- Apr 03, 2006 by the_angry_angel
..has changed somewhat, since 1.3.x days, which is when I last did anything along these lines. Since I only run Apache 2.2, there seems to be a bit of a learning curve to this whole thing - which can be tricky with the little documentation out there.
I have managed to come across apachetutor.org, which has given me a lot of think about. However, other than that the "official" docs are rather lacking and its taken me sometime just to find out that buckets and brigades are a little more complex than I first thought.
- Apr 01, 2006 by the_angry_angel
According to Slashdot:
Our marketing department has done extensive research over the last 3 quarters and discovered that our audience is strangely disproportionately skewed towards males. Like, 98.3% males to be precise. To correct this oversight, we have decided to subtly tweak Slashdot's design and content to widen our appeal to these less active demographics. Don't worry! We'll still continue to serve our core audience, but we hope you'll work with us as we try to find a balance that will work for all.
I can only hope its an april fool, but I'm starting to be comforted by that bright pink glow. Its much more relaxing than that green glow - although possibly going to cause more eye strain.
- Mar 26, 2006 by the_angry_angel
...basically the key to avoid getting "Can't set position pointer in file '/home/public/svn/testproject/db/revs/1': Invalid argument", is to remove the APR and APR-Util directories from the subversion source directory and then ./configure --with-apr=/path/to/apache/bin/apr-config --with-apr-util=/path/to/apache/bin/apu-config --with-apxs=/path/to/apache/bin/apxs
A big shout out goes sussman on #svn (irc.freenode.net) for pointing out that Apache 2.2 and SVN 1.3 use different APR versions.
- Mar 25, 2006 by the_angry_angel
Seriously, you can't make this stuff up.
- Mar 22, 2006 by the_angry_angel
In general I try to stay away from most Sage software as much as possible, but today we were handed a call with regards to ACT throwing an invalid pointer. We've been told its an error caused by a firewall issue and that we're to check it.
Needless to say, this has left me fuming. An invalid pointer reference occurs when a pointer's value is dereferenced, and whilst changing internal firewall rules MAY sort this, I don't feel its:
- Appropriate to "fix" problems in this manner,
- Be accused of essentially being party to causing the problem.
To be technical, one way to create this error is the following example; Take two char pointers (i.e char *p, *q; for this example). Say you do something with p, and q is left uninitialized. When the following is executed: p = q; pointer p will then become uninitialized as well, and any references made to p causes an invalid pointer reference. Hence this is called dereferencing pointers.
Either this error is giving a bad error message (i.e. its telling the user the wrong thing), or the program is poorly written and not checking for invalid pointers on reassignment. Either way, throwing this error message at a user and crashing the program is fucking terrible and shows the poor thinking behind the scenes, especially for a "corporate ready product". I'm fairly sure it's the right error message, and that doing a firewall change, somehow, bizarrely fixes it.
To further piss us off, said ACT! employee has linked to a KB article which requires a login (which we have no access to), and in the view of the customer called us basically fucking idiots because we don't have access to read it. Nice one.
