I've been a "passive user" of Stack Overflow and other Stack Exchange sites for years. I have derived enormous benefit from it (many thanks!!), and I finally decided to become more active. It seems difficult for a "new" user to get started.
A relatively short time ago, I finally created an account to start answering and editing and posting and commenting! I was full of excitement and vigor and immediately tried to upvote (nope!) and post a comment (nope!). I need (threshold) amount of rep to do comment on this or that, or even upvote certain things... which is totally reasonable (perhaps "necessary" is a better term).
So I browsed a bit (such as whats-reputation). Advice to new users seems to be: just ask, answer, and suggest edits! But there are so many questions and good answers, a truly good question and new question seems hard to create. To truly give justice to all previous questions on a topic requires as much effort (or more?) as posing a question. And there are so many users that to troll-and-pounce the new-questions board could be a full-time job. And BTW, you can only put 2 links in a question when you have <10 rep, so it's very difficult to show due-diligence and to pose a truly good question to begin with!
I didn't find any actual "question" on this topic of how to get started -- but found a few gems like six simple rules, walking a (presumably-intentional) delicate balance between productive debate and provocative cynicism.
So I decided to post a question on this topic! Meta.SO seemed like the right place. NOPE! I needed 5 rep to even post a question. Probably also for good reason.
Now that I have >5 rep (w00t), here I am. After all that background (sorry) --
How does anyone get started around here these days?
My understanding now boils down to the following:
You have no choice but to start slowly.
Be patient and try to do contribute where you can.
Be prepared to accept initial rejection and failure.
Learn how to edit and make stuff pretty.
What am I missing? Do I "get it"? Have I completely missed the point? How can The System encourage new users who are here for the "right reasons" to quickly start contributing meaningfully and harness their energy for the Common Good?
It seems like you get it. Long gone are the days of camping on the front page to gain fast reputation by quickly answering softball questions. The questions are coming in too quickly, site standards have changed drastically, and there's a lot more competition to either answer or close easy questions.
The one piece of advice that I'll give you that you haven't mentioned is to pick some favorite tags that you're an expert in and add them to your favorites list (in the main page right sidebar).
This will highlight questions with those tags when you view the list of Newest questions, and it will even filter the list of selected questions when you view the Stack Overflow home page so that you see more questions with your favorite tags. By focusing your attention on your favorite tags, you'll see more questions that you're interested in and may be able to answer. You'll also be better able to suggest good edits to questions in your area of expertise.
You can also block tags for languages that you don't know by adding them to your Ignored tags list. By default, questions with Ignored tags will just be greyed out, but you can hide them entirely from the Preferences tab in your profile.
If you need a little bit of inspiration, here are a few users that have gained a lot of reputation in a relatively short amount of time, despite not having joined the site at the very beginning:
akrun - Member for 2 years, 6 months with over 220,000 reputation
Wiktor Stribiżew - Member for 2 years, 5 months with over 150,000 reputation
Jean-François Fabre - Member for only 6 months, but already has over 24,000 reputation
What do they all have in common? They answer tons of questions!
There's an alternate route to obtaining basic privileges, if you're finding the competition here too intense.
Utilize the association bonus
If you hit 200 rep on any site, you'll automatically receive a +100 association bonus on all sites. In my experience earning reputation on the beta sites is extremely easy due to reduced competition. Find a topic you have some expertise on, become a valuable contributor there, and you'll quickly earn your basic privileges. Even better, now you're helping out two sites!
Think of it as someone else vouching for your trustworthiness, so don't let them down by coming back here and making a mess.
Take the tour, earn a badge.
If you are reading this, you are probably the sort of person who has the ability to succeed on Stack Overflow. Even so, the tour provides the big picture of how the site ought to work. It also gives you a badge. Another easy badge is Autobiographer, which has the advantage others can learn who you are as a person.
Consider editing.
The next easiest badge to earn is Editor. Anyone may submit a suggested edit for community review. If you find a mistake or outdated information on any post and you know how to fix it, click on the edit link and suggest a change. Editing is a good way to learn what the community expects from posts and also will familiarize you with how posts are formatted with Markdown. In addition, successfully suggesting edits earns a small amount of reputation.
Answering is often easier than asking.
It's almost certainly gotten exponentially harder to ask questions than when many of us earned our (now slightly dusty) beta badges. This chart tells the story:
year questions avg_score deleted_rate closed_rate dupes dupe_rate
---- --------- --------- ------------ ----------- ------ ---------
2008 70372 18.40 6.4 3.9 1145 1.63
2009 394567 6.19 4.5 3.6 4800 1.22
2010 820161 3.43 6.3 3.4 10162 1.24
2011 1445142 2.18 7.9 5.7 21103 1.46
2012 2065664 1.28 10.2 7.9 34471 1.67
2013 2759442 0.61 14.7 10.9 52002 1.88
2014 3040440 0.17 17.9 10.4 68500 2.25
2015 2061746 0.08 17.2 8.7 52759 2.56
New questions are more likely to be closed or deleted than in the past. It's gotten harder to ask questions that haven't already been asked. In the best of times, asking interesting questions is harder than answering them. So I'd recommend looking for questions you can try answering before starting to ask.
If you have a different way of looking at a question, it really doesn't hurt to add another answer even if there's an accepted answer. The goal isn't to just help the one person who asked the question, but to help anyone with that same general problem who might find the question via search. There's no guarantee that your answer will be upvoted, but as long as your answer is accurate, clear, and noticeably different than others, it's not likely to be downvoted.
Consider learning a new language.
There's a good chance that your question in C, C++, C#, Objective-C, Java, JavaScript, JavaFX, or JSF has already been asked. Less popular languages have less duplication and fewer grouchy grognards who have seen the same few questions asked over and over again. Newer languages tend to not reach that level of saturation, so it might be worthwhile to learn a new language for the purposes of getting started on Stack Overflow. Plus it's a great way to teach yourself programming in 10 years.
Debug before asking.
Sometimes, you just need some help solving a problem in your code at which point a question on Stack Overflow would be a good entry point. Don't make the mistake of posting your code verbatim. Instead, search for the handful of lines that seem to be buggy. Surprisingly, doing just that is often enough to discover the problem. If your goal is to participate on Stack Overflow, don't be afraid to ask and answer your own question. Be sure to check for duplicate questions before posting (in which case, consider posting your own answer), but don't feel as if your question is wasteful if you already know the answer. Remember that helping the initial asker is not the primary goal of Stack Overflow.
Learn from setbacks.
You will almost certainly be downvoted at some point using the site. You might get critical comments, have a question closed, or a post deleted. In those situations, it's important to know that:
it's not personal,
there's almost always something you could have done better, and
recovery won't be hard if you take a few minutes to understand what happened.
Far and away the biggest mistake people make when using the site is ignoring advice they don't immediately understand. When people continue to post without learning what those signals are trying to say, they start running into suspensions, blocks and rate limits.
Get help and get meta.
Most of the common problems people run into are answered in the help center. In particular, read how to ask and how to answer. If those don't help, look around on meta for other people who have had the same problem. If that still doesn't help go ahead and ask about your specific situation here on meta. Be aware that meta has very similar conventions to the main site. Much of the above advice applies here too.
Try answering at a time of day or day of week when there are fewer users on Stack Overflow and presumably less competition for answering questions. Yeah, I get that Stack Overflow is an international site and people are on it at all hours of the day, but there are times of the day with significantly less traffic as seen here:
Please note these times are GMT.
See this post for more details.
It appears that Stack Overflow's heaviest users are North America as seen here so the lightest times are when North Americans are sleeping.
I would imagine there are days of the week that are also that are lighter eg Friday
Perhaps it is just a coincidence, but I found this out the hard way. I was burning the midnight oil so to speak and posted a question at the lowest activity time and received no answers.
I started at the end of last year and it was easy enough to rack up a reputation score. I'm a Java expert so I just started browsing the latest Java questions and when I saw a question that looked interesting I posted an answer for it.
A lot of those questions only need a few lines or a paragraph to answer them. My first ever answer wasn't much over 3 lines but I was lucky and got 6 upvotes. My next few answers got 0 or 1 votes but I persisted and over time got better at answering - and as a result the number of votes I started getting for answers started going up.
Don't expect to get upvotes or accepts on all of your answers but it only takes a few upvotes to start removing the new user restrictions. One thing that does help a lot is replying both fast and accurately. With multiple correct answers generally the first one posted will get the upvotes.
It's actually much easier to get reputation on answers than it is on questions. There are always questions in your favorite topic waiting for you to answer them.
There are no limits on how many questions you can answer - so find a way to isolate the questions in an area where you have expertise and then focus on answering the new questions that do not have good answers yet or questions where the existing answers are incomplete. Duplicating existing answers will get you nowhere although sometimes people do post the same answer simultaneously but that cannot be avoided.
Bill the Lizard and Cupcake provide excellent answers. I would just add a few things.
Learn how to identify motivated question askers. If the asker has been responding to comments, he still needs an answer. If he hasn't, he's more likely to have abandoned the question, so you won't get rep for answer acceptance.
Consider looking at bounties, especially on tags where you actually are an expert. Anyone who is spending their own rep to get a question answered is likely pretty motivated to get that answer, and will probably be back to select a correct answer - and also to answer requests for clarification, which can help a lot in producing an accepted answer. It can be worth spending quite a bit of effort to answer these questions; for example, on my most successful bounty attempt, I learned parts of an unfamiliar library for a platform I don't write for, but I got 525 rep for it.
As you allude to, editing questions is a way to grind past the early newbie levels. In particular, lots of newbie questions have unformatted or poorly formatted code; edits that format the code properly are usually accepted and as a bonus are very helpful to anyone who subsequently reads the question.
There's something that isn't really touched on in the other answers, at least that I saw. I read them all but if this is covered in another post, well, oops.
The other answers seem to be focused on how to gain reputation and what not. And if that's your goal, cool, those are great answers.
But if your goal is to really contribute to the community, do your job (or hobby or whatever it is when you're programming) and when you run into something difficult, post a question. I know answers are much better than questions for rep, but for really learning Stack Exchange, you have to be personally invested.
If you go answer someone's question, you might have some attachment to it. Might. But when you ask your own question, it really brings it home. The question that really brought it home for me was this one. I had a problem at work I was given because I was a Java guy and that must mean I knew SQL. But I didn't. But I wasn't one to shy away from the task. So I sought to really understand the problem and thought to myself "I can describe this in English so easy... and this has to be a common problem... but I can't find the solution anywhere with the terms I'm using..."
And then this guy came along and completely saved the day. My manager was super happy, and when I told him I just made a Stack Overflow post, he about had to change his pants. He couldn't believe that someone out there just looked at my question and gave me the code snippet I needed to get my job done right. And so quick - it was about half an hour between asking and having it answered.
And ever since then, it's been personal. It's been "there's people out there with problems, and I want to help them like I was helped." I want to help get people out of a bind (when I can, I find I have stretches where there isn't much time). And if you want to feel driven to help people, you have to know what it's like to be helped. And that means you have to ask questions.
The only useful tip I can add, that I've found to be extremely convenient, in addition to following your favourite tags, is to make yourself a custom Stack Overflow bookmark; it really helps to weed out everything except whatever it is that you are interested in. Mine, for example:
Clicky
https://stackoverflow.com/questions/tagged/delphi+or+c%23+or+plc+or+.net+or+labview+or+assembly+or+x87+or+vb.net?sort=newest
This gives a landing page with posts curated to seven of my key tags and sorted with newest posts on top. Obviously you can customize as desired. It saves a lot of clicking around, and it lets you always drop in on new and active posts. In addition, I find that I'm always learning something new along the way, because nearly every question that shows up is automatically relevant. Also, regardless of how frequently people are posting in a given tag, newer posts almost always need answers more than older posts. This isn't to encourage bottom feeding, certainly, but all other things being equal... you still have to be mindful of what you shoot for.
Read through, pick things that are interesting to you, and just try to answer them. Even if you don't know the answer or even if there is already an answer, just do it anyway - pretend it's a test and you have to come up with a solution; like a personal challenge. Sooner or later you'll be the one coming up with the answer first, best, or both. It's excellent training for general problem solving skills even if you don't get the repuation points most of the time.
It's been seven hours and fifteen days*
It took me nearly three active months to get 2k rep and this thread helped me quite a lot, so I decided to contribute some findings.
I can do whatever I want, I can see whomever I choose*
Choose your favorite tags: Choose a topic you really know something about and have fun thinking and learning about. Be prepared to do some research in order to answer a question. You'll learn a lot. (And earn some rep along the way)
I go out every night and sleep all day*
Take advantage of timezones: You'll observe that most answers happen during certain hours a day, in my Tag when Europe or the States are working. At other times of the day or during weekends there is much more time to prepare a "fast" answer and less competition.
I could put my arms around every boy I see*
Be clear about your knowledge: Don't try to answer each question which sounds somewhat familiar. Only answer when you are sure you are right and you can contribute something useful. There are many smart people around here, you will get bad comments and downvotes when you say something 'silly'.
Tell me baby where did I go wrong*
Accept critique: It is unavoidable to do silly things in the beginning: bad answers, silly comments. You'll get downvotes and bad comments. Try to understand what they try to tell you and improve.
I went to the doctor and guess what he told me*
Imitate: Quite soon you'll discover that people from the same small gang tend to be faster, have better answers and get a lot of upvotes for the same questions you are working with. Try to find out what they are doing and try to do the same. In my tag it is #Jon Skeet: He is always well informed, gives very understandable answers with nice code examples, which are explained in layman's terms if necessary. Just study what he is doing and try to do the same.
I know that living with u baby was sometimes hard, but I'm willing to give it another try*
Don't give up: The very first active steps on SE are hard. In fact the first steps are the hardest. After your first upvoted answer things start to be fun and it becomes easier with every answer you write, later every comment, every edit. Try to survive the first few active days.
* Lyrics by Prince for Sinead O'Conner: Nothing Compares 2 U
Easy steps for getting started at SO.
Read the rules.
Learn the formatting.
Ask questions.
Understand that not even SO is immune from bullying, ignore the ones who are impolite, they are a very vocal minority, but JUST a minority
Enjoy the site
Contribute
Don't be a taker
Build your own rep, but remember, there are people who will vote you down no matter how good your question or answer is. Don't take it to heart. Keep trying until you get the feel for this place.
Use your up-vote power generously when you get it.
When someone gives you a good answer, choose it as the best answer. They get a reputation bump and it is the best way to say "thanks"
Don't take criticism too hard, to those of us who have been doing this for a while, it looks easy, we forget that it's not to a new
programmer, or to one who has switched disciplines.
Start slow, watch, and read. There are some helpful people in here, and there are some who are not, just like everywhere else.
The people here really care about the site. They may seem harsh at times, but it is out of a sincere concern for the site and for the people here.
Just like everywhere else, there are people here who are not helpful, while they are the most vocal, that does not make them the
most numerous.
Be patient. This place gets flooded with bad questions and by people who just want to take what they can get without contributing anything to the site itself. Because of that, they have created a "tiered system" to screen out people who are not serious.
You will not be cut any slack. This is a professional site, you will be treated as a professional.
You will encounter the occasional jerk, if this happens, flag for the staff to deal with, don't get into the mud.
As was already mentioned above, answering is the best way to gain reputation.
Here are some "pro" tips of how to maximize your reputation points per unit of time spent on Stack Overflow, based on personal experience (observing and answering):
Try not to answer questions that you do not understand. Chance is you will not get it correctly, and/or it will take unreasonable amount of time to argue with OP about "what they really wanted". If you feel like the question is unclear, consider leaving a comment, and skipping to the next question. If your comment is later answered, and the question becomes clear enough, and it's still unanswered - now it's your time to give an answer.
Try to only answer questions if you immediately know the answer, or can figure it out in 2-3 minutes by doing a simple Google search + maybe 1-2 minutes sandboxing in your development environment. This way even if you don't get any reputation points (for example, someone did it faster), you've only wasted 5 minutes of your time. It's very unrewarding to spend even 0.5 hour on someone's question only to find out they already accepted an answer, and never bothered to check other answers. This is relevant to the next point.
The faster you answer, the more reputation you can get. This is because other people visiting the question may upvote your answer. You posted late, they've already been to this question and are definitely not coming back just to upvote your answer. There is a caveat - you answer incorrectly - you may get a lot of downvotes. So your initial answer must be fast, precise and actually address the issue in full. You may later edit it and add links to documentation, relevant articles, other Stack Overflow answers, etc. to make it nicer. Don't even try to write a perfect answer from the first attempt. There is a high chance some other user will provide a "fast" answer, which will get upvotes, get accepted, and then you finally post your answer, to find out nobody's there to read it.
When answering a 1-2 hours old question, be prepared to waste your time. If a question was not answered immediately (within 10-15 minutes), and especially if it has no upvotes, or worse - a negative score, there is a high chance (I'd say 90%), you will not gain any reputation here (or get an accept 5 days after and that's it). Unless it takes you 5 minutes to answer (generally when it's a complicated subject, but you are an expert in this area), it's best to move on.
I stumbled on this Q&A and was surprised (or honored :)) to find me quoted in the accepted answer.
I feel I can share my experience in a detailed answer that I had written earlier but seemed to be off-topic for the question so I deleted it.
I think it will be more on topic here, and won't hurt people into thinking I'm providing techniques to accumulate rep unfairly. This still requires a lot of work on the site, and it's certainly not designed to game the system (I deleted some upvoted answers because they were wrong, so no, reputation is not the ultimate goal, it's just a consequence of being helpful)
A few hints to get started & get some reputation/badges on SO. Those are "techniques" I used, but I feel that those aren't gaming the system and are fair.
On the new questions:
You need to be ahead. Being to be one of the first to read the new questions is a real must have (to answer newer questions on popular tags like python, java, C++, C). That means you have to spend a lot of time on the site, or frequently check new questions all along your day.
To be ahead, tune your filters to avoid seeing all questions. You won't be able to follow, and you cannot know all the languages/technologies.
Don't lose time answering crap questions. A question with a score of -4 is very likely to be closed / ignored. You'll waste your time, and won't even get an acceptance from OP who doesn't have a clue (you might get 1 upvote, maybe or some downvotes). And in the meanwhile, you're missing better questions.
For some questions, you have to be a FGITW (be the fastest to answer), but your answer must be spot on. So stay sharp and drink coffee (with a straw so you can keep on typing)
For some questions, it's better to comment, ask clarifications, leave other FGITWs answer (and do it wrong because they actually didn't read the comments). While all bad answers are being posted, hone yours, make it better/more detailed/more performant than the others and post it afterwards. The combination of "a lot of comments" then "an answer" is appreciated by followers, because you took your time before answering properly.
Don't answer obvious duplicates. Instead, vote to close / hammer them if you can. You'll be punished by some (specially if you have a high reputation) by answering. You should know better. Instead, you can answer the "original" question if you feel something is missing. I did that once, and my answer now has a +10 score.
On the old questions:
There's a "new answers to old question" review queue. I think that's where I got my first +1, because I added a above average by answering to an old question and I was a newbie so someone wanted to encourage me.
Of course if you're a specialist of some obscure/less popular tags (like Ada) you'll get upvotes on older answers by followers of those tags / people who have the "active" setting in SO page to see not only new questions, but active ones (which is impossible to follow on the popular tags BTW)
On any question:
Once you posted, edit your answer to add details. If it's already good, you can get upvotes, but enhancing it makes it "active" again, and if it's better you may get more upvotes.
Answer the comments made on your answers. Some commenters upvote if you answer them (better: edit your answer to take their questions into account if worth it). Plus it means that you care.
If you feel it's wrong, delete it, edit it, undelete it. You'll save a stray downvote.
Upvote concurrent answers if they're good (you'll even get a "sportsmanship" silver badge for that eventually). It creates a gap between your score and the other(s) answers, which isn't necessarily bad. Some may even think that yours have not enough votes // the others and that could even play your way (don't do that just to achieve that result, though)
If the question is bad, but you still want to help, you can comment on what's wrong. Doesn't hurt, and you'll get known as a nice fellow.
Don't answer like you would comment. If you don't feel like answering, then don't, and just comment.
Also upvote the good questions. That'll make them visible, only if it's worth, not to indirectly promote your answer. A lot of people forget to do that. Good questions need love too.
A bonus: by keeping a spotless behaviour (asking for precisions in comments, be reactive to comments, helping some users with typo questions by commenting on the error "for free", not answering turds, not answering obvious dupes, closing as duplicates with a small personal note to the OP, creating excellent answers,being nice most of the time :)) you may get unrelated upvotes: people wanting to upvote you twice (not recommended, but not serial voting yet), people visiting your profile and finding other good stuff you wrote in the same style and upvoting it)
Asking (good) questions & answering on meta also proves that you care for the site, not only for the rep. That can have strange effects (I frequently get downvotes on my questions after posting on meta, but upvotes on some answers at the same time!!), but globally it has a positive effect on your "reputation" (the one you don't measure with points). Can't hurt.
Gaining a couple of reputation points is not all that hard if you know enough about a certain topic. Just filter for it and start helping people out. A lot of times there is plenty of stuff to add, even if a certain question is answered. Elaborating on a very old question is a good way to earn reputation points and improve Stack Overflow and Stack Exchange. Once you have those 10 reputation points you can edit and improve your own answers with more links.
If there are no more questions you can answer or improve and there are no more questions you can ask, then I'm wondering why you want to get started here. If you can not improve there is no reason to get started. On the other hand, I'm a novice hobby programmer, and I can still help out people here and earn some reputation points when I am active enough. I am sure everybody with some knowledge can improve Stack Overflow and Stack Exchange.
Finally, this site is about putting up good questions with good answers attached to them and not about earning reputation points or some kind of reputation points challenge. Just start, gaining the first 10 reputation points is a cakewalk and from there you can do everything Stack Overflow and Stack Exchange is intended for.
Edit question, +2 each time suggested edit is accepted => 25 edited questions for reaching the magic 50 threshold
Fact is if your domain of expertise is C++ or Java or any super well known domain , it is almost impossible to find a good enough question not answered/accepted, and if you pick up the newest, it will be answered before you have sent your response.
But there are thousands of questions out there that can be improved. Most of them have some Tags missing or some Tags not relevant.
Somes can be improved to help understand the issue.
I was full of excitement and vigor and immediately tried to upvote (nope!) and post a comment (nope!).
That's indeed the main problem with the blessed site of Stack Overflow. People take this site as fun, as a game, as anything but sharing knowledge.
So, I'd tell you how to really start.
Register.
Start answering.
NEVER read the question body, but only tags and title
Write an answer that just looks like a good one (preferably just copy and paste some code snippet from manual, or other answer, if you want to bring some explanation along), but has no real relation to the problem.
Get deserved and hard-earned upvotes a ton
Don't be afraid of getting some downvotes - as long as your answer looks like a good one, the only downvote you can get from someone who have a clue and time to bother, but such people are scarce. Yet for every downvote you will get a comforting upvote - this site is for fun and happiness - remember?
Start your desired "activity" with votes, comments and unicorns. That's the real fun and purpose of this site.
Enjoy!
Apparently my old advice was SO bad, it wasn't even good, it was just bad.
So, here is what NOT to do under any circumstances:
1. Write 'Any help would be greatly appreciated' at the end of each question, because that's obvious.
2. Do what I did, and pretend that someone's comment was helpful just to influence them into re-upvoting your question, even if it does give you better rep. Stand up for what you really think! It's better for the community.
3. Ask a question that you haven't researched, especially one that has a good answer on the very same website (you might embarrass yourself, or make people unreasonably angry).
Here is what you should do:
If someone answers well regarding a piece of code, but you realise that what you posted was a much simpler version of what you're actually attempting, and you now want active help for your HARDER piece of code, just post another question instead of editing your old one. It's not cheating! Someone told me this and said my EDIT was a completely different question, and more people would notice it if I dedicated it to a new question.
If programming, then post your precise error - it's easy to do, and it's really hard to get any sympathy without it.
Be concise with your English. It makes a difference: e.g, 'it is important to note that I have already tried X, Y, and Z' could be said as 'I have already tried X, Y and Z'. Or, 'overly complicated' could be said as 'too complicated'.
Also, don't include anything that DOESN'T help people answer your question. For example, backstory. No-one cares. This is an exaggeration, but e.g, 'I've been doing this really hard project at this workshop with an old version of X and we're not allowed to use imported modules for some reason, other than X, and it's taken me ages and I feel like I'm missing something totally obvious; lots of people I know seem to have managed it just fine, but by the way, I was never quite sure if it would be better to do X, Y, Z' will probably alienate your audience. Also, don't say 'I'm really a beginner, I only started python X months ago', because you may as well say 'I don't know anything, I'm so sorry, I'm completely wasting your time'. It's not going to make people answer your question any better.
I believe answering those questions which you feel comfortable with. Favorite tags will present you specific set of questions.
However, if you genuinely feel an urge to answer a question out of knowledge or interest, then you can go ahead. Don't worry about an up vote or reputation. If you are engaged in a programming language, you yourself got some errors at that particular time, but you resolved it with the help of Google or Stack Overflow. Try to answer such questions, which you are very sure of.
I don't think it is necessary to answer on a daily basis. Unless and until it's within your helping range, don't go for it.
You should first ask yourself why do you want to join this community.
Figure out weather this is a community you want to be part of. Spend some time and research the type of people that are active contributors here (especially the elitists that run this site). Make sure you take your information from sites that are not under the stack exchange umbrella, since the content of those sites is moderated.
In hindsight, that would've made a huge difference, at least in my case.
So, to sum up and answer your question, the first thing a new user that wants to join stackoverflow should do is to understand what he's getting into. Failing to do so will result in a lot of wasted hours.
Related
I'm seeing a lot of "meta" questions about software testing lately. These are questions like, "What do I need to know to become a tester?", "How do testers do their work?", and "How well do I need to be able to code to be a good automation tester?" These sorts of meta questions are not appropriate for Stack Overflow, but there isn't another Stack Exchange site where this really fits. Where should questions like this be directed?
I will accept an answer in about two days. I want to give people time to answer.
http://programmers.stackexchange.com
I upvoted programmers since I think it's the best choice right now. Several career testers hang out there, so you would probably get a good answer.
We have a SE 1.0 site (testing.stackexchange.com), but the traffic is pretty light, so you may not get a great answer. We're working through a 2.0 proposal combining a few communities, so hopefully there's a better answer in the long term.
Closed. This question is off-topic. It is not currently accepting answers.
Want to improve this question? Update the question so it's on-topic for Stack Overflow.
Closed 13 years ago.
Improve this question
We've all been there. You have written some code and unit tests, the tests all pass, and the code is decent (nothing's perfect, right?). Then, someone who is sure that they know better than you comes along and decides to change your code or the interfaces to your code just because he/she does not like the variable/class names that you used. No "real" refactorings, no real optimizations, no real improvement -- just different words -- not necessarily better words, just different.
My real problem with this is that (a) its a waste of time and (b) it shows a blatant disrespect for the fellow developer that wrote the code in the first place.
My visceral response is to lash out, but that's counter productive. Instead, I though that I might wright a paragraph or two as sort of a "Charter" or "Philosophy" that is adopted for the project. I'm wondering if anyone else has tried this, and if so, was it successful?
After looking at the initial comments below (which are appreciated), I think that my problem is primarily that this change broke the build for code that was already working. So time needed to be spent to fix the code for what was (in my opinion) a non value-added change.
-- Thanks
...decides to change your code or the
interfaces to your code just because
he/she does not like the
variable/class names that you used.
My real problem with this is that (a)
its a waste of time and (b) it shows a
blatant disrespect for the fellow
developer that wrote the code in the
first place.
My visceral response is to lash out...
I see some VERY CONCERNING things in those statements.
naming is REALLY, REALLY important. It is worth rewriting code to get it correct.
It is not YOUR code
How is it disrespectful?
You are taking it too personally.
I once worked with someone who freaked out when I made changes to "his" code. His code as horrible; it was buggy and unmaintainable. He was always staying late, fighting fires and breaking things - basically a negative contributor. I rewrote all his bad code for a big piece of functionality for a project one weekend and when he came back in on monday had a hissy fit. I am not saying your stuff is horrible, but maybe you need to calm down and be more objective about it.
Don't take it so personally. Step back and think about it - maybe "your" code needed fixing
We might be able to give better answers if you posted the code and changes, or at least some better idea of the situation with an example or two.
EDIT:
After seeing the code change and finding out that the build was broken, I am going to have to change the tone of this answer. I understand Steve's frustration - and i agree - that is not a good change. It makes a specific typedef more general and not very descriptive any more.
While I think some of my points are valid, in this case it looks like the changes were not appropriate.
The issue of code "ownership" is irrelevant. If the code changes are useless then everyone on the team should not be happy. If they are good changes then everyone should be happy about it. If there is a difference of opinion then you all need to find a common ground.
Breaking the build is not a good thing.
Steve, sorry if I came down harsh - it looks like in this instance you are justified in your frustration, but not because it is "your" code.
One thing that might help in this sort of situation is to require code reviews for all changes. People are less likely to make pointless changes if someone else has to review it first. If they can actually convince another developer that their change should go in then maybe it isn't so pointless after all.
Heey,
Guys! WE all need TO TALK!
Just sit together and TALK! There are always reasons to change and there are always reasons NOT to.
Decide together!
Don't just go to StackOverflow or a forum and say ask this kind of questions.
The new dev does it - he gets responses from community probably positive (yeahh, bad code should be refactored).
The current dev does it - he gets responses from the community too: "What an idiot could do such kind of changes!"
And the result is: Counterproductive, destructive, offensive environment for a long time.
Who wants it?
Just put your arguments on the table and that's it.
New dev needs some introduction too.
Old dev needs to listed TOO.
This should be collaborative work AND not pissing each other off.
Decide together, talk, as THE TEAM.
And... better ask questions like "How is it better to refactor this?"
Cheers.
In any software development team with a size > 1, standardization is key. Not only for each developer to understand the other's code, but so that the people that come along in 2 years, 5 years and 10 years can look at any part of the code and see a clear and consistent pattern. Will you... and the rest of the team... seriously still be there, working on this project, years down the line?
If you both "just have your way of doing things" and there is no formal standard for the project/company, I suggest you work with the team and/or your boss to suggest a formal standard be adopted. There are many standards already published for the various environments that you can use either as the standard, or as a starting point.
If there is a formal standard, everyone on the team is obligated to follow it... no matter how much "better" they think their way is.
So much for the hard skills.
Now on the soft skills side...
You and your colleague need to develop a healthy relationship, or decide to work in different places. Tit-for-tats that result in people feeling that they want to lash out will make everyone unhappy, not to mention gravely jeopardize the project everyone is being paid to complete. Look for a person you both respect (maybe your boss, maybe a respected and level-headed senior member of the team, maybe HR if you have a good HR department). Explain to that person what the problem is and that it makes you feel unvalued and disrespected. Ask for help talking through the situation with your colleague and agreeing to a better way of working together.
Finally, you need to be open to the possibility that your colleague may be making subjectively correct changes, even if the manner he's doing it in offends you. Separate the correct coding from the correct interpersonal interactions. Do the right thing for the project.
Well if that guy is going to maintain your code, let him do whatever he wants to.
Just remember that it is not "your" code. The code belongs to the company for which you work for. You wrote the code and you got paid for it. Let the Management do whatever they want to do with it.
Don't take things personally, move on.
Sometimes, changing names might be justified. It can be confusing if half the project refers to a person's sex, and then you check in some new code that refers to gender or something. Okay, this might be a bad example as technically they are two different things and their meaning is most likely still obvious. But if a project's code uses two different terms to refer to the same concept, it can be confusing.
Usually I try to leave people's code alone, unless I have some justification for refactoring. Luckily the same seems to go for my colleagues, so no, I have not had the need for writing such a charter yet.
How about using an automated build system, so when this person changes the code and breaks something the team will get an alert about it. This solves your problem with having to waste your time fixing something broken by someone elses change to your code. This way everyone will know that so and so made a change and broke the build, and can see for themself. The rule is "dont break the build".
You should be discussing this with the person who did it, in a non-threatening manner.
I believe every developer should take responsibility and hence own some of the code, but not all of the code. I understand the code that I've written better (irrespective of how good/bad it is) than any other guy that has ever seen it. Therefore the changes I make will be faster and less prone to error.
I don't mind anybody changing the code I've written later on, but I have a couple of conditions:
If you change the code and that causes something else to break, you are responsible for fixing it, not me.
If I don't agree with the changes you made I will change it back to the way I want it since I have to take responsibility for this piece of code in future.
Not all developers should be making changes to all the code, all the time. Only some of the time, for the purpose of getting to know the code (sharing knowledge).
I've never worked for an employer that endorses a "everyone can change anything any time" policy. Developers own certain parts of the code and they are asked specifically to make changes/refactor based on a development democracy.
You touch my code and break something, (1) you better have a good reason for the change that all developers agree with and (2) you better not leave broken things broken or ask me to go do the clean-up for you UNLESS you're my superior. I will humbly submit if that's the case.
I agree with Laurence that code reviews might help. This is an issue of how your team should work together. What might help is the notion of Egoless Programming - in a nutshell, considering the code as a joint product of the team, and trying to make decisions for the sake of the code rather than because of the programmer's ego. Your teammate violated the fourth commandment of egoless programming - "Don't rewrite code without consultation."
Maybe if your team is made aware of these principles, things will improve. I would try this.
Perhaps not completely on topic, but .... If you have developers who have the time to make changes to code just because they don't like the variable names used, then maybe the conversation should be about whether you have too many developers and which ones should be shown the door ... or how you're going to justify to management the bloated staff you have, especially in the current economic circumstances!
Closed. This question is opinion-based. It is not currently accepting answers.
Want to improve this question? Update the question so it can be answered with facts and citations by editing this post.
Closed 3 years ago.
Improve this question
We have a project hand over from on shore team to our team (off shore) not long ago. However we were having difficulties for the hand-over process.
We couldn't think of any questions to ask during their design walk-through, because we were overwhelmed by the sheer amount of information. We wanted to ask, but we didn't know what to ask. Since they got no question from us, the management think that the hand-over process was been done successfully.
We had tried to go through all the documentation from our company wiki page before attending the handover presentation, but there are too many documents, we don't even know where to start with.
I wonder, are there any rules or best practices that we can follow, to ensure a successful project hand-over, either from us, or to us.
Thanks.
In terms of reading the documentation, personally I'd go for this order:
Get a short overview of the basic function of the application - what is it meant to achieve. The business case is probably the best document which will already exist.
Then the functional specification. At this point you're not trying to understand any sort of how or technology, just what the app is meant to do. If it's massive, ask them what they key business processes are and focus on those.
Then the high level technical overview. This should include an architecture diagram, required platforms, versions, config and so on. List any questions you have.
Then skim any other useful looking technical documents - certainly a FAQ if there is one, test scripts can be good too as they outline detailed "how to" type scenarios. Maybe it's just me but I find reading technical documents before I've seen the system a waste - it's too academic and they're normally shockingly written. It's certainly an area I'd limit the time I spent on if I didn't feel I was getting a reasonable return for the time I was spending.
If there are several of you arrage structured reviews between you and discuss the documents you've read, making sure you've got what you need to out of it. If the system is big then each take an area and present to the others on it - give yourselves a reason to learn as much as possible and knowing you're going to be quizzed is a good motivator. Make a list of questions where you don't understand something. Having structured reviews between you will focus your minds and make it more of an interactive task, rather than just trawling through page after page of tedious document.
Once you get face to face with them:
Start with a full system demo. Ask questions as they come up, don't let them fob you off with unclear answers - if they can't answer something have it written down and task them with getting the answer.
Now get the code checked out and running on your machines. Do this on at least two machines - one they lead, one you lead. Document the whole process - this is the most important step. If you can't get the code running you're screwed.
Go through the build process. Ensure that you can build the app (including any automated build and unit tests they may have). Note that all unit tests should pass - if they don't or if they say "oh, that one always fails" then they need to fix that before final acceptance.
Go through the install process. Do this at least twice, one they lead, once you lead. Make sure that it's documented.
Now come up with a set of common business functions carried out with the application. Use this to walk the code with them. The code base will be too big to cover the whole thing but make sure you cover a representative sample.
If there is a database or an API do a similar exercise. Come up with some standard data you might need to extract or some basic tasks you might need to carry out using the API and spend some time working through these with them.
Ask them if there's anything they think you should know.
Make sure that any questions you've written down anywhere else are answered.
You may consider it worth going through the bug list (open and closed) - start with the high priority ones and talk through anything particularly worrying looking. Even if they've fixed it it may point at a bit of code which is troublesome.
And finally if the opportunity exists - if there are any outstanding bugs or changes, see if you can pair program a couple.
Do not finally accept the app unless you are 100% sure you can:
Get the code to compile
Get the code to build (including the database)
Get the application installed
Do not accept handover is complete until they have:
Documented anything you picked up on that wasn't covered to your satisfaction
Answered ALL of your questions - a question they won't answer after being asked repeatedly screams of something they're hiding
And grab their e-mail addresses and phone numbers. Even if it's only informal they'll probably be willing to help out if the shit really hits the fan...
Good luck.
My basic process for receiving a handover would be:
Get a general overview of the app, document it
Get a list of all future work that the client expects
... all known issues
... any implementation specifics
As much up-to-date documentation they have
If possible, have them write some tests for critical components of the system (or at least get them thoroughly documented)
If there is too much documentation (possible) just confirm that it is all up to date, and make sure you find out from them where to start, if it is not clear.
Ask as many question as possible; anything that comes to mind, because you may not have the chance again.
Most handovers, perhaps all of them, will cause a lot of information to be lost. The only effective way to perform a handover that I have seen is to do it gradually. One way to do it is to allow a few key people from phase One to stay on the project well into Phase Two.
The extreme solution is to get rid of all handovers, and start using an Agile mindset.
As a start, define the exit criteria for the handover. This should be discussed, negotiated and agreed with both parties and make sure higher management knows this. Then write up a checklist of all things needed to achieve the exit criteria and chase it.
Check out "Software Requirements" and Software Requirement Patterns for ideas on questions to ask when gathering information about a project. I think that just as they would work for new development, they would also help you to come to terms with an existing project.
Closed. This question is off-topic. It is not currently accepting answers.
Closed 11 years ago.
Locked. This question and its answers are locked because the question is off-topic but has historical significance. It is not currently accepting new answers or interactions.
When interviewing for a programming position, did you ever get an answer to your question that you didn't quite expect? The answer could've been quite smart that you didn't even know or it could've been a dumbest answer you never expected. I'm expecting technical type of questions but anything interesting is fine.
Q: "Do you have any weaknesses?"
A: "Kryptonite"
A light-hearted one...
Towards the end of an interview, which I thought went reasonably well with the candidate asking sensible questions, showing interest and a good general rapport I asked a closing question because I wanted to know whether to proceed, "so what do you think, would you be interested in the position?".
He replied "I think it is the most boring job I have ever heard of and there is no way I would come and work for you and I think your whole company is lame" (the company was a nasdaq listed software house turning over around a billion dollars). I looked at my co-interviewer, who was a seasoned development manager who worked for me. He smiled and we showed the guy out.
In almost 20 years of interviewing that is the most surprising response I have had.
We were looking for an application programmer. After putting up some pseudo code on a white board I ask the candidate a question about it. His answer "is this some type of programming thing?"
As an interviewer I never had a big surprise, but as a candidate I was shocked and appalled at one company where I was not allowed to ask questions... WTF? I left pretty quickly.
Not really an answer, and a long dead question, but anyway... I was once one of several people interviewing a candidate. It wasn't going very well for him, but not a complete disaster either. At one point during the interview, his phone rang.
At this point, I kind of felt bad for him - he was obviously nervous and now had forgotten to turn off his ringer.
What I didn't expect to happen was for him to TAKE THE CALL, speaking for several minutes. In a foreign language. My coworkers and I looked at each other incredulously.
When he was finished on the phone, he simply hung it up and proceeded with the interview as if nothing out of the ordinary had happened.
He didn't get the job.
Not any questions, but I was once told that I had to hold a presentation about the company after he had given me information about it. Made me really pay attention to what he told me about the job and organization (I got the job).
I have also been doing a lot of interviews and hired several of my co-worker, something I wrote about for this question:
https://stackoverflow.com/questions/194543/for-interviews-how-do-you-gauge-whether-the-candidate-would-be-a-good-coworker#285594
I remember once I was interviewing at Microsoft and the manager told me that he was on the fence with me about whether to hire me or not. I told him, "Then don't hire me," which I think may have been a bit of a surprise to him.
I asked someone to sketch some Java code on a whiteboard when interviewing and was surprised to see the candidate put some Python up instead. Turned out the person's Java skills were, shall we say, over inflated on the CV...
Not an answer I heard, but one that I gave.
An interviewer asked me if it was possible to write object-oriented programs "using only a C compiler". I was a little bit amused by the peculiar word choice, so I jokingly answered that, in C, you could write a compiler for an object-oriented language and do it that way.
The interviewer didn't give me a chance to say how I understood that structs aren't really objects, functions aren't really tied to their data strongly enough, and there are a bunch of other OO features missing, so you can't really do it in C, which was probably the answer he was going for. I thought I had screwed up until he ended the interview with an offer for the position.
As an interviewer I had not been surprised very often. Most of the people have been respectful and intelligent. I had only been surprised by how quiet and nervous some people get or how little they tried to answer some of the technical questions.
As an interviewee I've given answers that were not expected. There was one occasion where I gave an answer at an early stage interviewer for large multinational company- and they simply did not understand the solution. Lesson learned: never give a 'different' or 'out of the box' answer unless it is in writing and the person receiving it is technical in that area. Otherwise, they may simply regard it as incorrect. Give the answer you think will be the norm- because in most cases the early interviewer has a very short list of the 'possible' answers.
Update:
The interview I was referring to was a 4th or 5th stage phone screen- so I guess it is not as applicable. The question was one of those ones that involve 'assume you have infinite memory' so I played on that coming up with some strange 'what if' scenario. The 'best' answer was one that was more to the point using traditional methods.
This doesn't apply just to the single item I mention below.
We did ask an applicant for a C++-centric software engineering position one time to talk about classes and objects in C++, which he could not answer. The final thing that ended it all was "do you even have any experience with C++?" The response: "No."
"You had a school class and you list it on your resumé..."
I interviewed a guy for a C++ job. He had lots of C++ buzzwords on his resume, including 'smart pointers'.
I wrote a little example program that used a raw pointer in a loop, the pointer was never deallocated. There were several if() statements that had if() statements inside of them, so stuff was going in and out of scope a lot.
I told him that this program ran fine for a while, but eventually it would bog down my computer or throw some kind of error message from the OS. I asked him to please take a look at this and see if you can see a problem or suggest an improvement.
After what seemed like forever, but was really on five or ten minutes, he noticed that several of the if statements would reassign my pointer without deleting the thing it was currently pointed at. He went through and added delete statements before these lines. So far, so good, if a little slow.
I asked him if there was anyway to make this code cleaner and less dangerous. I tried for over ten minutes to get him to say 'use a smart pointer', but I just could not do it. At one point I even said 'the answer is on your resume'. Still, total brain lock.
I really expected that using a smart pointer would occur to a guy who put 'smart pointers' on his resume would. I had anticipated we would discuss the different flavors of smart points that exist in the C++ universe, I didn't expect a total vacuum on something listed on his resume.
After more talking the fellow it came out that in his current position in a defense contractor he spent almost all his time going to meetings and wrote almost no code at all.
I liked the guy, but didn't feel good about putting in a full time C++ job so we passed on him.
Sorry, I misread the question. But since I've also done interviewing I can adapt my answer.
Where I work, the first question we ask in every interview is "tell us a joke". (Yes, I know, it isn't technically a "question"). This tends to lead to some unexpected answers, but they'd only be entertaining in person.
Closed. This question is off-topic. It is not currently accepting answers.
Want to improve this question? Update the question so it's on-topic for Stack Overflow.
Closed 10 years ago.
Improve this question
During hiring a .NET web developer I give the candidate a coding test.
I tend to limit the candidate to MSDN installed on the test server - I think it holds everything the candidate needs to complete the task.
I admit, this is not the normal case as I don't expect the candidate to do his work without use of the web.
On the other hand I don't want the candidate to google for a complete example and copy-paste it, i want to evaluate his skills.
The question is do I need to allow free use of the web during the test?
If you think the whole coding test is wrong - I would like to hear alternatives you may have for me.
As you say, 'I don't expect the candidate to do his work without use of the web' why not allow it too during the test? And what if he does copy and paste? I do that too. Surely the key is to know where to look, be discerning with what you find and apply it intelligently. Do you want to hire someone with a terrific memory or someone who can develop software for you?
When I was at school, calculators were just becoming affordable. As their use was seen as unavoidable, the exams were changed. Simple number-crunching was no longer tested in the way it was before (it was important then). Rather problem-solving techniques were to be tested.
I usually allow candidates to use whatever resources they want. After they're done, I sit down with them and go through their code together, ask questions like why they chose that particular approach etc.
If a couple of minutes of Googling was enough to not just copypaste some code but to learn enough about it to be able to defend the decisions within, then he's intelligent enough!
There are tests, where web access can be given, and there are where it doesn't really make sense.
Case where its fine to allow web access
When its unlikely to find even 60 percent of the code over the net
When you will ask to explain the code after he/she completed the code
A very specific solution using SQL query, which is unlikely to be found on the web
Case where its fine to not allow web access
Some basic programs like, recurssion, fibonacci, factorial, string manipulation, small trick programs, etc. There is no need of computer even in some of these cases
I'm very sceptical about coding tests during interviews. I think that a lot of the test I have seen, represent very specific (artificial, non real-world) problems where you would use the internet to solve them.
I think it's not really important to know how to solve such problems by heart - often time it is much more important that you know how and where to search for answers.
If you want to test the persons during the interview, I think it is better to ask them some conceptual questions instead of a specific programming problem. E.g: questions about object orientation, polymorphism, design of n-tier application, etc. etc.
Or as an example from the ASP.NET world, ask the interviewed person question such as: what is ViewState, what is a postback, what is session-/application-state, etc.
If you want to get an idea of how a candidate will perform in a job, I think it's best to try and make the conditions of the test as close as possible to the actual working conditions.
It should be pretty easy to prevent copy-and-pasters from slipping through the cracks by asking the candidate to explain his/her code.
Well, one thing you want to be aware of is that the developer you hire might not know everything that he will be thrown during the time he is working for you. If you ask him a question that he doesn't know off the top of his head you would want and expect him to research it and come back to you with proof that he understood the concepts that he just learned.
I say let them use the web - but ask them to explain in their own words how their code works. Most of my knowledge comes from online resources. However, I make sure that every line of code I write I understand.
There is a baseline knowledge that developers in a particular field should know; but you also want to figure out how quickly he can learn new things. A good test IMO is to throw a question you know he doesn't know and see how long he can figure it out using the resources he would have if he were an employee of your company.
Is your goal to see what basic knowledge the candidate has and if he can code without copying solutions from the web, then don't allow internet access. If you want to see what strategies he employs to get to a solution, let him use the web if he wants to.
I personally find it more interesting if a candidate can solve problems on a larger scale than just solving a simple programming problem. So I tend to ask him about the methods he uses when programming (Unit testing? Ever worked with it? What do you think of it?). This gives me a better picture than coding in an interview situation.
Sometimes it helps if you ask the candidates beforehand to bring a one-page coding sample to take a look at their coding style. This also saves you time during the interview.
It's important to make sure a candidate is resourceful - you don't want your programmer sitting there when they get stuck, not moving forward; you want them to use whatever resources are at hand - be it MSDN, picking someone else's brains, using the web, etc - to get the job done. Cut-n-paste from the web does seem like cheating, but (a) if you design your task carefully then it will be unique enough for there not to be a standard answer they can copy from the web, and (b) isn't re-using existing code a key part of building software? It's not much different from using 3rd-party libraries, to avoid reinventing the wheel. On the downside, of course, you also want them to show they can develop algorithms, so the unique task needs to include some element that requires that without the solution already being on the web. Trouble is, forums are the achilles heel to all of that since they can simply ask for the solution and someone, somewhere, is going to hand over the answer unwittingly!
Allow the candidate to use the web but tell him beforehand that if he used the web, you will have to evaluate HOW he solved the problem.
If he used the web for something simple such as finding the syntax or parameters which he forgot, don't mark him down. This is normal.
If he used the web for something like look at how a specific function is used, don't mark him down. This is normal.
If he searched for a specific code and then copy-paste it, then ask him about how the code works. If he can explain how the code works, then there's no reason to mark him down. If he can't explain it without looking at the site where he got the code, you have to mark him down.
If he used stackoverflow.com, check his profile for questions, answers and badges. From there, you can check how good a programmer he is.
It all depends what you want out of your successful candidate. I contest the view that knowing how to google makes you a good programmer because the simple fact is that the internet is full of bad examples as well as good ones. You don't really want your codebase to reflect how lucky your googler was on the day he cut and pasted all his code off the web. You want it to demonstrate sound practices, proven methodologies & elegant, efficient solutions that your team understand and are enthusiastic about. Not a jumble of styles that don't resemble each other. There's a wealth of good to be gotten from knowing how to get help from the interweb but real knowledge and ancient wisdom is being lost every day that people who don't really understand what they are doing are given jobs because they appear to solve problems with their ability to "google it".
If you really want to give your candidates access to the web then by all means do, but make the questions hard and scrutinise the results to see if they've picked the first solution they found or if they've picked the best solution to the problem.
As do many other respondents, I'd rather employ a resourceful developer who know how to use the web to the fullest to draw on other's experiences and previous work, than a developer who limits himself and his applications to the MSDN way of doing things.
I copy other peoples code all the time - daily in fact. The knack of it depends on finding the right solution quickly and integrating it into your existing work.
So let your candidate use the web and ask him how he came to his solutions. You might learn more about him from his methods than from how will he can remember previous solutions.
Three things I'd do.
Let applicants send in a coding example along with their cv.
Let applicants produce some real-life code (maybe even pair-program with a developer on your team) this will show you if they can actually use the tools. Internet is a tool too so they should be able to use internet.
Let applicants solve a problem in pseudo code on a blackboard during the interview. In this case you can be their "internet" by helping them.
These three approaches will show you different things. The first is a good early warning mechanism but can easily be faked (they could just download oss code from the web somewhere). The second is good to see if they can actually code but they might score badly if they're unfamiliar with the tools you use. The third will show you if they can solve theoretical problems but won't show you if they actually are good team players or if they write maintainable code.
I recently had a friend start talking to me on IM, he was in a coding test job interview. He had a couple SQL questions. At first i thought, hell you've got to do this yourself. I'm not going to help you cheat during an interview.
Then i thought about it again. I've been answering questions and talking to him about various technical issues for years on IM as part of his work. So when he encounters problems in the real world with the job if he gets hired, he'll do the same thing.
We don't talk about it much, but having a good network of friends to ask questions, and knowing how to search out relevant answers on the net are a big part of being an effective programmer or sysadmin. I've met people who were super smart programmers, but didn't really know how to find information online. They missed a lot, were kind of out of the loop. Knowing how to use resources should be important.
When i do interviews i often ask people what websites they read, what development tools they use, and why. It's a similar thing. Sure it's not about how they write x line of code, but it's about how they work.
No how to get around somebody just copy and pasting "answers". Well first, don't ask questions which have pat answers. Secondly when i'm interviewing i like to give people some code, ask them to refactor it, have them talk through what they are thinking. Then ask them to write some new code which implements a feature. Pair program with them. It's hard to hide inability to code when pair programming. While they are pairing, it totally makes sense to say, "let's go look up the api on the date time library."