[Monetdb-developers] JDBC liberation: use ant instead of make

Stefan Manegold Stefan.Manegold at cwi.nl
Fri Dec 9 11:24:16 CET 2005


sounds reaonable to me, hence, there's a green light from me ;-)

Currently, we simply skip compiling (and hence testing) JDBC, in case there
is no java; hence, we should then check for ant only in case java was found,
and if there is java, but no ant, we simply skip JDBC (preferably, configure
should tell that it did not find java or ant ...)


On Fri, Dec 09, 2005 at 10:41:23AM +0100, Fabian Groffen wrote:
> Hi list,
> Tonight's testweb reveals that not a single JDBC test has run, due to
> incorrect packaging of the driver.  I have always disliked the was the
> driver was built and packaged, but I think this time again, it is proven
> that `make` is too ancient for Java, or at least too clumpsy to deal
> with it.
> This is not only my personal opinion.  That is, long ago other people
> already found out that make doesn't work that well with Java
> environments[1].  Although I understand the need to keep it 'simple'[2],
> I personally feel that fixing the current setup again to make it work is
> just another hack.  From the very start of JDBC development, I used
> `ant`[1] to build the driver.  Ant is, unlike make, a build tool geared
> towards Java projects, and the ant build scripts[3] that are currently
> in CVS have a much better understanding of the various files than make
> does.  Just to name a few:
> - ant has multiple build targets for JDBC, XML:DB, MCL, etc.
> - ant date stamps each build
> - ant generates documentation when required
> - ant understands Java
> - ant builds jar files much more flexible.
> I am strongly in favour of using ant in the future to build Java related
> files.  That is, let configure detect if ant is available or not, when
> Java is available.  If not, don't build, or complain.  I think most of
> our test platforms (especially since Sun and IRIX are RIP) have ant
> available (even Darwin has it!), as such availability of the build tools
> shouldn't be an issue.
> On a side note, testing Java on many platforms isn't that interesting,
> since the JVM itself has been tested on all platforms Sun supports.  For
> JDBC, currently only endianness, timezone and charset are important.
> Since Darwin is both different-endian as well as in another timezone and
> using another charset by default, it is the only machine that really
> is interesting from a testing perspective, next to a windows machine and
> a linux machine in our neigbourhood.
> [1] http://ant.apache.org/
> [2] that is, using tools known to those who deal with it
> [3] ant's build scripts are XML files called build.xml
> -------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
> for problems?  Stop!  Download the new AJAX search engine that makes
> searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
> http://ads.osdn.com/?ad_id=7637&alloc_id=16865&op=click
> _______________________________________________
> Monetdb-developers mailing list
> Monetdb-developers at lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/monetdb-developers

| Dr. Stefan Manegold | mailto:Stefan.Manegold at cwi.nl |
| CWI,  P.O.Box 94079 | http://www.cwi.nl/~manegold/  |
| 1090 GB Amsterdam   | Tel.: +31 (20) 592-4212       |
| The Netherlands     | Fax : +31 (20) 592-4312       |

More information about the developers-list mailing list