On Wed, Dec 21, 2016 at 4:36 PM, Iiro Laiho <ilmaisin@xxxxxxxxxxxxxxxxx> wrote: >> You've just offended three people without knowing anything about them. > > No. I am just describing the current QA procedure. The +1 comments from three people are seen as some kind of proof of stability of the update, and those commenters don't go through any screening of any manner. "Works for me" really is a common expression in those comments. That we often don't know anything about those random people that happen to comment the update and still trust their word *is* a part of the problem. They are not random people, it is definitely a non-scientific sample of users, with FAS accounts, who contribute feedback that varies along a continuum between subjective and objective. There is no practical way to completely test all faucets of a package before its release into the wild. Saying people should not be trusted merely because you say so is an amusing strategy. > > You should understand that falsely accusing people of insults isn't in any way respectful either. Nor is accusing contributors of being "total amateurs" nor is accusing them of "yelling". Looks to me like you can dish it out the insults but don't like being called out on it. Your language is inexact, I suggest cleaning it up and aim for better precision if you think you're being misunderstood. > I am helping. Without me investigating and reporting the problem, the package in question would maybe still be broken. But I am just one person with many things to do and cannot bend myself to substitute a whole QA system. But I have to agree in that the current situation can sometimes except the maintainer to know everything about their package, and such exceptation is unrealistic. Well it seems to me if a packaging change means some files are going to be removed upon updating that package with the new (and changed) package, that it's reasonable something somewhere is going to break, and that this should cause the build to fail with a warning so at least it can involve a human to assess the likelihood for breakage. But I'm pretty sure that automation doesn't happen right now and the process depends on people simply hitting the problem and reporting it. It's more correct to say the automated testing of package has a hole in it, rather than blaming people for whom the package did in fact work and caused no regressions. That feedback wasn't wrong and it's inappropriate for you to suggest they were wrong to up karma the package, that they are therefore total amateurs, and that they were yelling about anything. -- Chris Murphy _______________________________________________ devel mailing list -- devel@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe send an email to devel-leave@xxxxxxxxxxxxxxxxxxxxxxx
