Hi,

>> 1. Index and search actual digital content? (ie: can we index PDF's, Docs, and text files, and then search that content within CA?)

Yes. As long as you have the appropriate "helper" applications installed. All are open-source and freely available. These are pdfToText (for PDFs; part of the xpdf package) and LibreOffice (for Excel, PowerPoint and Word docs). XML files can be uploaded as well.

>> 2. Must all content of one type (eg: mp3s) reside in a single sub-folder like /media/mp3, or can it be spread across multiple folders like /media/mp3/folder1 & /media/mp3/folder2 & //server#2/media/sound etc...?

You can spread it across as many volumes as you like, if you want to mess with the default configuration. And within something like /media/mp3 the system automatically partitions it across subdirectories. It doesn't attempt to stick everything in a single directory. This is a legacy of olden times when Linux file systems sometimes did odd things when you tried to store 500,000 files in a single directory.

>> 3. Are there any upper limits to the >> a) number of records it can reference?

We use unsigned ints as internal ids so the theoretical limit is 2³² (₄.2 billion) records of any given type. As a practical matter the MySQL-based default search engine starts getting slow when you get into systems with several hundred thousand records. At that point you can swap out SQLSearch for SOLR or ElasticSearch, which require additional configuration but scale better.

>> b) individual file size it can reference?

This would be tied to limitation of the operating system you're running on. In general the only practical limit we've run into is a hard 2gig limit for files uploaded via a web browser and is a limitation of PHP and Apache. Uploading the file the server and ingesting it from the local file system bypasses this limit. It's usually more practical avoid uploading very large files via web browser in any event.

>> c) total collection size (in bytes) it can reference?

How would you calculate collection size? As the sum of metadata and media? Just metadata? Something else?