News for hackers. Stuff that matters.
Re:Diaspora Will Now Be A Community Project by chimeracoder on Monday August 27, 2012 @03:00PM (#4440242)
I think most of the lessons are non-technical, which makes them all the more valuable.

1. Diaspora had all the characteristics of vaporware for the longest time - I remember they were #1 in a HuffPo 'Top Ten Alternatives to Facebook' lineup in May 2010[1].

2. The source code released that fall seemed to fail Spolsky's second rule of being able to build in a single step, because I remember checking out the repo and then realizing that I'd need to dedicate more than ten minutes to figuring everything out, making a mental note to do that at some point,a and promptly forgetting about it entirely.

3. The officially-hosted version was a victim of poor timing - the first time I saw it was last summer (2011). Aside from the fact that this was more than a year after I had already seen it in a mainstream news source, this was immediately after the release of Google+ and at the height of the 'nymwars'. By that point, Google+ had already gained its reputation as a massive, overhyped soon-to-be-likely-failure[2]. Diaspora's UI was nearly identical to Google+, so it seemed less like an alternative to Facebook and more of an open-source alternative of an alternative to Facebook.

I don't think the technical spec of Diaspora as a product is broken at the core (though that doesn't mean that something like what you envision couldn't also be valuable as well). I do think that the end goal is worthwhile, and let's be clear: this does not kill Diaspora - it has every chance to be as much alive as Netscape is, just in a different form (ie, Mozilla/Firefox). That's the beauty of open-source projects, particularly community-driven ones.

[1] The lineup is of course indicative of nothing other than the fact that Diaspora was already hyped-up enough to be included in such a lineup!

[2] Whether or not you agree with that conclusion now, it was certainly a popular sentiment at the time.

Re:Diaspora Will Now Be A Community Project by epscylonb on Monday August 27, 2012 @05:50PM (#4441100)
Someone said the other day in the submission about Tent, they should have published a protocol first. It should have been designed before they wrote any code, app.net at least seems to have this in it's favour.
Re:Diaspora Will Now Be A Community Project by joe_the_user on Monday August 27, 2012 @03:33PM (#4440413)
It seems like you're flagging the results of bad development. The question we need to ask ourselves is, what leads these to unfortunate results. The decision to use two-stage build process might not be bad when a project is in that kind of shape. What leads a project to get into that kind of shape? Well, that is the question.

At the same time, I remember the first code-release of diaspora was before G+, it just wasn't usable code. Any project that takes a long time to create is going to have "timing problems". If Diaspora had smoothly working code right now, timing probably wouldn't be an issue.

Re:Diaspora Will Now Be A Community Project by taligent on Monday August 27, 2012 @05:15PM (#4440972)
Inexperience.

It was VERY clear that these were just a bunch of college kids who had no decent real world experience in software development. When you have a such a small project as Diaspora failing to get something the fundamentals sorted from day one is a bad sign IMHO.