@socrates maybe you've addressed this before so sorry if I've missed it, do DOI links go against the no shortened links clause of the community standards?

@drb Yeah good thing to think about

Not sure if they do the sort of problematic user-tracking that I'm worried about

I'll re-post with the link to the journal's site and do some digging, thanks!

@socrates, @drb: DOIs are supposed to be canonical pointers (is that the correct terminology?), so I don't think they should be thought of as "URL shorteners". Maybe more like 301 redirects or something?

I was thinking about the language from the community standards about obfuscation of the link target which occurs with a DOI. Most DOIs are reputable though I don't know about user tracking. However, I could start a random blog and use The Winnower or another service to point DOIs to it.

Yeah, good point, but I still think of DOIs as intrinsically different from data-harvesting URL shorteners.

@RaoOfPhysics @drb @socrates Agreed. I think HTTP DOIs should be allowed in a scholarly instance, unless we find malfeasance.

Also TIL about, which is a service offers to shorten DOIs that contain long expressive strings!

So returns a 301 to a URL that then returns a 302 to the actual resource.

Their example points at a Wiley resource, which gets angry at curl and eventually redirects to - heh

@socrates @drb Hmm. As the publisher of a few journals our library receives monthly reports from CrossRef of how many times each DOI was resolved. I suppose it's possible they do more problematic tracking (as they would have a good view of individual patterns across many journals). Will investigate.

@socrates @drb So notes that the HTTP request uses Cloudflare as a frontend, resolves the DOI, and sends a redirect (per @RaoOfPhysics ) to the target URL.

"curl" shows that Cloudflare sets a cookie that expires immediately with a 302 redirect (302 because the target URL could change as journal reorgs site or whatever)

CrossRef is the registration agency for and resolves most(?) DOIs and has a decent privacy policy:

@socrates @drb @RaoOfPhysics But currently it appears neither nor make any statements about avoiding the aggregation and resale of individual user data.

It seems like it would be a terrible idea for a non-profit to go that route and a perversion of the goal of providing a stable identifier for resources that change their "real" URLs over time, but...

@drb @dbs @RaoOfPhysics So based on the stuff in this thread and their policy, it looks like they're *probably* not doing anything nefarious in terms of tracking

There's still the issue of target obfuscation

Maybe a good policy would be to say that if one uses a DOI link, to clearly describe the destination in a manner that is human-readable (i.e. title, author, journal name, etc.), on pain of having the post deleted

And then if there's ever "DOI-gate," we change the policy


@bgcarlisle @drb @RaoOfPhysics That sounds reasonable. Hard to imagine a situation where at least some of the human-readable info (title / author) wouldn't naturally be present in context.

Maybe in all of our spare time (hahaha) we could teach masto to use the free CrossRef API to look up a DOI and provide the summary automatically? Small instances should be able to play nice at least.

@dbs @drb @RaoOfPhysics Oh I mean a post that's like:

> I love this paper!! dx.doi/NUMBER

^ This would get deleted

@bgcarlisle @drb @RaoOfPhysics Yep, a post like that would get my spidey-senses tingling. "Please repost with additional context"

Sure. I wasn't trying to advocate that DOIs should not be allowed. Moreso noting that DOIs seem to go against the current wording of the community standards.
@bgcarlisle @RaoOfPhysics

@drb Oh I understand & appreciate your identification of that tension in the policy!

It also led me to start thinking about how much user tracking may be associated with DOIs. More awareness is always a good thing.

And I didn't know about The Winnower before.

So today has already been very productive, although my comps prep hasn't gone anywhere yet.

On that note: *disappears for a while*

@bgcarlisle @RaoOfPhysics

Sign in to participate in the conversation
Scholar Social

The social network of the future: No ads, no corporate surveillance, ethical design, and decentralization! Own your data with Mastodon!