use planetmath without sign-up or log in

[holtzermann17]

From a recent discussion in the Planetary dev list, we've seen that requiring a user to sign up in order to use a public site is considered a little gauche, so we'll eventually want to think about how we can make it so that people can use PM without signing up, and then only create an account later.

This is more of a workflow design than a visual one, but Ray's proposed visual design for the frontpage (https://github.com/holtzermann17/planetmath-docs/issues/38) suggests speedy access and so brings the speedy workflow to mind.

[rspuzio]

Gauche or droite, I do not think that allowing people to do more that a few limited things without being signed up is a good idea. In particular, I think that alllowing people to post or edit content without an account is a mistake. Also, as the exerience of thousands of peoplle who signed up and only to leave a message or ask a quick question, if even that, shows, requiring accounts is an extremely low hurdle (more like a speed bump) and is unlikely to deter serious potential contributors.

[dginev]

Personally, I am on the other side of the fence. Requiring accounts is a deterrent for people who want to contribute content, as opposed to seek to become part of a community. I find this to be a very valuable low-entry point in Wikipedia, and a great incentive for people to get their hands dirty and try things out right away.

Registrations and log-ins are a hassle, even more so on mobile devices. Last Thursday I had DLMF mathematicians with no experience baffled at the high entry point of PlanetMath - they had expected to edit without an account, but even after they made an account they couldn't edit an article they wanted to correct, since they had no permissions. And they had no strong visual queues guiding them how to file corrections, so they felt lost and put off.

I don't think new serious contributors will just come to PlanetMath out of nowhere - as far as new users are concerned such contributors are "grown", as they gradually become immersed in the platform. The further the entry point, the more users will give up before having done anything with the system.

Mind you, the lack of registration and log-in doesn't have to equate lack of accountability - IP information, signed cookies and captchas can go a long way to ensure security and control.

[dginev]

Ah, here's the good compromise from services such as WriteLaTeX - create, edit, reply anonymously to your heart's content, but only until you hit the "Post" button. If you want your content published, you'd need to register and log-in, respectively.

The crucial point to be made is that a user should be able to experience all strong points of PlanetMath before logging in - the editor, the rendering, the encyclopedia, the community, and only when she is a step away from jumping in herself, would a registration seem as a "low hurdle" to jump over.

Maybe this is common ground?

[kohlhase]

I tend to agree with Deyan on this issue.

Delaying login until publishing seems sensible, but might leave PM with lots of unpublished content ,...

Michael On 23.4.13 05:30, Deyan Ginev wrote:

Ah, here's the good compromise from services such as WriteLaTeX - create, edit, reply anonymously to your heart's content, but only until you hit the "Post" button. If you want your content published, you'd need to register and log-in, respectively.

The crucial point to be made is that a user should be able to experience all strong points of PlanetMath before logging in - the editor, the rendering, the encyclopedia, the community, and only when she is a step away from jumping in herself, would a registration seem as a "low hurdle" to jump over.

Maybe this is common ground?

— Reply to this email directly or view it on GitHub https://github.com/holtzermann17/planetmath-docs/issues/39#issuecomment-16837371.

---

Prof. Dr. Michael Kohlhase, Office: Research 1, Room 168 Professor of Computer Science Campus Ring 1, Jacobs University Bremen D-28759 Bremen, Germany tel/fax: +49 421 200-3140/-493140 skype: m.kohlhase

m.kohlhase@jacobs-university.de http://kwarc.info/kohlhase

[rspuzio]

The basic issue for me here is accountabilty, especially in the event of plagiarism or infringement. As long as one has in place a system which allows the same level of accountability and possibility of identification as having a user account, it should be acceptable.

Inspired by your point about DLMF mathematicians, one thought that comes to mind is that we could allow contributions without logins from trusted domains like nist.gov or universities ----- if someone is breaking into such a domain or spoofing it or using it for unethical or illegal purposes, I am quite sure that the people in charge would be quite interested to know and take appropriate action!

Taking this a step further, here is a suggestion for how to compromise more: allow anonymous contributions under the conditions that a. the user click a box assenting to something simiar to the user agreeemnt and b. either be logged in from a trusted IP adress or else provide an e-mail address and not have the contribution appear until the reply to the confirmation e-mail arrives.

[holtzermann17]

This feature request should be considered in connection with another one: https://github.com/KWARC/planetary/issues/351 ("killer feature: use git to facilitate instant revisions"). No other site that I'm aware of has this feature, so it's slightly unfamiliar -- but I think the perfect solution. My thought here is to be very open to contributions, but very careful about how we process them.

My proposal:

1. People can edit without logging in; in order to do so, they must assent to the CC-By-SA and aspects of the other contributor agreement, like non-infringement.
2. Changes from people without editing permissions will not be merged into the articles until the owner or one of the designated co-authors agrees -- but will rather be stored in a branch. In particular, a new set of changes (diff) will play the role of a "correction".
3. In order to own an article, a contributor must have an account.
4. In the unique case of world-writeable articles, changes are merged directly, but we keep an audit trail.

This means anyone can edit every article -- not just world-writeable ones -- whether or not they have an account. They will start to build up a site profile that they can "claim" wheneven they do create an account. They will also get positive feedback and encouragement if their changes are accepted (e.g. an automailer will say "thanks, your change is accepted").

IMO it should also be possible to solve problems anonymously (some people might prefer that) -- and probably create other kinds of new content (like new problems or new questions) anonymously. Of course, this content would have to be moderated before it appears on the site, and a lot of it would be likely to be spam. So I think this is a big additional can of worms. Even the less risky phase of anonymous edits to articles (which won't show up on public URLs until they are accepted!) is likely to draw some spam from stupid spammers, so we'll want to proceed carefully.

[rspuzio]

Editing an article modulo owner's permission is isomorphic to filing a correction, so I suggest it be handled that way. Of course, should we one day implement corrections via patches, this will amount to what you are describing.

As for spam, that could be taken care of with CAPTCHA's. The combination of a CAPTCHA, asking for an e-mail address, and agreeing to site policy is rather common protocol for posting to blogs, so I find it hard to imagine that it would turn many people off.

Providing we have some such protocol, I don't think that there is any need for a special moderation policy; they could be moderated just like any other new submissions once they appear on the site.

The automated letter provides another opportunity to ask people to consider joining the site and point out the benefit of not having to fill out CAPTCHA's and getting credit for one's contributions. It should also contain a link to the page where one can complete the registration process by providing a screen name and password.

Done right, this could be a solution to the problem we have had with numerous people who register and maybe only post a quick comment; under this sytem, such people can submit a quick note without having to create a new account which lingers about unused,