Re: FOSS-as-a-service
Anthony Towns <[email protected]>
| Newsgroups | gmane.linux.australia |
|---|---|
| Message-ID | <[email protected]> |
On Wed, Jan 20, 2016 at 10:47:38PM +1100, Tennessee Leeuwenburg wrote: > Essentially central-server designs seem to be meeting network/mesh/p2p > designs. > This is in some ways antithetical to FOSS principles, because whoever > provides the services effectively controls the project. To me this applies to most SaaS versus deploy-your-own. SaaS means you don't have control over the software -- if gmail's spam filter gets things wrong, you have effectively no options but to drop gmail entirely; if slack decides to send all your conversations to the FBI without a warrant, they're sent. Many good services let you export your data, which would let you spin up a similar service you do control and reimport your data, but since it's not open source similar isn't the same, so this could be a lot of work. The upsides of SaaS are (IMO) twofold: - it's straightforward to charge subscriptions and make money, which can fund further R&D, so SaaS gets more features and slicker packaging, which pretty quickly translates to "user-friendliness" and "easy for total newbies to set up and use" and so forth - as centralised services, it's easier to get network effects -- you join twitter, and immediately can participate in lots of groups that use twitter (ditto for slack); you join github, and you can easily browse other projects and what other people are working on > For some reason, we have opted to remove cost and risk from the individual > by moving the infrastructure responsibility and legal hosting onto private > companies. Do we need LA or EFF to host a gitlab instance in place of > github and move FOSS hosting onto a truly libre platform? I think hosting a copy of repositories at github is a good thing; it makes it easier for people to find the code and hack on it, and if you're only using the git hosting part, rather than issues, pull requests or the wiki, there's no lock-in, or even much loss if github goes down entirely. > Should we all pay > a tithe to an agnostic infrastructure hosting context in order to reduce > the influence of money? I (personally) don't think it's about reducing the influence of money. (Others do, of course) I think it's just about supporting the basic open source / free software ideal that as a user, you should be the one in control of what your software does, and as long as you're willing to actually change the source code (or hire someone to do so), there shouldn't be anything else limiting what you can do with the software you use. "The freedom to run the program as you wish, for any purpose (freedom 0)." -- http://www.gnu.org/philosophy/free-sw.en.html "must allow modifications and derived works, and must allow them to be distributed under the same terms" -- http://opensource.org/osd > Or, is the ability to draw a rent from hosting > funding innovation into gratis tools, providing a genuine commercial > challenge to the axioms of FOSS software in providing a gratis solution to > users? I do think this is (more or less) the case. It's easier to make money with SaaS (and you don't even have to worry about piracy unlike distributing proprietary software), and it's easier to get people to spend time working on software when you've got money to pay them. And working on software is how you end up with software that gets stuff done. Finally, people are trivially willing to give up "fundamental freedoms" in order to get stuff done (especially in the short term). Cheers, aj