Hi Yahya,

When I fetch the URL of your bibtex file I get this redirect response with a cache control line. I can't say for sure, but I suspect that's where things are being cached.

> curl -i <a href="https://research.seas.ucla.edu/arni/files/2019/08/arni_db.bib" class="redactor-autoparser-object">https://research.seas.ucla.edu/arni/files/2019/08/arni_db.bib</a>
HTTP/2 302 
server: openresty
date: Tue, 01 Dec 2020 17:40:45 GMT
content-type: image/bib
location: <a href="https://cpb-us-w2.wpmucdn.com/research.seas.ucla.edu/dist/3/23/files/2019/08/arni_db.bib" class="redactor-autoparser-object">https://cpb-us-w2.wpmucdn.com/research.seas.ucla.edu/dist/3/23/files/2019/08/arni_db.bib</a>
cache-control: public, max-age=31536000
etag: 53fac6d6bed1e03f4313341be5aedf76
x-robots-tag: noindex, nofollow
x-cache: BYPASS
x-cache-bypass-reason: Arguments found

Even after deleting that bibtex file manually from BibBase's cache, the result is still the same, meaning the server that is serving the file (the one that bibbase is redirected to), still serves the cached file when bibbase is requesting it. I assume that is a policy by this host, which seems to be a wordpress provider, can that be?

This issue should resolve itself eventually when the cache of your host expires. Alternatively you could use a different server to host your bibtex file, or sign up for one of our premium plans and store your file on bibbase.org directly, if that's an option for your group.

Hi Oscar,

Thanks for finding out! It seems that we would need to implement a new integration with a different part of their API to make this work as these attachments are not already included in their bibtex format export. Based on what you found this seems possible, but unfortunately our priorities right now need to be on our premium features. After 15 years of free service we finally need to find a way to make BibBase financially sustainable, at least to cover our costs.

Do you think your research group or the Max Planck Institute for Art History would be interested in signing up for one of our premium plans?  It would be easier to justify custom development for a paying customer.

Thanks for understanding.

No, it's not currently possible to change the sort-order of groups. You can however use no grouping at all and then sort by year: "&groupby=&sort=year".

Regarding the 1540 entry, the issues seems to be the dot in "Theatra.de_PDF" used in the url field. This trips up our database, since dots are not allowed in keys.

@book{la_perriere_theatre_1540,
    address = {Paris},
    title = {Le théatre des bons engins: auquel sont contenuz cent emblemes moraulx},
    shorttitle = {Le théatre des bons engins},
    url = {Virginia_IIIF=https://v3.lib.virginia.edu/catalog/uva-lib:830320/iiif/manifest.json Theatra.de_PDF=http://www.theatra.de/repertorium/ed000134.pdf}},
    language = {French},
    urldate = {2020-07-21},
    publisher = {Denis Janot},
    author = {La Perrière, Guillaume de},
    year = {1540},
    note = {OCLC: 889853146},
    keywords = {43AA42 triumphal entry, 49M3213 emblem book, Moral education},
}

I took a look but unfortunately it doesn't look like the Zotero API is exporting those fields. This item (from my hometown Cologne!) has IIIF manifests in the library, but all the Zotero API gives us in the bibtex is:

@book{adrichem_theatrum_1582,
 address = {Cologne},
 title = {Theatrum {Terrae} {Sanctae} et {Biblicarum} {Historiarum} cum tabulis geographicis aere expressis},
 url = {Paper=mypdfs/gau-hum-1830.pdf Link=http://example.com/1830/measurements-of-the-world.html Slides=talks/gau-hum-1830.ppt}},
 language = {Latin},
 urldate = {2020-08-28},
 publisher = {Birckmannica officina, Arnold Mylius},
 author = {Adrichem, Christiaan van},
 year = {1582},
 note = {OCLC: 778667079},
}

So no, I'm afraid this won't be possible until they change their API. They might though if you ask.

Hi Oscar,

Can you point me to the page where you are embedding BibBase? Then I can take a look and see whether these Link to URI fields are already exported by the Zotero API and show up in the data we are getting from them. If that is the case then it should be pretty easy for us to support that.

Thanks,

Christian

Not per se, no, but I actually don't think that that's the best approach for accomplishing what you have in mind. Since you don't seem to want to use the abstract as an abstract but rather to provide additional information (annotations), you can just use the bibbase_note field instead. See this example. That field accepts HTML and is always rendered. So in your case you may be able to just rename all those abstract fields to bibbase_note. You may also want to use a div rather than a span, so that it rendered on a new line (like the abstract now) instead of inline.

Does that work for you?

It's not directly supported, but I think you can get it done using an additional field, with values 1_preprint, 2_misc, 3_..., grouping by that, and then using CSS to hide the group headers and show custom headers instead:

div.bibbase_group > span:nth-child(2) {
  display: none;
}

#group_1_preprint .bibbase_group_count:before {
      content: 'Pre-prints';
}
#group_2_misc .bibbase_group_count:before {
   content: 'Other Publications';
}

I know it's not ideal, but it should work.