LWN: Comments on "GNU Mediagoblin Project launches (NetworkWorld)"

GNU Mediagoblin Project launches (NetworkWorld)

ersi — Thu, 12 May 2011 20:29:27 +0000

There is a bunch of people that are trying to put together a software solution/bundle and support it on some kind of embedded hardware (as low-power as possible, for the always-on) which would support these kinds of peer-to-peer services and storage.

Check out the FreedomBox wiki-page (where it sort of started): http://wiki.debian.org/FreedomBox and https://freedomboxfoundation.org/ which is the newly founded FreedomBox Foundation page.

GNU Mediagoblin Project launches (NetworkWorld)

dlang — Mon, 09 May 2011 00:37:15 +0000

given that home DSL bandwidth is usually asymmetrical with much lower upstream capacity than downstream capacity, serving large images from a home server doesn't work especially well, videos are even worse.

GNU Mediagoblin Project launches (NetworkWorld)

martinfick — Sun, 08 May 2011 18:20:20 +0000

> ... (the AGPL means they can't make money on feature differentiation -- they have to compete on actual service instead.)

Which is unfotrtunate, since I suspect that this will make this project much harder to succeed. It is not truley open source (or free) if users cannot install software on their own site and use it as they please. I am having a hard time figuring out who this is aimed at with the AGPL if it is not for home users, and not for busineses?

GNU Mediagoblin Project launches (NetworkWorld)

danieldk — Sun, 08 May 2011 12:13:33 +0000

> Er, it's an open source project, so there is no need for "funding" or a business plan.

I think that's a bit naive, at least if you want to grow a sizable user group. First, the network effect of Flickr and especially Facebook should not be underestimated. People don't just post photo's there because they offer lots of disk space, but primarily because their friends use the same website. I had a Facebook account previously, and ended it. Sadly, this cuts you out of a lot of social interaction as well. Many dear friends keep other friends up to date where they used to use e-mail, SMS or phone before. Using some external site doesn't integrate photos in social interaction as well as using Facebook. So, you'll have to come up with some incentive for people to swap Facebook or Flickr for something new.

Second, 99% of the world doesn't run their own server. So, people have to resort to third-party hosting. Now, since competitors can obviously not compete on features (as you indicate), they have to compete on price. Ultimately, this will lead to near-zero margins. No entrepreneur in would want to be in that situation in their right mind. You could see this in web hosting. Nearly everyone was running Linux and Apache, but nearly every successful player differentiated with their control panels. The same thing is happening in the VPS space, many players use Linux plus Xen, but most of them use a proprietary management system. Third example: source code hosting. Years ago, everyone used Trac. Nowadays Github is far more popular, having their own proprietary web interface. Conclusion: running an AGPLed project with no differentiator other than price is not a good prospect for hosters.

Your Facebook example is not to the point, since they are almost the anti-thesis of Mediagoblin. Sure, it is free, but every user pays by handing over a lot of personal information for very directed advertising.

GNU Mediagoblin Project launches (NetworkWorld)

spaetz — Sun, 08 May 2011 10:55:59 +0000

> i just don't get it. the only differentiator i can see from the article is the promise of free software, and the ability to have federated servers/services. otherwise its just another flickr

Yes. :-)

> so how does this differ from stacking up debian, apache, and one of the
> bajillion gallery programs? lots of these already provide for privacy
> hooks. sharing is easy, you just make photos public and anyone can view
> and save them.

Checklist:
- The bazillion of gallery products scale to thousands of users and offer video upload too.
- Can subscribe to remote instances of gallery instances to follow pics of users on those remote instances.
- Can integrate with identi.ca and other systems implemening ostatus.net
- Are actively maintained by enthusiastic developers.

Cool project, bad name

Velmont — Sun, 08 May 2011 09:54:57 +0000

Cool. Been waiting for it.

But too bad about the name Mediagoblin, it sounds horrible! I really think GNU should become better at branding things. Most of my GNU stickers isn't very well designed, and some of the names here are really bad.

"Daisychain" was a working name for GNU Social (or a service based on GNU Social), and it sounds really nice. More good names, less bad ones.

I'm actually also building a photo managing+sharing site, and want it to use OStatus. However, at this point, I just need to build the features that none of the other FOSS photo managers have (I've tried them all, Lifebox was the best fit, but it was code rotten to the core, sadly - hard to build upon so bad code).

Anyway, I'll be following Mediagoblin, and when I build OStatus-support for my gallery, I guess it will be the first project I want to interface with (and use the same solutions, because I guess they will think it more thorougly through than me).

GNU Mediagoblin Project launches (NetworkWorld)

b7j0c — Sun, 08 May 2011 05:13:39 +0000

i just don't get it. the only differentiator i can see from the article is the promise of free software, and the ability to have federated servers/services. otherwise its just another flickr

so how does this differ from stacking up debian, apache, and one of the bajillion gallery programs? lots of these already provide for privacy hooks. sharing is easy, you just make photos public and anyone can view and save them.

GNU Mediagoblin Project launches (NetworkWorld)

dmag — Sun, 08 May 2011 01:20:28 +0000

> How will Mediagoblin deal with funding?

Er, it's an open source project, so there is no need for "funding" or a business plan.

The idea is:
- you can run your own server (from your bedroom or wherever)
- you can pay someone to run a server for you. (the AGPL means they can't make money on feature differentiation -- they have to compete on actual service instead.)
- disk space is getting cheaper, so there will be "free" services (If the product is free, then the "product" is you. They are selling your eyeballs and/or your data.) FaceBook is the largest photo sharing site on the planet and they don't charge a dime!

Unlike FaceBook (there can be only one), the MG protocol will be like e-mail -- you don't care who "hosts" other people's e-mail, only your own. In the same vein, everyone will have to "pay" for their own photo hosting.

GNU Mediagoblin Project launches (NetworkWorld)

danieldk — Sat, 07 May 2011 09:30:38 +0000

What is sorely lacking from the linked article and Mediagoblin website is a business plan. Since photos and videos require a lot of disk space and bandwidth, someone has to pay. If they will provide a subscription model (a la Flickr), there will probably some time where most people just test the waters (use a free account). Usually, a startup would seek VC funding to bridge this period.

How will Mediagoblin deal with funding?

GNU Mediagoblin Project launches (NetworkWorld)

jku — Sat, 07 May 2011 08:53:28 +0000

Yeah, that was my thought as well: media sharing is a fairly clearly defined problem, unlike "Libre Facebook"... There's one common problem both of these projects need to solve though: Who will do the hosting and why? Will they integrate advertisement services into the web UI or what's the plan there? How will users be able to select reliable hosting?

I think this is not one of the cases where "if you build it, they will come": Reliable, trust-worthy data and service hosting is a massively important piece of the problem that Mediagoblin and Diaspora intend to solve, and to be successful there needs to be a solution in place fairly early. I guess Mediagoblin/diaspora need an entity that's like wordpress.com to Wordpress...

GNU Mediagoblin Project launches (NetworkWorld)

mlinksva — Fri, 06 May 2011 22:01:37 +0000

Chris Webber, the lead MediaGoblin, says "If I had an LWN account I would respond to https://lwn.net/Articles/441844/ and say that's why we're going to use OStatus :)"

GNU Mediagoblin Project launches (NetworkWorld)

elanthis — Fri, 06 May 2011 21:25:04 +0000

In fact, services like Diaspora really shouldn't worry about photos at all. Build up a network of services to handle different tasks very well. Make integration smooth and seamless, of course.

But sriously, I want a dedicated kick-ass photo sharing server, and a dedicated kick-ass contacts server, and a dedicated kick-ass status/micro-blogging server, and I want them to just work together. I don't want one server that does all of them half-asses.

GNU Mediagoblin Project launches (NetworkWorld)

AlexHudson — Fri, 06 May 2011 20:14:35 +0000

Diaspora is only really tangentially aimed at sharing photos. Doing something a lot more specific like Mediagoblin is aimed at doing is, IMHO, much more likely to be useful quickly and more successful.

GNU Mediagoblin Project launches (NetworkWorld)

rfunk — Fri, 06 May 2011 20:07:25 +0000

Sounds very much like Diaspora, except not as mature. And Diaspora isn't very mature itself. https://joindiaspora.com/