Does SuSE-8.0 come with the ODBC libraries for MySQL? There should be an RPM package called MyODBC which contains the libraries to link unixODBC to MySQL, but I can't find it. Specifically I want the libmyodbc.so file, normally found in /usr/lib on Redhat. -- There are 10 types of people: those who understand binary, and those who don't
On Tuesday 18 June 2002 10:43 am, Derek Fountain wrote:
Does SuSE-8.0 come with the ODBC libraries for MySQL? There should be an RPM package called MyODBC which contains the libraries to link unixODBC to MySQL, but I can't find it.
Presumably you wish to explore OpenOffice's ability to connect to a MySQL db, a la Access. If you are not already using the excellent and high-quality howto from OpenOffice, written by John McCreesh , can I suggest that you download that first (2.1Mb PDF) from www.unixodbc.org/doc/. Richard had a very cunning way of getting at MyODBC, but unless there is some way to pull it out of the package, it would probably be faster to download it direct from the MySQL site (it's only 226Kb) - the RH rpm v2.50.39-1 will do. This installs into /usr/local/lib, whereas the rest of the libraries are in /usr/lib, so run: sux (+password) cd /usr/lib ln -s /usr/local/lib/libmyodbc-2.50.38.so libmyodbc.so to create a symlink (you could also move the libraries, I suppose, at the risk of confusing your package manager). I'm not sure why there is a discrepancy in the numbering. Next, install the package unixODBC-gui-qt, which is on the disks (unixODBC itself and libiODBC are already installed on a default Office setup of SuSE 8.0). Then set up the drivers - run: cd /usr/bin ./ODBCConfig This will open a GUI (the OO howto p6 tells you how to do this using the /etc/odbcinst.ini text config file, but if we're emulating Access, it's best to get used to point-and-click!). On the Drivers tab, click Add. Put in a name (eg mysql) and description (eg MySQL drivers). For the driver file, browse to /usr/local/lib/myodbc.so, and for the setup file, browse to /usr/local/lib/libodbcmyS.so - note the helpfully confusing names. Leave the other fields at their default settings. Click on the tickmark. On the System DSN tab, click Add, then click on the MySQL driver, and click OK. Enter a name for the data source (essential), set the server to "localhost" (assuming you have MySQL on the same box) using the drop-down list, and select a database. Click on the tickmark, and then on OK. Your data source should now be listed. Again, p6 of the howto shows how to do this using the /etc/odbc.ini text config file. At the bottom of that page, it also suggests running: isql <name of datasource> to test the connection. I found that this did not work if the name of the datasource was different from the name of the database, ie the datasource name "MySQL-test" is used in the howto to refer to the MySQL database "test", but this gave me an error-message: Unknown database 'MySQL-test' There is also a misprint in the howto - "select version();" should NOT have a final semicolon. Thereafter, just follow the howto. All in all, the OO/SOfunctionality is very impressive: although still a little clunky compared to Access, it's very useable (though a bit slow), and it's pretty amazing given that ODBC is more widely used in Windows than in Unix. Desktop database access was the one thing that was still lacking on Linux desktop software (the closest competitor, theKompany's Rekall, I have not yet managed to get running), and this has the big benefit that you can encourage people to use a network server for their databases, which can then be backed up centrally. Moreover, the data in them is probably a good deal easier to get at than it is in .mdb files. There are some shortcomings: printing address labels from a database is not exactly intuitive, and the incremental auto-filter can be a bit confusing. My early testing also spawned lots of mysql processes (99 at one point), but I think that was because of various configuration options I was trying out - keep an eye on it before deployment, though. HTH Kevin Donnelly
Op dinsdag 18 juni 2002 21:31, schreef Kevin Donnelly:
Richard had a very cunning way of getting at MyODBC, but unless there is
Thank you! This list does not only teach how to use linux, but also new words like cunning.
but unless there is some way to pull it out of the package, it would probably be faster to download it direct from the MySQL site (it's only 226Kb) - the RH rpm v2.50.39-1 will do.
Alright she is quite big 3.5MB. Version vshop-3.0-189. Installing the package will indeed keep your system package wise consistent. -- Richard Bos Without a home the journey is endless
On Tuesday 18 June 2002 8:31 pm, you wrote:
Presumably you wish to explore OpenOffice's ability to connect to a MySQL db, a la Access. If you are not already using the excellent and high-quality howto from OpenOffice, written by John McCreesh , can I suggest that you download that first (2.1Mb PDF) from www.unixodbc.org/doc/.
Yup, that's what I want, and that's where I started.
Richard had a very cunning way of getting at MyODBC, but unless there is some way to pull it out of the package, it would probably be faster to download it direct from the MySQL site (it's only 226Kb) - the RH rpm v2.50.39-1 will do. This installs into /usr/local/lib, whereas the rest of
Yup, that's what I did. The RH RPM went in without fuss.
Next, install the package unixODBC-gui-qt, which is on the disks (unixODBC
I never found that package on my SuSE-8.0 disks. I did my config using the text files. Now you've named it specifically, I've installed it, though it's a bit late now. :)
Thereafter, just follow the howto. All in all, the OO/SOfunctionality is very impressive
I'm not sure what I was expecting since I've never seen Access. The last database front end I worked with was dBase, years ago. I got the OO/unixODBC working when I'd figured out the OBDC issues, but the OO interface doesn't seem to work properly. I run a simple query to list a single field from all the records of a table. The list shows records "1 of 38", even though there's 285 rows in the table. Paging down seems to pull in a few more records each time. Is this normal? Also, filters. I click the default filter and set up "Procedure = 'Accept'", then click OK. The dialog box disappears, but nothing else happens. I'd expect to be left looking at a single record, but they're all still there. I've been ignored. I pull up the default filter dialog again and what I entered has disappeared - it's all back to the default values. Basically, it doesn't seem to work. Am I doing something wrong? -- There are 10 types of people: those who understand binary, and those who don't
participants (3)
-
Derek Fountain
-
Kevin Donnelly
-
Richard Bos