Save files via http headers / maintain files on server?

[xcislav]

I didn't found any nice materials comparing the two ways of information unload to user. Neither have I a firm understanding of how high my learning curve should be in both cases, which of which to choose for better way. I'm a beginner asking for your professional word of advice. Help me to better formulate for search what I want, Please.
I know keeping files and images in fs is better served by os/fs-drivers but i'm just learning and just trying to remember select * for "poking around". + i do hate images text analyzing it'll will be!
Is it better to save files from MySQL fields via http headers, or it's even more better to perform the saving of the kept txt files (in one specific folder) and all the maintenance of tempfiles on server (archiving, indexing, cron/planning)?

So:
1. Learning & get experience in file maintenance?
vs
2. Finding out how to deal with protocols (download resuming)

What I've found so far:

header('Content-Description: File Transfer');
header('Content-Type: application/octet-stream');
header('Content-Disposition: attachment');
header('Content-Transfer-Encoding: binary');
header('Expires: 0');
header('Cache-Control: must-revalidate');
header('Pragma: public');
header('Content-Length:');

There's a textfield full of appended edits of a specific user (for example list-ordering/or article appending). Headers should deliver it as a file, while skiping the file creation on a server.
Or do I just learn how to do it with files and save just file link then accomplish the persistent file-links cleaning (because there would be lots of them)?

One one side there is a certain straightforwardness in dealing with files, but then again I'd prefer the both ways while considering continuing to study MySQL-approach. I understand protocols/headers with PHP are no easy task that's why I'm hopefully leaving the last word for the professional....

Thanks in advance!