Attachments Inconsistency

xinru1414 commented 2 years ago

Two problems:

Naming convention:
- for acl 2022, I've seen attachments named as papernumber.zip in the attachments folder, e.g. 1.zip ; however in the papers.yaml file they are named as papernumber_zip , e.g, 1_zip.
- for naacl 2022, I've seen the attachments named as papernumber_LaTeXSource.zip or papernumber_OptionalSupplementaryMaterial.zip or papernumber_LaTeXSource.tex in the attachments folder, e.g. 40_LaTeXSource.tex and 40_OptionalSupplementaryMaterial.zip. Some papers have both papernumber_OptionalSupplementaryMaterial.zip and papernumber_LaTeXSource.zip as the attachments with Supplementary Material as their type (I wonder if this is a mistake)
Attachment type:
- for acl 2022, attachments (normally end in .zip) have type software in the papers.yaml file; however for naacl 2022, attachments have type Supplementary Material for zip file and tex file.

crux82 commented 2 years ago

Hi @xinru1414

I try to clarify this point.

All files should have the classical format: filename.fileextension. During ACL the exporting script had a bug so it exported the files as filename_fileextension This bug is now fixed and (I really hope) you should do not receive any file in the filename_fileextension.

Regarding the second point. The instructions at https://github.com/rycolab/aclpub2#papersyml specify that the format should be:

  attachments:
    # A list of additional files associated with the paper.
    # The type, along with one of file must be specified.
    - type: dataset | note | poster | presentation | software | attachment
      file: Local file path, e.g. attachments/5.zip

Unfortunately the organizers of that workshop were not aware of it. Moroever, they were exporting the files from softconf and they believed they need to send also the latex source material. I told them not to upload any source in latex and they should solved the problem of the type.

Obviously, feel free to suggest any possible way to improve our documentation. It was not simple to foresee all possible cases and maybe we were not clear enough.

xinru1414 commented 1 year ago

Keeping this open for now since we are still seeing wrong attachments in EMNLP 2022 submission.

anthology-assist commented 1 year ago

For EMNLP 2022 and EACL 2023, attachments have been consistently having issues:

EMNLP 2022 main and findings volume: attachments and paper mapping is wrong
Some EMNLP 2022 workshops: metadata are mistaken as attachments
Some EACL 2023 workshops: metadata are mistaken as attachments

Not sure if these are aclpub2 issue or human error.

ryancotterell commented 1 year ago

It's human error. The publication chairs are not taking responsibility for the EACL 2023 attachments, so we should CC the workshop chair Zeerak.

crux82 commented 7 months ago

It should be solved by https://github.com/rycolab/aclpub2/pull/166

rycolab / aclpub2

Attachments Inconsistency #111