[Openstreetmap-dev] again issues with server rounding

[Immanuel Scholz]

Hi,

> it's very specific so it's not really something Steve or Mikel have to do.
> You, me or whoever can do that.

I have no access to the database scheme nor I want to fiddle with MySql or
the real database.

If solving involves any changes to the database, I won't do it.

BTW: The trac is not good for discussion. I will file a bug report when we
got a solution.


> But IMHO your problem is not with rounding algorithms but it feels
> more like a versioning issue, this will happen whether it's the server
> which "moves" the nodes or another user.

No. My problem is with the server implicit change every data uploaded.

Your problem is a very different one, since then other users intentionally
changed data.


A versioning system that creates automatic new versions of tons of objects
now and then when user *don't* change anything (but just forgot to resave
their data) is crap!
This is not a problem of two users changing the same but a broken system
that autogenerate versions..


>> My concrete problem:
>> - User enters a new node in JOSM
>> - save the data to disk
>> - upload the data.
>> Now the node location is rounded in the server.
>> - User restart JOSM without saving
> * Someone edits the nodes the offline user uploaded.
>> - reload the data from disk.
> * offline user changes data
>> (Without server communication, JOSM must think that the node was not
>> uploaded before.)
>> - User upload the node again.
>
> How do you solve these extra points?

This is an acceptable issue. Now the node gets uploaded twice and it is
the fault of this user here. But this issue is several orders of magnitude
less important.


Some analogy to text based wiki:

- User edits its entry in his favourite text editor
- save to "foo.txt".
- User upload foo.txt to the wiki, creating a new entry "Foo".
The wiki server changes the name to "Foo_56423" (The cunning server
changes every upload to <name>_randomnumber)
- User close its browser without reloading his new entry
- User reload the browser
The browser doesn't know anything about _56423, so he cannot but assume
that Foo does not exist yet.
- User upload foo.txt again.


It sounds a bit posed, but I think, in such a system, other users won't
blame the user if they see multiple Foo entries but blame the stupid
system that silently changes their stuff.



> But it sure would be a good idea to force your users to have a copy of OSM
> data before they edit,

No.

I had a post some time ago about this, but to summary it again: Not the
developer decide what user want, but the user decide. The developer
implement what user wants. This is a fundamental true.

And since I have several requests to even provide a "New Empty Document"
function, JOSM will definetly NOT go in the direction of forcing the user
to have some server downloaded data first.


> and then check for changes each time you upload to the server.
> Actually that should be an requirement.

If it is a requirement, it should be enforced in the server.

Notice that this does not affect the server-is-rounding issue. The issue
is, that the offline clients CANNOT!!! detect the problem, even if they
download the data before uploading. The user would have to look on the
doubled stuff. (But then, the user could have saved the uploaded data
anyway..)


Ciao, Imi.