Slashdot Mirror
Google Wave and the Difficulty of Radical Change
cedarhillbilly writes "An article by Matt Asay in the Register takes on Google Wave from the perspective of visionary change versus incremental change. He suggests that visionaries should focus on smaller transformations of our day-to-day lives rather than leapfrogging. 'Much as it may want to radically change the world for users and developers, radical change generally happens over time, through a series of incremental, unexceptional edits to existing technology and processes.' Perhaps Google sensed this when they famously said they were worried about having too many geniuses. Asay revisits the point that the open source development model necessarily builds on a community of contributors and users, and not the mad scientist in an ivory tower."
>> He suggests that visionaries should focus on smaller transformations of our day-to-day lives rather than leapfrogging.
Why can't they make something radical, then add on compatibility stepping stones for a transition period? Would Wave have been so unused if you could read your normal POP3 mail in it and intercommunicate with traditional IM systems?
Nobody who spent any time using Wave thought that the problems were due to too many geniuses in the mix.
A real genius doesn't just show you a vision. A real genius creates a useful artifact that solves a problem of importance. We're not talking about art.
Am I part of the core demographic for Swedish Fish?
Google Wave didn't fail because it was "too innovative" or "too radical." History is jam packed full of inventions and technologies that succeeded precisely because they were drastically better than what came before them (lightbulb versus candle, car versus horse, calculator versus abacus, GUI versus CLI). Google Wave failed for a combination of reasons. It wasn't marketed well, it didn't really solve any problems, and it just wasn't "better" enough over the standard ways of browsing the web.
Google Wave was a cool engineering project, but never should have been taken to market.
Warning: Contents May Be Flammable. Keep Out Of Reach Of Children.
It has been known for quite awhile that it takes 20 years for new/radical technology to be adopted. They expected it to take less time because they are Google and when you are Google and have so many geniuses working for you people will just do what you think is great right. The real world works much differently. Although the world may one day yet adopted the Googly way we won't know for another 20.
Incremental, gradual change is not radical change. The problem is that incremental, gradual, and radical have definitions, and those definitions are not synonymous.
Oh stop it already, Google Wave failed because it was a piece of crap and badly engineered software, not because it was "too radical". It was an awful idea trying to clone unix talk and no hype in the world could have saved it. Stop trying to pretend it's the users fault that it failed, it was barely even functional after a conversation had gone on for more than 100 lines (something that we fill up on IRC in about 10 minutes), what did they expect?!
A revolutionary rethinking of how we communicate will always take time to gain inertia. Real people have busy schedules, and you can't just tell everyone you are ditching email etc and moving onto the Wave: You have to get reluctant collaborators onboard and lineup a good project or two with which to get the hang of it at the start. This is never going to happen in 3 months, and i think google know this. I can't help but feel that they cancelled for some fundamental failing that they are not talking about.
Still, i hope it doesn't go away. It has so much potential that it deserves to be developed.
Yes, all those glorious geniuses who never found a way to make Wave work... more like pretentious geniuses.
I hate the word genius. There isn't a thing alive on Earth known as a genius. A true genius would be capable of doing anything thrown at them given the right things to do it with.
Google just has people specialized in certain areas of knowledge, with decent intelligence. Yes, there might be some people who can give you an IQ of 200, but that doesn't mean a damn thing since said IQ can vary in meaning between everyone.
Some can be great with numbers but awful with equations, great with spacial awareness but awful at remembering where to go in said space.
Want to know why Wave failed? Google, an advertising company, never advertised it enough. They never gave it enough time either.
It wasn't an issue with UI, yes, it was awful, but it worked. People use bloody Facebook and Microsoft Project every day and they have to have the worst UI annoyances in existence.
Not only that, their expectations were set WAY TOO HIGH.
They never made it that accessible from the beginning.
They released its existence way too early.
Google are always too focused on stupid shiny bells and whistles at the expense of speed, WRONG WRONG WRONG.
They also had way too much going on in the UI, event handlers flying out the nose especially.
Too many event handlers were the main reason for Wave slowdowns when they get larger. Worse yet is the fact they had fade-in animations, and real-time message updates pretty much just kicked it in the nuts when it came to speed.
If they integrated it in to Gmail, or even replaced the Gmail UI with something based on Wave (conversation system in Gmail would benefit from that), THAT could have worked.
Who knows, maybe they might still get it working. But at the moment, they have been a massive failure when it came to dealing with Wave.
It was a failure before it even got a chance. The doctor couldn't save the poor kid.
Google Wave was a collaboration tool, and that made it nearly useless during its limited preview. It was available generally for less than three months before Google killed it. That would be a ridiculously short time for any new service, let alone for one that actually requires network effects to become useful.
I don't know whether Google Wave would have replaced E-mail or chat; it had the potential to do that, but that was far off. But it was an excellent collaboration tool. It could have been Google's replacement for Sharepoint, Lotus Notes, and systems like that, and it looked like it was on track for that. Incremental changes to GMail are not going to cut it.
With killing Wave, Google killed something that could have become quite important for them in the future. And they also killed the good will and trust of a lot of developers and users.
Google should have given Wave three years, not three months, of general availability.
... focus on smaller transformations of our day-to-day lives rather than leapfrogging
Because the power base of the current controlling interests is too large and cannot leapfrog. The current power base will only take steps forward which do not require them to surrender their stability. Leapfrogging is not a stable type movement for large power bases and organizations.
The only way to get my department away from proprietary Unix based infrastructure to LInux based infrastructure was by replacing the people. Now Unix admins from other departments are asking how to change to Linux. But they are unwilling to leapfrog their skillset; despite the minimal change required. Because it's easier to defend their infrastructure with their large budget.
Basic people power games.
I think this is one instance where Google's limited release method failed spectacularly. When they started to release Wave, I had a bunch of people in mind to collaborate with, but only one or two of us had it. By the time it was available to the majority of us, we had already gone back to using other means of communication, including Google's own docs. For all its potential, we ended up only having two active waves of substance. Hopefully they'll be able to incorporate some of the more interesting concepts into Gmail or Gtalk, and I think Docs already has some simultaneous editing features. So wave may live on, just not as wave.
Be quick to listen, slow to speak, and slow to anger.
Probably was being too radical more than the initial stability problems and bugs what hurt the grow that KDE was having by the time the version 4 was introduced. Still, as was basically "the" direction to follow with the entire platform (you could leave it going to gnome, stay with kde 3.x while all the apps move forward, or adapt to the new approach) it survived, and now is growing (not having hard numbers of gnome, kde and other linux desktops, but i think it went that way)
One of the purposes of Wave was to unify the various means we have of communicating. You wouldn't need forums, Facebook, email, IM, IRC, mailing lists, etc, because Wave does it all.
Unfortunately, the implementation bogged down when we had too many messages, it wasn't nearly streamlined enough, and -- not entirely Google's fault, since there was enough of an API for people to do this -- but since it didn't wire into any of those systems, and since everyone wasn't trying it all at once (partly because of the semi-closed beta), you now had forums, Facebook, email, IM, IRC, mailing lists, Wave, etc, which isn't an improvement.
Don't thank God, thank a doctor!
I'm reminded of Plan 9 from Bell labs failing to overthrow the Linux community, because so much time and effort has already been invested in the thing that works just good enough (plain old e-mail in this case).
Obviously I can't speak for others but in my daily work I have no need for adhoc collaboration of this sort. I can certainly see it being useful in some contexts.
There have been similiar systems providing "shared workspaces" in existance for quite some time so I don't understand the "its new" or "bleeding edge" crap. Technology is about solving problems. If there is no market need then it doesn't matter how "advanced" or how much time was put into a product. This is not a product failure - its a market research failure.
The real reason that Wave has failed to gather the following of so many other great ideas is that they launched an experimental product that they hoped would go viral and become wildly successful, but they restricted it to the smallest group of people possible: those that don't use Internet Explorer. They probably thought that the "explosion" of popularity of Wave would help them grab browser market share (or at least take it away from IE), but they failed miserably at that as IE usage is actually up now. The claim was that they were focusing their efforts on the quality of the product or some crap like that, so they didn't want to waste time on IE compatibility, but not supporting the browser that had and still has the largest share of the market is just suicide. And let's face it, it's not like it's that hard to get an AJAX app to work in IE - I did it before it was cool and that app still purrs on IE, Firefox & Safari without me really trying to support them. So it seems to me that Google deliberately "broke" Wave on IE. Which means they did the opposite of what they were claiming to do: not wasting time on IE; instead they wasted time on making it not work and wated time on trying to get browser share. Lesson learned? I doubt it. And I'm pretty sure no one here will agree or learn from the lesson, but at least I've tried. The lesson, in case you didn't get it, is: don't launch a product that people might have a hard time seeing the need for to the most restricted segment of a population: those not running IE (you can apply this to other markets, too).
And perhaps you are proving TFA's point. In all those cases the development of the new alternative didn't happen at once and was a succession of incremental changes over the previous technology.
Take GUIs; OSes like Windows 1 ~ 3.11 made heavy use of CLI as well (and you might remember that Windows didn't really take off until 3); Windows 95 was still mostly a front-end for a particular version of DOS (7, IIRC); even nowadays there are a number of things that are done through the command line. Same with Mac OSes and let's not forget Linux there.
Horses weren't replaced by Ferraris either. Primitive cars were hardly an improvement over horses, and in fact the development of engines happened over the course of centuries (seriously, experimental steam-powered vehicles existed as early as 1672).
Calculators have been in development for centuries as well. Think of the Antikythera mechanism, but also of Pascal's mechanical calculators from the XVII century, and so on.
The lightbulb itself has a history going back at least 80 years until Edison made it work well enough to be a commercially viable alternative.
In short, judging form these examples, it would appear that slow, gradual change is exactly what allowed these technologies and inventions to succeed. We might look back and say "oh, the car is totally better than the horse", but it was a long time since cars first started being developed until this became true.
I know I'm talking to /.ers, but don't you people get it, yet? "Too many geniuses" that's rich. How about: it is the natural problem of developers that your focus is on *the tech* (it's what spins your beanies, after all) and not on the sloppy, incomplete, confusing human/social/aspirational context and nature of your end users? It's all part and parcel of why open source won't be an answer to this kind of thing either (did anybody spit out at least some breakfast cereal when FF named their skins "personas"?). It's not about "not being so genius-y" or "incremental innovation" or "a hive approach". You just need to recognize that you need other members on your team - radically other members. You need folks that are more psychologists and ethnologists than they are technologists (an aversion to technology might well be a strength for someone on this part of the team) to provide a framework and focus for you to go to town inventing the technical means to achieve the well described ends of your end users (end users who are not frickin' YOU -- and most often not AT ALL like you). Why is this taking so long?
Google did a great job creating an open protocol. But they made two mistakes:
1. They were not open enough. Although they had suggested that people would be able to build their own clients (and demoed a curses based client) they never opened an API for writing a wave client. They wanted it to be a flagship web application - but just as people like all sorts of different clients for email (even if many now like web clients), they would probably have liked client choice for wave - especially if 3rd party clients had shown waves along side email and the like.
2. They were too open. Their programming model for wave (web-hosted applications with read and write access to your wave) had huge security implications. It was not clear from the UI who would have access to your data and when.
Both of these were things that slowed adoption of wave.
Instead of Wave, what Google should do with XMPP is evolve it into a replacement for SMTP/POP3 and probably IMAP. At the same time, evolve Atom (the format, not the protocol, obviously) to replace parts of RFC 5322 that are not covered in XMPP. Properly done the transition could be gradual and invisible to the end users. Then IM, multi-user chat, email, and feeds would all have the same underpinnings.
Climate Progress - Hell and High Water
In regards to the original topic, "Revolutionary" change, especially in software, is often remarkably... effective in sweeping away the ghosts of the past which weigh upon the minds of the present.
As example, a gem from the days of Wang which I just came across:
As an example of this strategy, a frustrated developer wrote Wang’s second generation e-mail system (Wang Office) over a long weekend. In his view–and he was right–the official spec meetings were taking too long. So he decided to cut through the bullshit and just code the thing (he’d designed Wang’s first generation e-mail system, Mailway, so he knew what he was doing). He sent out the new code to several large accounts, they loved it, and started calling headquarters asking, “We have the checkbook out–how do we buy this great e-mail system?” Back at headquarters, everyone (except for Steve) was going, “Huh, what are you talking about?” Once management realized that (1) customers wanted to buy it now and (2) doing it the “official” way would take another 18 months, they swallowed their pride, shot the official project, and gave Steve a small official slap while privately lauding his initiative.
/me files Matt Asay in the [bullshit|?|clueless|lost|confused] category.
Yes, it was a cool concept.
The implementation however was not good enough, simply too unwieldy.
I see no reason to discourage either radical new hardware nor radical new software. It will flower or perish on its own merits.
Should have been, "Google Wave and the difficulty of flogging stuff that's shit"
Are you sure about that?
My impression was that it was intended to be a replacement for email, im, and other realtime communication systems on the net. It wasn't an app to help with projects or conversations; it was a protocol/server platform for messaging, just like the SMTP protocol and mailserver that makes up email, but more flexible. I think the idea was to replace email, IM, web forums, twitter, etc. all with one flexible, scalable platform that could handle new kinds of data, provide gateways to disparate systems (connect your IM to your SMS, or your webcam to your audio-only phone, for instance), and to make it all expandable by bots which could do automated processing of messages.
It actually could have been very cool, but it was too big for the PITIFUL amount of weight google threw behind it. They didn't believe in their product. If they had, they would have built an exchange-killing open source mail/groupware server on top of it, which was fully backwards-compatible with Email, IM (including MSN as part of the exchange-killing thing), etc.
THAT needed to be a radical product launch. None of this beta crap; a SOLID, powerful, game-changing release of free server code for everyone to install and use. Where the gradual change comes in is integrating their translation engine to make global communication possible, integrating google voice, integrating reader, and generally taking the world by storm by combining all their existing products into one great solution that had ZERO competition.
Now that would have been radical. Launching a half-baked idea with a horrible web-ui and some code for a cut-down version that no one cared enough to look at? Not so much.
I once suggested the Linux distributions try to mirror some Windows conventions and naming to make it easier for users to adjust, and was modded to oblivion on Slashdot. Windows is the de-facto standard whether you want it to be or not. I realize one has to strike a delicate balance between security and being Windows-like, but that's where the hard work lies, not some new trinket or feature. And there's still some low-hanging fruit that doesn't involve security compromises, but rather things like vocabulary and placement. I'm just the messenger, I didn't create the de-facto standards and human nature.
Table-ized A.I.
When Ford released his first vehicle, he had ad writers tell people what it was for,
WTF? Everyone knew what a car was. They had been around for 100 years before Ford. What Ford did was successful mass production.
Deleted
What would happen if this was some anonymous project hosted at sourceforge instead? Some really untraditional open source software, looking to problem in different (very different!) angle has been successful. While "old school fans" hate what it has become, I can give Azureus. Think about it and remember the product it was racing against, compare both. The "simple application" it was racing against has become so problematic so they ended up acquiring something that was supposed to compete against them, as alternative.
Perhaps Google fools themselves thinking everyone buys their "don't be evil" slogan and trust them. Why would I trust some side project of an advertising giant? You know, if you are a Chrome extension developer, prepare your $5, just at front page today. Now, if some miracle happens and actual open source developers say "oh really? here is your account revocation", will Chrome extension project called "too innovative"?
It didn't have time to fail. The article correctly points to the fact that e-mail took 40 years to become as widespread as it is today. When it began it was an unrecognizable form of communication, a huge sea-change if it were to ever be adopted. People didn't even really know what computers were, let alone understand networks. Yes, it provided a nice electronic metaphor for the regular mail letter, which let people grasp it more easily. But everything else about it was still fundamentally alien.
But something happened. First people began to get used to email, and then people began to prefer email to communicate. Over the next few decades it spread like wildfire, to the point where now over a billion people use it every day.
Wave didn't even have time to begin such a process. It was announced as a collaboration tool on May 27, 2009. On August 4, 2010 it was canceled. A little more than a year and a half. Compared to 40 years for e-mail?
Give it time to breath. It seems stupid to invest so much money and effort into a product and then drop it if it doesn't acquire a million users right away. I think Google, structurally speaking, might have ADD - it is so used to overnight success that it's not willing to accept a slower rate of adoption, even if the pay off could be huge.
There is a time for incremental changes and a time for radical changes. The trick is figuring when that is.
I spend 8+ hours a day programming in front of 3 screens with about 10 tabs open in each and I've never even heard of it. Maybe, just maybe, that's why it failed.
Brilliant concept, well executed from the engineering standpoint, but never could really get legs under it. In the end Wave was shut down because the adoption rate was considered too low & slow by those responsible for overseeing its budget.
Post mortem
- the initial release was cautious and tentative to avoid risk of being seen as responsible for any high profile snafus.
- much misplaced focus on the "behold the coolness of what we can do with HTML 5 someday"
- missed huge wins by not launching with existing communication methods pre-wired, eg, incoming email|SMS|chat|Tweet -> Wave -> outgoing email|SMS|chat|Tweet
- Buzz came along to give the whole team a distracting "WtF are we working for the same company?" moment
Obviously Slashdot and Facebook have a much better models. Being able to edit the whole thing, go back in time just leads to a mess. Let everyone edit their little piece.
It was an engineering pet project that nobody else really wanted, but the wildly pro-Google media is working hard to spin this as just "too innovative" for the public, which is both misleading and condescending. Look, this isn't Google's first failure or the last. It failed because it just wasn't that good or useful.
the ivory scientist in the mad tower.
[signature]
1. It failed because it had bad user interface design. There's nothing radical about that. In fact, it's all too common in projects run by most hackers and code monkeys aiming to make a "cool engineering project." The difference between those projects and Google Wave is there are people crammed up Google's bum, willing to call Google developers visionary geniuses whose efforts are beyond what mere mortals can comprehend instead of lambasting them as they would anyone else for lacking usability in their software.
2. More importantly, it failed because Google intentionally made it fail. It was axed in less than three months of being public. Something very weird happened there, but who knows if we'll ever really know why.
3. The best, most successful advancements in computing were done in leaps and bounds, not the safe, incremental nonsense brought on by the commercialization of computing in the 80's.
Wave was a dumb idea because it was not finished and totally separated from all the tools used today. The features in wave all belong in other tools, not clumped together in something that was less than all other tools. For example annotating documents in the wave way. Very nice idea for google docs. Not good inside wave where the editor was crude. So a new feature inside junk, or a new feature inside a very good tool. Also google has some off idea that we want conversations all glued together, and that is stubbornly shown in gmail where you have to look at huge threaded views instead of individual replies. This is fine for some but many businesses hate it and it mars communication. Luckily I read one of the new higher ups is finally changing that feature. Finally the day I saw the wave presentation and I saw that in place character by character editing, I said, that's a developer toy, not practical. Just because something is possible doesn't make it useful! People hate the idea of that document editing playback too, how could they think that was a good thing? Another gee look what we can do, but is as desirable as acne! "this is an crappy great idea". That kind of playback is not really what people want. It's like someone spying on you as you type. Creeped out users.
Was all it was.
Like many here, I thought the possibilities immense, but considered the implementation poor.
I feel for the engineers who brought it to us. So much work for such a poor return.
Let's be clear. The idea and functionality are pretty darn rocking. Some very bright people did a great job solving a problem that was largely unspoken, let alone properly addressed. Yes, IBM, Microsoft, and others provide alternatives, but the open and distributed nature of the protocol was exemplary.
I predict two short-term likelihoods.
Firstly, Google will return to this technology, properly integrated with gmail and all their other googly goodness. It is a paradigm shift. Facebook will integrate the protocol very quickly. So will everyone else.
Secondly, properly distributed and heavily encrypted implementations will appear in the open source world, and will then infect everything else. I wouldn't be surprised if a limited and purpose-dedicated version makes it into the Linux kernel itself, if even only as an automated git/bug-tracking interface.
This is only the beginning however. Google's plans are unknown, but their use of Peter Norvig as director of research suggests considerable focus on AI, avatars, and ubiquitous integration.
The really interesting stuff will occur as the Internet itself begins to replicate Google's functionality. Search and Rescue built into the protocols themselves? It's coming.
Google Wave is not only not the first wave, it's not even the best or strongest or even the most coherent. But Google are on the case, and the rest of us will be grateful*.
* I am in no way affiliated with Google, and seek alternatives to their hegemony, but I give credit where it's due.
science in government
... that gets you.
The only reason that programs are still running on von Newman, SISD, architecture is because programmers don't like thinking, including von Neumann.
He figured he could simplify the expression of problems by a SISD reduction of the problem space and it screwed up the thinking of every programmer since.
The fact that every single CPU and GPU built with ICs is fundamentally a MIMD processor (even if constrained through a single clock pulse,) and runs lots of processes in parallel is completely opaque to the average programmer.
They keep going on about how many transistors go into a chip without realizing that it is just a bunch of dirty sand without the ability to connect all those components.
Noyce and Kilby got their Nobel prizes not for the number of transistors they could put on a chip but for figuring out how to connect them all together with traces.
The trace is what is behind the silicon and germanium empire of Intel and the other foundries.
Its the only component that has been working reliably and not giving the chip fabs problems.
Just because its not sexy, (read: troublesome,) it gets no respect.
But it IS essential.
MSBPodcast.com The opinions expressed here are my own. If you don't like 'em... Think up your own stuff.
Are you telling me an iPhone isn't radical? A desktop Web browser with a display 1/4 the pixels of the original Web browser and 1/8th the size, no mouse, no keyboard, no windows (the document floats inside the viewport), and holistic zooming is not radical? Yes, it is radical, but it is also usable.
With Wave, Google got blinded by how pretty it is under the hood and forgot to design a user interface. Wave should have looked like Microsoft Word v5.1, it should have shown the user "replace your Word, Email, and Content Management System publishing workflow with me." Have you seen how a company puts up a typical Web page? Word documents going around in email, eventually being pasted into a CMS, it is a joke. If MS Office didn't suck it would have had Wave-like features by now and no CMS would be needed by Office users.
Google needs humility. People think Apple is arrogant because they are great, but the truth is, they are great because they are humble. They did only one phone, and it took them 4-5 years to do it, working away totally in secret, iterating and testing and innovating. If Google was not so impressed with whatever they poop out, they would actually finish projects and make usable and successful products.
Even in ads, why did Apple have to do iAds? Why didn't Google offer ads that don't look like shit, ads that take advantage of HTML5?
How has Search improved in the past 5 years? How has it been made easier for the 90% of users who do not know how to use all the options like -term and site:foo.com and will never know that?
So call a failure a failure. Wave is a failure.
The real problem with wave, from my perspective were bad design decisions.
1. To make things familiar, they made it look like a threaded Slashdot discussion, and made each element of text a big block with a lot of decoration, framing, etc. This was way too chunky, as there was no way to highlight or otherwise mark up someone else's text.
2. There was no way to prune or trim a discussion, which made them stretch on to infinity.
3. You really couldn't collaborate on a piece of text, you could only have a conversation about it.
4. There was no desktop native client, which means you always had the slow and buggy web interface to deal with. It should have been smooth and fast, like a Word document open in more than one place in real time... but it wasn't.
5. Did they really have to waste all that screen space so you knew who owned every word?
I think the criticisms to date of Wave really miss the mark. Had it been more interactive and less clunky, it would have taken off. The protocol is open, and someone could fix the above mentioned problems. I'd be willing to help.
Just because they killed the Wave project doesn't necessarily mean it's dead and buried. Wave started out as an experiment project, which was allowed to continue up to the point of deployment, and maybe cost two or three million in developer hours, etc. At the end, they probably had a meeting to discuss what they learnt and what the best strategy was for putting things to good use. In this meeting they may have decided to take certain components and integrate them into Gmail, Docs, Talk, Chrome, etc. It doesn't mean they abandoned all work and gave up. Wave was just R&D. Every company spends millions on R&D that either eventuates to nothing or leads in different directions. The whole point of research is that you won't know what the end result will be. A lot of the time you'll hit dead ends, but the lessons learnt are always valuable and on very rare occasions you might even strike a breakthrough. The only difference here was that Google made their R&D public and built up geek-hype (the mass majority of people wouldn't even know about Wave).