> Most of the users I know could care less about the promised features. > Even the so called "power users" really don't care about their file > systems (or file formats). Maybe I have a dark and pessimistic view of > those who exist in the middle of the bell curve. That probably is more > than just possible :)
I'll just comment that this is largely due to not understanding the benefits of it. If you put a million files in a single directory, and then start working with them, you'll understand one of the easiest to explain areas where the problems lie. This may not be a big issue today, but it will become one in the foreseeable future, as the amount of data created grows, the speed of it's creation grows, and the expectation of it's retention also grows. Even apart from having a large directory listing, look at the solution you'll face if you run a FS out of inodes. This is still a more or less rare event, but it's becoming more common all the time. Note that this happens on EXT2/3 but doesn't happen back with Reiser3. Having other data stored in the FS? Maybe that'll be resolved by the various indexers. We'll see. But even there, how much faster would that indexing be, if it was indexing a database rather than flat files. For the record, this isn't a new idea either. I did development work on an AS/400 in the 90s, and it had used a database for it's filesystem for years, that was it's unique calling, and a large part of the reason it was so incredibly efficient at what it did. On Linux, this was first addressed by Reiser4, which was directly modelled after the 400, unfortunately this project seems to have gone by the wayside now. Maybe it'll continue as a project in 20 years to life... Kev. _______________________________________________ clug-talk mailing list [email protected] http://clug.ca/mailman/listinfo/clug-talk_clug.ca Mailing List Guidelines (http://clug.ca/ml_guidelines.php) **Please remove these lines when replying
