2020/10/12: Tech Brainstorming [HHC M of N]


This is just some brainstorming for projects in the front of my mind.

I always find that doing this frees my mind up to take an idea a little bit farther.  Plus, the more different things I mix in the same session, the more seemingly strange connections I end up making. Eventually there’s a point past synthesis when I get in to real creativity.  So if instead of just blurting this out to a word doc in a cigar lounge I actually intentionally added these kinds of things to my personal desktop wiki or another blackboard style format, I’m sure I’d come up with both more simplifications and more interesting ideas. 

The general scope here is “managing my data.”  So I started with a few headings and just started going from there, with an overall goal of keeping everything in one loosely-coupled but still cohesive system.

Have fun trying to figure out wtf I’m talking about or hell, give me some ideas if you’ve got any kicking around.

  • Images and Music

So this nonsense with the iPhone image format (heif?  Dafuq Apple?) has me rolling my eyes on one hand, but it reminds me that I’ve been wanting to write a media import script to keep my pictures, documents, and such in a well-defined tree. 

Sure the /root/* stuff on the external drives is a reasonable structure. But frankly even that’s become a mildly well partitioned dumping ground.

If I can find a python library for handling that nonsense I’ll add something to it that saves a ./original/ copy then converts it to jpg. 

While I’m there, something that traverses the jpg files and rips out the exif data will do my sense of sanity some good.  Actually…shit, no.  Actually what I’ll do is extract a copy of the exif data as a json file so I can keep a database and maybe use some of what’s actually in there (not sure what all it contains) to categorize the images themselves.

Yeah I like that.

I could probably do the same thing with mp3 files as well.

  • Text content

Take the wiki engine and merge it with something like bloxsom (or some generic text/markdown file storage) to merge my writing (import from word?) blog entries and such in a single library with a couple different front-ends. Then I could rip a bunch of code against it for higher-level processing, tag clouds, etc. Maintain a ‘publish’ tag for things that would be mirrored to a largely-static external website. 

I really like that idea, since nominally different types of content could still be easily represented. 

Shit I could add a JSON -> markdown module so I could view just about anything from a single console.

I wonder how easy it would be to build something like that that was sufficiently flexible without being too brittle.

  • Emails

I think an alert system that ran against Thunderbird’s mbox files, spliced into the old import scripts would be really useful.  Take, for instance, newsletters from AJAC or Bevans could be promoted and shuttled off to a reading inbox.

I could use a similar system to capture my clippings directory, where I save things like twitter or reddit threads.

  • SIP

Again, if I can figure out how to build a pdf or ps file in python that will print duplex, then print it automatically in the morning I’d be able to take my todo list and generate a simple hipster pda sheet that’d be waiting for me every morning when I got up. 

Hell, the sound of the printer might even be a sufficient alarm.

Either way it would dovetail well with a todo list/project management system that traversed the contents and built priority listings and dependency trees.  THAT would be exceptional if I could get a diagraming package (something like graphviz/dot, with less ‘auto-layout’ and more programmatic positional layout) to build something like actual tech trees for ease of visualization.

  • Contacts

Import vCard and csv files into a standard json format and get that synced with either iCloud or something else that’d sync down to my phone.  Getting real tired of having half assed contact management.  Deduping becomes the primary issue. 

IF I had a UI on the front of it I could present a side-by-side based on common key fields (either email, last & first name, or normalized phone number) so I could clean that damned list up.  With so many imports and exports I’ve got 3 copies of Joe’s contact card on my phone and none of them have his phone number.  That horseshit has to stop.

  • Frameworks and front-end

Most of this could technically happen without any UI at all. But a simple web front-end or Electron app (but I repeat myself) would be really nice.  The guys seem to like Vuetify as a framework.  I’ll give it a shot. Surely there’s a tutorial kicking around somewhere I can follow.  Everybody loves writing a todo list tutorial for their pet framework.

Problem is I’ve got an image in my mind of what I want from a web framework and I’ve got a really bad feeling that I know what’s going to happen.

A framework and generator script pair that would build a ‘module’ (whatever the fuck that was) based on a spec of a back-end script and some information about required parameters that could be plugged in to a page/app easily would prevent me from having to hand roll (even with a rich framework) each new addition.

  • Log viewing for a ux

A framework that had modules as discussed above could have an added facility to hunt around for a well-known log file name which it could display on an auto-refresh of some kind.  Then add some kind of ‘expect’ style output scanning and it could re-sync an essentially async process.

If I broke these all down to unix design philosophy command line scripts (probably a metric fuckton easier than dealing with it any other way) then I could bang out functional components at a considerable rate.

 Welp, I’m finished with my 2nd cigar and have to hit the stupidmarket for…well, cheese and water.

So I’m going to post this as is just for fun.

I didn’t cover RSS feeds or using RSS as a format at all. But that’s somewhere in my head with all this stuff as well.