From TLUG Wiki

Jump to: navigation, search



TLUG will be holding a meeting to discuss and work on the migration of the TLUG web services.

TLUG Server Migration Hacking Meetup


May 11, 2019 (Saturday)



- if you would like to come earlier please let the reception know you are with the TLUG event


Working Space Co-Edo
PA Building (5F reception),
1-3-4 Shinkawa, Chuo-ku,
〒104-0033 Tokyo

About the Space Google Map Directions

What needs doing

As far as the web site goes, off-hand I see four basic groups of things that need to be done to get most of the site up and running, only one of which involves any Haskell.

  1. Static pages.
    1. Copy into `docroot/` static pages and other files from the old site.
    2. Remove the PHP and rename to `.html` any PHP pages.
    3. Update the CSS and the like as necessary fix any brokenness in the pages when served from the new server.
  1. Wiki content:
    1. Copy the wiki content (markup) files for the wiki pages we want to keep into `wiki/`. (This is only on the dev/cjs/190321/wiki-pages branch right now.)
  1. Serving wiki pages:
  2. Figure out what we need to do to get Netlify to serve these "nicely," i.e., under the original URLs (without `.html` extensions) but still with the correct content-type. I think I know how to do this, but it involves a bunch of testing and playing with Netlify.
  1. Building wiki pages:
    1. Build templates, CSS, etc. so that the wiki pages look like they should when served. This involves only minor bits of Haskell if any at all, and wiki pages without transcludes will then be working.
  2. Finish the Haskell transclude parser, use it to build page markup with transcluded text substituted in the right places, then pass that to pandoc to render into HTML, to be served the same way as non-transcluded pages. This is the vast majority of the Haskell work, and I think Justin has made a fair amount of progress on this.
  1. Old mailing list archives:
    1. If the new mailing list provider can serve our old archives, great, we're done.
    2. If not, we need to collect them up, clean them and serve them from the website. Probably not a huge deal.


Personal tools