NOTICE: is no longer in use from 1st March 2024. Please use the OpenStreetMap Community Forum


Hi, I made some more changes to a camp ground today in the web editor but trying to save I get an error about a mismatch for a way ID. This is because someone already corrected my telephone number prefixes 0039 into the correct form +39 for Italy numbers. But now I cannot save my changes because I cannot undo his version 3 back to version 2 which my edit session was using.

This way is affected and I already fixed the telephone number in my submission as well but I am unable to undo the shape changes which I did pretty early in my currently pending session. Can this be recolved by someone or how can I undo this one changes in the 77 changes of my submissions in the web editor ? Error: "Version mismatch: Provided 2, server had: 3 of Way 177901194"

I know that I should have probably saved more often but now I do not want to loose all the work if possible. regards, Copro

asked 25 Jun '14, 17:06

Copro's gravatar image

accept rate: 0%

edited 25 Jun '14, 23:05

aseerel4c26's gravatar image

aseerel4c26 ♦

Hi all,

Just because I could not go to bed I used the web debugging proxy Fiddler2 with the Decrypt HTTPS traffic option and a filter for all POSTs to to modify the changeset and it was really as easy as replacing the version="2" attribute of one changeset to version="4" of this one way node in my uploaded XML. I might have also been able to import the XML with another tool but since it worked I just wanted to let you know about the possibility.

Thanks for the encouragement to also look at other tools like JOSM which I might have just ignored otherwise. Thank you and good night !


permanent link

answered 25 Jun '14, 23:17

Copro's gravatar image

accept rate: 0%


That's an interesting solution and might help others, too. But everybody using this solution should know that this overwrites the other version(s), loosing all changes made by other contributors in the meantime. Always take a look at the history of the element afterwards in order to see which intermediate changes got lost and if they need to be re-applied.

(26 Jun '14, 07:19) scai ♦

Good point scai - maybe I should have mentioned this more clearly. In that case I examined the changesset before and even notified the author and then incorporated the changes. So yes ... as every conflict resolution might invalidate other peoples changes you should be careful and know what you do. Thank you for mentioning that.

(26 Jun '14, 08:56) Copro

I guess the group of people who could make use of this method is not that big (many who could do those quite technically advanced steps likely use JOSM anyway). However, thank you very much for providing this info! A last resort for conflicts in iD (if you do not want to undo until before the conflicting edit).

It might be a bit easier to use a in-browser tool like Firefox' (similar for other browsers) built-in "web console" (no need for a proxy and for a HTTPS decryption even if you are POSTing to the API via HTTPS).

(26 Jun '14, 12:36) aseerel4c26 ♦

As SomeoneElse stated in the open issue on github most users will be caught by that issue with the browser open and like me afraid to close it not relying on the feature that unsaved changes will be restored. The "Save Changes" button idea is already available in a experimental branch here: That would allow the user to save the changes locally in case of a merge failure and then decide whether to 3-way merge, deal with the changes in another program like Potlatch2 or JOSM or send the file to another advanced user easily instead of losing all changes.

(28 Jun '14, 10:01) Copro

Instead of overwriting the version manually, I'd recommend saving the XML as an osc file and resolve the conflicts in JOSM, which has a built-in conflict resolution workflow. I've written up a workaround here: Of course you can use Fiddler as you suggest instead of Chrome's Developer Tools to get the body of the XML. Cheers.

(29 Apr '16, 10:40) carciofo

Ahemmm iD has had conflict resolution for quite a while now ... this discussion seems to be a bit moot.

(29 Apr '16, 16:11) SimonPoole ♦

Hm,I just ran into this issue a couple of days ago, with ID offering no help or options to help resolve the conflict.

(30 Apr '16, 09:37) carciofo
showing 5 of 7 show 2 more comments

As described in a previous answer here, you'll need to undo back past the problem, save, and then re-edit.

In future you'll have to save more often to prevent the problem happening, unfortunately.

permanent link

answered 25 Jun '14, 17:29

SomeoneElse's gravatar image

SomeoneElse ♦
accept rate: 16%


Damn. Okay if that is the only way at least 2 hours of awesome mapping is gone and I am not sure if I am up for doing it again ;-) I thought somebody with more rights could go back to version 2 of that way by discarding the change made in version 3. I will leave the iD editor window open to see if there might be a better solution and if not maybe try JOSM next time for bigger edits.

(25 Jun '14, 17:35) Copro

Sorry - I don't know of a way to do that. Even when data is redacted from the database, the version number only goes forward.

If you'd been using JOSM, you would have been able save your edits to a file or resolve the conflict locally; if you'd been using Potlatch 2 you'd have been able to save your edits to a file and then correct and upload that manually.

(25 Jun '14, 17:39) SomeoneElse ♦

It's actually been an open issue against iD since March last year.

(25 Jun '14, 17:48) SomeoneElse ♦

Even in JOSM this sort of thing can happen and be a bitch to sort out. Frequent Saves and Updates help a bunch but the Conflict Resolution process is tedious and IMO tricky to get through satisfactorily.

@Copro - I can understand your frustration. I've lost editing that way too. Mark it off as a learning experience and don't give up mapping.

(25 Jun '14, 19:38) AlaskaDave

Thank you for all the answers ... I was already really happy to have received an answer from the mapper (Marco) that reverted the change. Sadly that does not help and the mentioned open issue is a really annoying one.

(25 Jun '14, 22:22) Copro

Save every 10 minutes. I am serious. You are mad if you entrust two hours of work to the security of a web browser and an Internet connection. This is exactly why Potlatch 2 has a little "unsaved minutes" counter and pops up a "ffs, save!" warning at 20 minutes and thereafter.

(25 Jun '14, 23:17) Richard ♦
showing 5 of 6 show 1 more comments

done a revert of my edit... i hope you can save yours edit... now...


permanent link

answered 25 Jun '14, 21:44

mcheck's gravatar image

accept rate: 0%


Thank you Marco ... I just checked and as SomeoneElse mentioned the version counter was increased even by the revert and I still cannot upload. "Version mismatch: Provided 2, server had: 4 of Way 177901194" Thanks a lot for the effort tho ... I will definately check out JOSM.

(25 Jun '14, 22:13) Copro

@Copro - at the risk of getting some angry replies I would encourage you to leave iD and Potlatch behind and move to JOSM. Both are very limited and Potlatch maintenance and updates will probably be soon discontinued. JOSM involves a bit of a learning curve but it's well worth the effort. Kudos to its developers but I must say that IMO the new iD Editor, touted as a more modern HTML5 based replacement for Potlatch, needs a lot more work before it's acceptable.

My 2cents

(25 Jun '14, 22:23) AlaskaDave

Potlatch is not going to be discontinued.

(25 Jun '14, 23:18) Richard ♦

"JOSM vs Potlatch" and "JOSM vs iD" is OSM's very own "vi vs emacs" question - there's no one right answer for everyone, and each editor has advantages for some users in some situations. I tend to use Potlatch 2 or JOSM depending on the kind of editing that I'm doing - P2 wins at some things (such as mapping from a GPX track with waypoints in it, relation visualisation) JOSM at others (excellent plugin support for e.g. a "to do" list, imagery overzoom, changeset reversion).

iD has had a lot of criticism (including from me) but two things its got right are that new users on average make fewer mistakes (according to some analysis that I did for UK new users last year) and they also map more POI address data - something that we definitely need more of in OSM.

(oh, and emacs - obviously!) :-)

(26 Jun '14, 10:30) SomeoneElse ♦

Follow this question

By Email:

Once you sign in you will be able to subscribe for any updates here



Answers and Comments

Markdown Basics

  • *italic* or _italic_
  • **bold** or __bold__
  • link:[text]( "title")
  • image?![alt text](/path/img.jpg "title")
  • numbered list: 1. Foo 2. Bar
  • to add a line break simply add two spaces to where you would like the new line to be.
  • basic HTML tags are also supported

Question tags:


question asked: 25 Jun '14, 17:06

question was seen: 7,108 times

last updated: 30 Apr '16, 09:37

NOTICE: is no longer in use from 1st March 2024. Please use the OpenStreetMap Community Forum