Speed of CD tree /directory tree

[buckley]

Hi,

I would love to use this feature (alt-F10) but it currently takes too long (about 5secs on a 2.4 Gig quad machine) to load and use it all the time.

Are there any plans to make it fast (please do!) or tips?

Regards, Tom

[Flint]

Are there any plans to make it fast (please do!) or tips?

How do you imagine scanning your whole several (or several hundreds?) gigabytes hard drive to be fast?
Actually, TC saves the directory tree it read, you just need to wait till it finishes (and not remove the treeinfo.wc files in the roots of the drives), then further calls of Alt+F10 should be fast.

Also, you can use separate panel trees. They are faster, because the do not read whole list of dirs and subdirs at once, but process them on request.

[buckley]

"How do you imagine scanning your whole several (or several hundreds?) gigabytes hard drive to be fast?"

I can imagine it I didn't say it was easy.

Google search billions of pages and gives back a result in milliseconds using moderate hardware but I don't want to start a discussion about whether is possible or not. AFAIK adding a good database backend with the right indexes should speed it up.

Its slow the second time I ask for it (also 5 secs) so it does not perform better when asked many times, at least not on my PC.

A function that would search through all the filenames on my pc (disregarding contents of course) and does this very fast would be very usable at my part.

[Flint]

Its slow the second time I ask for it

I see. I thought, you were talking about the first building the tree when TC reads all the disk.
Well, the slowness I suppose is explained by the fact that the treeinfo file is a plain text which should be read and parsed and converted into internal directory structures. Some speed improvements would be nice, of course…

[JohnFredC]

Google search billions of pages and gives back a result in milliseconds using moderate hardware.

Actually Google is not searching all those pages when you initiate a search. Part of the Google magic is how it pre-processes (i.e., searches and correlates) all those pages in advance and continuously, an activity which is a form of indexing, and whose obviously efficacious algorithm is proprietary to Google.

Google search is of a structure of a structure of a structure... etc.

So: not in any way comparable to TC displaying a tree of your hard drive.

[buckley]

Yep, here we go about the google comparison. I am a programmer of profession so I know where the comparison fails. My point was that the data, once scanned, can be efficiently searched through even if the size is immense. Even on not so powerful PC's which maens 1 Ghz (!) these days.

But if you realy kow how Google does it searched please share