Hello,
I recently read an article on object prevalence at developerWorks.
I have been experimenting with an implementation of object prevalence in
Ruby, called Mnemonic.
I was wondering if anybody is using Mnemonic (or Prevalayer) as a storage
mechanism in any of their current projects? Also, can it be used within a
apache/mod_ruby environment for web-based applications?
Regards, Oliver.
A link to the article:
I have tried Mnemonic a couple of months ago. It seems very promising,
but I could not use it for web-based applications because it doesn’t
handle mulitple processes/threads/users very well.
It would be ideal if it is implemented with a strategy like Suneido
(multiple version of a datastore, to provide a simple but reliable
transaction mechanism):
http://www.advogato.org/article/250.html
Cheers,
Erik.
Oliver Beddows wrote:
Hello,
I recently read an article on object prevalence at developerWorks.
I have been experimenting with an implementation of object prevalence in
Ruby, called Mnemonic.
I was wondering if anybody is using Mnemonic (or Prevalayer) as a storage
mechanism in any of their current projects? Also, can it be used within a
apache/mod_ruby environment for web-based applications?
Regards, Oliver.A link to the article:
Erik Terpstra erik@solidcode.net writes:
but I could not use it for web-based applications because it doesn’t
handle mulitple processes/threads/users very well.
Pattern of Mnemonic usage:
multiple processes: use a single mnemonic process for, what’s the
buzzword?, the business logic? Of course this complicates things and
is usually implemented as if accessing a centralised database.
multiple threads: use mutex
multiple users: make mnemonic file location adjustable on per user
basis.
YS.
Hello,
I have done much research on object / relational mapping patterns. After all
that effort to get a clean object design, it just feels like I’m wasting my
energy. I end up reverting to some hybrid relational model.
I understand the concept of multiple threads, and the mutex class.
I have only experimented with a few examples of using Mnemonic.
How can I avoid using a relational database when Ruby is concerned?
Cheers, Oliver.
On Friday 20 September 2002 02:48 pm, you wrote:
Pattern of Mnemonic usage:
multiple processes: use a single mnemonic process for, what’s the
buzzword?, the business logic? Of course this complicates things and
is usually implemented as if accessing a centralised database.multiple threads: use mutex
multiple users: make mnemonic file location adjustable on per user
basis.YS.
Hello,
I am sorry for the lack of clarity in my last message, it had been a long day.
Basically I would really like to use Mnemonic within an Apache or Webrick
environment. I guess I do not understand WHY there would be a problem with
concurrent access (is that the correct term?). I am also a little confused
about the issue of multiple processes, threads, and users.
I know as a first step I should probably gather some more information on
POSIX threads. This may help me to understand the subject of threads more
clearly. Apart from that what else may I need?
Thank you, Oliver.
On Friday 20 September 2002 02:48 pm, you wrote:
Pattern of Mnemonic usage:
multiple processes: use a single mnemonic process for, what’s the
buzzword?, the business logic? Of course this complicates things and
is usually implemented as if accessing a centralised database.multiple threads: use mutex
multiple users: make mnemonic file location adjustable on per user
basis.YS.
Oliver
Basically I would really like to use Mnemonic within an Apache or Webrick
environment. I guess I do not understand WHY there would be a problem with
I’m using Mnemonics in conjunction with Apache. The connection between the
two is made in my case using DRbSessions (I’m not sure whether a
FastCGI-thingy would be feasible, but it might be worth investigating)
In any case the Business-Logic is running as an independent Service which
provides session- and user-management via some kind of
Interprocess-Communication.
Hannes