Speakers

Speaker Images

Speaker images will be provided by the conference and workshop committees, they will deposit them in the "Speaker Images" folder of the website committee's shared Google Drive folder. Once we have images, we follow this process to optimize them for the web and utilize on the site:

• Crop the image to be square (1x1 aspect ratio)
• Reduce the size & quality of the image, aiming to get it under 100kb file size, dimensions don't particularly matter (most from past years were 400px square)
• Export as a Progressive JPEG (e.g. from Pixlr or other image editor)
• Add the optimized image to the GitHub repo under the assets/img/speakers folder
• These images will be referenced in the speakers_final.yml data; if we do not have an image for someone, use a assets/img/nopic image at random (nopic1.jpg through nopic11.jpg)

Speaker data

Speaker data will be present in the spreadsheets provided by the Preconference and Program Committees. It will likely be present on the same worksheets as program talk and workshop information. The main pieces of information we need are:

• full name
• title
• institution
• bio (can be HTML)

Once again, fromSheetsToJekyll is your friend. You want to upload a CSV of speaker data, use the "Jekyll Data File" tab, and select settings such as these:

• name
• id - you can use a duplicate of the full name column with Slugify selected (we've also used hashed email addresses as ID in the past but I find this easier since the IDs will be human readable—whatever method you use, it must match the IDs in the workshop/talk posts' "speakers" array in their frontmatter )
• last name (you can use an equation like =INDEX(SPLIT(${name}, " "),1,2) to obtain last [sic] name from full name, where ${name} is the full name column, but you'll need to double check as some people provide additional names that throw off this reductive formula)
• bio (select Text)
• title
• institution
• twitter_handle (not including the "@")
• slack_handle (not used currently)
• github_handle (not used currently)
• keynote (static value of false for most speakers)

Sometimes title and institution are in the same field and another equation like =INDEX(SPLIT(...)) using a comma delimiter can automate some of the work of splitting these up but, as with names, it will need to be double checked.

Speaker images have to be manually added to each entry as the "image_src" property, though we try to use a consistent file naming convention. You can bulk add the string image_src: /assets/img/nopics/nopic#.jpg to every entry and then replace the "#" with a random integer from 1 to 10 to provide randomized placeholder images for everyone.

Once these steps are complete, rename the downloaded file to "speakers.yml" and add it to the "_data" directory of the website repo. You can test the outcome by visiting /speakers/ to see everyone listed. This is also a good way to catch duplicates—sometimes a speaker is doing both a preconference and a program talk, so there's potential for us to list their data twice.