I discovered today that the Automatic Bug Reporter stopped working sometime
near toward the end of May, due to a change in the tracker setup at SF.
It looks like the tracker at SF now requires a cookie in order to submit to
it. I can’t estimate how much time it will take to get this working again.
Getting this working in the first place took a long time, due to having to
treat SF like a black box.
On the upside, we haven’t lost any of the bug reports, since the script which
receives them logs them all. I’d be happy to hear suggestions for how you guys
think we should proceed with this.
Despite that the new website will not be live for awhile, perhaps you can redirect bug reports to the bugzilla portion of site and host it temporarily on your server till the vassal site goes live - A good way to start testing out that portion at least?
I discovered today that the Automatic Bug Reporter stopped working sometime
near toward the end of May, due to a change in the tracker setup at SF.
It looks like the tracker at SF now requires a cookie in order to submit to
it. I can’t estimate how much time it will take to get this working again.
Getting this working in the first place took a long time, due to having to
treat SF like a black box.
On the upside, we haven’t lost any of the bug reports, since the script which
receives them logs them all. I’d be happy to hear suggestions for how you guys
think we should proceed with this.
If it really just wants a cookie to be set, the workaround is probably not too hard, but if SourceForge is requiring the client to be authenticated before submitting a bug then we’re pretty much forced into hosting our own bug tracker.
It’s not an authentication issue. It’s still possible to report bugs at
SF without logging in.
The problem is with where the cookies come from: All of the cookies are
being set by JavaScript which is called from the bug report page. I know
how to extract cookies sent in the headers of the original page using PHP.
I don’t know how to run the embedded JavaScript from PHP in order to get
the headers from the HTTP requests those scripts make.
After screwing with it for a whole day, I finally was able to convince SF
to accept bugs from the bug reporter again.
Also, I’m in the process of uploading the bugs which were logged but not
reported to SF during the period when the bug reporter wasn’t working.
There are 249 new bugs to sift through (not all of which are uploaded
yet). Help!
A lot of the collected bug reports were for modules—I’ve sent those over to the Support Requests tracker you’re tending. Could you take care of those when you get a chance? The primary thing is to let the V40K guys know that their most recent release was not built for Java 1.5, and as a result people using older Macs can’t use it. We’ve had dozens of bug reports due to this.
BTW, I see that you closed a lot of bugs. (Good.) Did you also email
the reporters letting them know that their problems are fixed in the
current version?
Bug [2816275] NPE or ArrayIndx in Embellishment adding layer with no image (swampwallaby-3.1@5801)
Bug [2813765] Can’t always Peek at a card, even though you can unmask it (swampwallaby-3.1@5802)
I think I’ve picked up all of your fixes now. Do you have more you want
to get into 3.1.8, or should I go ahead with putting up a 3.1.8 test
build? (I’d really like to stop the reports from that AreaOfEffect as
soon as we can…)
Just give me 20 minutes and I will finish annotating BasicLogger and the will be it. I’m just chugging through the bug reports and feeling a bit overwhelmed.
