I started watching the first video, and it said "Check out the remake". "Okay," I say, "if it's better, why isn't it on the main site? If it's not better...why is it so prominent?" And then it turned out that both videos were, in a word, awful. You lost me after the first 100 seconds or so of each one. I don't know if they ever advertised your website--I didn't get that far. PS: The one on the main site has misspellings in the lyrics.
I also looked at the main description:
>Let your candidates code in their own IDE where they
>feel comfortable and check the results in Google Docs.
"Ah," I think, "he's about to pull some kind of voodoo so that I can show my interviewer what I can do with Vim and SLIME." Then I actually investigated, and found out that by "their own IDE" you mean a "common IDE", and by "common IDE" you actually were referring to only two IDEs, neither of which I have used or want to use.
The web version is actually cool, and I'll probably use it with my friends (I actually just finished typing some code into Facebook to help a friend out--it would've been great to use this website instead). But I had my hopes built up about my "own IDE", so using the web interface turned out to be disappointing.
So my humble suggestions:
1) Trash both videos.
2) Change some wording to make it clear that most people will use a web interface that niftily allows real-time editing by two people.
3) Since your product is so cool, mention that if the customer uses Eclipse or IDEA, there's a plugin so he can use his own IDE!
Why is this free? B/c you're not charging for your product, I'm skeptical that you'll be around long enough for it to make sense to learn how to use your product. Remember you're target is businesses, who are much less price sensitive, and are much more interested in stability. PLEASE CHARGE.
Or just use a screen sharing tool like Skype. That's how I got interviewed by the company I'm working for right now, and no browser based tool can match it. I could use my favorite editor(Vim) and show off my command line skills too. It was one of the best interview experiences I had.
I can't decide if I'm happy or not that non-coding skills can leave such an impression.
I'm still at my first job (I'm only a HS senior), and for the interview I was asked to bring in some code. I brought my laptop, whose splash screen proudly displayed "Crunchbang Linux". I believe this was the moment that gave the most "points" in their eyes. Then I opened up my code, did some basic-but-impressive-looking flying around with Vim (it's not like there was actual editing involved), popped over to a terminal to `cd' around and then run it, and left a very good impression.
And my PM has a habit of asking me for help with bash. I have a rudimentary-at-best knowledge of it, but he's happy when I kludge together something with pipes and grep and sometimes "after that command, vim will open. You should then press the following to find and replace the text as desired." I'll then include as a footnote, "There's a much better way to do this as a one-liner with sed and awk...I just don't know how."
So of {beautiful code, working product, knowledge of tools} they only value two...and I'm just surprised it's the last two, not the first two. I guess I can't complain though.
If it works then it is great, if it does not then it sucks. Skype auto changes quality based on the connection meaning that your text might be readable and it might not be.
The other part is bitmap vs text. If you have use for the text (clever solution for something) or maybe some analysis on it then text wins. Will depend on the case.
Screensharing wins for showing off everything and not just a single editor text. Commandline, keyboard foo, google foo etc.
Google Docs is great, especially because everyone knows about it and it just works. We started off by writing IDE plugins that posted to Google Docs but latency was horrible. Should give it another shot in a year.
Cool idea! I have been going through the dev hiring process currently, and it is a huge pain in the ass to write up a test and email it to everybody, then review the answers and try to keep them straight. It would be great if the interviewers could create specific questions and keep track of individual candidates answers, as well as comment on specific parts of the answer and (maybe) even rate the answers or sort them by 'best' answer for internal use.
Honestly I wouldn't be very happy doing live coding in a stressful situation in some random web interface (emacs not supported? heretic) - half of the time would be wasted on writing import statements, non essential functions, etc.
I understand why companies would want to see somebody code, but if they can critique some code that they have never seen before that is properly as good a signal and it is easier to give the same test to the candidates.
I ask a lot of coding questions in my interviews - if the candidate starts writing imports, they've already missed the entire point of an interview. I'm not interested in knowing your intensely deep knowledge of Java boilerplate.
Interviewers who demand compile-ready code on-the-spot are frankly, insane.
I use typewith.me (an etherpad clone, I think), and it's great. The big win for using real-time typing apps is evaluating fluency. Someone can probably bang out some reasonably looking code if you give them homework (assuming they did it themselves!), but there's no faking the keystrokes of someone who's intimately familiar with the syntax of a language and knows WTF they're doing.
Other related projects/sites/products include Blueberrytree[1], Trollim[2], CodeYourIdea[3], HappyJobSearch[4], OneDayOneJob[5], and the aforementioned iSeekMikeCode[6].
I'd have to be at my home desktop and check my history for a good list of links, but I got to use a handful of competitors when I was looking for work a couple months ago. I'd have to try this one out to be sure, but the extra polish plus the ability to use the interviewee's own IDE could be huge advantages.
Wow you can create your own challenges? That's awesome. I've seen some sites like this before, but they usually have a predetermined amount of algorithms.
I always wanted to make my own challenges that would have to be automatically solved by potential candidates, and here it is! Glad someone else did this already.
How does etherpad (acquired by google) compare to just using Google docs? Recently went through an interview with Google where we used a shared document, it was a horrible experience. I remember thinking that no sane person would code with a plain text editor and there should be a better way to do it. The ideal service to me would be something like a web based collaborative emacs.
We've tried using google docs for interviewing, and found that the code will not always sync. The remote developer would tell us "here's my code, it's ready", and we would see a blank doc for like 5 minutes.
Etherpad (and its clone instances) is awesome. Never had a problem.
I also looked at the main description:
>Let your candidates code in their own IDE where they >feel comfortable and check the results in Google Docs.
"Ah," I think, "he's about to pull some kind of voodoo so that I can show my interviewer what I can do with Vim and SLIME." Then I actually investigated, and found out that by "their own IDE" you mean a "common IDE", and by "common IDE" you actually were referring to only two IDEs, neither of which I have used or want to use.
The web version is actually cool, and I'll probably use it with my friends (I actually just finished typing some code into Facebook to help a friend out--it would've been great to use this website instead). But I had my hopes built up about my "own IDE", so using the web interface turned out to be disappointing.
So my humble suggestions:
1) Trash both videos.
2) Change some wording to make it clear that most people will use a web interface that niftily allows real-time editing by two people.
3) Since your product is so cool, mention that if the customer uses Eclipse or IDEA, there's a plugin so he can use his own IDE!