Key Functions of Project

[jeffshumphreys]

I forget, I tangentialize, I lose the plot, I build a rocketship to nowhere.

So what is the point of filmcab in terms of what it means to me, assists me in some personal function, goal, that I can at a later date go: Hey, that really helped.

[jeffshumphreys]

The first major thing that had to be done was capture the file detail in tables and views that tell what is. Not a bunch of cool function tables or apps, but just what is. What file is where and what type it is, how it got there, etc.

That took a while, and probably went down some rabbit holes. But the "files" table is the bomb. Too many columns? Probably. directory_hash instead of directory_id may be stupid. file_hash is good, and really uniquely identifies files regardless of their name or location.

But it works. There's performance issues, yes, primarily around regenerating hashes when files are moved or renamed. That can be fixed.

This is a usable state, and file details are captured nightly. Next, then, the issues.

[jeffshumphreys]

The function, or one of them, is the facilitation of collecting more film than I could ever watch. That's a real function, though the other function is to collect movies that my wife will watch with me, and enjoy, as well as things I want to watch. Finding stuff I want to watch is hard, and as the collection grows, I find I easily switch to "what about that other thing?" This is the too many channels and nothin' on problem. Three channels + UHF kept me busy as a kid.
Another issue is the Completist addiction. I have S09 of The Brokenwood Mysteries, but I automatically want S01-S08, even though I'll probably never watch them. Accumulating things I'll never watch creates new issues.

So to support this engulfing addiction, which in ways ends up with an inability to settle on one thing to watch, I had to add first an app that regains more space so I could download more!

So that's what offload_published_directories_selecting_using_gui.ps1 is, and it's critical. Once a drive runs out of space, with torrents still downloading, it's going to create stress and panic. So start the clean up now, before it runs out. It's working, and that has become a key function supporting the infinite collection process.

I can move directories off spindles, and i can move individual movies off, and I can leave links and junction points so they're not forgotten. I can also NOT leave a link for things I have no desire to remember fondly.

[jeffshumphreys]

There's no real way yet to automate the reclamation work. I would need an inline review system, knowing what was watched, what wasn't. Not there yet. Currently I'm moving out all the Battlestar Galactica I have, even though some goofy stuff wasn't watched. That's 44 GB, which is a huge savings relative to the size of some individual stuff. But if I move in that many series I won't ever watch, I can't keep up.

Ideally this would all be moved to a ZFS drive that just hid the details of massive underlying drives as one drive. But the current machine is drive heavy.

[jeffshumphreys]

A big goal is to make me constantly aware of things I want to watch. The folder "O:\Video AllInOne__Jeff wants to watch" was a stab at that, and it helped, but then I refuse to use it consistently, constantly reverting to the hunt and peck through folders system.

One problem is that these are usually hard links since VLC or some player I had on FireTV couldn't follow symbolic links. These mean when the offload app moves stuff, the copy is still present in this folder! So space is not reclaimed unless the wants to watch is cleaned out.

Also, the folder is alphabetic rather than by desire, so I get bored just looking at things I "thought" I wanted to watch. "But" what if I miss something I wanted to watch more??

[jeffshumphreys]

Tracking the progress and performance of sourcing video. Are these sources being blocked? Is Qbittorrent configured best? Is my ISP throttling? Is Defender blocking? Port forwarding?

What about archive.org: are things getting harder?
One thing I grabbed is snapshots of Qbittorrent's activity. Over time, is there an increase in stalls or missing metadata? Are seeders going up?