Open soerenthomsen opened 2 years ago
I've done a bit of digging. Looks like the culprit is the div name of sections. UItterances identifies sections presumably to create a seperate set of comments for each one.
Inspect element shows this block is throwing an error
sections = document.querySelectorAll("div.section");
if (sections !== null) {
section = sections[sections.length-1];
section.appendChild(script);
}
It's looking for divs with class "section" In the Nitrate SOP, where comments are working, we do see these divs, e.g.
<div class="section" id="questions">
meanwhile, over in the Oxygen SOP, the same object is called a section, rather than being a div with class section
<section id="questions">
I'm sure there's an easy way to solve this and I'm having a go, but my html-fu is rudimentary at best.
Even changing the JS to search for section, it still throws an error later on. Can't figure out why jupyter-books changed to using section rather than div.section :/
Thanks @callumrollo for digging!
It is strange but maybe helps to find the error that indeed the commentary function works for the nitrate SOP but not for the oxygen and salinity. And even more confusing not for the Depth Average Current SOP as we have not changed anything there for a long time.
@jbusecke and myself pushed quite some PR's lately to the Oxygen SOP to create the pdf (just for Oxygen SOP so far) and improve the citation style (both for oxygen and salinity).
@jbusecke any idea what might have happened?
I cannot say since when exactly the commentary function is off unfortunately. Realised it yesterday during the workshop!
Here some info: https://utteranc.es
In the
_toc.yml @jbusecke changed to "jb-book" and "chapters" to build the release v1.0.0 pdf. Don't remember why.
`format: jb-book root: README options: # The options key will be applied to all chapters, but not sub-sections numbered: true chapters:
it is called "sections" in the _toc.yml, which works
`format: jb-article root: README options: # The options key will be applied to all chapters, but not sub-sections numbered: True sections:
_toc.yml, which does not work...
`format: jb-article root: README options: # The options key will be applied to all chapters, but not sub-sections numbered: True sections:
Seeing as the _toc
files for nitrate and salinity are identical, bar section names, I don't think it's the source. That was the first thing I tested. My suspicion is that something in the build process has changed. nitrate SOP hasn't been updated since May. Perhaps we need to pin some earlier package versions if the behavior of the jupyter books build has changed since then
Nitrate was last built with Jupyter-Book v0.12.3, the others are using v0.13. Coud've been a breaking change. I will try to pin the environment and build locally with the earlier jupyter-books
Here @callumrollo fixed it quickly for the salinity SOP ahead of the workshop starting now. https://github.com/OceanGlidersCommunity/Salinity_SOP/pull/139
Sorry for the late comments here.
I am having a bit of trouble identifying the issue here.
I initially thought the same as @soerenthomsen , but the _toc.yml
and and the _config.yml
are also different on the Salinity SOP (which is currently down it seems 😜).
Perhaps we need to pin some earlier package versions if the behavior of the jupyter books build has changed since then
If this ultimately fixes the issue this should be definitely brought to the attention of the jupyter-book folks (maybehere?)
Ill try the version pin on the oxygen SOP right now, lets see if that works
Even changing the JS to search for section, it still throws an error later on. Can't figure out why jupyter-books changed to using section rather than div.section :/
This would be an excellent question to ask over there! Great job digging out the reason for the failure!
It might be useful to make a minimally reproducible example (maybe some dummy 1-page repo?) to check if this persists in a less complex case.
I've filed an Issue with jupyter-book. It's late in the day here, but I'll try making a minimum reproducible example tomorrow
Amazing. Thanks for the work. Glad there was an intermediate fix!
Thanks @callumrollo and @jbusecke for your support on this!
So we keep this open for now until we find a long term fix?
Yeah, let's see what comes out of the jupyter-books Issue. Pinning on a previous version isn't ideal and could create problems down the line. It's the best fix we've got for now though.
Hi @jbusecke,
during a workshop we just realised that the commentary function below our beloved SOPs is gone? Any idea what happened?
Cheers