Time and Time Again
Gee, this bug looks awfully familiar.
Now, maybe if I had had a company paying me to get software working, I could’ve spent the time to produce such detailed test cases. But I didn’t. And Murray insisted the bug was our fault anyway. He did no such thing at any point to the VMWare people in the link above, however.
Temas and I are fairly certain that this bug was the bug we were encountering with Gabber with the newer libsigc++/gtkmm. So maybe now Gabber 2 will work with the latest libsigc++/gtkmm without crashing all the time. Too little, too late.
Edit: Our friendly library maintainer responds by calling me an idiot. Gotta love ad hominem arguments. See comments.
I remember reading an old blog post of yours on this issue, but didn’t associate it with our problems until now. If it helps any, this isn’t exactly a new problem we’ve dealt with.. We’ve been living with the frustration and pain and immense amount of workarounds needed to overcome this bug every since before I started work at VMware 8 months ago. We just finally got tired of it last week and started spending time figuring out what exactly was causing it and how to reproduce it using a simple test case. At least now any future work you do will have that fixed, once it’s in CVS?
Well, I don’t plan on ever doing anything Gnome-related again. Ever. So that won’t be much of an issue.
Idiot. I asked them to make a test case, just like I asked you. They did. I fixed the bug. I didn’t just debug their application for them like you wanted me to. And I didn’t admit to adding secret code to make the bug happen on purpose, as you insisted I should.
Temas and I never insisted that you debug our application for us, nor did we insist that you added secret code. You always assume that application developers want you to debug their application when all they’re doing is trying to figure out what’s going on with gtkmm.
When we explained what we were trying to do with Gabber, we were told that what we were doing was not an appropriate way to do things and never should have worked in the first place. We responded by pointing out that it worked quite well until we switched to newer gtkmm, thus leading to our belief that something in libsigc++ or gtkmm was broken.
Let’s review the logic again: 1. Gabber + old libsigc++ + old gtkmm = works fine. 2. Gabber + new libsigc++ + new gtkmm = crash. This is why we believed there was a libsigc++ or gtkmm bug and this is why we approached you asking what might be wrong.
You, however, continued to insist that there was nothing broken and that it was most likely a bug in Gabber. You refused to even look into it (even from the libsigc++/gtkmm side) without a complete test case.
Also, according to the post by chipx86 right here in these comments, the VMware people had been experiencing this bug for a long time before they finally gave up and made a test case at your insistence. The difference between them and me is that they were getting paid for the hundreds of man-hours they wasted trying to figure out what was going on. I wasn’t about to waste a ton of time to make a test case for someone who called my friend an “ungrateful fuck” and who repeatedly insisted that the problem was mine and that I was doing something bad to begin with.
I’m only an “idiot” to you because you have not once bothered to figure out what I was trying to say. You always make a crapload of assumptions and attribute the worst parts of them to me. If you had once, just once been at all understanding at any point in my interactions with you, I would have worked on a test case. I didn’t just not work on a test case because I’m an idiot—I didn’t work on a test case because of the way you treated us.
The way I figure it, contributing my time is like contributing my money. I’m not going to spend time working on software for a project whose library maintainers treat me like this. There’s no reason I should. I’m sure the great majority of Gnome hackers out there couldn’t give a damn about whether I work on software for Gnome or not, but it is my intention to make great software regardless. Gnome simply won’t benefit from my time anymore.
To Murray:
Think twice before you call somebody “idiot” again. Maybe you’re the idiot here because you refuse to admit that you’re wrong.
Even to people that do not understand what’s going on, your attitude is surely going to lose you quite a few friends.