OpenStreetMap

1 Million Mappers Soon ... No, Not Really

Posted by jremillard on 13 December 2012 in English (English)

When you log onto the OSM web site, you get this really nice map of all of the users hometowns, click on the bubble, you get the user name. This is a great! Disappointingly, I have discovered at least 80% of the "users" nearby are in fact people that made an account but never managed to make a single change on the map. Also, many of the accounts are several years old.

I propose that OSM should delete any account that is over a year old and does not have any actual map edits. The account is dead.

This will make the "find nearby users" feature signal to noise go way up and will keep everybody honest about the actual size of the OSM community, which is probably 20% of the registered user count.

Comment from aseerel4c26 on 13 December 2012 at 23:51

The "find nearby users" could be set up/programmed to show only users who are active (probably is is just: send the patch) - it is not needed to delete user accounts. I would support such a change (maybe user configurable?). Also some people may come back after a long inactivity. Sure, one should not confuse the number of user accounts with the number of active users (however "active" is defined) - and I think it usually is quite clear about which figure one is talking.

To see active users (who are mapping nearby) you can use WhoDidIt (but it currently seems to be not displaying edits - worked some hours ago). That is a bit more useful, IMHO.

Hide this comment

Comment from Hjart on 14 December 2012 at 08:11

I too would very much appreciate an option to filter out nearby users who never made an edit or those that wasn't active for a given amount of time.

Hide this comment

Comment from Richard on 14 December 2012 at 09:21

All of this pretty much depends on the 'OWL' technology and there's a lot of progress being made on integrating this into the main site.

Hide this comment

Comment from robert on 14 December 2012 at 13:12

We don't necessarily want to delete these peoples accounts as there's nothing stopping them becoming actual users at some point in the future. Them already having an account lowers the annoyance barrier to getting them to make an edit.

If anything we should be poking these users somehow encouraging them to make their first edit.

Hide this comment

Comment from Tordanik on 14 December 2012 at 14:23

Like other commenters, I disagree with the idea to delete accounts without map edits. For one thing, it is known that some people only start mapping months after they have registered their account. Furthermore, OSM accounts are needed for other services than mapping: Posting in the forum and on help, uploading gps traces, or even just storing your preferences for osm.org. Thus someone who is not a mapper might still be actively using their account - e.g. a developer using OSM who participates in forum discussions.

If the problem is that "find nearby users" is too cluttered with inactive accounts (and I actually agree with that), then the solution is to filter the accounts showing up there, not to delete them entirely.

In communication, reporting the number of people who have actually mapped something, rather than the number of registered users, would indeed be preferable. But that would be possible even today.

Hide this comment

Comment from dcp on 14 December 2012 at 19:04

When someone wants to leave the OSM community then it should be made easier for them to do so. The settings page should be made to cover this option. Signing out should be as easy as signing in!

Is it possible that the OSMF want to have all those redundant users (80%) just for propaganda purposes?

How many people sign on and want, for example, just routing? As they cannot see how to get it, they leave and don't come back. This is because the OSM Homepage is unfortunately inadequate as it does not readily reflect the full capabilities available.

Also I think we should differentiate between users and contributors: Users use the data. This would include for example OsmAnd, openmtbmap etc. Contributors fill the database, do the backroom services, etc. Of course you can be both!

Hide this comment

Comment from robert on 15 December 2012 at 01:36

"When someone wants to leave the OSM community then it should be made easier for them to do so. The settings page should be made to cover this option."

Huh? So what happens to any data they've contributed? The database needs to keep a paper trail of everything that users have done. And we preferably need a way of contacting everyone who's made a contribution. So what are versions of objects a "deleted" user has made associated with from then on? Do you propose we just have these mysterious "blanks"?

"Also I think we should differentiate between users and contributors"

I disagree.

Also, bonus point for OSMF conspiracy theory.

Hide this comment

Comment from dcp on 15 December 2012 at 07:10

"Huh? So what happens to any data they've contributed?"

I purposely did not go into this issue as it seems to be obvious that those that have signed on have already agreed to the public domain concept. Signing off would not change this at all.

The only easy way for someone to cut-off themselves off from their OSM account is to eliminate their email address with which they have initially registered. They can also "falsely" modify their "My Settings" beforehand.

Unless it is for satisfying statistics I can see no reasons why an OSM account must be kept open when it fulfills not one single purpose.

Surely it is up to the user to decide if he/she wants to quit and the OSMF should make it easy for them to do so.

As one famous historical figure once said;

Don't trust any statistic that you yourself have not manipulated.

Hide this comment

Comment from jremillard on 15 December 2012 at 15:54

The find nearby users feature should show new accounts without any map changes. If somebody opens up an account nearby and does not manage to edit the map, I would like to contact them and offer to help. If it is more than a year old, we probably lost them already.

For the 0.05% that open up an account and 2 years later make their first edit, all of the other issues don't seem worth saving them 1 step. I don't have data, but it seems like this is an extremely rare occurrence, and we should not be degrading important features because of this.

For users without any map edits, if we get our password hashes stolen, then we have 800,000 people that are at risk for no reason.

Not being able to close an account seems kind of rude.

Hide this comment

Comment from butrus_butrus on 17 December 2012 at 00:37

Hi!

I created my account on Sep 5th 2007 and my first edit occured in Jun 2012. Since then contributed 187 edits.

I strongly disagree with your proposal ;-)

Hide this comment

Comment from tmcw on 17 December 2012 at 16:10

As butrus_butrus said, this isn't a very reasonable solution - my account was idle for more than two years, and now I'm rather deeply involved in OSM.

If the problem you're trying to fix is misleading 'users numbers', the solution is to publish 'active users' numbers alongside.

Hide this comment

Comment from jremillard on 18 December 2012 at 02:55

butrus_butrus, tmcw, I did not suggest that nobody would come back, but rather these dead accounts have a real cost on project. They are not "free". Also, I bet both of you would have made an account when you came back if needed. Anyway, no worries, I'm glad you both came back :-)

Hide this comment

Comment from dcp on 18 December 2012 at 06:38

@butrus_butrus and tmcw

I am not quite sure with whom you disagree. Just for clarity, I don't want an automatic account closure concept.

What I would like to see is an easy opt-out for those who sign-in, don't want it, and want to sign-out: THEY JUST CANNOT DO IT!

Their numbers are probably in the hundreds of thousands.

Hide this comment

Comment from Vincent de Phily on 18 December 2012 at 12:14

I support a policy of deleting "dead, never used" accounts, if only for the purpose of data cleanup. I't becoming embarassing to see the "1M accounts" number being anounced, when it is really meaningless. The "nearby users" feature is a completelydifferent thing, something based on whodidit would indeed be better.

But deleting accounts needs to be very conservative. How about :

  • The account should not have contributed any map edit, help question, diary entry, or anything else.
  • The account should not have logged in in the past 12 months.
  • After 12 months, send an automated email once a month warning about impending deletion (at this stage, simply logging in would reset the counter).
  • After 6 warning emails, go ahead and delete the account.

We can argue and tweak those numbers, but I think that with those kind of guidelines, account deletion becomes reasonable. Now we just need to agree that it is usefull.

Hide this comment

Comment from porjo on 18 December 2012 at 20:55

Don't automatically delete anything, just disable it. Any frontend system that deals with accounts can be set to ignore disabled accounts. When/if the user ever comes back then account gets 're-enabled', perhaps with a quick email verification step, and then returns to normal view.

Of course, if someone really wants to delete their own account, then they should have that option too.

Hide this comment

Leave a comment

Parsed with Markdown

  • Headings

    # Heading
    ## Subheading

  • Unordered list

    * First item
    * Second item

  • Ordered list

    1. First item
    2. Second item

  • Link

    [Text](URL)
  • Image

    ![Alt text](URL)

Login to leave a comment