Posted on

Fixed a search bug on FilkArchive!

Searching for “Atlanta, GA” was giving 0 results, despite there being several GAFilk items tagged “Atlanta, GA”. Turns out my schema.xml was a bit off. So, I fixed it on my local machine and rebooted Solr to reload the schema. It was glorious. I simultaneously fixed the related bug of creators not being searchable.

But then, when I uploaded the new configuration to my server, there was no change in behavior. I restarted Solr. Still no change. Tried restarting Solr in a different way. Didn’t start. Long story short, my production server stores its schema in a different folder than my local machine does. I copied the file over to the proper directory, restarted Solr the original way, and ta-da! Glorious, functional searches.

Posted on

Filk Archive update

About 70% of the spreadsheet is in EAD format now. It’s on the github repository but I haven’t indexed it on the website because it needs extensive cleanup. Names especially are not consistent.

I’ve installed ArchivesSpace locally to help me with that. That installation was straightforward, but when I tried to load the 70%-done EAD I found out that AS has different standards for EAD validity than Arclight does. Naturally, a hand-coded 1.5 mb file had many rought spots and AS choked on those. I spent the evening fixing and, by mdinight, had a parsable file to load.

I’ll finish up the last 30%, load it in, and use AS’s handy user interface to clean up the names and subjects. It’ll take a while, but it’ll be straightforward, and much easier than by hand.

Posted on

Side project! A Filk Archive

A friend of a friend, Harold Stein, was very interested in filk (science fiction folk) music. He recorded filkers at conventions, he collected filk songbooks, he produced CDs. He amassed a large collection of filk recordings, filk songs, and filk-related objects. Harold sadly died in 2018 of cancer.

Our mutual friend, Merav Hoffman, asked me for help describing this collection to make it more accessible to the wider filk community. It’s a large collection–at least 52 boxes of paper, CDs, hard drive. So far, Merav and I have gone through and listed almost all of the paper items, in a Google spreadsheet. I’ve started putting together an EAD2 finding aid, which I’m using as an excuse to set up an Arclight server. (Arclight is a newish archival discover tool based on Blacklight.) So far, half the objects in the spreadsheet are in the EAD2, but the EAD2 is inelegant and needs cleaning up, since it’s mostly made through spaghetti Google Sheets concatenation code and hope.

Please see Merav’s website at http://filk.meravhoffman.com/ for more information.