Just focus on being a maximally useful media hosting server, with a really simple accounts system for admins and so forth.

Make it easy for me to stand up a server, upload a video of my kid, then share the link to that video on the internet.

Leave all the community/social/commenting stuff to community/social/commenting software.

I'd rather have both an open-source media server and an open-source equivalent to twitter, than have both of them rolled poorly into one.

Not OP. But the argument seems to be: Start by being maximally useful to one person. When you have enough people using your software that they want to use it together, that's the time to start investing a lot of effort in helping them do so.

The Wordpress model - only better?

Put together a distro with a complete solution - mail server, web server, media and social servers.

Make it one-click easy to set up. Keep security tight. Maybe do some deals with cheap hosting providers.

Then federate and connect.

Isn't that what Sandstorm[1] was trying to do? It was a very cool project, but got hung up on enterprise sales and ran out of steam.

[1] https://sandstorm.io/

And still is a very cool project, just moving a bit slower now.

I just moved my calendar from a large email provider to a self-hosted Radicale server on Sandstorm (syncing with Thunderbird, and DAVDroid on my phone). Small steps, but it was surprisingly easy.

MediaGoblin is available through Sandstorm: https://apps.sandstorm.io/app/70awyqss6jq2gkz7dwzsnvumzr0725...

Oddly enough, this is what I was trying to do with my Defiant project (http://BeDefiant.io). I wanted it to be a one-step installation framework to combat proprietary systems (better description on the website).

The bad news is that I am the only person working on it, and while I am an experienced developer, I am also a grad student.

In other words it is not ready for production at all. I have presented it to quite a few people for the purpose of finding help in the development, and everyone thinks it's a good idea, but everyone also wants to see a finished version of it before seriously considering helping out. Unfortunately, I can only work on Defiant in my free time, so developing the basic functionality is steady, but slow, progress.

I love the motivation, but this is way too generic - I don't see anything that distinguishes it, for this use case or honestly for any other, from any other JS application framework I might choose, and there are a lot of other ones out there that are much more thoroughly baked than this one. (Another way to say this might be "it's a good idea, but I want to see a finished version of it before seriously considering helping out.")

Sorry to be that guy, and I know it's not a nice thing to hear. But I'd hate to see you burn a ton of time on something that's born under the cloud of "there are a million other ones just like it and no reason anyone can see why this one is worth investing effort into". I'd rather see you take all that effort and skill, find a project that's already got some momentum behind it, and get behind that and push - everybody saying "I'll build my own!" is half the reason we've got where we are in the first place.

On the other hand, maybe your framework has a genuine USP that I'm just not seeing. In that case, you'll have an easier time getting interest if you find a way to surface that, front and center, so that even the most disinterested professional engineer who's skimming HN by way of procrastination before the next code review can't help but see it.

Either way, hope this helps, and good luck in your future endeavors!

> Put together a distro

I think this is kind of a good point. It's too hard to set up these 'solves-one-thing-well'-suits to work together in the right way. Having collections of services that are super simple to set up and works well together is probably a good way to do it.

Like one media platform, one search engine, one social platform. Or perhaps having a couple of alternatives but having integrations in between up and running in a plug and play way.

Then again, if you want all this to work in a decentralised fashion you probably have to think about how to achieve that early on.

> I feel like I want to strongly agree on this sentiment but what is the concrete point here? What should be the focus of projects like these??

You could say that Mediagoblin has 2 stated goals:

Stated Goal 1) Organize and share user media, which is what people actually want it for.

Stated Goal 2) Federated social something mumble, which no real world users actually want but which free software projects seem to get big e-rections over anyway.

I think that Mediagoblin should have 1 stated goal, the first one.

All news for the past several years has been about goal 2, and I just don't think that anyone in the real world actually wants or cares about that. Most people don't even know what the hell federation is supposed to mean. People just want to be able to stop relying on Flickr to organize their damn photos. Chasing this federated social something mumble dragon has seriously gotten in the way of making the thing that people would actually use, and it's frustrating to watch because there was so much potential.

Their official linked-from-the-landing-page "Easy Deployment" page (god, I'm already terrified by the word "deployment") has bold "WARNING: OUT OF DATE" messages all over it.

Their official site for (Not-Easy) "Deployment", right at the very top, says "GNU MediaGoblin is fairly new, and so at the time of writing there aren’t easy package-manager-friendly methods to install it."

The project is now 6 years old. 6 years isn't new anymore. There isn't an installer?

Federation is important because the average web user today is less likely to regularly visit individual sites than in the past. People tend to only visit new content if it shows up in their social network's feed. So, if you create content with a tool, but that tool doesn't post the content across social news feeds where people will actually find it, what's the point?

> Federation is important because ...<something entirely unrelated to what mediagoblin does>

What problem is mediagoblin actually trying to solve? Is it organizing your media or is it being a new social network? Because one of those makes sense and one of them does not.

> So, if you create content with a...

Mediagoblin isn't for creating anything, and it certainly isn't a network client. It is (ugh, should be) a tool for organizing your media.

Know what's a really great technology for sharing media? Hyperlinks. We built a visual web that is literally world-wide using them. Know what people don't have any trouble sharing media to using hyperlinks already? All of their social networks. Know what hyperlinks don't need? Years of effort spent chasing a white whale.

Mediagoblin still has a chance to be OpenFlickr if they would just stop faffing about with things that don't have any connection to it and start focusing on core competencies and important things like installability.

I think there can be a useful interpretation for the federation idea though. There is sort of a problem with there being an over abundance of communication channels that makes the social media landscape very unweildy. Not sure if this actually address this kind of problem though.

Can be, but I just cannot see how a media organization platform has any sane role in that.

Instagram is both a "media organization platform" and social network. 700 million people each month consider that a reasonable combination of features in a single app. The efforts behind Mediagoblin think similarly.

One use case I imagine is to add media files to my local Mediagoblin instance, and then have them show up on my friends's federated social media feeds on various platforms (GNU Social, Mastodon, Pump.io, etc.).

> Instagram is both a "media organization platform"

In what universe is that? Certainly not the Instagram I've ever seen. Mediagoblin's core compentency is 10000x more like Flickr than like Instagram. Now maybe that's not what they want to make, but damnit it's what they present themselves as trying to make and failing to make major progress on.

> One use case I imagine is to add media files to my local Mediagoblin instance, and then have them show up on my friends's federated social media feeds on various platforms (GNU Social, Mastodon, Pump.io, etc.).

Look, that would be a great secondary feature to add onto an already functional and installable tool. But you make the darn tool first and try not to invent imaginary problems! Because GNU Social and Mastodon are social networks, and Mediagoblin is a tool. Do you know what you can already use to post images to social networks with tools? An authentication API and hyperlinks. Not a database redesign.