The current IL WIki guidelines state:
So Arabic is correct, unless this policy changes.
Agreed, but as Harel said, let’s not jump to conclusions. It was probably a mistake / uninformed user, not a troll vandal.
Thank you for your answers.
This is a map, that can be maybe used for orientation:
ألبرشت شونكا,
The Data Working Group (DWG) recommendation for suspected vandalism is to contact the editor first. Contacting through the changeset comments provides an open discussion platform that also enables the DWG to view the entire discussion and decide if and what intervention is needed on their behalf, in cases of repeated vandalism.
Did you contact BMM994 in any way before posting here? He/she may or may not be reading the posts here.
Message deleted. My misunderstanding.
Albarshat shunka, Marhaba (and for all the others too)
Notice that before you started changing names of Arab places to arabic, all the arab places inside the State of Israel was named in Hebrew, so it looks to me like your editions were vandalist. but if it’s agreed to name arab places in Israel in arabic only, it will be fine.
You also changed the name of Judea Desert Nature Reserve to Arabic, but I changed it back to English because it is not an Arab place. Originaly i named it in English and not in Hebrew or Arabic to prevent argument about who is “the owner” of it. In ground it is on area B, but both of the Israeli Authorities and The Palestinian Authority doing what they want there.
I agree with dsh. Let’s not use the word “vandalism” lightly and assume good faith, this is probably just a mistake or a misunderstanding of conventions.
Some remarks:
-
Some people seem to try taking this too far and renaming places where there’s obviously a Hebrew speaking majority to Arabic. (Recently Ashkilon’s name was changed to Arabic and was reverted). This is not acceptable.
-
Albarshat shunka, please take care to preserve name:he and not destroy data regardless, if only “name” is present, please make surecopy it to “name:he” before putting the Arabic name.
-
Are there any objections to the following Wiki guideline?
- Are there any objections to this related guideline?
Disclaimer: This is a guideline written by me, but there was a discussion with the community.
The rationale is: The name tag can get messy and multilingual due to edits similar to Albarshat shunka’s, but this guideline guarantees name:he is always present and is Hebrew, and name:ar is often present and is Arabic. SafwatHalaby_bot enforces this guideline by copying over name to name:he if “name” is detected to be Hebrew, (or to name:ar if “name” is detected to be Arabic)
Related previous discussion: https://forum.openstreetmap.org/viewtopic.php?id=55156 “bilingual names in Israel”
I think this makes sense whenever there is a “most common” language in the area, but that isn’t always the case; for example, wiki recommends that border features have the “name” tag contain the names in all bordering languages: https://wiki.openstreetmap.org/wiki/Multilingual_names#Shared_boundary_features The Dead Sea is tagged this way.
I agree that we should encourage both name:he and name:ar to be set with the appropriate values on any feature that has a name at all. (Not just in Israel, by the way; some major streets in Berlin have name:he tags, and that’s absolutely fine.)
Aside: perhaps we should document our conventions on that wiki page too? So tourists who map in Israel know what our conventions are?
Not an objection to the guideline, but a question: due to this guideline, when I add, say, a convenience store, I have to add its name twice. Could we perhaps document to mappers that they can set either name or name:he, and the other one will be set automatically by the bot? The idea is to reduce the workload on the volunteer human mappers.
In my opinion, humans can easily copy-paste a name to the appropriate language. I wish all editing tasks were so simple…
A bot can easily make mistakes: copy the wrong language to the “name” tag or copy the “name” tag to the wrong language.
That almost sounds like a joke 
A properly debugged bot never makes mistakes. Humans, on the other hand…
The bot can set the other one automatically, yes. The bigger problem is when name and name:he already exist, but a human changes one of them and forgets the other. The bot is capable of detecting this and changing the forgotten one, but some didn’t agree to this feature, so it’s disabled.
The main argument against it is: “what if the new value is wrong? Humans should manually review it and a bot shouldn’t auto copy it”. But I think humans should manually review all edits regardless, via Osmcha or the like. Any new edit could be wrong.
I’ll document this.
Could the bot create a new map note (or set the fixme tag) to flag the mismatch for human review?
Thanks 
I’m planning an off-map log script, perhaps combined with a “fixme: name/name:he should be the same”. But I need to finish the GTFS project first.
There is always one more bug.
I think this is not a realistic requirements. Have anyone done a review for the bot runs so far?
Yes, it’s unrealistic; regardless of what the bot does, some bad human edits will make it through.
It doesn’t matter anymore since I’ve settled on just letting the bot print logs that are manually inspected (in the case of name/name:he mismatch)
Note that with the proper filter*, reviewing all Israel changes is possible, and rather easy via Osmcha and I regularly do it. (It has useful quick links to other viewers like Achavi, Deep History, etc). Though for experienced users, I only skim their edits, which are the majority of edits.
- Bbcode seems to dislike the filter URL. copy it from here:
https://osmcha.mapbox.com/?filters=%7B%22date__gte%22%3A[%7B%22label%22%3A%222017-09-27%22%2C%22value%22%3A%222017-09-27%22%7D]%2C%22in_bbox%22%3A[%7B%22label%22%3A%2234.16%2C29.57%2C35.93%2C33.42%22%2C%22value%22%3A%2234.16%2C29.57%2C35.93%2C33.42%22%7D]%7D
I always review a sample of the bot’s changes before upload. It’s not a fully automatic bot. Feel free to make further reviews, all changesets are logged.
We’re a bit off topic and I think this should continue here: https://forum.openstreetmap.org/viewtopic.php?pid=667870#p667870
It seems sivan00 is still changing names back to Arabic. See this:
https://osmlab.github.io/osm-deep-history/#/way/91408674