Also, some tests of this type would have saved me some time.  First I was 
looking for a problem in my code, and it took a bit too long for me to question 
whether the ODBC package still worked.  It does, but just differently than when 
I originally tested it.  Tests would have not made that very obvious, both by 
verifying functionality and serving as coding examples.

Bill



________________________________
From: [email protected] 
[mailto:[email protected]] On Behalf Of Mariano 
Martinez Peck
Sent: Friday, September 04, 2009 3:13 AM
To: [email protected]
Subject: Re: [Pharo-project] ODBCWarning



2009/9/3 Benoit St-Jean <[email protected]<mailto:[email protected]>>
Cool!

If that's of any help, I volunteer to test your stuff on Windows XP (SP3) for 
both clients & RDBMS.  At the moment, I can test for MySQL, InterBase, SQL 
Server 2000.  But if you need it, I could take some time and set up a DB2 
server and an Oracle server for the same platform.



That's why I love the community :)

Cool!  Very very cool!

Question : I know it's usually not recommended to test stuff that requires 
external output (such as setting up user, password, DSN, etc) but I think we 
should design a very simple database, on all those RDBMS and use SUnit to 
automate testing.  What do you guys think ?


I do that. I have 3 or 4 tables and all of the tests use those tables. They are 
created and drop in every test so that not to have a side affect in other tests.



-----------------
Benoit St-Jean
Yahoo! Messenger: bstjean
Blog: lamneth.wordpress.com<http://lamneth.wordpress.com>
A standpoint is an intellectual horizon of radius zero.
(Albert Einstein)


________________________________
From: "Schwab,Wilhelm K" <[email protected]<mailto:[email protected]>>
To: 
"[email protected]<mailto:[email protected]>"
 
<[email protected]<mailto:[email protected]>>
Sent: Wednesday, September 2, 2009 8:56:59 PM

Subject: Re: [Pharo-project] ODBCWarning

Hello all,

I think I have the catalog (tables and field name/type given a table name) 
working.  The latest ODBC offering appears to work, but in fairness to me, it 
does not work the same way it did (some time ago) when I crossed database 
access off of my potential-show-stopper list.

Re the warnings about change of language, for some strange reason, I find it 
very easy to believe that this is a Microsoft idiosyncrasy and so I added a 
suitable #on:do: to capture the warning.  That could be bad though; better 
ideas are welcome.

The problem I had going from Linux to Windows will probably persist.  Of 
course, I would like to see the startup/shutdown system revised to make such 
things easier to get right, after which it would hopefully become short work to 
fix.

The ODBC package derives ODBCRow from IdentityDictionary and ODBCResultSet from 
Stream.  IMHO, it always seems cleaner to compose rather than inherit the 
limited functionality that the base class provides.  However, the inheritance 
is not really hurting anything here; socket addresses being inherited from byte 
arrays *is* hurting us.  I brushed up against that at one point, but was able 
to work around it.  When it gets between me and something I need, I will 
introduce an InternetAddress class and try to lazily resolve names and 
addresses, fix the printing, etc.

Bill




-----Original Message-----
From: 
[email protected]<mailto:[email protected]>
 
[mailto:[email protected]<mailto:[email protected]>]
 On Behalf Of Schwab,Wilhelm K
Sent: Tuesday, September 01, 2009 3:15 PM
To: 
[email protected]<mailto:[email protected]>
Subject: [Pharo-project] ODBCWarning

The porting continues...  It is getting closer to doing something.  It just 
raised an ODBCWarning about the server changing the language to us_english.  
Any ideas?  Either I do not care or should set it up front??

One thing really surprised me, and note that I might have brought this on 
myself (though I think that is unlikely) by messing with startup and shutdown: 
the ODBCLibrary was stubbornly clinging to the Linux library from my home 
machine.  I finally used

    ODBCLibrary default initialize.

which seemed to put an end to the separation anxiety.  Is there a better fix?  
I suppose I should try this in a clean image to see if there is really a 
problem therein, but that will be a little hard to do without a lot of my code, 
almost any of which either depends on changes I have made or could have caused 
the problem.

Maybe I should ask whether some of you use ODBC on Windows and Linux and have 
no problems moving between the machines.

Bill

_______________________________________________
Pharo-project mailing list
[email protected]<mailto:[email protected]>
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

_______________________________________________
Pharo-project mailing list
[email protected]<mailto:[email protected]>
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

________________________________
The new Internet Explorer(r) 8 - Faster, safer, easier. Optimized for Yahoo! 
Get it Now for Free!<http://downloads.yahoo.com/ca/internetexplorer/>

_______________________________________________
Pharo-project mailing list
[email protected]<mailto:[email protected]>
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

_______________________________________________
Pharo-project mailing list
[email protected]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

Reply via email to