm yet its leaky, buggy and low-feature. Useful for casual use but nowhere near professional. When will they drop it, or break it? Like some gmail features got broken this week.
Lesson: don't hang anything important on Google apps.
Google docs is fantastic. I use it exclusively (docs and sheets) and haven't seen anything buggy about it. I have never found myself a wanting a feature in google docs that wasn't also missing from Office. The only limitation I came across in sheets was the limited scope of conditional formatting. That's pretty much it.
FWIW there are a few features I would love to have in docs: different headers/footers for title pages pops out to me. Sheets is a LOT less featured then excel but the scripting is pretty excellent so that actually makes it a better product for most of what I want to do. Plus who really uses the advanced features of excel?
As you suggest, just because some idea was dropped doesn't mean that Google did the wrong thing. But, it also doesn't mean that they didn't abandon the idea, so it ends up here in a list. I don't see any general Google bashing yet.
Similarly, I'd put a phonograph in a museum, even though it's features have been replaced by other things.
I'm not a google apologist, but the list definitely has a few undeserved products. The most obvious is Writely, which was simply rebranded to Google Docs, then further rebranded to Google Drive. The writely.com domain even redirects to Drive.
I agree. Consumers can't lambast companies for not being innovative or daring enough, if they turn around and rip them a new one if they release anything other than something they're fully behind for 20 years, and either keep everything other than that to themselves or kill it before it gets to that stage.
Of course we can. It's just that companies are now so geared to quarterly roadmaps that nothing has time to gain momentum unless it becomes a phenomenon.
I see this with airlines as well; they launch a route in the middle of winter, or on a business route with schedules that are awkward for meetings, and then can it two months later because 'no-one uses it'. Which dissuades other airlines from trying a more sensible approach.
No, we really can't have it both ways. Obviously (clearly) they will keep products that have sufficient momentum.
With a little thought, however, it will be obvious to you that the consumer's definition of "sufficient" cannot possibly match a company's definition in 100% of cases: there will be a grey area where at least some people will like mouthwash-flavored bubble gum, but not enough people do that the company can afford to keep it. (And see if it does take off later, as you say would have happened in many cases.)
Consumers have two alternatives: either allow companies to test products like this without backlash, and accept that some products will be discontinued due to lack of adoption, or to have stagnancy with only upstart companies trying new things and the existing brands sitting on the bylines looking for success built elsewhere that they can copy.
I support innovation and encourage companies to go out on a limb, without any backlash if they discontinue these tests. (Even if I loved the product.)
You have this unstated assumption that they have to quickly abandon products that don't gain traction. But that's not true. It's short-term thinking over long-term.
When a company spends 30 man years developing a product, they can set aside 1 or 2 extra for long term maintenance. By promising or consistently providing this, they can build customer trust and get more users.
Keeping digital services up is cheap. It's shameful whenever DRM or video game servers go down after just a couple years because the company ran out of money or doesn't care anymore. These things could have been funded upfront with the tiniest fraction of the budget.
To add one more in the list, Google Talk eventually became Hangouts. The list states that it was abandoned in 2013, but it was actually still working properly at the end of 2014 when I last used it. I don't know if it works any more, since I eventually switched to hangouts (although I was reluctant to do so - https://twitter.com/StathisG/status/530740696719060992).
no, talk literally became hangouts. For the android and chrome apps at least, the change was pushed as an update to the existing talk app. There's differences between what it is now and what it was before, but it was an update.
Odd, I have a memory of my phone telling me that hangouts was a replacement for talk..
Not that it matters. Pushing a change "as an update" is a detail of how the app database works. It has no bearing on whether it's a replacement or not.
Wikipedia sure seems to think they're different things, and they love to merge articles.
Also if it's not a replacement they sure could have chosen better wording for: "Google Talk app for Windows will stop working on Feb 16 2015. It is replaced by the new Hangouts Chrome app."
> If the idea doesn't work they pivot and drop it.
Which is nice for Google, but screws over the users. If you're a user, you should be well aware that Google's products are unreliable and are there as an experiment.
Really? Can you explain more why. Is there an article accompanying it that shames Google or says anything bad about it. To me it just looks like a company that iterates and is trying new things.
> A lot of these projects ended up as other projects or built into other projects.
Well it would be bad if they completely threw away the code, and then hired a whole new group of people to work on a similar project. That would make them look worse.
Nah, the whole reason it exists, and is linked here is to try and pile on that Google is terrible and shuts down products that everyone in the world loves. But the reality is the minor minor subset of the world that is HN loved some of these projects and were upset no one else used them, thus they were shut down.
I haven't seen any accompanying MS articles like this, mainly because there haven't been angry posts here of MS shutting down products we think they should keep running forever that no one uses.
Have you not seen how much outrage appears whenever Microsoft change the Windows desktop?
Microsoft also gets the reverse treatment: criticism for "excessive" backwards compatibility^ and leaving old products in use that people want to kill (IE6).
^ Except when they were sabotaging Lotus, Quarterdeck, OS/2, and Samba back in the 90s.
I volunteer time for a project called The Shotbow Network - it's a Minecraft game network. We've got lots of projects, and there are features etc that we hype up and then either cancel or don't release.
Inevitably, though, everything that went into building either those mini-games or that feature for another game ends up in all of our future projects.
This sort of thing makes it look like those projects are completely dead, despite the fact that they live on either in full or in part under rebranded names or driving Google's next big feature.
Because not everyone here is the teenage girl version of a programmer who only goes after the newest fad without taking in account the actual features of a system and practical implications of promoting a monopoly.
RIP Google Code Search, aka the best code search ever. No credible competitor has yet emerged, and this is a real loss for all programmers: no more easy searching for sample code in obscure WIN32 Api, easy searching for arch-specific POSIX implementation details used by various open source projects, global error message grepping, and countless other little things which are not incredibly painful and difficult. I am still puzzled that we suffered such a tremendous regression.
I heard rumours that it was shut down because of license reasons. i.e. displaying the code to a user was making a copy in a legal sense, this has differing implications depending on the original software license.
Github search is fine, do not get me wrong. But this is like telling someone lamenting on the hypothetical end of Google search: "Just use the search function in Facebook, you don't need more"
I'm pretty sure Joel Spolsky and Jeff Atwood, and probably a lot of other github employees, lurk around here... if you harp on it enough they'll probably make it happen.
My primary use of Google Code search was, searching for this rare function (like a POSIX one) whose man page I can't understand and is largely undocumented elsewhere. I would often try to understand how other projects are using this function via internet-wide code search. :D
While Github is great, I think it still hosts only a small fraction of all the code that is public.
I was one of the (apparently few) people who used Google Sets a lot. Even though its results were often hit-or-miss, there were enough hits for it to be a valuable resource.
And unfortunately, I don't know of any comparable alternative. (But please tell me if you do!)
Here is the problem it solved:
It let you discover members of a set by entering a few members that you already know.
For instance, if you enter "terrier," "bulldog," "German shepherd," it might return, "beagle," "poodle," "chihuahua."
Obviously, in this example, it would probably be just as easy to Google "list dog breeds."
But what if you know a few members of a set ... but don't actually know what the set is, or the proper way to describe the set to a search engine?
Or what if the search engine results take you to pages that list some related terms, but you've got to sift through a lot of other content to get to them?
That's when Google Sets really came in handy.
Edit: I admit, Google has gotten a lot better in recent years at returning this sort of information in search results. And Wolfram does a good job, too. So my need for Sets has decreased, but I still miss it.
What were some real-world situations where you used Google Sets? It always seemed to me like a cool toy without much practical use, and certainly not the most worthy sink of a good engineer's limited time.
Oh, was that E_NOTIMPL, E_NOT_IMPLEMENTED, E_NOIMPL...? I forgot again. Google sets to the rescue! S_OK, E_FAIL, submit. I did that particular one at least 20 times.
I worry that a truly great technology has died with Sets: the way it would correctly intuit the scope of your set. Give it cities in Ohio, it spits out other Ohio cities. But give it London and Tokyo and you get all the major metropolises across the globe. Kudos to those who made it, and I hope they can do it again when applications for it arise.
Naming servers. Doesn't matter if you use colors, or star trek characters, or nautical terms: plug the existing servers into a set and get a couple new name suggestions.
This was my only valid use for sets too, I can kind of see why it wasn't a project worth serious time maintaining. I always thought it would eventually be integrated into Google Proper via a comma separated syntax or something.
I didn't say it's not possible to find a list of characters. Wikipedia has a list page for everything. But the specific claim was that one could plug set elements into a regular search to find more set elements.
Remember this before you rely too much on a Google product. Google isn't Microsoft and that's okay, but one of Microsoft's strong points is that they keep things around forever. Google still has the start-up culture where you can throw things away that aren't working. The difference is, in most start-ups, you don't have millions of people using the product you just threw away.
I doubt Google would kill Gmail outright but they might go the Twitter route and remove generic IMAP access. I would be surprised if IMAP was still an option after the next 2-3 years.
Google Wave was simply too ahead of it's time. It's funny to see how Slack and HipChat are so similar to it. I recall it's performance being poor and people being a little confused by how much it could do.
Google Answers was destroyed by Yahoo Answers - you could see the approach they were going for though - a properly answered question can be shown again and again. SO seemed to strike a balance between the two.
Google Wave was a collaborative editor. It was nothing like Slack or HipChat (unless I've managed to really misunderstand Slack despite building a few things on top of it). It's problem was that it didn't have a purpose - no one really knew what to do with it because a "Wave" was presented an amorphous blob of "knowledge" that people could contribute to at the same time. It was an online notepad, sketchbook, and spreadsheet all at the same time, yet it was also none of those things because you couldn't easily get the document out of Wave.
Wave was a textbook example of how you can fail if people don't understand why they should use your product. They gave up on it (by handing it to the Apache Foundation where it continues to be developed), rolled the clever collaboration technology into Google Docs, and told everyone they'd made it so you could edit a doc at the same time as someone else. Everyone continues to love it to this day.
Google Wave was an XMPP-based protocol¹. It allowed communication and collaboration with revisioning. It could have been embedded into applications such as office suites, giving you collaboration and communication right inside the app.
Wave.google.com was an example product using said protocol. Since the protocol was barely spoken of, even by developers, everyone assumed the website was the (only) product. Since the website did a lot of things, and none of it well, often being compared to email, it flopped and the entire project folded..
¹ Technically, the "Wave Federation Protocol", but anyone that really looked into Wave would know that the protocol was the most important element of the project.
It was certainly very like Slack in that it's like a IRC/Forum/IM combo that can integrate with third parties and display almost any content pics, images, vids etc with channels and private messaging for disseminating content by topic. You could describe Slack as a "collaborative" tool as well.
The purpose of Google Wave was basically to be a fancy do all messaging service. Many critics have compared Slack to Google Wave including Hunter Walk an ex YTer and Googler. You can see that Google Wave was probably even more powerful - so powerful that it looked daunting and confusing and people didn't know how to use it.
Google pitched it as a replacement for email (another reason for people being confused about it's purpose) but the reality was that it was absolutely nothing like email. A wave was a document that you could edit with other people.
Actually, looking at the screenshot on that page, I can see why s_dev sees it as something like Slack. It has a much more 'chat'-based interface than I remember. Sorry s_dev.
Yeah I always saw it as a mix between an instant message, a wiki page, and an email with real time collab.
I really loved Wave because for interactions that were not well defined (such as building a knowledge base) I believe you could start waves from parts of a wave, and playback how each component of the information came together in a form of primitive version control.
Simplifying the interface, inviting more people to the beta when it was originally out, and increasing the ability to federate easily were three big things that were missing from wave for more people to adopt it imo.
As a Google Answers verified researcher (gee, am I breaking my NDA by stating this?), and this is speaking from my perspective, the project closed/failed for reasons other than Yahoo!'s rise in the area.
GA worked quite well at first. There were some very active users that did excellent work and got paid quite nicely for their efforts. But a vague set of rule changes started to migrate in.
For example, a non-verified observer could comment on a question and pretty much give the entire answer. An official GA Researcher would answer the the question above the line, and the response from Google would be "Well, you didn't add any information to this topic than what was already on the page, so no payment for you". Riiiiight.
There were other little quirks and then eventually Google announced one day it was over. That's just how it rolled with them. I'm chalking it up to shiny things elsewhere that took attention away from GA. I did get a nice fleecy blanket as a Christmas gift one year. Not sure why.
The funny part is how for years we pretended Jabber was good and we couldn't understand why no-one was using it. There are a few critical pieces of functionality that you seemingly can't get with that protocol - being able to sign in from multiple devices and have messages sent to all of them, groupchats as a first-class thing (highlighting when you're mentioned but not otherwise), mobile support.
> Google Wave was simply too ahead of it's time. It's funny to see how Slack and HipChat are so similar to it. I recall it's performance being poor and people being a little confused by how much it could do.
That's somewhat true for certain technical aspects but it's too quick to ignore the mistakes Google made and minimize the value of work which more successful products put into the user experience.
Google never made the case for why using Wave would benefit you; as with Buzz, Plus, etc. Googlers reported liking it because they had both critical mass and pressure to use it long enough to see the benefits but few people outside did.
Instead, what we did hear about a lot was the technology which was going to change the world right up until it failed. This is a common failure mode for Google where the strong engineering culture crowds out UX, forgetting that for almost any product that's an implementation detail which most people shouldn't need to care about.
Wave tried to be a lot of things -- a social network, a collaborative editor, and a chat client at once. It did none of these things easily as a result of the mishmashed use cases. It's good technology searching for a problem that doesn't exist.
Wave has been collecting dust as an Apache project ever since.
Google Wave was a collaborative editor and communication tool that didn't let you collaborate with the people you wanted (invite needed); IMHO that's what killed it.
I've found need for a note taking mobile app recently, but after trying to use Keep and seeing that it has zero organizational tools such as tags or folders, I just started to use Evernote instead.
Frankly, it is nearly impossible to recommend Keep when Evernote exists. Until they have, at minimum, some way to organize notes, it's just a pile of notes sitting in one window. But it would still be a hard sell when compared to all the Evernote features, such as the desktop clients that allow you to take screenshots and annotate them.
Edit: Oh, and there's also Geeknote¹ for CLI usage, as well as Vim plugins².
I've been using OneNote, solely because I can use a keyboard shortcuts to click+drag screenshots or pop up a new ad-hoc, unfiled note immediately. Does evernote do this?
If you disable the annotation tool popup after taking a screenshot in the settings, then yea, it creates a "Screen clip" note with the image immediately after you finish dragging the screenshot range. (Of course, you can annotate later.)
I never really use it for anything important anyway but I can't imagine it going anywhere other than getting rolled into some aspect of Drive. It's already sort of a simplified note-centric version of Google Docs and mostly works the same way. I just use it since it's easy to keep on my phone's home screen for quick notes and snapshots I want to be able to access elsewhere if needed. I know something like Evernote is more fully featured but since Keep came out, I don't really use Evernote anymore. I think my needs in that space are just simple.
The chances of it being "gone" tomorrow are basically zero. The worst thing that could happen tomorrow would be an announcement that in 60 days, no new Keep items could be added, and you'll have a year to download all your old ones as text, HTML, or JSON.
The hassle of having to switch is a giant burden though. Imagine grep was replaced every 3-4 years on your machine with some new, slightly different regex tool.
You're right, I meant "gone" in the sense of deprecated more than vanished. The data is safe but the disruption to my workflow would be annoying enough that I'd rather not get too used to it.
They forgot Google Code[1]. It's still "up" but it's obvious it's been abandoned by the amount of spam issues that have been building up. I know people like github more than google code - but I personally liked Google Code for it's simplicity.
Wow...killed the same year Uber was founded. Missed out on a nice market opportunity...probably just needed to be refined, remarketed and keep it going. (I never even heard about it.)
Can't believe nobody mentioned Google Reader yet - that was such a great product and they killed it, still haven't found anything that comes close (and I'm paying for a theoldreader.com subscription, they're good but no Google Reader).
I tried half a dozen replacements after the Readerpocalyse and finally settled on https://kouio.com/ - been very happy with it ever since. The only real gripe I have is the lack of sane sidebar feed list ordering by default.
I find Newsblur.com has passed Reader in almost every way – feed management, social, etc. – and it has a viable business model, not to mention being open-source (https://github.com/samuelclay/NewsBlur).
The one advantage Reader arguably still had was what came from their search infrastructure (caching nearly everything forever & near-realtime updates) but that's not something I notice very often.
Wow, I've forgotten how many Google products I used that were abandoned at some point. I counted 15 that I've used.
To be fair, some of these didn't die - they were renamed, graduated, or folded into another project. Google Talk actually still works on Google Chat if you have the old client. Google Listen became Google Music, Checkout was mostly folded into Wallet, etc.
Sidenote: if you are the developer, that yellow-on-white is almost completely unreadable. I know flat pastels are in, but I have to either strain or open the box to highlight.
I always wondered why Google didn't spin off some of this as open source projects.
It seems to me that the bottleneck wasn't the hosting, but the man hours being wasted on them. Google could have reduced their involvement to providing an API and the development could have been managed by community.
It's easier said than done, obviously. Maybe the bottleneck was psychological and focus related, and they wanted to wipe the slate clean?
I suspect that most used large amounts of internal Google libraries and was structured to work on Google's storage backends. Consequently it would be a phenomenal effort to strip that out, or make it sufficiently abstract so it has a hope of running elsewhere.
It is notable that Google doesn't internally use services. If for example they wrote everything to work on appengine/cloud in exactly the same way and APIs as outsiders do, then spinning out would be a lot easier. http://steverant.pen.io/
How many of these were companies acquired by Google, and then dismantled after the project didn't gain traction?
Not bashing Google, but there's always posts on Medium about how some startup got bought by Google, promised they wouldn't change anything and within a few months, simply dropped the project and laid off the employees.
Some, but while being an acquisition is a small risk factor, Google is pretty ruthless about killing its own stuff as well. In my dataset from 2013 ( http://www.gwern.net/Google%20shutdowns ) there were 63 acquisitions vs 287 internal products, and being an acquisition increased the odds of death by an OR of 1.13 (small and not statistically-significant).
I liked Google Squared a lot and it could have been a great comparison tool. For those who never tried it, you could search for "Ultrabook" for example, and have a table with different models, the screen size, autonomy, price etc ...
I remember this, it was very short lived but it indeed existed; I think it wasn't pre-YouTube though, I remember it as a direct YouTube competitor prior to the take over by Google.
I am still super unhappy about Google removing the "Discussions" filter as a search option over a year later. It was insanely useful. I think you can still do it by adding something to the URL, but the quality of the search filter has dropped drastically (which is maybe why they removed it). I don't know of any other search engines that have a Discussions filter.
Interestingly, the HTML spec indicates that only hyperlinks should change the cursor to a hand when hovering. Nowadays though, everyone expects that hand == clickable.
This flakiness is really hurting Google's B2B business. Google tends to do things in a particularly proprietary Googly way that locks companies in, but the history of killing off products makes nobody want to commit to that.
Thus the B2B offerings underperform, making it more likely they'll get killed off.
I follow google reasonably closely and I haven't even heard of a good 40% of these.
One of the worst designed products I've ever used was the Google Search Appliance. I'd never recommend it to another developer, the user interface is just terrible. We'd pay a bunch of money and what we got was a hard to set up box that was kind of buggy.
Not sure who you're talking to (probably 20-somethings) but Microsoft is the champion of backward compatibility. They may have floundered recently but traditionally their adherence to old models has actually been a hindrance.
My first computer was a Timex 2068. Like many I eventually went through a love/hate relationship with Microsoft products, to eventually realise, that when compared with many enterprise companies, Microsoft is quite good.
Google checkout turned into Google Wallet. Google latitude ended up as part of the Google+ Identity overhaul.
There's also a lot of projects that are "labs" or "beta." Clearly they didn't gain the traction Google wanted from them.
Google's like any other startup on a project-by-project basis. If the idea doesn't work they pivot and drop it.