Weekly Edition
Return to the Development page
| |||||||||||||
Diaspora source released
The Diaspora project, working on privacy-aware social networking, has made
its first source release. "Much of our focus this summer was
centered around publishing content to groups of your friends, wherever
their seed may live. It is by no means bug free or feature complete, but it
an important step for putting us, the users, in control. Developers, our
code is on github, our tracker is public, we have a developer mailing list,
and we are happily accepting patches." (Thanks to Tom Arnold).
(Log in to post comments)
Diaspora source released Posted Sep 16, 2010 14:35 UTC (Thu) by kragil (subscriber, #34373) [Link]
So much for all the people who called this a money scam.
That said, I for one am not totally happy with the implementation first, standards later ruby on rails implementation that uses a lot of gems (some of which are one man shows).
But I will try to install it on my Debian server and play around with it, after all from all the FB alternatives this has IMO the most potential.
Diaspora source released Posted Sep 16, 2010 16:18 UTC (Thu) by Trelane (subscriber, #56877) [Link] So much for all the people who called this a money scam.Would that be me you're referring to?
Diaspora source released Posted Sep 16, 2010 16:49 UTC (Thu) by Trelane (subscriber, #56877) [Link] Relevant link to the discussion: http://lwn.net/Articles/402551/ Please once again note the bolded, all-capsed, italicized text in that comment that contradicts your claims here, if indeed you're referring to me as "all the people." I don't know how to be more clear that what you're saying I'm claiming isn't what I'm claiming.
Diaspora source released Posted Sep 17, 2010 8:32 UTC (Fri) by vv (guest, #69729) [Link]
I know you've got a rather large ego, Andrew, but you're not the only person to have labelled this project a scam...
Diaspora source released Posted Sep 17, 2010 11:22 UTC (Fri) by Trelane (subscriber, #56877) [Link]
who's Andrew?
Diaspora source released Posted Sep 16, 2010 16:19 UTC (Thu) by Trelane (subscriber, #56877) [Link]
Great to see code finally materialize!! Now it's a real project.
Diaspora source released Posted Sep 16, 2010 16:51 UTC (Thu) by SEJeff (subscriber, #51588) [Link]
But mongodb? Seems like a very poor choice for data consistency.
Diaspora source released Posted Sep 17, 2010 8:53 UTC (Fri) by intgr (subscriber, #39733) [Link]
But it's a web-scale database!!!! It doesn't allow joins so it must be fast!
*gg*
Diaspora source released Posted Sep 24, 2010 13:18 UTC (Fri) by dmag (subscriber, #17775) [Link]
> But mongodb? Seems like a very poor choice for data consistency.
First, to clarify: MongoDB doesn't have eventual consistency. So the data is always consistent.
I think you're talking about "single-server durability" (don't loose data if you pull the plug). It's not in the current MongoDB, but it will be in the next release.
(And for those who think "single-server durability" is important: It is, but only if you *ONLY* have a single server. If you have multiple servers, it's better for availability to fail over to a hot standby. The alternative is to be down while the broken server reboots, does a BIOS check, does a disk fsck, and then does a database repair.)
Diaspora source released Posted Sep 16, 2010 17:37 UTC (Thu) by Velmont (guest, #46433) [Link]
Great, great!
It has to fix the great scale-problems now. It is written in Ruby, and although the Ruby community hates everyone saying it is slow, the first open diaspora-node (woops, seed) is already crawling. And was, doing that already when there were two persons in there.
But it truly looks grrreat, and fun to use (which is most important, after actually working). So maybe someone choses to write an interoperable PHP+MySQL one-user FAST version that uses the great UI? :]
Anyway, this is great. I hope the development and bug fixing speeds up now (although Ruby may, sadly, also be a hinder in getting contributors...)
Diaspora source released Posted Sep 16, 2010 20:33 UTC (Thu) by Trelane (subscriber, #56877) [Link]
Interesting post regarding Haystack: http://blog.jgc.org/2010/09/myth-of-boy-wizard.html
It's interesting IMHO (others may well disagree) to compare Haystack and Diaspora.
Diaspora source released Posted Sep 16, 2010 21:28 UTC (Thu) by kragil (subscriber, #34373) [Link]
What is your problem? It is not a scam (and never seemed that way, like you said) and it was never vaporware. They delivered exactly on the date they intended.
And there is no need to compare them to every possible failure. They don't promise as much as haystack. What they want to provide is much simpler IMO. They got a lot of money, so what? Sometimes people get lucky, deal with it.
Diaspora source released Posted Sep 17, 2010 0:40 UTC (Fri) by Trelane (subscriber, #56877) [Link] What is your problem?What problem do you think I have? Please be explicit, and cite your grounds for this assertion(quotation and link). It is not a scam (and never seemed that way, like you said)Actually, what I said is that it had some of the characteristics of a scam, but that did not necessarily mean that I thought it was a scam, let alone that it was. it was never vaporware That depends upon your definition, as we discussed in the other thread. Synopsis: We also disagree on the definition of "vaporware." Vive le difference! And there is no need to compare them to every possible failurePlease show where I've compared them to a few vaporware projects, let alone "every possible" one. They don't promise as much as haystack. What they want to provide is much simpler IMO.They were promoted as an open source (nee, Free Software) Facebook. I think that's similar. Not that the size of the project actually matters for the purpose of this discussion, but rather it's the secret development with lots of prominently-positioned promises and lots of media coverage (in a similar david vs goliath mode that was discussed in the article, see http://www.brandchannel.com/home/post/2010/08/30/Diaspora-Playing-David-to-Facebooke28099s-Goliath.aspx http://www.medienhandbuch.de/news/david-diaspora-gegen-goliath-facebook-41643.html and google for "diaspora goliath") that I thought was the parallel. Yes, as we both know and I am glad to find out, they released working code. Yay! Up until they did, they'd not, and it wasn't clear that it was necessarily going to happen, since it was, you know, developed in secret. Even haystack distributed sneak previews to select people. (see the linked article). They got a lot of moneyI don't see where the money enters into the argument here. Sometimes people get lucky, deal with it.What is this supposed to mean? Please be explicit as to what you're trying to state.
contributor agreement Posted Sep 16, 2010 21:46 UTC (Thu) by johill (subscriber, #25196) [Link]
So they have a contributor agreement: https://spreadsheets.google.com/a/joindiaspora.com/viewfo...
Anyone look at it yet?
contributor agreement Posted Sep 16, 2010 22:43 UTC (Thu) by corbet (editor, #1) [Link] It's a full copyright-assignment deal - they get the right to take your work proprietary or do anything they want with it. The patent grant looks...rather broad...it's hard to imagine many companies signing up for that.
contributor agreement Posted Sep 17, 2010 3:50 UTC (Fri) by jspaleta (subscriber, #50639) [Link]
Actually.... its a co-ownership arrangement between the contributor and the project...and not a straight up assignment of all rights to a single entity. So its a little more complicated a situation than just a straight copyright assignment because now multiple parties have the ability to re-license a particular contribution. Though I think your right in that another corporate entity would avoid contributing manpower to even if there was some marginal benefit to the rights maintained in a co-owership arrangement.
And actually... I would image violation enforcement actually gets much more complicated with co-ownership of a particular contribution because you'd have to check with all co-owners to make sure they haven't potentially licensed the work to the alleged violator. Even the more typical argument for assignment to a benevolent project entity to make enforcement easier doesn't necessarily hold with co-ownership.
-jef
contributor agreement Posted Sep 17, 2010 17:22 UTC (Fri) by codewiz (subscriber, #63050) [Link]
All contributor agreements for free software actually grant you co-ownership of your own code. Do they grant the contributor co-ownership of the entire codebase? Of course not.
So the agreement creates a huge disparity between community contributors and one organization that retains 100% of the code ownership, including the right to relicense it and sell proprietary versions of it.
Developers who care about retaining control of their online identity should think twice before signing a legal paper that makes them loose control of their own code.
contributor agreement Posted Sep 17, 2010 18:05 UTC (Fri) by jspaleta (subscriber, #50639) [Link]
No actually.. they typically don't typically use co-ownership mechanisms. Typically when assignment is required one entity is given full copyright ownership and the contributor is given a broad _license_ to continue to make use of the work and in some cases sub-license it (but not always). Ownership is very specifically bound up with who has standing to enforce breaches of copyright licensing terms. Typically in FOSS copyright assignment situations, the goal is to transfer that ownership to a single central authority.
For example, the FSF's assignment requirement for contribution to GNU codebases is not co-ownership. Its a full transfer with a broad license back and some additional language putting limits on what the FSF can do in terms of re-licensing your contribution. One you assign copyright ownership over to the FSF for a contribution for GNU, you no longer have standing to bring a lawsuit for violation of the licensing terms for that contribution if anyone violates the copyright license. Only the owner can do that, which in this case is the FSF. If you make use of that code in another project, you are only acting as a licensee of that code..not the owner.
I'm actually not aware of another project at the moment which uses a co-ownership mechanism in its contributor agreement.
-jef
contributor agreement Posted Sep 23, 2010 7:50 UTC (Thu) by ekj (guest, #1524) [Link]
That's jurisdiction-dependant, offcourse. For example many European countries have non-transferable "ideal rights" connected to copyright. And since these are not (even in principle) transferable, you can't possible sign away this fraction of your rights.
contributor agreement Posted Sep 23, 2010 8:27 UTC (Thu) by codewiz (subscriber, #63050) [Link] I think the right term is moral rights and, yes, these are generally inalienable.
contributor agreement Posted Sep 17, 2010 3:58 UTC (Fri) by wmf (guest, #33791) [Link]
The part that concerns me is "Diaspora Inc." Is this a non-profit or did people just donate $200K to a for-profit company?
contributor agreement Posted Sep 23, 2010 15:09 UTC (Thu) by cowsandmilk (guest, #55475) [Link]
Incorporation has nothing to do with type of corporation.
Bottom of fsf.org :
OH NOES, RICHARD STALLMAN TRICKED US ALL INTO DONATING TO A FOR-PROFIT COMPANY
Why? How? Posted Sep 17, 2010 6:12 UTC (Fri) by kripkenstein (subscriber, #43281) [Link]
I don't get it.
There are thousands of interesting open source projects, and lots of them in important areas. But Diaspora gets more attention than almost all of them, and got that attention even before a single line of code was released. Why? How?
Why? How? Posted Sep 17, 2010 6:22 UTC (Fri) by spaetz (subscriber, #32870) [Link]
The right time at the right place with the right message.
I agree, it is a mystery, but so are marketing departments.
Why? How? Posted Sep 17, 2010 10:46 UTC (Fri) by gowen (guest, #23914) [Link]
The right message, certainly.
The right place, well I guess lwn is a good place to discuss free software in general.
The right time? I doubt it. Facebook and twitter have reached the state where their own success makes them self sustaining. I can't use a social networking website different from the one that my (non-geek) friends use. That defeats the very object of us all being in one (virtual) place.
The privacy concerns haven't really dented that success, and for social websites, like auction websites, that mass of people becomes the drawer for others. Facebook has become a defacto standard for letting the world know you're making an omelette, and to displace it you'll need either full interoperability with it (unlikely) or an extremely large-scale, concrete privacy/personal disaster on FB's behalf.
Why? How? Posted Sep 17, 2010 12:20 UTC (Fri) by arctanx (subscriber, #59239) [Link]
Really, it was the right time. Either by planning or serendipity, the project was announced on kickstarter in the aftermath of some forced privacy setting changes on Facebook.
If even 0.01% of Facebook users say "this is kind of rubbish" and hear Diaspora's plan and like it... that's a lot of money that can be raised.
It was the right time not in the sense that they're going to _kill_ Facebook, but absolutely the right time for generating a huge amount of interest and funding as an open source project.
Why? How? Posted Sep 17, 2010 13:20 UTC (Fri) by gowen (guest, #23914) [Link]
You're absolutely right.
Why? How? Posted Sep 17, 2010 16:32 UTC (Fri) by elanthis (guest, #6227) [Link]
One of the points of Diaspora is interaction with Facebook and Twitter. Assuming those services don't shut down the integration hooks that Diaspora will eventually use.
Integration in general is pretty important for these sites, though. If you use Twitter, you're not going to want to also use Facebook statuses (usually), but rather just have your twitters go to Facebook. Likewise, people like to integrate Facebook chat with their other IM accounts.
At some point, the social sites either need to become all-in-one (and then suffer the fate of Google, which is to be hated by all the geeks despite not having done anything wrong besides being successful and ubiquitous) or they need to federate through open APIs allowing mash-ups of social content and interactions the way the users want them.
Diaspora is just another cog in that eventual great social website wheel.
Initial public source, not ready for prime time Posted Sep 17, 2010 14:03 UTC (Fri) by rfunk (subscriber, #4054) [Link]
I'm amused that people are complaining about implementation details of an initial code release. Nobody claimed that this was anywhere close to a final production release. It's not even an alpha release. It's a start, which people can build on, modify, or whatever.
And as for it building on Ruby/Rails, that seems like a great way to make the initial coding go quickly. Making it run fast can come later, piecemeal, and maybe outside contributors will help with that now. Remember the wisdom of Knuth: "Premature optimization is the root of all evil."
Initial public source, not ready for prime time Posted Sep 17, 2010 15:22 UTC (Fri) by pboddie (subscriber, #50784) [Link] I'm amused that people are complaining about implementation details of an initial code release. Nobody claimed that this was anywhere close to a final production release. It's not even an alpha release. It's a start, which people can build on, modify, or whatever. Sure, it's a noble goal and everything, but $200K later and we get to see a code drop that may only be replicating a bunch of existing projects? Someone's got a future in PR and/or marketing, that's for sure. (Seriously, that it's licensed under the AGPLv3 is enough for me as a Free Software advocate, but there are people of the "bazaar" persuasion who aren't particularly happy about the "initial code offering" tactics going on here. I can certainly see the point of their criticisms.)
Warning! Do not run this code! Posted Sep 24, 2010 7:22 UTC (Fri) by job (guest, #670) [Link]
This code does stuff like check your permissions but then ignore them when you actually do stuff, passing unescaped strings to the database, and so on.
You can read a few of these examples over at this article: http://www.kalzumeus.com/2010/09/22/security-lessons-lear...
If you are a bit on the cynical side you may say that apparently is was more important for this project to play with exotic frameworks and databases than produce working code. It's really that bad.
Diaspora source released Posted Sep 25, 2010 7:46 UTC (Sat) by steffen780 (guest, #68142) [Link]
I'm not qualified to comment on the code, but just out of curiosity, am I the only one confused by the fact that a project whose main purpose of existence is privacy would require lots of personal details for the privilege of helping them out? Seems a bit strange, especially given many governments' view of secure networks.
Diaspora source released Posted Sep 25, 2010 7:46 UTC (Sat) by steffen780 (guest, #68142) [Link]
Sorry I should've been more explicit, I'm referring to the contributor's agreement.
|
|||||||||||||