reporting build problems

[David Relson]

G'day,

Many of you build bogofilter from source.  From the SourceForge statistics, 
many more people gt the .tgz file than get the rpm files.  Doesn't matter 
which way you get bogofilter, it's great that you do get it and use it.

Part of the bogofilter distribution is a set of regression tests and their 
expected outputs.  We developers use the tests to help ensure that we don't 
break the program as we add to it and revise it.  The tests have also shown 
themselves of value to folks building from source since they can verify 
that the success of the build.  The tests have also revealed a variety of 
portability issues (due to differing shells, compilers, cpu types, 
etc).  We didn't write the tests for portability testing, but they've been 
very valuable in that area and have revealed a number of problems.

After running "make" to build bogofilter, it is a good idea to run the 
regression tests.  "make check" is the command to use.  If you want less 
verbosity, use "make -s check".

Here's what the output of a successful run:

[relson at osage cvs]$ make -s check
Making check in dcdflib
Making check in src
Making check in .
Making check in tests
Making check in .
PASS: t.env
PASS: t.ignore_spam_header
PASS: t.integrity
PASS: t.integrity2
PASS: t.lexer
PASS: t.lock1
PASS: t.lock2
PASS: t.nullstatsprefix
PASS: t.score1
PASS: t.score2
===================
All 10 tests passed
===================
Making check in bogoutil
Running test 1
Running test 2
Running test 3
Found a message count of [4] in db. Throwing away text file count of [1]
Running test 4
This database appears to have been upgraded already.
But there's no harm in doing it again.
PASS: driver.sh
PASS: t.dump.load
==================
All 2 tests passed
==================
Making check in bogofilter
PASS: t.lexer.mbx
PASS: t.robx
PASS: t.split
PASS: t.systest
PASS: t.grftest
==================
All 5 tests passed
==================
Making check in doc
Making check in contrib


When running "make check" all test outputs are discarded, so they won't 
clutter up your hard drive.  If any test fails in your environment, we 
developers are very interested in seeing the output from that test.  You'll 
know when a test fails because you'll see "FAIL: t.test" (instead of "PASS: 
t.test")  and the summary line will indicate how many tests failed, for 
example "2 of 5 tests failed".

IF this happens to you, run each failing test from the command line and let 
us know that you have had a problem and have the saved test 
outputs.  Running tests from the command line and saving the outputs is 
very easy.  For each test that fails you should "cd" to its directory 
(which will be tests, tests/bogoutil, or tests/bogofilter) and run the 
test.  The output of test t.xxxx will be placed in a directory named 
xxxx.yyyymmdd.  Then report to bogofilter-dev that you've got the 
files.  We'll likely ask for a .tgz file of the test output directories.

Happy building and testing!

David