But it makes me wonder. What Maciej is effectively doing here is asking for donations. Which is great. But the fact that many Pinboard users would willingly upgrade to a "paid" plan as a token of their support indicates to me that while at the 10,000 foot level signing up for Pinboard is just a market exchange like any other, at a closer level Pinboard users are members and not just customers. This despite the fact that Pinboard eschews the social features that are common among other bookmarking platforms!
I bring this up because this kind of thing arises in open source software development as well. For instance, when the developer of htop disappeared for a while, and the community forked it. But we (Internet culture) have not developed the same approaches to handling administration of services that are useful to a group of people. This surprises me. I think there's room for some movement in this direction, where a group of people can maintain a service that is useful to them and made available to the whole group. Perhaps various chat servers / Mastodon approximate this, but even in this case they're often run by individuals and susceptible to the same kinds of outages.
First I have a feeling that many Pinboard users are like myself and partly pay for Pinboard because Maciej runs it. After all there are many bookmarking options; I do get some utility out of not having to maintain a self-hosted service but all in all, I think we shouldn't discount the subscription-as-patronage effect here.
Regarding member-controlled administration of services, there is an obvious model here: worker co-ops. They don't seem to be that common in tech though . Pubnixes (classic example, SDF) are kind of close perhaps?
While closing tabs recently I stumbled on a particularly interesting (to me) HN discussion (https://news.ycombinator.com/item?id=25506007) from a couple months ago.
I'm taking the following comment (https://news.ycombinator.com/item?id=25506951) out of the context it was posted in, but I don't believe this particular case takes liberties with the original intent:
---
> analog31
> ... Economics is largely about an emergent phenomenon, thus falling squarely on the "is" side of things.
> Emergence is a relatively recent concept in the linguistic timeline. For most of human history, everything seemed to be either of human action and human intention (turning on a light), or neither of human action nor of human intention (the weather). As such, languages did not emerge (ha!) that easily expressed emergent behaviors, things that were of human action but not of human intention (traffic jams).
> Similarly, one can see this with the subject of biological evolution (another emergent phenomenon). It's been nearly two centuries since Darwin got that ball rolling, and still almost any discussion about evolution is rife with inaccurately implied intention.
> One other consequence of this linguistic difficulty is that it leads people to propose solutions to perceived problems as if they were just about human action/human intention...
---
I think emergence is perhaps the most intuitively accessible way of reasoning through the impact of bureaucratic network effects. We've simply never before in history had to deal with human social cohesion at the scale we do today, so we've never reached the point where we've *had* to understand the dynamics correctly in order to make any progress.
Compare it to a common sentiment I saw on HN recently, which is that selling a self-driving car kit is irresponsible even if you technically have some shrink wrap agreement that says “this isn’t a real product, use at your own risk.”
I don’t know where the line is in each case, but I think that there is some responsibility one takes on when one accepts money from the public for a product, and no amount of cute warnings makes that responsibility go away.
I don't think this example is comparable. Most obviously, the effects of failure in Pinboard vs a self driving car kit are vastly different.
As much as I love maciej's blog and twitter, Pinboard has been pretty flaky for the time I've used it. I signed up for Pinboard in 2015, have an archiving account prepaid to 2024.
The API isn't reliable, archiving is lucky if it gets most of the page, it's not been a great experience so far. I used to use Pinboard with IFTTT and Wordpress to autopost favorite links with a description; two years ago I removed the Pinboard link to my IFTTT account because it just wasn't reliable.
Happy to hear of upgrades, one thing I would like to see is getrevue.co integration and full page screenshots of bookmarked pages (also accessible from API). Full page screenshots alone would justify my money.
If I had been paying $22/year at the time I would have been a lot less happy and would probably have closed my account.
The crux of the issue is that Pinboard's centralized architecture is unsustainable in the long run. It's essentially an on-demand archive.org service with the same ever-increasing storage requirements, which becomes a bigger problem as more users join.
It must be difficult for a one (or two) person team to manage, especially since they also chose to manage their own hardware. I assume this is to avoid cloud storage and hosting costs, and I respect the decision, but this can add a lot of overhead on top of just keeping the software performing optimally.
The sustainable model for this type of service is self-hosted. https://archivebox.io/ seems like a nice solution I've been meaning to try.
It's clear from this thread that there are users who have thousands of bookmarks, using dozens of GB of storage, and those who use it much less and whose storage costs are negligible. Having either group pay a flat rate doesn't make sense to me, and you end up losing revenue either way.
A tiered model could allow you to offset the cloud storage costs, maybe migrate things into the cloud and ease some of the maintenance burden on yourself.
Larder is so small, there’s no real community around it. I think there’s a cli on GitHub.
Now, the question is, what type of SaaS service can you make, to find 10,000 people paying $25/year. One that hasn’t already been consumed by the big players.
The business we’re discussing here is a good example of this process.
If half of them has even a basic question, that's already 5.000 answers to deal with.
In the past year things have happened like the API returning invalid JSON for weeks[0] and the search returning an error whenever I wanted to use full-text search (one of the selling points for the extra archival fee.) Not to mention the ongoing issues with page archiving in general (things frequently fail and need to be manually re-queued).
I really like Pinboard, I do, but the reliability issues have made me want to look somewhere else. My subscription is set to renew at the end of March and after seeing this I guess I'll give it another year. Here's hoping things trend upwards.
[0]: https://twitter.com/Pinboard/status/1274055573902774272