LONG FileNames use 'em for curation and more

[SpectateSwampBANNED]

I make use of LONG FileNames for cataloging (<=260chrs) My app can use the filename to determine where to start video playback, for how long and at what speed…
And more a lot more.
This is a far better spot than being buried in the Metadata.
Nothing is like having all your info at your fingertips… NO matter how much you have..

With all the mpg jpg mp3 having long file names.. I can do a quick catalog of all of them..
Then randomly or by specific key show / play them.
The filename is even more important than long program names for this... Wow

 

[SpectateSwampBANNED]

squirrel at cabin green mountain road 20190329123657

This is a recent example of my video naming (for now)
My app can search for all mpg files and create a catalog that I can search and then Play 'em
I can search for "squi/roa/2019" easy easy. curation and play...

Give 'em long file names you'll be glad you did.

 

[Vayra86]

1. date sorting can be done by standard OS sorting methods.
2. long file names tend to fall off on many display/UI sizes.

Its not a bad idea, its just highly situational I think. There is another way, you can use affix/suffix language, or 'tags', categories. Long descriptions also take time to read, while tags are simple and tend to cover the content; its easier/faster to skim through files that way and you can still search on tags.

I do use dates a lot in my filenames. Its great for version management as well.

For those videos for example; tags like "outdoor" "wildlife" would be appropriate. Or, if you are a fan of filming squirrels, 'squirrel' is automatically also a tag for you. If you do it like this you can even sort within the entire map and get a logical order.

Also, what app are you referring to?

 

[Aquinus]

If you need to query data about files like that, why aren't you using some sort of database? Something like SQLite is great for this sort of single-user application that needs to be highly portable. Metadata can be stored in a way that you can query upon and it's likely faster than scanning the file system. The problem is that at scale, performance may be suboptimal on rotational media drives and this might not work well because you're essentially doing random I/O to determine file names, whereas using something like SQLite is far more likely to do sequential reads to find the same data. I suspect you don't want to store all of your video on an SSD. In addition to that, if you use something like SQLite, you can put the catalog on an SSD while keeping the content on a spinny disk, giving you both excellent performance, and good scaling as your media library grows.

Personally, I think using the file name for anything other than unique identification is a mistake and is possibly error prone. Using an external catalog with hashes could let you detect duplicates and renamed files as well. All things to consider.

Edit: A database could also let you record several tagged times on a single video, which could themselves have their own unique names, so that's a thing too.

 

[EarthDog]

Looks like you reinvented the wheel even though there are things out there for this already.

 

[Aquinus]

Looks like you reinvented the wheel even though there are things out there for this already.

Reinventing the wheel is okay if it's a better wheel. I personally see value in things like data deduplication and being able to search your files with your own metadata. I think there is value in that, but there isn't value in trying to cram all of that information in a filename. I think the idea is fine, I just question the implementation.

 

[SpectateSwampBANNED]

Also, what app are you referring to?

My app is a vb5 program that was started in 1999..
11,000 lines of spaghetti code.
http://www.telusplanet.net/public/stonedan/source.txt

If you need to query data about files like that, why aren't you using some sort of database? Something like SQLite is great for this sort of single-user application that needs to be highly portable. Metadata can be stored in a way that you can query upon and it's likely faster than scanning the file system. The problem is that at scale, performance may be suboptimal on rotational media drives and this might not work well because you're essentially doing random I/O to determine file names, whereas using something like SQLite is far more likely to do sequential reads to find the same data. I suspect you don't want to store all of your video on an SSD. In addition to that, if you use something like SQLite, you can put the catalog on an SSD while keeping the content on a spinny disk, giving you both excellent performance, and good scaling as your media library grows.

Personally, I think using the file name for anything other than unique identification is a mistake and is possibly error prone. Using an external catalog with hashes could let you detect duplicates and renamed files as well. All things to consider.

Edit: A database could also let you record several tagged times on a single video, which could themselves have their own unique names, so that's a thing too.

Random is just fun.. exciting... beautiful...
To catalog all my mpg videos... one simple function.. and it's created in 2 or 3 minutes
a text file with 2 lines per video is created..
The first is the "search criteria line" and the 2nd is the path to said video.
3000 videos would result in 6000 lines in the text file...
The app counts the number of lines in the file and generates a RANDOM number
It reads to that point and starts the search.. If no match found in 3 lines then another random is generated..

The date time stamp on videos can be very useful...
We did a golf tournament video.. 2 camcorders at a beautiful hole..
by having the camcorder times set to 15 or 20 seconds apart.
We could download both camcorders clips for each golfer...
And they would be automatically sorted into order.
I created a specific catalog function that set these 2 lines up...
one set to play the full 4 or 5 second clip
a next set to play the last second in slow motion (the swing)

Looks like you reinvented the wheel even though there are things out there for this already.

This is the THING you can use forever.. 20 years old already and just getting better... What a Wheel.

Reinventing the wheel is okay if it's a better wheel. I personally see value in things like data deduplication and being able to search your files with your own metadata. I think there is value in that, but there isn't value in trying to cram all of that information in a filename. I think the idea is fine, I just question the implementation.

When we originally did the family albums.. scanned in around 2000.. we put the metadata into the hand made catalog file.
Now.. I'll want to change the CP feature that copies files to another folder... so that it will change the name to contain all the catalog details..
I don't worry much about data duplication.. I worry about losing stuff.. and that doesn't happen when everything I have is in 1 huge folder..
I just drag and drop that to an external 3TB drive.. Duplicates and all..

Back in the Day the Wheel only had 8 characters for a file name... now now it's <=260

Being able to explain how Search works.. And the source code right there.. People can make use of it..

Reinventing the wheel is okay if it's a better wheel. I personally see value in things like data deduplication and being able to search your files with your own metadata. I think there is value in that, but there isn't value in trying to cram all of that information in a filename. I think the idea is fine, I just question the implementation.

Yup having my own metadata allows me to see if any files have gone corrupt or missing.
Same with video. play the first few seconds and last few of all clips.. They can be corrupted..
You wouldn't know unless you had a master list.. or compared backups to backups... Handy in the long run.. for sure

With filenames now.... One doesn't have to type them in more than once.
the rename / copy / paste makes it easy to have those names at hand..
Curation without random ... I don't want it..

 

[FordGT90Concept]

...just install something like Emby and make sure to keep the metadata in the files correct. It does all of the database building itself.

"Your Stepson Tried to Burn & Don't Feed the Squirrels"

 

[SpectateSwampBANNED]

...just install something like Emby and make sure to keep the metadata in the files correct. It does all of the database building itself.
View attachment 119964
"Your Stepson Tried to Burn & Don't Feed the Squirrels"

The only reason there is such a thing as MetaData is because filenames were limited to 8 characters...
Having to have a special program.. right off the bat to change and see the details is rubbish.

 

[Aquinus]

My app is a vb5 program that was started in 1999..
11,000 lines of spaghetti code.
http://www.telusplanet.net/public/stonedan/source.txt

Someone needs to introduce you to a modern programming language. You can do the same thing with far less code and with a language like Clojure (my favorite,) you could make it multithreaded very easily as well.

Random is just fun.. exciting... beautiful...
To catalog all my mpg videos... one simple function.. and it's created in 2 or 3 minutes
a text file with 2 lines per video is created..
The first is the "search criteria line" and the 2nd is the path to said video.
3000 videos would result in 6000 lines in the text file...
The app counts the number of lines in the file and generates a RANDOM number
It reads to that point and starts the search.. If no match found in 3 lines then another random is generated..

Random numbers are not a great way to store unique identifiers. You should use a sequence of some sort where you know the last value because incrementing a counter is constant run time. Scanning everything is linear time at best, assuming you don't encounter a duplicate, worst case run time is likely polynomial since it's possible that you could hit every value for every file you have. It's unlikely but, linear run time for an operation like that for best case is terrible. That's another good reason to use something like SQLite, it can do that for you.

The date time stamp on videos can be very useful...
We did a golf tournament video.. 2 camcorders at a beautiful hole..
by having the camcorder times set to 15 or 20 seconds apart.
We could download both camcorders clips for each golfer...
And they would be automatically sorted into order.
I created a specific catalog function that set these 2 lines up...
one set to play the full 4 or 5 second clip
a next set to play the last second in slow motion (the swing)

That's neat, but it's implementation independant. A better implementation could do the same thing, but I see what you're saying. As I said before, my issue isn't with what you're doing. It makes sense. I just question the implementation. It's very rudimentary.

When we originally did the family albums.. scanned in around 2000.. we put the metadata into the hand made catalog file.
Now.. I'll want to change the CP feature that copies files to another folder... so that it will change the name to contain all the catalog details..
I don't worry much about data duplication.. I worry about losing stuff.. and that doesn't happen when everything I have is in 1 huge folder..
I just drag and drop that to an external 3TB drive.. Duplicates and all..

Being able to explain how Search works.. And the source code right there.. People can make use of it..

I would rather have this kind of data in a PostgreSQL database that I can do a localized (language specific,) full text search as opposed to a simple string matching search. You also shouldn't need to copy the files, you should only have to keep track of where they are. You also should care about data duplication. I have almost 700GB of video (< 1300 files,) and 53GB of pictures (19k files.) Duplicating the pictures isn't terrible, but video is not feasible without another drive, not to mention that copying that much data takes a lot of time.

All I'm saying is that good software solutions exist and that people like me, who do this kind of thing for a living, are going to take a different approach for good reasons and when you open your software for critique, expect some honest feedback.

Edit:

The only reason there is such a thing as MetaData is because filenames were limited to 8 characters...
Having to have a special program.. right off the bat to change and see the details is rubbish.

No, it most definitely is not. Metadata is important because it enables you to do things like sort and filter on data that's not related to a file's name; data that is optional.

 

[FordGT90Concept]

The only reason there is such a thing as MetaData is because filenames were limited to 8 characters...
Having to have a special program.. right off the bat to change and see the details is rubbish.

Creation Date, Modified Date, Accessed Date, Hidden attribute, Read-Only attribute, etc. these are all meta data for NTFS has for files (and most are for directories too). Filename is pretty minor in the grand scheme of things.

There's also metadata embedded in virtually all file types as well. Photos can have information about the camera that took it. Music can have album art, track, and artist information. TV recordings can have season, series, and episode data. Films can have information about the original format that it was converted from. The list is virtually endless.

A program like Emby not only sorts it all and makes it searchable, it makes it network/web accessible, keeps track of consumption based on user accounts, and in situations that requires it, converts it on the fly to a device supported format/container.

 

[Aquinus]

Filename is pretty minor in the grand scheme of things.

Well, it's minor in the sense that it's a really simple thing. It's huge in the sense that it represents a particular file on the file system in the context of a particular directory. Don't undersell unique identifiers.

 

[SpectateSwampBANNED]

Someone needs to introduce you to a modern programming language. You can do the same thing with far less code and with a language like Clojure (my favorite,) you could make it multithreaded very easily as well.

Random numbers are not a great way to store unique identifiers. You should use a sequence of some sort where you know the last value because incrementing a counter is constant run time. Scanning everything is linear time at best, assuming you don't encounter a duplicate, worst case run time is likely polynomial since it's possible that you could hit every value for every file you have. It's unlikely but, linear run time for an operation like that for best case is terrible. That's another good reason to use something like SQLite, it can do that for you.

That's neat, but it's implementation independant. A better implementation could do the same thing, but I see what you're saying. As I said before, my issue isn't with what you're doing. It makes sense. I just question the implementation. It's very rudimentary.

I would rather have this kind of data in a PostgreSQL database that I can do a localized (language specific,) full text search as opposed to a simple string matching search. You also shouldn't need to copy the files, you should only have to keep track of where they are. You also should care about data duplication. I have almost 700GB of video (< 1300 files,) and 53GB of pictures (19k files.) Duplicating the pictures isn't terrible, but video is not feasible without another drive, not to mention that copying that much data takes a lot of time.

All I'm saying is that good software solutions exist and that people like me, who do this kind of thing for a living, are going to take a different approach for good reasons and when you open your software for critique, expect some honest feedback.

Edit:

No, it most definitely is not. Metadata is important because it enables you to do things like sort and filter on data that's not related to a file's name; data that is optional.

I don't worry about computer usage.. I love it when random is running hard and busy busy.
Full text search is what this app is all about.
to do that ... I pad the detail text with a leading and trailing space.
then reduce multiple spaces and tabs to 1 space.
and
show all results highlighted
show proximity matches in a different color.
put displayed results to an extract text file.
show large font text ... with slowed print.. a minor pause between characters...
and more a lot more..
You do have a lot of stuff.. Good going..
I'd like to randomly see all your data.. Unless it's weirdo material...

Oh I love the feed back.. Way more is learned from someone that disagrees.. that's for sure.

 

[Vayra86]

The only reason there is such a thing as MetaData is because filenames were limited to 8 characters...
Having to have a special program.. right off the bat to change and see the details is rubbish.

Sorry but no, wrong on every possible level. Filename limitations were never the reason metadata was considered a good idea.

Its fine you've found a solution that works for you, but its grossly inefficient. Its not better in any way than existing solutions - even the simplest ones... You've devised something that takes far too much time per file to implement. 2-3 minutes might be fine for home use fiddling with video in free time, but explain this to any company and they'll say you're a nutcase. Again, its fine it works for you, but this is the perspective it faces in the real world.

I've done Functional IT management and querying was my number one tool to probe a database for problems but also to ensure or make visible data quality, the amount of data, and provide the business side with detailed reports and impact analyses. You can't do any of that without metadata, without it, all your files are just huge walls of text that only serve as the search string you placed there, manually. In terms of data quality and management that is completely invalid. The larger your data pool gets, the more problems you will face.

In conclusion, I'd suggest, like others have, that you explore SQL as a language and learn a thing or two about querying. Its actually really quite fun to get a handle on and being so well equipped to extract information from a huge pile of files / a database. While you learn SQL, you will immediately get a handle on the types of metadata you'd want to have for your use case, as you start thinking about and looking at your data in a different way. You can figure out things long before others even notice it. You can see trends, for example, and in your practical use case, that would allow you to see that your interests have shifted, for example by querying a subject matter over a period of time. You'd be able to see that you no longer make 'squirrel' videos and have turned to 'moose' or 'mountain ranges'. And at the same time, that same query gives you the file ID's or locations you need to open them.

 

[SpectateSwampBANNED]

MetaData I could and have lived without it..
My app puts audio / mp3 metadata into the catalog file.. but I didn't and don't care for that approach. nope
I don't care about the file internals until there is a problem..
When doing 50 plus telco and cablesystem billing system ETL's
Search was handy when plowing through toll files and other unstructured data... Billing files.. and more.

I keep everything I need with this app..
I place the shortcut in the upper right hand corner... the hot spot.

Nobody has better access to their video, audio, Pictures and text than this.

 

[SpectateSwampBANNED]

www.youtube.com/watch?v=MJZfY6lxBaQ

3 videos in this reshoot.
Their filenames are set to control the video segment thumbnails.
The first starts at normal speed as a jet boat approaches. Then switches to slow motion as it passes AND freezes on last frame for a Sec
Then the 2nd has the ToKin Stone blasting out some smoke.
Then the 3rd is the start of the Jet Boat race.. with a repeat play.. They point up river to start and head down river.
the it loops on these 3.

Now those are THUMBNAILS..

Now I quickly catalog video as I shoot 'em and offload.. And shoot some more.

 

[SpectateSwampBANNED]

Sky Creature moons me in segment of demo85 pict7 9 clips 20191105114917

by creating multiple copies of the special file.. I can easily have these segments play time and again.
All by these controls in the filename.

Tobacco Leaf Flyer clip1 left side 23 way up again begin==40000 start==43750 wait=.65 speed=125 len=262260 demo85_pict7.mpg
Tobacco Leaf Flyer clip2 lower right to center and back again begin==58000 start==61800 wait=1.1 speed=125 len=262260 demo85_pict7.mpg
Tobacco Leaf Flyer clip3 the biggie again begin==59000 start==63600 wait=.6 speed=125 len=262260 demo85_pict7.mpg
Tobacco Leaf Flyer clip3.5 up to the cloud from lower right again begin==62100 start==64800 wait=.6 speed=125 len=262260 demo85_pict7.mpg
Tobacco Leaf Flyer clip3.7 poking out of cloud again begin==64000 start==69000 wait=.6 speed=125 len=262260 demo85_pict7.mpg
Tobacco Leaf Flyer clip3.8 up from bottom off to left again begin==70000 start==73000 wait=.6 speed=125 len=262260 demo85_pict7.mpg
Tobacco Leaf Flyer clip4 just barely bottom left again begin==125000 start==129800 wait=.5 speed=125 demo85_pict7.mpg
Tobacco Leaf Flyer clip5 from right arcoss then down again begin==160000 start==162200 wait=1 speed=125 demo85_pict7.mpg
Tobacco Leaf Flyer clip6 lower right side again begin==205000 start==208250 wait=.5 speed=125 demo85_pict7.mpg

 

[R-T-B]

I'd like to randomly see all your data.. Unless it's weirdo material...

...

I like frogs, mkay?

 

[SpectateSwampBANNED]

...

I like frogs, mkay?

Grab a copy of the original file

http://www.telusplanet.net/public/stonedan/demo85_pict7.mpg

See the segments at regular speed
No creature known to Man or Woman moves like that... Nope.

 

[SpectateSwampBANNED]

I manage all my video with Long FileNames
Cataloging the thumbnail details in the FN
Start point, Duration, Speed, Repeat, Freeze and more
unbreakable thumbnail collage

Am I good or what?

 

[SpectateSwampBANNED]

Geo Location in the filename is extremely useful
Macaroni Flats bridge and ford 49°14'58.8N 118°41'44.2W 20200909123741.mpg

 

[SpectateSwampBANNED]

Been going over all my decades of video and pics...
I've changed my app so I can easily use long Program Names as well as File Names to pass parameters to the app
I was tired of renaming hundreds of pics that I had and changed my "CP" feature to copy the files to another with
more details in front... It Is easy.. Just a golden noodle or two and shazam. all copied to another folder with a nice name.

check this location on google maps What a great feature for filenames... YUP
49°14'58 8N 118°41'44 2W

 

[SpectateSwampBANNED]

What rubbish most metadata is.
Most of the time I could care less about; type, size and date modified...
Give me the long names.

 

[Frick]

VisualBasic is the devil.

 

[SpectateSwampBANNED]

Like jaundra
it is stored as a code and needs to be translated.
What a stupid name on top of it all.
A million jaundra's that nobody cares about.....