Slashdot Mirror
Ray Ozzie Calls Google Wave "Anti-Web"
TropicalCoder writes "Ray Ozzie says that Google Wave is 'anti-Web,' by which he seems to mean that it is too complex for its own good. In the video he complains about its complexity in relation to Microsoft's Live Mesh: 'If you have something, that by its very nature is very complex, with many goals... then you need open source to have many instances of it because nobody will be able to do an independent implementation of it.' That's its weakness to Ozzie, apparently — that this complexity that can only be overcome by open source. While he heaps high praise on the Google team that came up with this, he feels that the advantage of Microsoft's approach is that '...by decomposing things to be simpler, you don't need open source.' The Register's author summarizes it like this: 'In a way, this is classic Microsoft meets what is emerging as classic Google. Microsoft gives you an integrated stack but all the moving parts are anchored on a single company's vision. Google frees you to work out the bits yourself, but you must rely on your own smarts or those of your chosen tools.'"
Calm down. The editor is our friend kdawson. When he choses to not troll directly he just choses some indirect form of trolling. Like putting out an obvious article of a MS employee praising their own products above the competition. Just to get this same reaction. Kdawson is just at work, doing what he does best: trolling.
It's time to realise that Abble's products are the biggest abomination these days. Just say NO to the dumb iAbble way!!
Talk about being anti-web. The wave google tool is something you can use on the web. The Microsoft tool is something you have to download and then install before you can even start using. The wave google tool can be used with anyone with an email address. And the Microsoft tool can be used only with other people if those other people registered, downloaded, and installed their software. Yeah, I really wonder who's anti-web now.
Lotus Notes = Ray Ozzie
if this man is speaking, I am not listening
That's not the issue at hand here. The site linked from the summary, Live Mesh (Beta), supports sharing and discussing documents. It does not do it in real-time, but, realistically, the real-time part of Google Wave's colloborative document editing is not that important.
The real issues are design and openness. I am a bit confused about where Ray Ozzie is coming from: I think he means that the problem with Google Wave is that it is too simple and web-like, not that it is too complex. That is, Google Wave has a lot of potiential, but much of that potiential depends on people writing gadgets/add-ons for it, as opposed to its features being limited to those Google/Microsoft can think up but already layed out in a structured way. The same issue is often referenced as one of the web's greatest strengths -- and weaknesses.
There is another large issue related to openness: privacy. With Google Wave, you can get all of the features running it on your own server, fully controling the software and hardware. Live Mesh is just yet another web service like Dropbox, etc. which depends on Microsoft's Live Mesh servers. Then again, Microsoft may plan on making it part of Windows Server, which gets rid of the privacy issue.
I think the web has shown quite clearly that leaving a protocol open allows for wide-ranged, unexpected innovations to be based on it. Google has shown off some of its ideas on what Wave is useful for. The Wave groups and various blogs have plenty more. Most likely, if Wave actually catches on, at least some of the common/mainstream uses 5 years from now will bare only passing resemblance to the ideas being thrown around today.
Centralization breaks the internet.
The solution that Microsoft was pursuing was good, and attempted to fit the RSS model blogs use to push content
I think that FeedSync is great...if you think of it as a "improved RSS/Atom", but nothing more. I mean, using it as synchronization protocol for any kind of data flowing to/from the cloud looks stupid.
And this whole synchronization thing seems to be oriented, in the Microsoft side, to sync data between storage devices and computers. Google however seems want put most of the data in their servers. Just "upload" them one time, and the rest of the time access and share that data with the browser. No need to sync - most of the time. Microsoft is all focused in building a "synchronization protocol" that is not really going to be neccesary if we move all/most of our data to the cloud...
So I was wondering who Ray Ozzie is, and how about that, he's a software architect for Microsoft.
Ray Ozzie is the Chief Software Architect of Microsoft. He replaced Bill Gates as the person who drives Microsoft's technological decisions.
Live Mesh is Ray's brainchild. Why is it important to listen to what Ray says? Because he directs the future of Microsoft's development in the space. He controls billions of Microsoft dollars. The point is that he's not some random Microsoft shill - he's the guy in charge.
An ajax web app that tries to ape a simple desktop app is built with:
HTTP
HTML
CSS
XML
SQL
JavaScript
PHP/Python/Ruby/other scripting language
That's 7 different text-based (aka "simple") languages/syntaxes a developer has to learn just to be able just to get the same basic functionality as a simple desktop application. The current system as it is isn't simple.
Sure, if you include only a snippet of what I said, carefully excluding the rest, then I guess you could have a point. That being said, even if I accept your straw-man, at least if I download Firefox, and if my mom downloads IE, in most cases (barring a few exceptions), we'll be able to browse the same web pages.
You seem to imply that Android is closed source?
Did "imply" come to mean "say the exact opposite"?
The hardest part of the search technology, the processing of massive amounts of data and the indexing of that was open sourced as well.
Wow, that's very wrong in at least two respects.
First, what company is the largest contributor to Hadoop? (Hint: not Google. Their MapReduce implementation is still unreleased.)
Second, MapReduce itself is "merely" a tool, albeit a nifty one, not "the hardest part of the search technology". The hardest part would be coming up with the applications that run on MapReduce and actually handle the data. What does Google index on? How does pagerank actually work? These are questions that, to my knowledge, are still Google trade secrets.
If you gave me a few months I could write a fairly unoptimized, fairly poorly-performing MapReduce implementation, but one that still got fairly decent scaleup. If you then gave me a few more months, couple Google engineers, access to their code base, and four times the processing power of Google, I could probably more or less duplicate Google with my MapReduce implementation as a replacement for theirs.
If you gave me a server farm and a few years, I could come up with a crappy search engine, but I suspect no better. Certainly not anything that I could put on Google's MapReduce implementation and have anything that produced something close to the quality of results Google, MSN, or Yahoo produces nowadays.
There are already good solutions to this problem: it is called revision control and the Subversion system is a high-quality open source solution to most common version control / sharing scenarios. Visual Source Safe wishes that it could be as good as Subversion, but the open source crowd beat them to it.
That misses why Google Docs was actually popular. If two people edit the same document at once, using a revision control scheme, then there's a significant potential of a merge conflict or of a nasty "someone else has the lock on this document" message, both of which are a usability nightmare if your users are non-technical -- the user is stopped in their tracks, gives up, and goes away. Google Docs does use a revision control method behind the scenes (google-diff-match-patch), but because the commits and updates are happening automatically every 30 seconds, the changes are kept very small and the chance of a merge conflict is very much lower. To show just how simple it is technically, Docwit is a very small hobby open source project that ties TinyMCE to google-diff-match-patch to do the same thing, but because you can run your own server you don't have to give Google your data.
Google Wave essentially just goes "Hmm, why don't we shrink the update period even further, and (like SubEthaEdit, and also quite like a few other projects that have involved working on XML documents remotely) send operational changes when they happen rather than polling every 30 seconds?". The change size gets even smaller, and with it the chances of having to show a user a "merge conflict" or "lock conflict" scary box are also reduced.
You see, it turns out not many people use Google Docs for "proper" documents (of the corporate kind) but a heck of a lot use it for collaborative note taking, as a cheap-and-easy wiki, and for lots of other "low-fuss" tasks.
Oh come on! You're living in a fantasy world!
Notes 6.5.1 (the last version I have extensive experience with) did not come out a decade ago. It came out in 2004, and it a gigantic piece of crap. The only thing good I can say about it is that it *finally* worked correctly on NT with multiple users-- only 10 full years after every other piece of software on Earth did!
Are you really complaining that a version of software a over a decade out of date was unstable? If your apps were flimsy, you should have talked to your developers. The Domino system is and has been for a long time a very robust system.
Dude, reality check:
IBM sells Lotus Domino/Notes as an email system, "groupware" if you want to use that term. Look: http://www-01.ibm.com/software/lotus/ Right there on the website, it says the top two features are Email and Calendaring.
Email and Calendaring. Lotus Notes may work for many tasks, but two tasks is *does not* work for is Email and Calendaring. Not even close. Hell, I had to reset Palms at my workplace 3 times a week when Notes would reliably bug-out create appointments that ended before they began-- which of course confused the poor Palm software to no end.
The amount of lost data due to Notes' failure of a UI is legendary. Deleting a copy of an email filed into a folder *also* deleted any other copy in any other folder. Amazingly retarded design. Notes didn't open attachments in the Temp folder as Read Only, so it encouraged users to edit them and save their changes. While, at the same time, it was super-aggressive about cleaning up the Temp folder. I can't even guess at how many documents were lost that way by poor, understandably confused, users.
Yes, Lotus Notes can do all that and a bag of crap, but it's sold as groupware and that is how it shall be judged. I'd go as far as saying that I don't even give a shit what else it can do: it's sold as groupware, and it *sucks* as groupware, and thus it's a failure of a product. (It also costs twice as much per-seat as Outlook, for a far inferior product.)
The anti-Notes trolls always crack me up. They basically say "I once saw a badly implemented application in Notes a decade ago, and it didn't compare to applications that are being written today." It makes about as much sense as complaining about Windows because you didn't like WindowsME.
Oh please. Compare Notes 6.5.1 with Outlook 2003. NIGHT AND DAY. (Notes being "night.")
Here's what I'll acknowledge: there is a certain subclass of human being, you included among them, that are not only blind to Notes' downsides, but actually are huge fans of the program. I won't attempt to change your mind, because I know from experience that your brainwashing is total and complete. But I'd really appreciate it if you didn't just dismiss all criticism of Notes out-of-hand.
Comment of the year