open-source network database

#11

Thanks! It looks very promising.

#12

In article <I_GVa.14167$Hr.571713 (AT) twister (DOT) socal.rr.com>, Steven C.
<nospam (AT) xxx (DOT) com> writes

Quote:

Andy,

This sounds very interesting. I'm looking for something simple. My
database will just be key value pairs. The key is a string and the value is
binary data under 32K in length. I plan on having less than 100K records.
I will be doing mostly reading. One thing I'm curious about is lets say a
value field for a given key is exactly 10K and I updated the value and it is
again exactly 10K will it over write what was there which is the way I want
it to work. I don't want the DB to grow in size. Right now I'm using the
MS Jet engine and it seems to grow and grow and grow.

Thanks,
Steven

This looks almost perfect for my C code ...

Basically, this stores key-value pairs in a hashed file. The "value" bit
is random length (stored with the key if it's small enough), and when
using dynamic hashing the file is resized "on the fly" so fast the user
is unlikely to notice.

Given your example, given the key it would take two head movements to
retrieve your data (one for the key and data location, the second for
the data itself). And yes - if you updated the data and it did not grow,
it would overwrite the previous data.

Cheers,
Wol
--
Anthony W. Youngman - wol at thewolery dot demon dot co dot uk
Witches are curious by definition and inquisitive by nature. She moved in. "Let
me through. I'm a nosey person.", she said, employing both elbows.
Maskerade : (c) 1995 Terry Pratchett

#13

"Anthony W. Youngman" <thewolery (AT) nospam (DOT) demon.co.uk> wrote

Quote:

In article <I_GVa.14167$Hr.571713 (AT) twister (DOT) socal.rr.com>, Steven C.
nospam (AT) xxx (DOT) com> writes
Andy,

This sounds very interesting. I'm looking for something simple. My
database will just be key value pairs. The key is a string and the value is
binary data under 32K in length. I plan on having less than 100K records.
I will be doing mostly reading. One thing I'm curious about is lets say a
value field for a given key is exactly 10K and I updated the value and it is
again exactly 10K will it over write what was there which is the way I want
it to work. I don't want the DB to grow in size. Right now I'm using the
MS Jet engine and it seems to grow and grow and grow.

Thanks,
Steven

This looks almost perfect for my C code ...

Basically, this stores key-value pairs in a hashed file. The "value" bit
is random length (stored with the key if it's small enough), and when
using dynamic hashing the file is resized "on the fly" so fast the user
is unlikely to notice.

Given your example, given the key it would take two head movements to
retrieve your data (one for the key and data location, the second for
the data itself). And yes - if you updated the data and it did not grow,
it would overwrite the previous data.

Cheers,
Wol

Wol,

Where do I find your C code?

Thanks,
Steven

#14

In article <767f8298.0308011208.43dbb43e (AT) posting (DOT) google.com>, Steven C
<steven1832 (AT) hotmail (DOT) com> writes

Quote:

Wol,

Where do I find your C code?

Have you got my emails? I've altered the "reply-to" to my office, as

hotmail falls foul of my spam filters at home ...

Cheers,
Wol
--
Anthony W. Youngman - wol at thewolery dot demon dot co dot uk
Witches are curious by definition and inquisitive by nature. She moved in. "Let
me through. I'm a nosey person.", she said, employing both elbows.
Maskerade : (c) 1995 Terry Pratchett

#15

In article <I_GVa.14167$Hr.571713 (AT) twister (DOT) socal.rr.com>,
"Steven C." <nospam (AT) xxx (DOT) com> wrote:

Quote:

Andy,

This sounds very interesting. I'm looking for something simple. My
database will just be key value pairs. The key is a string and the value is
binary data under 32K in length. I plan on having less than 100K records.
I will be doing mostly reading. One thing I'm curious about is lets say a
value field for a given key is exactly 10K and I updated the value and it is
again exactly 10K will it over write what was there
If you use the OOFILE Professional version, which uses the c-tree Plus

backend, it shouldn't grow in this circumstance (characteristic of
c-tree variable length storage).

The dBase backends (OOFILE Personal) won't grow either if memos are
replaced by memos of the same size.

I wouldn't recommend the RAM-based streamable backend in the open source
version for that many records (although it should be fine given enough
RAM). Hmmm. Interesting test - I haven't got much time right now but if
there's interest I'll modify a sample and try it. Would hammer VM a bit!

As a general issue, given the XML support already in OOFILE
(http://www.expatpp.com/), it's trivial to extend the free version so it
streams to and from an XML file.

Is this something you think worth me spending a few hours on?

BTW I may miss newsgroup-only replies to messages so if you were
expecting a response to a posting, please email me.

--
Andy Dent BSc MACS AACM http://www.oofile.com.au/
OOFILE - Database, Reports, Graphs, GUI for c++ on Mac, Unix & Windows
PP2MFC - PowerPlant->MFC portability

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

open-source network database

comp.databases comp.databases

Re: open-source network database - 07-30-2003 , 11:40 AM

Re: open-source network database - 07-30-2003 , 05:42 PM

Re: open-source network database - 08-01-2003 , 03:08 PM

Re: open-source network database - 08-07-2003 , 04:57 PM

Re: open-source network database - 08-12-2003 , 09:44 PM