Re: hi, newbie coming

[Jörg Beyer]

Hi again. 

<OffTopic>
I just thought about ODBC yesterday, but had to hold back some comments on
this. I am on OS X, and while I'm able to build PostgreSQL without a hitch,
I had quite some troubles to get PostgreSQL's own ODBC driver working. Good
news is, I found a problem in the Makefile last night, and psqlodbc seems to
work now. 
</OffTopic>

So +1 for ODBC support first, instead of discussing which particular RDBMS
to choose. While generic PostgreSQL support would have quite some
advantages, I think it has to be taken into account that some people would
rather prefer MySQL (especially in Web environments), some MS-SQL Server, or
Oracle, .... 

But I think it could pay back if the functionalities to connect to a RDBMS
leave some room for future extensions, e.g. by adding generic support to
particular RDBMSes, too.

Back to ODBC. AFAIK, there are at least two options regarding the driver
manager (iODBC, unixODBC). The question which to support could be crucial.
As a nice side effect, ODBC should also solve the spreadsheet question, at
least OpenOffice and Excel (not sure about Gnumeric).

Cheers, 
Jörg



Am 17.10.2007 3:10 Uhr schrieb Laurent Valdes (<address@hidden>):

> Hi everybody,
> 
> PostgreSQL is a great DBMS,
> it appears that now good and free ODBC packages exist.
> I suggest it would be better to implement ODBC integration before any
> special kind of database.
> After that, integration with PostgreSQL could be made in a breathe.
> 
> Best regards,
> Laurent
> 
> 
> 
> 2007/10/16, Jörg Beyer <address@hidden>:
>> Hi all!
>> 
>> Interesting thread. I'm watching pspp since January or so, and this seems a
>> good time to chime in for the first time. I'd like to give some feedback
>> that I hope is of some use. Sorry that I can't offer more at the moment.
>> 
>> 
>> +1 for PostgreSQL-support.
>> (Or better, +10^10, PostgreSQL is a fantastic RDBMS IMHO :-)
>> It may be of interest to note that the forthcoming 8.3 release adds two new
>> generic data types, UUID and XML.
>> 
>> +1 for being able to read spreadsheets, especially OpenOffice ones, which
>> should give us an opportunity to read Excel docs, too. Sometimes you don't
>> have a choice.
>> 
>> 
>> In general, I see pspp not only as a clone of 'that other software', but as
>> a very welcome and integral part of a larger ecosystem for data analysis.
>> Thanks for all your efforts so far!
>> 
>> Jörg
>> 
>> 
>> 
>> On Tue, 16 Oct 2007 17:30:39 +0800, John Darrington wrote:
>>> 
>>> On Mon, Oct 15, 2007 at 03:01:04PM +0200, Laurent Valdes wrote:
>>>> 
>>>> I think I will focus on database integration,
>>> 
>>> Database integration would be nice.   It should be fairly
>>> straightforward to make a casereader that reads from a postgres
>>> database, using the libpq library :
>>>  http://www.postgresql.org/docs/8.3/static/libpq-connect.html
>>> 
>>> Having done that a writing a GET TRANSLATE command is a matter of
>>> rountine.
>>> 
>>> It'd also be nice to have an option which reads/writes gnumeric and/or
>>> openoffice.org spreadsheets too.
>>> 
>>> J'
>> 

============================================================

Jörg Beyer
PHILIPPS-University Marburg
Dept. of Psychology