Chris Messina's Thoughts on Opera Unite | Hacker News Reader

Chris Messina's Thoughts on Opera Unite

factoryjoe.com

45 pointschris2417 comments

I'm surprised anyone (outside of Opera) went to this much trouble over Opera Unite. It's just not interesting. It won't be particularly popular, very few people will switch browsers in order to use it, and the people that do will find that it just isn't all that great having your own web server running on your PC.

I like Opera, and wish them the best of luck with their business and their projects, but Unite doesn't change anything. It will be a barely used feature in a barely used browser.

Ever heard of Dropbox? Potentially obsolete with Unite.

Ever wanted to share all the photos/videos on your 120GB hard-drive? Unite does that too, bye-bye flickr.

Ever wanted to host online chat - it's already there in your browser now.

Ever wanted to listen to your music when all the files are on some other computer? Relay via Unite, it's in your browser.

Want to host web-pages on your local computer? Don't want to install a web-server .. no problem.

Really I find it awesome, I think we've only just scratched the surface here too. A genuinely new paradigm IMO, http://alicious.com/2009/opera-about-to-change-the-world/.

Welcome to the Peerweb.

I think we'll just have to wait and see who has the properly functioning crystal ball. I think you're entirely missing the point of Dropbox, Flickr, online chat, etc. And you're also imagining a very unlikely future in which anyone, other than a handful of nerds that already know how to run a real web server, thinks, "Ooh, I want to host web pages on my own computer!" The idea that this will become dramatically more common, rather than as quaint as growing all of your own food, is, to me, absolutely absurd.

Even ignoring the fact that the idea of "your computer" is becoming less and less a big box that sits on your desk and more and more a tiny device that you have in your pocket or your bag; and ignoring the fact that we'd really like for computers, even the big ones sitting on our desk, to sleep when not in use so we don't chew up so much power and produce so much waste; and forgetting about the fact that someone has to actually write the code for all of the magic functions to address the problems you've proposed Unite can solve (and for a proprietary platform that about 1% of users use); and discounting the chicken and egg problem that in order for this to be easily useful across all of your devices you have to have Opera on all of your devices. Even forgetting all of these, huge, probably insurmountable even to a company the size of Microsoft, problems, Opera Unite still doesn't solve any big problems that really effect users.

What happens if my computer is turned off?

Sorry, not quite catching you there, you do what to your computer? /sarcasm

Yes that is one of the down falls, pretty major too, but if you're family wants to share from your computer they can just call, or you can look and see the connections are 0 before turning it off. I can also imagine a caching/on-line backup system (like Dropbox) - eg for your fridge notes.

Peerweb is not the www they have a lot of overlap but this is going to change the way we use computers IMO.

Dropbox and flickr? Are you serious? In order for those to be obsolete, even supposing Unite is technically superior, this would require people actually using Opera. I think this is a bit naive.

As far as "Welcome to the Peerweb." Wow, talk about drinking marketing Kool-Aid.

As far as a new paradigm is concerned .. not really. I think a new paradigm would be more like a rich technology that doesn't use the Web/HTML/HTTP at all.

I agree with him, but he doesn't seem to understand how Opera's proxies are intended to work -- they don't want to have to tunnel all your traffic except as a last resort: their main purpose is as a backchannel for coordinating NAT hole-punching.

As dumb as Opera is, they're not stupid enough to pay for all your live filesharing bandwidth, symmetrically.

Actually, I do understand that. That's why I said that they were using a "P2P-like" network... Given that it's an alpha, it's pretty awesome that it works as it does, but this deficiency — tying you to an "operaunite.com" domain — is primarily where my criticism lies.

And, that they used such confusing and contradictory terms in their EULA.

Still, the claim in the next sentence that "you must push all your traffic through Opera’s proxy service" isn't true.

I understand that you were primarily criticizing/debunking the reason d'etre and social crap in the copypastad press-release -- not the technical handwaving.

How are you prevented from buying your own domain and using it for your Unite services?

What is your proposed solution?

AllPeers required an AllPeers client on both ends, right? But then you undermine the whole Unite concept, don't you?

Before unite, I thought of opera as an irrelevant company that did good work. now I just see them as irrelevant.

One point that Chris didn't make in his post is that the problem they identified as the reason for building Unite (or so they say), the loss of ownership of one's data to third party services, is the same dilemma that inspired DiSo.

The reason DiSo is relevant, and unite is not, is because the good folks (like Chris) behind DiSo recognized that it's not just about where your data is stored. Just as important is that the format is open and interoperable. Although Unite has this quasi distributed model, it's formats/APIs are (AFAIK) unknown, not based on open formats, and for all intents and purposes, as good as proprietary at at this point.

Pushing the 3rd party into the browser is not a solution to the problem.

Huh? Opera Unite is based entirely on open standards, AFAIK. All APIs are public.

http://unite.opera.com/support/#services_tech http://dev.opera.com/articles/view/opera-unite-developer-pri...

My point is that having a public accessible API or docs are no longer enough, you need to adopt open standards whenever possible. Adopting open standards ensures that the end user is free and able to move their data in and out of your service. One of the ways Microsoft torpedoes open standards is by adding creative extensions to them, so while 95% of your data is contained in an open format, it's not portable b/c the extension makes it unusable to services adhering to only the standard.

It's using XMPP vs. rolling your own chat protocol. Supporting oAuth for file access vs. a custom authentication API.

Everything I saw browsing through the docs is about how to develop for the unite platform. I don't see anything about interacting with a unite service outside of an opera browser.

In this sense, unite is a creative extension to an open standard, browsers with common functionality like firefox, ie, safari, etc.

Following this line of reasoning, Twitter is horrible because it doesn't adopt an open standard. Likewise Flickr. And delicious. And anything else with a homegrown REST API.

If their API is well documented, who cares that it wasn't designed and rubber stamped by a committee.

Twitter has adopted open standards, they use oAuth. There's nothing wrong with home grown APIs, but I think APIs that ignore open standards will be a problem as services become increasingly interdependent.

I think if Opera provided an explanation of how to set up a "Unite server" on your own server, than a lot of this controversy would be moot. That's why Google was so epic with their launch of Wave.

What kind of explanation are you looking for? You just download Opera, change a few settings, and leave it running.

Reply on news.ycombinator.com