At my old job, we used the "15 minute rule" which worked pretty well for us. When you are coming to the realization that you're stuck, and don't know how to do something, you give it 15 more minutes to figure it out. If you haven't made progress in that 15 minutes, you must go get help.
It works both ways - for people that tend to ask for help too soon, without actually trying to figure it out for themselves, and for people that tend to try to figure it out themselves for too long, when asking another person may get it figured out quickly.
> I don't know how to come up with task estimates that have any relationship with reality. I've said "I don't know how to give software timeline estimates", but often get pushed to give a number anyway.
This is often difficult in that it seems like what we do as software developers is always novel and new. But, the techniques for estimation are the same, whether the project is months long, or hours long.
Break the task down into smaller and smaller bits until you hit bits that you CAN estimate the time of - then add them up.. and then probably double it...
In the beginning, you won't be able to do this off the cuff in a meeting, but the correct answer should be, "I don't know right now, but I'll have that estimate for you by Xpm today"