A long time ago I reported an issue with a Catalyst plugin.
The POD indicates it is possible to set a temporary directory for use with
Catalyst::Request::Upload in Catalyst::Request::Upload.
This does however not seem to work.
jonasbn
A few days ago I received a response:
Can you provide us with a failing test case?
The mail caught me at a very bad time and I was a very few keystrokes from responding:
Can you provide me with a succeeding test case?
But luckily I did not.
My problem is simply the response time from when I sent the request in the first place.
I created the ticket: Oct. 24th. 2006 and I received the response: May 26th. 2008
That is 1 1/2 year ago, the project I was working on is not longer relevant, I have the source code yes, but I cannot dig up the details of what I tried and what I did not try, back then I just read the POD and examined the code and I could not get it to work - so just for the record I created the RT.
I guess that due to time constraints I should just respond:
No
But it bothers me, that there might be a bug, which is not dealt with and perhaps this RT is just closed as irrelevant in order to close 5 bugs a day or something like that, I might however put too much into this since this might not be the case, but I have worked in a place where statistics got the boss high and the 5 a day trend no matter how nice it sounds does come with a pitfall of things being closed too rapidly.
I just wrote a journal entry on a problem with test reports on old distributions and in my opinion time is a large part of the problem.
If we have bugs and issues deal with them while they are a problem and not 1 1/2 after, I would not expect many people would be ready to give you feedback on their bug reports after such a long time, they might have changed jobs, language, name... whatever, they have moved on, time goes by...
So if we have stuff queuing up for years and years I guess 5 a day might be a way to flush things so we can move on. So perhaps this 5 a day is not such a bad thing after all if handled responsively.
Which brings me to my next problem - Tests.
I was asked whether I had a failing test case - well I would love to have one, but testing this might not be as easy as it sound since the configuration part in Catalyst is somewhat weird. Back then I simply looked at the code after several unsuccessful attempts and since I could not locate where the problem might be, I thought I would let the more knowledgeable author examine the issue.
So with all this testing and stuff it has all of a sudden become the users job to carry the Burden of Proof. Would should this not be the authors job after all, he/she might now the code
better and should have test cases covering the usage scenario or at least the code statements.
But who am I to complain, this is open source, I could choose to fix it myself, take the time to write the test case or just get of the horse and drink my milk.
