Check me out:

http://www.c-sharpcorner.com/Code/20...adedXmlDoc.asp

Multithreaded XML Document for Read/Write Access
by John Bailo

["Followup-To:" header set to comp.os.linux.advocacy.]
On Wed, 21 Jul 2004 08:26:58 -0700, Praxis Happenstance wrote:
>
> Check me out:
>
> http://www.c-sharpcorner.com/Code/20...adedXmlDoc.asp
>
> Multithreaded XML Document for Read/Write Access
> by John Bailo

That's utterly retarded. Why not just bind the XmlDocument to a DataSet
if you want it to work like a database?

The XML classes are already thread-safe, and so is the DataSet.

Just create a XSD schema, then a master DataSet from it. Whenever a doc
comes in, do a DataSet.Merge() to bring in the new record.

Why would they publish that?!?

--
FreeBSD 4.8-RELEASE i386
9:55AM up 123 days, 1:58, 1 user, load averages: 0.01, 0.02, 0.00

Hi Democritus Rising,

I think General Protection Fault is write about one thing,
C#'s routines for writing to ( unstructured ) XML fields
( e.g. DataSet.Merge() ? )
are Very probably threaded/mutexed already,

Hence, no need to check for collisions.

Jeff Relf wrote:

> Hi Democritus Rising,
>
> I think General Protection Fault is write about one thing,
> C#'s routines for writing to ( unstructured ) XML fields
> ( e.g. DataSet.Merge() ? )
> are Very probably threaded/mutexed already,
>
> Hence, no need to check for collisions.

That is the stupidest conjecture I have ever heard.

--
incognito http://kentpsychedelic.blogspot.com/
fairgrounds http://home.earthlink.net/~jabailo/fairgounds/id6.html
storyteller http://home.earthlink.net/~jabailo/
latest article
http://www.c-sharpcorner.com/Code/20...adedXmlDoc.asp

Jeff Relf wrote:

> Hi Democritus Rising,
>
> I think General Protection Fault is write about one thing,
> C#'s routines for writing to ( unstructured ) XML fields
> ( e.g. DataSet.Merge() ? )
> are Very probably threaded/mutexed already,
>
> Hence, no need to check for collisions.

PS -- My next feat is to:

Apply the Mutex at the Node level ( field level locking )

--
incognito http://kentpsychedelic.blogspot.com/
fairgrounds http://home.earthlink.net/~jabailo/fairgounds/id6.html
storyteller http://home.earthlink.net/~jabailo/
latest article
http://www.c-sharpcorner.com/Code/20...adedXmlDoc.asp

Hi Democritus Rising ( John B. ),

Re: My suggestion that C#'s routines for writing to XML fields
are probably threaded/mutexed already,
denecessitating collision checks.

You replied, <<
That is the stupidest conjecture I have ever heard. >>

John John John,

How could you ever know
if you fixed this ( imaginary ? ) collision problem
if you're not even sure whether it exists or not ! ?

["Followup-To:" header set to comp.os.linux.advocacy.]
On 23 Jul 2004 01:44:18 GMT, Jeff Relf wrote:
> Hi Democritus Rising,
>
> I think General Protection Fault is write about one thing,
> C#'s routines for writing to ( unstructured ) XML fields
> ( e.g. DataSet.Merge() ? )
> are Very probably threaded/mutexed already,
>
> Hence, no need to check for collisions.

My MCSD in .NET says I'm definitely right.

--
FreeBSD 4.8-RELEASE i386
10:20AM up 124 days, 2:23, 1 user, load averages: 0.01, 0.02, 0.00

General Protection Fault wrote:

> On Fri, 23 Jul 2004 03:59:18 GMT, Democritus Rising wrote:
>> General Protection Fault wrote:
>>
>>> On Thu, 22 Jul 2004 15:40:15 GMT, Democritus Rising wrote:
>>
>> Data-base. A 'base' or storage of data. It does not specify the
>> structure
>> of topology of the data. A text file can be a database, for instance, a
>> comma delimited textfile /is/ a database.
>
> A database needs a structure. You can't just keep adding crap with
> different formats into a repository and call it a database.
>
> And, an XML document does not meet the definition of a hierarchical
> database. It's a hierarchy, but not an HDBMS.
>

Here is someone thinking along the same lines ( though he is approaching it
from extending the idea of XML as a database, not so much the
threaded/multiuse aspect ).


Creating an In-Memory Database Using XML and XPath -- Part 1
http://www.15seconds.com/issue/010409.htm

Part II
http://www.15seconds.com/issue/010410.htm

Some interesting stats on performance in Part II:

1) SQL Query - New Connection per Query 58.953 ms

2) 8) Branched XML - Keywords + IDs as Tags - Grouped Keywords 2.340