Unclear on why you cannot use 'neighbor_history' and skip EPost for section 9.15.3 with history?

[fomightez]

It looks as though you probably felt from reading here you could use neighbor_history and skip the second EPost under section 9.15.3 where you use history.
I would have thought that was the case too given it reads under ELink:

cmd=neighbor_history
ELink posts the output UIDs to the Entrez History server and returns a query_key and WebEnv corresponding to the location of the output set.

Example: Link from protein to gene and post the results on the Entrez History

http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?dbfrom=protein&db=gene&id=15718680,157427902&cmd=neighbor_history

Did you find or learn the explanation why you cannot? It seems to be exactly what neighbor_history would be for. I don't understand why they'd add the need for an extra step?

[gumption]

I did not find or learn why.

It does appear that I am / we are not the only ones who have been baffled by the correct use of neighbor_history.

[fomightez]

Thanks for that information. Since it goes back to 2009, I'd say it won't be getting fixed any time soon then.

Thanks very much for working out a way around it.

[fomightez]

I found this here which maybe suggests the description for ELink's neighbor_history is misleading:

ELink and the History Server

Although ELink can accept a stored set of UIDs from the History server as input, this eUtil cannot load its output onto the History server. The consequence of this is that the linked UIDs found by ELink must be parsed out of the XML output and then provided as input to another eUtil, either directly using the &id parameter or by using EPost to store them explicitly on the History server. Then they can be passed to other eUtils.

Maybe that is as much explanation as we'll ever find.