Slashdot Mirror

You don't need special features like you suggest in a filesystem to manage your files properly.

I doubt that any single person, let alone "Grandma", will ever manage "gigabytes and gigabytes" of highly structured data.

I'm an awfully untidy person (my home directory looks just as cluttered as my workroom), yet even I don't have any problems locating my project files or my photo collection.

If you really need to manage huge amounts of structured data, you may as well use an SQL database.

And performing search operations on such databases will only be fast if there is an index for the key(s) the search is based on. That is, at some point you'll have to anticipate the ways your data will eventually be used. Structuring data is a non-trivial task.

Btw, does anybody use the "indexing" feature of MS Office?

I wonder how many years it will take for Linux to play catchup to these kinds of features that I imagine will be commonplace by the time 2005 rolls around.

You can have database-like filesystems on Unix today. You can't on Windows.

And these features are in no way new; they have been "commonplace" in other OSes years ago.

Heck, I'm still holding my breath for a hardware accelerated X replacement, but the Linux zealots are too afraid of change for that to happen anytime soon...

WTF are you talking about? X has its problems, but it supports hardware acceleration.

Gnutella
Bit Torrent
Freenet
Reiserfs
Linux Kernel
Open SSH
Encrypted Filesystems
GnuPG

At least in my opinion p2p and crypto are the edges in coding right now. Both can be hugely successful if you succeed in writing them properly. They can also be a huge failure if done improperly. Personally, I'm amazed that there aren't more p2p worms/remote exploits out there. Every now and then there are a few breaks in crypto from a weird angle, but in general they have been very successful as well.

One of the main points in M$'s argumentation is that the GPL hurts the industry, because you cannot write commercial apps based on GPL software, but the GPL is not the only Open Source license and most reasonable OS libraries are licensed under the LGPL or similar licenses that allow developing commercial software.

Open source developers simply have to choose an appropriate license for their project when they start. And if they find out that they chose wrong there is still the possibility to change the licensing terms. A very prominent example for such a license change is the Wine project that changed it's license from X11-like to LGPL recently.

If a company finds an OS library useful for their own project, but they cannot use it, 'cause it's GPL, they can still contact the author and ask for different licensing terms. They'll probably have to pay for that then, but they'd have to pay for a commercial product, too. So even GPL'd libraries are not really a hurdle for commercial software development. A good example for such dual licensing is ReiserFS, which is published under the GPL, but sold under different licensing terms to companies that want to use it commercially.

True, if you know a priori that you are dealing with character data, your file metadata are highly stable, and the recursive nature of a file hierarchy is the scariest piece, one would suppose that a customized file system database engine might not be too huge.
But I speak well above my skill level, having just written my first little bit using a RegEx. Go, boost.
Maybe keeping an eye on Hans is a good idea.

Well, in that vein, I found these ramblings about ReiserFS rather interesting...

I regret using reiserfs on my box due to the ugly hack for bad block handling, see here.

1. It's extremely buggy. For example, it's supposed to run only when the system is idle, but it usually chooses to run right when I'm doing some big IO-intensive process. It periodically also consumes all available memory and CPU cycles, so I usually leave it turned off.
2. It doesn't index very much metadata. For most files, it only indexes the standard file system attributes (size, date, extension, etc.) For MS Office documents, it can index the Summary Information property stream, which usually includes things like Author, Title, etc. If you have IIS running on a box, it can also index the META tags in HTML documents. Other than that, not much. It does have some extensibility , by means of DLLs that implement the IFilter COM interface, but no 3rd-party vendor seems to have done much with that, perhaps because of #1.

You may ask what common in ReiserFS and database. Well, just look at it its whitepaper. The word 'database' is everywhere in it. Hans Reiser started to think about it long time ago. And they already are working on it.

You may ask what common in ReiserFS and database. Well, just look at it its whitepaper. The word 'database' is everywhere in it. Hans Reiser started to think about it long time ago. And they already are working on it.

Do some research on ReiserFS for how OSS projects should be run in this respect. PS, companies don't always want 'exciting' they just want results.