Hacker News new | ask | show | jobs
by eterm 4665 days ago
The best way "in" as a newcomer is to have some knowledge that no one else has. In my case my first good answer (although it took months to get any rep at all) was an obscure bit of paypal sandbox knowledge. I actually came across the knowledge first, and since the problem I had hadn't been solved via StackOverflow I searched SO for the same problem I had been having and updated the old question with my new found solution.

That has since earned me 50 rep. This along with a few other bits took me to the 200 threshold which then gives 100 rep across all stackexchange sites. (Enough to upvote/comment.)

But the fact that "newbies" can't contribute what in all honestly is largely going to be misinformation or "me too" answers is also part of the genius, it gives real answers breathing room and keeps spam out.

My question asking has been less successful. I've asked 2 questions, one of which had no response whatsoever, and the other had an answer which answered the question in the community's mind but because I had been slack with my terminology didn't answer the issue I was really having.

4 comments

Rep on SO is unpredictable.

For example I randomly answered a question outside of my primary domain because it was a relatively easy issue that I had encountered myself. I have since earned close to 2,000 rep from that simple answer alone and get 20-50 points a week from it despite it being over 2 years old.

On the other hand answers that I spent way too much time researching have hardly attracted any points at all.

Same experience here. Quick, one or two line answers to simple, common questions seem to do the best.

Although, those questions that I've poured the most time into and really tried to answer thoroughly are my favorites. Especially if it's something I had to do a bunch of research and/or testing to confirm. The lack of feedback is supplanted by the quenching of the thirst to learn something new. In fact, this quenching is my favorite part of the SE sites.

Definitely. Most of my rep is from Prolog answers. I can spend an hour on a lengthy, detailed answer, and the best I can hope for is <4 votes and (maybe) an acceptance. Average return on an answer is probably 20 rep. OTOH, if I squeak a Haskell answer in, average is more like 50-100 rep because there's simply a lot more eyes on Haskell answers. At the same time, this means while there's a good hour or two window between a Prolog question being asked, it's more like 2 minutes for a Haskell question. The community is just really good like that. So there's a limiting factor problem there. Competition for MySQL and Postgres answers is also really high so you have to be really fast with your answering, but all my "greatest hits" were in Postgres.

Question-askers benefit the most from this competitiveness, but without a niche it would be really hard to break in. For a little while it seemed like people were starting to treat S.O. activity as a test of your virtue or capabilities as a programmer. Today it seems absurd to judge someone for not having an account. It's just too hard to get started.

Interesting, you talk about the competition, breaking in, monitoring unanswered questions so you can answer first, etc... - What's the motivation behind it?

I guess it is not just contributing to the common knowledge, but maybe to show off your profile to potential employers or similar?

For me, I just like being helpful. The rep is nice too. Stupid internet points are pretty motivational, but I also review edits, and you don't get any rep for that. I would be flattered if an employer had noticed me on there, but I don't think it's likely to happen. I've been active on SO for longer than I've been active here, but I have gotten many friendly recruitment emails from here and none from there.

Prolog is an extremely small niche. Most of the questions are pretty basic and from students. I don't like seeing students being told odd things about Prolog, with a dismissive air of "Prolog makes no sense, you just have to feed it this nonsense to make it go." So another part of my motivation is to keep this sad religion alive and inviting. Haskell has a much larger community of people working very hard to make it inviting.

The hiring angle probably works better for other niches. Every Java EE question is answered by BalusC. I imagine that pays dividends for his consulting: he's the most helpful guy in that area and I'd hire him in a second if I were in charge of that kind of thing.

My last two job offers came from people who had checked out my SO answers among other things.
Wow, I feel so disproven.
> it's more like 2 minutes for a Haskell question

My experience: it is still possible to come in an hour and beat the quick answers, although I agree with the other comments about the rep being somewhat unpredictable.

it's got harder for everyone. i'm over 10k and mine isn't going up quickly any more. there are fewer interesting questions (partly because many have been asked, but also because the sites have fragmented and the policing is more strict) and more competition answering.

i'm less sure, but i think answers are getting worse too. it feels like the smarter people have disappeared.

I don't get much rep for answers. I get it for asking the right questions. My rep still isn't fantastic (a few hundred), but it's good enough to give a bounty when I need it.
If you are looking for something, and find older questions, try creating an updated answer for something, or even re-visiting your old answers. A lot of time an answer can change over time, new frameworks/tools etc... about half of my points come from the graveyard.
Try giving more general, more in-depth answers. Such answers, if correct, get dug up by search and keep bringing upvotes months, or even years, after having been written.
I don't get much rep for answers because usually someone else has already given a better answer than I'm able to give. I'm just saying you don't actually have to know anything to get rep. You can get it for asking good questions too. (But first search to check if someone else has already asked the same thing.)