2 ms·
Excellent article. The part about "setting boundaries" is worth the "price" of reading it, all by itself. Documenting bugs so they can be fixed: now that is pr
by AlbertCory 5y ago
Excellent article. The part about "setting boundaries" is worth the "price" of reading it, all by itself.
Documenting bugs so they can be fixed: now that is probably no different from a commercial software setting, where I worked. I have mixed experiences with that, so some (probably) intensely controversial observations:
If the user just reports "it broke when I did X" and you do X and it doesn't break, is that the end of the story? What if lots of people report similar-sounding bugs?
Too many developers and PMs are happy to stamp it "not reproducible" and move on. I think the dev should take it as an opportunity to insert more diagnostics to catch it next time. If it's an online service, then you will have the logs available.
If it's client software, you have to ask the user to save them for you (but if they're so entitled that they refuse, you have every right to write them off). Some people will get into the adventure and try to help you out, and you can be glad you gave them a good experience!
Columbo should be the role model. Someone got murdered and I'm going to solve the crime, no matter how long it takes. Obviously you can't do this for every random bug that comes in, but if you never do it, you're in the wrong profession.
All the time in one job at Google, Drive and Docs would f&ck up (and we all used them all day). No one I worked with would ever report them, because "Why bother? They're just going to stamp it 'not reproducible' anyway."
I had one memorable exchange where I suggested upping the log level so they could diagnose it. The PM said that was impossible -- the logs would be too big. Maybe they could up the level just when I'm the user? Nah, too much trouble.