Thank you, Winsling for spinning off my thread, which was itself a spinoff of one of our perennial Who Is A Real Engineer threads.
As for titles, we don’t have SE’s, only Programmer Analysts I and II, Sr. PA’s I and II, and Principal Analysts. But we’re working currently with a software vendor for a particular project, and they seem to call all their developers engineers. Based on that, and reading the above, it seems like the title “Software Engineer” is spreading in popularity.
I agree with what’s been posted; only optimize what needs to be optimized. Anyone who’s been in the field for awhile has a story where they speculated on what the bottle neck was, only to be completely off base.
Javaman, I look forward to the day when every open thread can be traced back to one ancestral thread, and several dissertations will be written, “Constructing the Ur Thread.”
Hal
The Ur-Thread would be one of infinitesmal negative ignorance distributed over an infinite domain.
And I should add, knowing when bubblesort sucks is important. For nearly sorted input, it can be best.
Moderator’s Note: I think this particular thread is probably more suited to IMHO.
Yes, but it’s probably simpler to just use quicksort with a randomized partition choice than to try and figure out whether input is nearly sorted.
THese things widely vary. My first job we had a “release” every few weeks. We just added feature by feature, and kept giving company X (our one big customer) every new release. Dangerous, but it was necessary. At my current company we have a release every 6-12 months or so, with point releases every 2 months or so.
Due to layoffs, everyone is roughtly the same level - from experienced to “very experienced”. I don’t fall in the “very” category, but we are all Senior Software Engineers.
I die. I weep in my shame. Gallons upon gallons of tears flow out from me, leaving a dessicated corpse. Still the tears flow, covering the land in brine. No crops can grow, no animals can graze. All die. O the embarassment.
Except if you’re doing something like inserting into a sorted structure, which I find more common than sorting random data.
I can’t imagine figuring out at runtime if the data is close to sorted, and then switchng algorithms. Off the cuff the switching logic would be O(n) which blows away any savings. Does this really happen?
Hal
With the sorting algorithms known today? Not bloody likely. You can do mergesort and be done in O(n * log(n)), even if you don’t use an extra storage array. I had to code that for an algorithms class–fortunately it was a partner assignment, and my partner still deserves mad props.
Now, now. It’s OK. No need to beat yourself up over it.
Did I mention the Moved Thread Fee? Cash, checks, or major credit cards accepted.