[Imports] Uganda Bureau of Statistics educational facilities import

[Paul Norman]

On 12/29/2015 4:15 PM, Rafael Avila Coya wrote:
>>> The import has been communicated to the Uganda OSM community, and is
>>> being submitted here for final validation by the wider OSM community.
>> Can you provide a link to the Ugandan discussion? It's also wrong to
>> refer to this as a final validation - it's discussion just like needed
>> to be done with the Ugandan community, and neither is more a final
>> validation than the other.
> https://lists.openstreetmap.org/pipermail/talk-ug/2015-December/000069.html
>
> I just follow the imports guidelines.

But my point is that this isn't a final validation, it's a consultation 
with the community, just as with talk-ug.

Since talk-ug@ is a very quiet list, is there anywhere else the Ugandan 
community uses to talk? e.g. forums, a group elsewhere, FB, etc. With no 
response to the talk-ug@ message, it's hard to know if the locals want 
this or not.


>> Looking at the suggested changeset comment, you should not mandate a 
>> particular comment, but have users use a good changeset comment[1]. 
> I agree completely. Thank you. I will add that to the workflow wiki.

Don't forget the Import_Uganda_Bureau_Of_Statistics_Education_Facilities 
page too.

>> You've also listed the source tag twice.
> If you mean that source and source:date are both in the nodes and the
> changeset, it's true.

No, you're requiring people to use two source tags on the changeset, 
which is impossible.

>> Past experience has shown that importing with a fixme doesn't work. They
>> don't generally get fixed.
> The fixme tag would affect an small subset of those 24,000+ nodes.

How large a subset? Also, it doesn't really address what I raised.

>> The workflow mentions abbreviations. Are these common enough to be worth
>> expanding in the conversion script?
> Yes. They are indeed very common. I will add the most complete set of
> abbreviations to the final script so users will find as few as possible
> of those abbreviations. Being 24,000+ nodes, it's really difficult to
> correct all of them.

What  kind of portion will be left?