Hide subscribe and unsubscribe buttons for own changesets

[scaidermern]

Currently there is a subscribe button for your own changesets although you are automatically subscribed to them. After someone commented on your own changeset this button gets replaced by an unsubscribe button.

In my opinion there should be neither a subscribe nor an unsubscribe button for your own changesets as the changeset author is automatically subscribed and should stay subscribed.

There has been a related question on help.

[tomhughes]

I'm not sure we have any right to force a changeset author to receive email about their changeset? I can even see an argument that it would break data protection law to do so...

[scaidermern]

I see but in that case there should be an option for disabling changeset comment notifications completely. And there is still the bug of the existing subscribe button for own, not yet commented changesets to which authors are automatically subscribed.

[tomhughes]

Well there shouldn't be a subscribe button if you are already subscribed, there should be an unsubscribe button...

Are you sure you aren't looking at an old changeset from before comments existed, to which you won't be subscribed?

[scaidermern]

That's a good question. When looking at my own new changesets there is only an unsubscribe button. But the author of the question at help did see only a subscribe button, yet he received a mail about me commenting his changeset. Maybe he looked at the wrong changeset and there is no bug at all.

[tomhughes]

There is nothing stopping you commenting on old changesets, but the author won't be subscribed so won't get emailed (this is issue #823).

[jase88]

I was the one asking on help.osm.org
Yeah @tomhughes You're right it probably was an old changeset that I was viewing. I didn't know you're not automatically subscribing to the "old" changesets (wasn't mentioned in a blog post or somewhere public - except github). So it is pretty confusing to have a subscribe or unsubscribe button on changesets (of course depending on the date). So I guess issue #823 should be fixed.

I'm sure I hit the unsubscribe on the one changeset and received an email notification. Tried to reproduce it but had no luck. I guess there is no log or history to verify the status of subscription that, right? Just got that mails...

[tomhughes]

So you're saying you got an email after you had unsubscribed? If you give me the number of the changeset then I can look into that.

[jase88]

Hi,
scai has left a comment on one of your changesets created at 2014-11-06 11:38:07 UTC without comment

A new comment.

More details about the changeset can be found at http://www.openstreetmap.org/changeset/26581590.

Hopefully the error wasn't on my side and I misclicked the button or something. Thanks

[tomhughes]

Well you're definitely not subscribed to that changeset currently...

[tomhughes]

It looks like you unsubscribed twice, at 11:35:49 and again at 14:09:35 but that shouldn't be possible...

Can you explain exactly what sequence of actions was taken after this changeset was created?

[jase88]

Okay, I will try to remember:

1. changeset created
2. scai commented -> got email
3. wrote own comment
4. clicked unsubscribe
5. wrote own comment
6. scai commented -> email again
7. unsubscribed again

[tomhughes]

I can only think that the first unsubscribe failed somehow, but I can't find any sign of a failure in the logs...

[tomhughes]

OK So the answer was staring me in the face all along...

The "problem" is that when you wrote a comment at step 5 that resubscribed you - anybody that comments on a changeset is subscribed to it if they aren't already subscribed. So I think this is all working as designed and there is no problem here.