DEPCNT on area OBSTRN or WRECKS

[Christian-Shom]

The UOC (§6.3) indicates that the danger line should be encoded by OBSTRN or WRECKS objects. There is no indication that no DEPCNT should be encoded adjacent to the OBSTRN or WRECKS. We have found instances of such encoding (which is not "wrong", useless) in our ENCs and we suspect it might also be the case for other HOs. There is no S-58 validation check in regard to this (probably because there is no guidance in the UOC). I suggest adding guidance in the UOC (France could suggest it as a comment in the review process of edition 4.4.0) and creating a new validation check:

[FrankHippmann]

I would amend Christian's proposal to:

For each DEPCNT feature object which is COINCIDENT with an OBSTRN or WRECKS feature object of geometric primitive area.

[JukkaHelminen]

In our rocky archipelago we have hundreds of places where the limit of the OBSTRN areas share edges with DEPAREs and therefor also with DEPCNTs (see image). We don't see anything wrong or problematic with this encoding. If the OBSTRN area happens to share an edge with DEPARE then we don't see any problem having DEPCNT there. Splitting the longer DEPCNTs into small segments avoiding all OBSTRN edges would seem unnecessary. In our data it would increase the feature count quite a lot and probably the size of the cell. It would also require a lot more editing which takes more resources.

So for us this validation message would create hundreds of unnecessary validation messages even though there is nothing wrong with the data. So I don't see a need to amend S-58 or UOC.