Closed ronaldtse closed 1 year ago
@manuelfuenmayor @anermina could you please help review this document one last time before we hand over to BIPM?
The particular points to look at are the meeting / decision / resolutions references of the CIPM and CGPM.
The worry is that the changes to the references caused the rendering to differ greatly (especially if the French reference output is no longer French but English).
There are not so many changes, so please use the diff as a guide on what has changed:
Thanks!
@ronaldtse, @Intelligent2013
There are several issues in the equations that were not before. Here below, I mostly show examples from French version, but they apply to English version as well:
Look at the in the image below and compare the highlighted regions with original version: Generated French version:
Original:
(Instances like shown above are everywhere throughout the document.)
There are a few cases where parentheses are not shown, for instance: Generated French version:
Original:
(Please also note the second case, (X) g/mol
, that I missed to highlight.)
Also, there is this case that I think could be considered as a minor issue:
In the ToC of Annex 1, there is a horizontal misalignment between the first column and the second: Generated French version:
On the English version, the misalignment occurs in a different manner -- there are random line breaks in the first column: Generated English version:
In regard to the references, I noticed practically no issue. There is a difference in the presentation compared to the original document.
Generated:
Original:
But I assume that that is the expected result.
On the other hand, there is this unique detail:
In French version, references that refer to an external website (i.e. hyperlinks), they do it to the English version of the website instead of the French version.
For example, in the highlighted reference below:
The hyperlink of CIPM — Recommandation 5 (1989) leads to: https://www.bipm.org/en/committees/ci/cipm/78-1989/resolution-5 instead of: https://www.bipm.org/fr/committees/ci/cipm/78-1989/resolution-5
But maybe this is not an issue.
I have converted the comments above in tickets.
Thank you @manuelfuenmayor ! I've also posted the Plurimath ticket.
I have found new issues that were not before at the beginning of this review.
@manuelfuenmayor this means we need to have tests to ensure the PDF layout is correct. We can’t keep doing manual reviews of an automatically produced artifact. Ping @Intelligent2013
@manuelfuenmayor this means we need to have tests to ensure the PDF layout is correct. We can’t keep doing manual reviews of an automatically produced artifact. Ping @Intelligent2013
@ronaldtse I don't figure out how to ensure the PDF layout programmatically. If we've checked and made sure once that generated PDF is identical to original PDF, then on the next iterations I propose to compare previously generated PDF with new PDF in such manner (visually comparison):
Or use the compare feature in the Adobe Acrobat (full version).
Merged. Thank you all!
Metanorma PR checklist