|
|
| Line 1: |
Line 1: |
| {| | | {{autonomous|Meeting notes (websites)}} |
| |-
| |
| ! Service
| |
| ! personal data?
| |
| ! public/collaboratively-created data?
| |
| ! source code?
| |
| ! permanence?
| |
| ! identity?
| |
| ! comments?
| |
| |-
| |
| | Wikipedia || personal data? || public data? || source code? || permanence? || identity? || data and software are freely available. How are APIs licensed?
| |
| |-
| |
| | geonames.org || personal data? || public data? || source code? || permanence? || identity? || comments?
| |
| |-
| |
| | hostip.info || personal data? || public data? || source code? || permanence? || identity? || comments?
| |
| |-
| |
| | gmail || personal data? || public data? || source code? || permanence? || identity? || server source is proprietary, but all (email) data is available via standardized/open protocols, and users can make the service transparent (from an identity perspective) by using other email addresses.
| |
| |-
| |
| | ebay || personal data? || public data? || source code? || permanence? || identity? || comments?
| |
| |-
| |
| | Amazon EC2: || personal data? || public data? || source code? || permanence? || identity? || server images are ~ standard xen images. If you can move your data and computation transparently between S3 and completely Free xen-based servers, is it open/free/autonomous service?
| |
| |-
| |
| | GNU Savannah || personal data? || public data? || source code? || permanence? || identity? || comments? | |
| |-
| |
| | Second Life || personal data? || public data? || source code? || permanence? || identity? || client is Free; service is not.
| |
| |-
| |
| | audioscrobbler/last.fm || personal data? || public data? || source code? || permanence? || identity? || comments?
| |
| |-
| |
| | altlaw.org: || personal data? || public data? || source code? || permanence? || identity? || comments?
| |
| |-
| |
| | chicagocrime.org (defunct, mashed up government data (free) with google maps (non-free) || personal data? || public data? || source code? || permanence? || identity? || what happens when you mash up Free data with non-free services like google maps?
| |
| |-
| |
| | [http://www.openstreetmap.org/ OpenStreetMap] || personal data? || public data? || source code? || permanence? || identity? || comments?
| |
| |}
| |
| | |
| == Current services and how they rate ==
| |
| | |
| === Wikipedia ===
| |
| | |
| Data and software are freely available. How are APIs licensed?
| |
| | |
| * users have very little individual autonomy in the context of wikipedia (you can't reorganize things unless "the community" agrees
| |
| * many of the images are itself not free and cannot be downloaded in total (this is a technical issue -- even text hasn't been available for the past year because exporting the data takes three weeks). To what extent is technical failure in implementation a freedom problem?
| |
| * wikipedia doesn't do as good a job as it should or it could
| |
| * users of wikipedia can't add things to the site or change things
| |
| * people can't take their reputation with them and move it there: if you moev to a different place, it's hard to link your edits on wikipedia to the things you've done in a different place
| |
| | |
| Kragen says these issues are created by centralized nature of WP. There's an opportunity to charge rent. Providing dumps is not a priority because it's not the normal way of using the site.
| |
| | |
| But decentralization has its own costs.
| |
| | |
| Some of the obstacles here are purely technical. Wikimedia sometimes has trouble running normal Wikipedia. This is why Kragen wants to focus on technical aspects.
| |
| | |
| Sometimes it's hard to know exactly why any service provider isn't providing a certain service: technical difficulty, financial incentive, just not a priority, etc. Mako says it's okay to impose inconvenience (freedom standards) if it's worth it for the community, but of course deciding that is hard.
| |
| | |
| === Geonames ===
| |
| | |
| http://www.geonames.org/
| |
| | |
| Mostly a data server for geographical names. All data available under the
| |
| CC Attribution license, version 2.x. Unknown whether the software is Free.
| |
| | |
| === Hostip.info ===
| |
| | |
| ''skipped until someone knows those''
| |
| | |
| === GMail ===
| |
| | |
| Server software is proprietary (but we assume it is based at least in part on some free libraries)
| |
| Data is available via free protocols (imap and pop)
| |
| You can send with SMTP -- you can use GMail as your e-mail provider without using their web client at all
| |
| You can relay data through GMail as well.
| |
| Contacts data is exportable.
| |
| | |
| Unlike HOTMAIL and Yahoo, it runs mostly on the client side (via JavaScript) that allows you to customize the way it works in deeper ways (i.e., [http://googlesystem.blogspot.com/2007/11/gmail-api-for-greasemonkey.html with greasemonkey and similar systems]) that you can with previous system. To a large extent gmail is already a distributed application but it is set up so that all of hte code can be compromised by Google.
| |
| | |
| Kragen: In that sense, a Gmail has been a step forward in practical software freedom and many people have been distributing modified releases and Google is sending an XML object that it can manipulate.
| |
| | |
| 15:21 <@AaronSw> I want to qualify Kragen's praise of Gmail -- Google controls
| |
| the API in such a way that their particular interface style is
| |
| fairly tightly coupled to the server
| |
| 15:22 < luisv> agreed
| |
| | |
| James: People talk about having privacy from Google. When you use Gmail you give up privacy google but what you get is privacy from their employeer or institution. Google might sell your marketing data but your boss is might fire you for using your email for disapprovedb purposes.
| |
| | |
| Luis: Their underlying OS is highly customized GNU/Linux; Google engineers swear that the source is not really usable to people in the outside world without that OS. If we assume there are thousands of man-years invested in the difference, how valuable would a source release be? Traditionally we assume we have a copy of the OS.
| |
| | |
| Things Gmail is doing right:
| |
| | |
| * object code is semi-distributed -- the JavaScript is compiled
| |
| * modification of the client code itself, which is quite clunky
| |
| * some identity portability (if you want it and are willing to put in some work)
| |
| * they use a standard api
| |
| | |
| Sets of wrongs:
| |
| | |
| * access to code might not give us much because the OS itself is not something that we have access to
| |
| * there is some data that is in there that is hard to get out (some of things that the address book can getout)
| |
| * arbitrary account cancellations (your account may be cancelled at any time and its hard to fine out why)
| |
| * security problems that allow people to get into your account and change your password
| |
| * it was rumored that google was crawling urls
| |
| ** the above is a rumor and we don't know whether or not it's true -- this reflects an accountability problem
| |
| * javascript client code is obfuscated
| |
| | |
| === eBay ===
| |
| | |
| Reputation and identity are tied to the platform.
| |
| | |
| you can take your name out
| |
| | |
| how could a group move out of ebay?
| |
| | |
| ebay's south american operation has much stricter control over one's identity. attempts to block transactions outside the system to block use
| |
| | |
| No code released, lots of restrictions on user actions, user communication.
| |
| | |
| How could we replicate this service in a free sense? Especially in a world where there's a lot of network effects behind eBay. You can sell your stuff on your own site but it's a lot harder to get buyers. This even influences the market -- prices tend to be better on eBay because there are more sellers, etc. eBay interacts with PayPal, which gives them a lot of competitive advantages.
| |
| | |
| Network effect gives a huge benefit.
| |
| | |
| Could you have a free alternative: there are issues of financial services; credit card services and such requires a high degree of trust which is based on a high degree of control
| |
| | |
| === LiveJournal ===
| |
| | |
| LiveJournal was the blog before blogging was popular. Much of their code is released under the GPL. Small groups of users have been frusterated with governance and forked (e.g., dead journal). You can export all of your posts but it's not possible to export comments.
| |
| | |
| If you are a talented enough programmer, you can mirror everything including all of your comments. However, because that is not an easy to use application it is a technology have/have-not applications. (Luis and James argue that this part of it must/might be out of scope for the goals of this committee)
| |
| | |
| * you can't take your social network with you (no ability to export relationships)
| |
| * Changing terms of service over time, even after promises to the contrary exacerbates the problem.
| |
| | |
| however, it is possible to compete with livejournal because we have the code and are able to do it
| |
| there are documented cases of the livejournal account changing
| |
| there are also issues of dead livejournal users or just people who have moved on
| |
|
| |
| === Second Life ===
| |
| | |
| kragen:
| |
| | |
| * unlike most other network services: they declare that you own everything that you make in the service
| |
| * they sell virtual land in second life that you can continue to use without continuing to pay them (except for the monthly fee)
| |
| * they are trying to turn second life into a decentralized service (they want to be able to put land on different servers around the net which is part of their strategy to ensure that their users have the types of freedom that we are used to in free software)
| |
| * they promised to let users sell copies of scripts, textures, etc. without letting other users copy those -- but now that's technically impossible with a free software client
| |
| | |
| Mako: one big diff. b/w SL and other things: people own land. It's basically theirs. It's easier to say "You have your own server where you have control." Analogs to more traditional forms of property make the libertarian position more feasible. But managing the community spaces is harder. (I'm having trouble here; mako should flesh this out.) If people don't like your island it may be possible to move it to a different server (at least in the future) -- while still interacting with your old friends, etc.
| |
| | |
| Mako: If SL succeeds in all their wildest dreams, are we free?
| |
| | |
| Kragen: Person who owns your rack can spy on it, etc. (Kragen should flesh this out too.)
| |
| | |
| === Savannah/SourceForge ===
| |
| | |
| Mako: There are particular problems with using non-free services to develop free software. SF.net goes back and forth between releasing their code as free software. Currently it is not free. Savannah is based on a fork of previous SF.net code.
| |
| | |
| Luis: Savannah is interesting because all value there is communally owned in a peer sense. There are individual copyright holders, etc., but all the data is communally owned, if source is available, to what extent do I need to be able to copy everything practically?
| |
| | |
| Kragen: There was a security breach, services were unavailable, and that caused forks.
| |
| | |
| bkuhn: We had basic consensus on the GNU side to keep things down, but probably hurt the broader (hosted non-gnu projects) community as a consequence. We couldn't get things to them because of lack of resources. This is an inevitable consequence of a centralized service, even with the best of intentions. In centralized systems, the consequences of failure scale with the project, even when intentions were clear that a desire to export to the community in the right way existed.
| |
