Hacker News Reader

Comment by olliej | Hacker News Reader

Elsevier requires that authors do not do that - they can only provide copies of their papers in response to requests.

The last time I published in an Elsevier journal, I was allowed to deposit it in my institutional repository with a 2-year embargo period. Still sucks, but if your repository software supports embargo periods (my university's did), it means you can deposit now and it'll be automatically switched to open public access when the embargo times out, which is at least better than never (and less error-prone than hoping I remember to post it in two years).

I also posted the PDF on my personal website immediately, on the theory that they are probably not going to sue me personally over my own paper. Not impossible, but I haven't gotten a C&D.

Does that include HTTP requests? :)

This could be an extremely interesting legal argument, but I suppose the language used requires a request in 'writing' of some sort, which I suppose HTTP requests would fail.

They write your a nice letter with “GET /paper.pdf HTTP/1.1” ;)

how about a request form that ostensibly takes an email address and a name, and an optional text field and then automates the process of sending the paper (or a hidden (unique) link to that paper) to that email address. it need not store those details anywhere, except that it optionally sends an email to the author (if there is a message). the only concern is spam.

One of our academics at Harvard does exactly this via their lab website.

Or better... have an email address like papers@authorname.com which receives requests with the title of the paper in the subject line, and a piece of software then receives the request, waits for a random amount of time (anywhere between 10 mins to a couple hours), and then sends the paper to the requester.

It would seem that the author personally received the email and replied with an attachment as a "professional courtesy", while completely automating the process in the background.

that would be to hard to use. the idea is to make the experience as similar as possible to clicking a link to download without making the document directly available for download, but make it look like a personal request that could be handled manually. by submitting a form, that is achieved. the form ensures that the right document is selected, and the process can be automated without error. but for all we know, it just sends an email to the author who might handle the request personally.

> This could be an extremely interesting legal argument, but I suppose the language used requires a request in 'writing' of some sort, which I suppose HTTP requests would fail.

Are emails not considered communications "in writing"?

Attach a printer to the access logging of your web site ;-)

How about we have an autoresponder framework, like MailMain, that can handle mail with messages like

[paper-request] bacterial population analysis under exposure to fox news

And it would respond with the proper paper or papers attached.

Papers have dois so it’s way easier than asking for by name.

And some kind of index that lists possible "request" locations? Perhaps even with some kind of "hyperlinks". I think we may be onto something big here.

https://github.com/Zingle/smtp2http

That's not exactly true -- they allow authors to post their papers, e.g., on arXiv; and on institutional repositories such as eScholarship after the end of an embargo period. Here is the page with the (very complicated) terms: https://www.elsevier.com/about/policies/sharing

Elsevier 100% does allow authors to post preprints (most due). However, there are some journals published by Elsevier that don't.

Reply on news.ycombinator.com