Garry's Posthaven | Hacker News Reader

Garry's Posthaven

garry.posthaven.com

41 pointsdmor38 comments

I hate sounding like a broken record here but if you don't meet the following criteria then stop saying you're going to preserve users' data forever.

* Have a full featured API right out of the gates

* Outline how users get their data out of your system

* Open source the code to ensure the ability for it to live on

Pledges are cheap.

Not as cheap as dismissive comments on forums.

How's this dismissive?

I've blogged and posted about this topic for a while [1]. I've also spent the last 2 years full time on a project that relates to data ownership and portability [2][3]. For those of us who write software specifically designed to solve this problem it's unfortunate when others make the same claims based simply on promises.

[1] http://blog.theopenphotoproject.org/post/10537443380/namespa...

[2] https://github.com/photo

[3] http://www.shuttleworthfoundation.org/fellows/jaisen-mathai/

Your suggestions are absolutely valid, and I appreciate them Jaisen. Your efforts with the Open Photo Project are noble, and I respect them a ton.

Give us some time -- Posthaven was a vague idea just a few months ago. We are under the gun to ship software at the moment.

Appreciate the response. I do hope you get around to adding some of these features. The Internet needs services like that desperately.

But until those features are released I have to remain a cynic :).

Ping me on Twitter (@jmathai) if you wanted to grab coffee. I never pass an opportunity to chat with someone about the preservation of digital archives.

Providing fully featured API out of the gates for a service which got launched in a hurry (at least that's the impression I got) would probably be a bad idea considering the early nature of the service. I'm confident that Garry will provide one but API is something people shouldn't rush too much with. I'm a firm believer of shipping fast but with APIs I tend to be cautious.

As for feedback to Garry: it would be nice if the line width would be shorter or text being larger, I find it myself playing tennis a bit (fan of large article type here)

I just think when you make a pledge to users that you'll safeguard their data forever that you need to think through a few things that their team hasn't yet.

Postponing the forever message until the underlying features to support it are built would go a long way.

Personally, I don't see the issue in launching with an API early. We launched the API for OpenPhoto well before we had any UI.

I wonder if the posthaven founders have decided to make this into an open source like "foundation" (garrytan had mentioned interest in an earlier comment [1])

I think a lot of people could get behind the idea of a permanent entity that wasn't a for profit corporation. There is a strong case for building something that endures, something that can be counted on to last for decades if not centuries. Fiduciary responsibility means that corporations must sell their assets or shut down service if it's the most profitable thing to do. Not to mention that companies can wind up selling all kinds of data and usage information to whoever in a fire sale scenario.

The world needs more services with incentives aligned to long term investment and global benefit as opposed to quarterly earnings numbers. Plus posthaven founders are a rare type of people with a record of making cool things. They have the ability to put together something that will be as useful in 2065 as it is today.

A sort of "Posthaven Permanence Foundation" would be interesting. The key insight is that it might be valuable/profitable immediately, or even in the next 10 or 20 years. But if my back-of-the-envelopes are correct, posthaven could run for a very long time on a relatively small amount of capital. Organizations that are built around long term value have a way of winning in the end.

[1] http://news.ycombinator.com/item?id=5229862

I have some suggestions for Garry:

http://www.onebigfluke.com/2013/02/blog-platforms-will-not-l...

"To actually commit to the claim of being 'the most durable blog platform ever created' and 'made to last forever', a blog hosting service should either:

A) Require new users to bring their existing domain.

OR

B) Have users register for a new domain during sign-up. "

OR they could agree to 301 your URLs even after you leave (maybe for a small, one-time fee).

301s assumes their host is up. Services disappear within 24 hours:

"The shutdown will be quite abrupt: Memolane accounts and content are set to be deleted tomorrow"

http://techcrunch.com/2013/02/21/memolane-the-internet-time-...

Seems like 1.) they let their users know about a month ahead of time and 2.) they were only a secondary aggregator of existing content that wouldn't be lost by them going away.

Your point still stands, that's just a poor example.

This is something Brett and I have discussed already. It's a good idea.

If you've ever accidentally let a domain lapse, it's pretty clear using your own domain is not necessarily more durable. Unless of course we figure out some way to auto-renew the domain, which is a possibility.

My domain is registered until 2021, and it only cost me as much as it would cost for two years of PostHaven.

The significant problem with long-lived domain registrations is not the price, but the potential that people tend to forget about the renewal date, move email address and credit card number over that 10 year period, and lose the domain when it comes to renewal time.

If you're talking "forever", or at least a lifetime, a 10 year cliff is a long time but not that long.

You could do things, like actually turn off the domain service after 9.5 years, giving people a 6 month retention reminder of the blunt sort, but there's not much clever that I can think of.

Well, I'm going to extend it to 2022 by the end of this year. I'm not advocating forgetting it for half a decade, but having a backup plan in case you do.

That's my suggestion to you. Make people sign up for your service and domain in a single, auto-renewing transaction. Hold the domain for them if it's going to lapse. Make it really hard to mess up as a user.

The day you do finally shut down Posthaven-- which will happen-- this will give all your customers a built-in migration strategy for their URLs. I think that's a stronger offer.

Does forever mean for as long as I'm paying my monthly fee? Or does cancelling my account- eg, due to death - simply stop update - previous posts will still be accessible to readers?

App.net has evolved into a very good micro-blogging service, so why not use App.net instead? You can upload photos and post long essays on App.net. And App.net costs $3 a month, instead of $5 a month.

But I would not choose App.net simply because it is a little cheaper than Posthaven. I would choose App.net because the guy who founded it, Dalton, seems to have some kind of long term vision. There is a heavy focus on the API as the future to a platform that goes beyond what Twitter is, and goes beyond what Posterous was. That longer term vision is enticing, because it also offers slightly more concrete possibility of longevity than the mere verbal promise offered by Posthaven. After all, the only way any of these services can offer longevity is by surviving, and the only way they can survive is to have an idea that really is adapted to what the world needs next.

BTW, I'm looking at App.net right now and it says it's $5/month. Price increase?

Or $36/year which works out to $3/month.

I love Dalton and App.net and am a supporter myself.

We certainly have a road ahead of us to show how we will make it to forever. For now, we're building software.

When this first came up I was trying to remember the last time I visited a blog that is hosted on Posterus, then I discovered that Ycombinator's blog[1] is still on Posterus, as is AirBnb's[2].

[1] http://blog.ycombinator.com/

[2] http://blog.airbnb.com/

I'm going to move the YC blog shortly.

I'm sorry but their pledge does not say anything about what if they shut down?

If they are trying to improve on Posterous they need to be up front with the fact that Posterous is closing and people don't want to move again in an year year when Posthaven shuts down.

Businesses close, people die. If you are going to make a promise, make it worth it.

I just built Urgeous, a (very beta) Posterous clone, and describe my motivations here: http://urgeous.com/p72t3aaa40h-mourning-posterous-how-and-wh...

An API is coming (later today!) that will let you serve the posts from your own server, while Urgeous deals with storage and the like.

The (current) plan is to stay free for things hosted on the Urgeous domain, and to charge for the API (other suggestions welcome).

But since it's still very beta, the API will also be free when it opens.

Check it out! Send a message to post@urgeous.com (supports Markdown)

</plug>

... and here's the API: http://urgeous.com/p83t3aaa410

If you have your own domain, you can always export your blog to another service, so what's the big deal? Also, if you want a reliable hosted service, you can always use the free Wordpress.com or Blogger instead of some new startup's offering. I don't think Wordpress or Blogger will be closing down anytime soon.

Note that we're not a startup. Our goals are to serve users and make software. If it made no money at all, I would still work on it.

It would be great if they can be successful at making a company where you can be confident that your data will not die; but flexible enough for the inevitable sampling of future 'bloggish' platforms that come out.

Question: How easy will it be to move the data out of the service at any given point in the future?

Never as easy as copying files off of a web host.

Curiously, the links on the top left (Funny Stuff, Photoblog...) still point to Posterous...?

The pricing reads like it will cost $5/month forever. I suggest at least adding inflation.

From the pledge (https://posthaven.com/ourpledge)

>Will it be $5 a month forever?

>As long as we can! We'll do our best to keep it as reasonable as possible. If costs go up, we'll have to raise prices. If they go down, we'll lower them.

>We believe fair's fair -- and we'll always be fair.

I suggest they remove the "forever" next to the price to make it clear the service will last forever, not the price.

$5.00 in 2006 had about the same buying power as $0.43 in 1930, so $5 in 1930 is equivalent to $58 today.

The cost of bandwidth and storage is dropping at a faster rate than inflation. Besides, I'd say 90%+ of the blogs on a service like this would be mostly text and low traffic, and $5/month can cover that today several times over.

That said, I'm not sure I'd trust another blogging service by the people that created Posterous. They already had a chance to demonstrate that they would keep the service running for the long haul and they didn't. I'd suggest people take their $5/month and spend it on Wordpress.com.

Reply on news.ycombinator.com