manubot / catalog Goto Github PK

due to crosscite/content-negotiation#104

Creating a nice screenshot takes a bit of artistic touch. How to crop it, what to include in the shot, and the "composition" of the shot are all things that can't necessarily be quantified.

For the short term, I can do all of the screenshots for papers submitted to the catalog. But we should come up with a more robust, longer-term solution.

Creating this issue to remind me to come up with a list of guidelines for how to create a screenshot that is both nice looking, and also consistent with the rest of the screenshots in the catalog.

Looking at https://travis-ci.com/github/manubot/catalog I see:

Builds have been temporarily disabled for public repositories due to a negative credit balance. Please go to the Plan page to replenish your credit balance or alter your Consume paid credits for OSS setting.

Would be nice to switch to GitHub actions, but we will have to look into having this action trigger the website action.

@sbonaretti's manuscript titled "Two real use cases of FAIR maturity indicators in the life sciences" is now on bioRxiv at https://doi.org/10.1101/739334. The source repo is at https://github.com/sbonaretti/FAIR_metrics.

Serena used a modified manubot setup where the manuscript was built locally rather than using CI. The repo doesn't host an HTML version of the manuscript via GitHub pages. Our software to populate the catalog from catalog.yml requires a html_url I believe to retrieve metadata. Possibly we could use https://rawcdn.githack.com/sbonaretti/FAIR_metrics/c12a837e1b0505ffa7a3cd311e3fa69b5b940d2e/paper/manubot/webpage/v/local/index.html, which seems to produce correct metadata:

manubot cite url:https://rawcdn.githack.com/sbonaretti/FAIR_metrics/c12a837e1b0505ffa7a3cd311e3fa69b5b940d2e/paper/manubot/webpage/v/local/index.html

[
  {
    "id": "1Hl6qxW9W",
    "type": "webpage",
    "title": "Two real use cases of FAIR maturity indicators in the life sciences",
    "URL": "https://rawcdn.githack.com/sbonaretti/FAIR_metrics/c12a837e1b0505ffa7a3cd311e3fa69b5b940d2e/paper/manubot/webpage/v/local/index.html",
    "language": "en-US",
    "author": [
      {
        "family": "Bonaretti",
        "given": "Serena"
      },
      {
        "family": "Willighagen",
        "given": "Egon"
      }
    ],
    "issued": {
      "date-parts": [
        [
          "2019",
          8,
          11
        ]
      ]
    },
    "accessed": {
      "date-parts": [
        [
          "2019",
          9,
          3
        ]
      ]
    },
    "note": "This CSL JSON Item was automatically generated by Manubot v0.2.3 using citation-by-identifier.\nstandard_id: url:https://rawcdn.githack.com/sbonaretti/FAIR_metrics/c12a837e1b0505ffa7a3cd311e3fa69b5b940d2e/paper/manubot/webpage/v/local/index.html"
  }
]

Also there are some nice figures in the manuscript. @vincerubinetti did you want to create a PNG screenshot that we could drag into this issue and use as the thumbnail?

Replace instances of irrelevant if submitting a figure directly with only relevant if taking a screenshot of the manuscript.

Replace instance of If your figure has to extend beyond the vertical boundaries of the image, leave this much horizontal padding, and vice versa. with If your figure has to extend beyond the top & bottom boundaries of the image, leave this much left & right padding, and vice versa.

Once manubot/manubot.org#24 and manubot/rootstock#267 are resolved, this can be done. Hopefully it's simple to do if pulling a known filename from a known location.

Related: #9

I just noticed that however the manubot catalog page parses/displays surnames incorrectly (and with a anglo-centric bias).

I have two surnames: 'Fellows Yates' (it is not hyphenated), but only Yates is displayed. While this is unusual e.g. in the UK, another co-author from Spain is displayed as only Valtueña whereas the actual surname is 'Andrades Valtueña' - this is problematic as dual surnames is extremely common in both Spain and also hispanic cultures in S. America (and elsewhere). This is also relatively important for specifity during indexing due to realtively common single components of surnames.

It would be great if this would be corrected, e.g. only the first 'word' being listed as a first name (whereby single first names is less common than multiple surnames), any initials in names of the metadata.yaml file being assumed as middle names, and everything after as surname, or more ideally (although harder), the metadata.yaml to be updated to have explicit forename, middle name and surname fields.

We should add this new @greenelab review to the Manubot catalog:
Repo: https://github.com/greenelab/biopriors-review
Preprint: https://arxiv.org/abs/1910.06738