Use correct Semibold font weight in static chart titles

[samizdatco]

Description

The design calls for chart titles to be set in Playfair Semibold (weight 600), but the PNG-generation code does not preload that weight and was not properly rendering exports when the style was updated. As a stopgap, the title weight was reverted to 500.

The actual fix would involve:

• [ ] adding the semibold weight to the list of pre-loaded fonts in the rasterizer (i.e., add a 600 to the front of the playfair font definition)
• [ ] replacing PlayfairDisplay-Medium with PlayfairDisplay-SemiBold in the script that generates embed.css
• [ ] regenerate the embed.css file by running make && make install in the devTools/fonts directory
• [ ] updating the title weight to 600 as in #2970

There's also the issue of creating fresh exports after making these changes since currently the baking process only generates new ones if the config has changed since the previous bake.

To get around this, try:

1. hard-coding the chartIsSameVersion function in GrapherBaker.ts to return false
2. triggering a bake
3. restoring the normal lazy behavior of chartIsSameVersion

Slack discussions

• updating the font weights used by the rasterizer and embed.css
• fixing existing static charts that use medium or bold weights for their titles

[sophiamersmann]

@samizdatco Running make && make install errors out for me and gives the following error:

unzip -jo .env/build/lato.zip -d .env/build/lato '*/*/*/Lato-*.woff2' && touch .env/build/lato/*
Archive:  .env/build/lato.zip
  End-of-central-directory signature not found.  Either this file is not
  a zipfile, or it constitutes one disk of a multi-part archive.  In the
  latter case the central directory and zipfile comment will be found on
  the last disk(s) of this archive.
unzip:  cannot find zipfile directory in one of .env/build/lato.zip or
        .env/build/lato.zip.zip, and cannot find .env/build/lato.zip.ZIP, period.
make[1]: *** [.env/build/lato] Error 9

When I look at .env/build/lato.zip, it contains a plain text HTML file (uncompressed). Is it possible we're accidentally unzipping twice?