ABI versioning

I am working on packaging metakit for Debian, and I noticed that there
is no SONAME in the default library provided. I was wondering what
rules you guys have for ABI compatibility - will you not break ABI
compatibiliy, or is the 4 in mk4 the ABI/API version?

In other words, is there any easy way to tell when a new release of
metakit will break programs that link against it dynamically?
(recompilation is not an issue)

Sean Hunt
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the "metakit" group.
To post to this group, send email to metakit <at> googlegroups.com
To unsubscribe from this group, send email to metakit-unsubscribe <at> googlegroups.com
For more options, visit this group at http://groups.google.com/group/metakit?hl=en
-~----------~----~----~----~------~----~------~--~---

metakit and python 2.5 on windows

Hi, i need to make a windows build of a python program that uses
metakit. I have read that i should use the latest version of metakit
to be able to use python 2.5. However i have been unable to find
binaries for download. I ask one of 2 things:

1) Someone share their binaries (from what i understand, a single DLL)
if they have made them

2) Someone point me to instructions on how to make them myself. I do
not have a C compilation environment set up. I have python, wx and
py2exe setup that i have used before, but this time i would need
metakit too.

I'm on a bit of a tough spot here so i'd really appreciate any help.

Hugo
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the "metakit" group.
To post to this group, send email to metakit <at> googlegroups.com
To unsubscribe from this group, send email to metakit-unsubscribe <at> googlegroups.com
For more options, visit this group at http://groups.google.com/group/metakit?hl=en
-~----------~----~----~----~------~----~------~--~---

Question - Memory Usage Patterns

Hi,

Has anyone made a study of Metakit's memory usage patterns?

I'm asking because I've created an 'associative database' layer on top
of Metakit, which I plan to deploy on a server with harsh memory
limitations.

My database over time is likely to end up with tens or hundreds of
thousands of rows across 3 views. Use of derived rows will be kept
fairly minimal.

Metakit appeals to me because of its nested views. But for such a size
of database, I'm worried that memory usage during the lifetime of
'storage' objects will grow in proportion to the size of the database.

For such a size of database, would I be better off using my MySQL
version of the layer?

Cheers
Dave

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the "metakit" group.
To post to this group, send email to metakit <at> googlegroups.com
To unsubscribe from this group, send email to metakit-unsubscribe <at> googlegroups.com
For more options, visit this group at http://groups.google.com/group/metakit?hl=en
-~----------~----~----~----~------~----~------~--~---

corrupted data

I'm trying to use metakit with mingw C++ under windows XP, but my data
seems to be getting curupted. this is my code:

#include "mk4.h"
#include "mk4str.h"

#include <iostream>
#include <string>

const std::string test_table_name("tst");
const std::string ID_field_name("ID");
const std::string result_field_name("result");
const std::string test_number_field_name("testnum");

c4_StringProp pID (ID_field_name.c_str());
c4_DoubleProp pResult (result_field_name.c_str());
c4_LongProp   pTestNum (test_number_field_name.c_str());

int main()
{
    std::cout << "testing MetaKit" << std::endl;

    c4_Storage storage ("testmetakit.dat", true);

    std::cout << storage.Description() << std::endl;

    // Start with an empty view of the proper structure.
    c4_View record = storage.GetAs((test_table_name + "["
            + ID_field_name + ":S,"

./libMk4tcl2.4.9.7.so: undefined symbol: __cxa_pure_virtual

I recently compiled 2.4.9.7 and I am running into the following
situation. Any idea where to turn?

% load ./libMk4tcl2.4.9.7.so
couldn't load file "./libMk4tcl2.4.9.7.so": ./libMk4tcl2.4.9.7.so:
undefined symbol: __cxa_pure_virtual

My TCL is a binary build straight out of the fedora distribution.
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the "metakit" group.
To post to this group, send email to metakit <at> googlegroups.com
To unsubscribe from this group, send email to metakit-unsubscribe <at> googlegroups.com
For more options, visit this group at http://groups.google.com/group/metakit?hl=en
-~----------~----~----~----~------~----~------~--~---

C#/Swig bindings

Hi,
  I was looking for a C# binding for Metakit. I found some 2 year old
discussions about someone working on a swig interface for Metakit. But
I found nothing after that.
  Does anyone know if there are swig interface files or C# bindings
made by anyone ?

Thanks,
- dBera
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the "metakit" group.
To post to this group, send email to metakit <at> googlegroups.com
To unsubscribe from this group, send email to metakit-unsubscribe <at> googlegroups.com
For more options, visit this group at http://groups.google.com/group/metakit?hl=en
-~----------~----~----~----~------~----~------~--~---

freeze when deleting storage (2.4.9.7 win32 debug)

Problems with file names containing latin accent characters

Hello all

I've used Metakit for a while, and recently I had some issues with
file names containing accent characters, like á or é. It only occurs
in Windows XP. In Linux all works ok.
But in Windows, when the file name had latin accent characters,
Metakit suppress them and thus it tried to open a non existing file.
I debug into Metakit source code and I found that the character
missing happened in this line:

MultiByteToWideChar(CP_UTF8, 0, fname_,  - 1, wName, MAX_PATH);

I'm not an expert in Win32 API, but after some googling I decided to
change the CP_UTF8 parameter for CP_ACP. And then all began to work ok
as in Linux. The new version of the previous line now looks like that:

MultiByteToWideChar(CP_ACP, 0, fname_,  - 1, wName, MAX_PATH);

I'm not sure if it make sense to change the Metakit sources that way,
but in any case I want to share this issue and maybe get some feedback
about the patch I applied.

Greetings, German

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the "metakit" group.
To post to this group, send email to metakit <at> googlegroups.com
To unsubscribe from this group, send email to metakit-unsubscribe <at> googlegroups.com
For more options, visit this group at http://groups.google.com/group/metakit?hl=en
-~----------~----~----~----~------~----~------~--~---

you got your database in my Python remote objects server

just want to throw this idea out and see it in appeal to anyone.  I would work 
on it myself except for the fact that my wrists have given out (yet again) and 
writing code with speech recognition isn't the most efficient method.

I have been working on a project where I am using Pyro (Python remote objects) 
as a way of providing network access to a variety of operations on top of 
database services.  I have been struggling with a variety of database techniques 
and I keep coming back to metakit as a solution that just feels right in terms 
of level of effort etc.

The main problem is that Python remote object servers are multithreaded.  They 
are either threaded on an object basis or connection basis.  Everything I've 
read so far says that metakit is maybe yes, maybe no, happy with threads in the 
Python environment.

My wish: update/validate metakit to work with threads in general and Pyro 
servers specifically.  I'm not asking for a lot. Metakit access could be single 
threaded which still allows for server applications to be multithreaded.  I 
believe that something this "simple" would be good because, in theory, it's 
either present but on validated or relatively easy to accomplish.  in addition 
to this wish, I would like to see a release of a thread capable/okay metakit in 
conjunction with Pyro 4.0.  A joint project/release would bring good, positive 
attention to both projects.

If folks think it's a good idea, what would it take to validate metakit in a 
threaded environment like the one I propose?  Is there a simpler model for 
dealing with threaded environments that I'm not aware of?

I have templates for Pyro servers that I could use to help build a testbed but 
I'm missing some critical knowledge to be able to complete the project.

Ever hopeful:

---eric

Abused by speech recognition since 1995

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the "metakit" group.
To post to this group, send email to metakit <at> googlegroups.com
To unsubscribe from this group, send email to metakit-unsubscribe <at> googlegroups.com
For more options, visit this group at http://groups.google.com/group/metakit?hl=en
-~----------~----~----~----~------~----~------~--~---

pb with macosx ppc

Hi
I compiled metakit 2,4,7,9 like this :

in builds directory :
export CXXFLAGS="-arch i386 -arch ppc"
../unix/configure --disable-shared
make
make test
This seems to work, but when executing on a PowerPC mac, it fails :

n14 - Insert in non-mapped position
         Failed: A(fp)
I tried removing -O3, but no change. I tried to compile with only -arch 
ppc, but it is the same...

Any help would be appreciated. I use i686-apple-darwin8-g++-4.0.1
Thanks

-- 
Très cordialement,

Riccardo Cohen
ligne directe : 02-47-49-63-24
-------------------------------------------
Articque
http://www.articque.com
149 av Général de Gaulle
37230 Fondettes - France
tel : 02-47-49-90-49
fax : 02-47-49-91-49

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the "metakit" group.
To post to this group, send email to metakit <at> googlegroups.com
To unsubscribe from this group, send email to metakit-unsubscribe <at> googlegroups.com
For more options, visit this group at http://groups.google.com/group/metakit?hl=en
-~----------~----~----~----~------~----~------~--~---

fleeing sql

If I remember correctly, metakit is multi-reader single writer right?
since I need multi-writer, guess I'm best solving that with embedding
metakit in a server using internal access locks?  does metakit work in
a python multi threaded environment either as multi thread access or a
single thread fed by multiple request threads/

thanks for any feedback

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the "metakit" group.
To post to this group, send email to metakit <at> googlegroups.com
To unsubscribe from this group, send email to metakit-unsubscribe <at> googlegroups.com
For more options, visit this group at http://groups.google.com/group/metakit?hl=en
-~----------~----~----~----~------~----~------~--~---