@skquinn @jalefkowit @natecull
Chandler choked and died on this concept, spawned a great book Dreaming In Code. They wanted this super flexible object database model that would allow people to recontextualize any bit of information intelligently within their workflow (e.g. turn a text memo into an email into a calendar event by manipulation of the fields) and organize that information based in object inherited implicit tags as well as user tags.
@natecull @jalefkowit @skquinn
the scope got gradually narrowed down, and eventually they gave up the interesting universality of those ideas. originally it was going to support a fully peer to peer database of objects with full peer to peer sharing, etc etc. It was some crazy shit. As the project progressed they got a lot more traditional.
@ntk @Irick @natecull @jalefkowit If all the important stuff is the file's actual data as opposed to a fork or stream or whatever... then what are we supposed to do with these extra pieces on operating systems/environments that don't have a clue about them? And what good are they, really, if they aren't preserved across a transfer between Windows, Mac OS X, and GNU/Linux?
@skquinn @Irick @natecull @jalefkowit
I don't know how much the HFS design was built directly around Mac OS requirements. The resource fork is used for noncompiled data. Why a fork and not another file section or another file? I don't know; it seems foolish to me. Named forks later got used by extended attributes as well.
NTFS supported streams originally for Mac OS compatibility support! I think they are mostly used by malware.
@ntk @Irick @natecull @jalefkowit I have used Windows versions after 98. They have not gotten any better. In fact, it seems like Microsoft has moved things around in later versions purely for the sake of change; it's most obvious in Office, but I had to ask my mom how to get to the Control Panel in Windows 10. And Windows is supposed to be the most "user friendly" of the consumer operating systems, I've been told.
@skquinn @natecull @jalefkowit
IIRC it used URI for external reference.
I was 'aware' of files and directories but it wasn't a native part of their data model. They definitely had more of an idea about defining the data along with how it was supposed to be used into a single object that got passed around. It could generate documents, etc, but those were not native paradigms.
The Vulpine Club is a friendly and welcoming community of foxes and their associates, friends, and fans! =^^=