Posted Oct 6, 2008 0:30 UTC (Mon) by njs (guest, #40338)
[Link]
Because for software that's just coming into prominence, 90% of user pain is usually caused by hitting stupid little snags that are trivial to fix but the developers hadn't bumped into themselves. So you easily end up in a situation where users think the software sucks and the developers think that it's great, and both positions are perfectly reasonable -- the bugs are real and cause real pain, but the developers start out unaware of them, and then see they don't reflect the true state of the software anyway.
That seems to be the case here, anyway. The complaints boil down to "I tried running this software in a weird untested configuration, and it didn't work, plus the error messages didn't tell me why" -- hardly a shocker, or proof that systemtap is fundamentally flawed. Just stuff that the systemtap people have to fix now, and examples of the kind of stuff they have to become more aware of as they transition into a being a real-world tool.