Repository files navigation

This repo contains the source files for the bufferbloat.net web site(https://www.bufferbloat.net). The site is built using the Hugo staticsite generator (https://gohugo.io) and most of the content isautomatically converted from the old bufferbloat.net redmineinstallation.

Contributions to improve the content are very welcome. See below forsome pointers on how the content is organised and how to contribute.

The content shown on the site is all in thecontent directory. Eachsubdirectory is a project, which in turn has subdirectoriescorresponding to the different types of content. Currently, these arewiki for Wiki pages,news for news items andissues for theexported bug reports from the old site. The latter is mainly kept forarchival purposes, and is probably not worth editing.

The scripts used to convert the content from the old site live in theexport directory, which also contains the csv files with the data.Hopefully, there shouldn't be a need to run these anymore. Theold-projects directory contains exported data for projects that it wasnot deemed relevant to move over to the new site. This data is here forarchival purposes and is not used when generating the site.

Finally, thestatic directory contains resources such as images andCSS files used for the site layout, and thelayouts directory containsthe templates Hugo uses to build the site.For more information on how the latter works,consult the documentation athttps://gohugo.io.

To contribute to the site, simply clone the repository, make yourchanges to the markdown file corresponding to the page you wish to edit,commit the results and open a pull request against this repository.It is also possible to file issues to point out errors or omissions, orgive suggestions for the content.

The site is build using theHugo static site generator.Version 0.16 (not higher!) is required to build the site.See theUsing Docker for development section below.

When making changes, you can see a local version of the site byinstalling Hugo and runninghugo serve in this directory. That willrun a web server on localhost, which you can navigate to and see a liveversion of the site.

You can use the hugo binary to create a new news item with appropriatefront matter (the stuff in the top that tells Hugo that this is a newsitem). First, decide which project the news item should belong to, thenissue a command like

hugo new -k news bloat/news/2016-06-12-my-news-heading.md

to create the new file (the-k news tells Hugo to create a news item).

The new item will be marked as a draft and appear in thecontent/bloat/news directory. You can then edit the file (don't forgetto the title and add your name as author). While you are editing, youcan runhugo -D serve to see the changes in real-time (the-D means'show drafts'). When you are done you can either manually change thedraft status at the top of the file, or you can use thehugo undraftcommand to do it for you. Then add the file to git and submit a pullrequest to have your news item included on the site.

The Bufferbloat.net site relies on Hugo 0.16 capabilities.(Newer Hugo versions giveWARNING: blackfriday'ssourceRelativeLinksEval is deprecated ... error messagesfor the hundreds of internal links in the repo.)

Docker provides an easy way to run Hugo 0.16 reliably on any OS.Thecodycraven/docker-hugorepo has a container for creating and running a Hugo server in Docker.

The Docker instance operates on the current directory,so normal editing,git commands, etc. work as desired.

This command runs the Hugo server, listening on port 8080.To use it,cd <directory-containing-your-Hugo-files>then issue thedocker run... command each time you edit the site.

cd <directory-containing-your-Hugo-files>docker run --rm -it \    -v /tmp:/tmp \    -v $(pwd):/site \    -w /site \    -p 8080:8080 \    codycraven/hugo:0.16 \    server \    --bind=0.0.0.0 \    --port=8080 \    -w

Follow the link to localhost:8080/projects,and you'll see the blog as it appears.As you edit the underlying pages, the web display will update live.