Hacking Developer Auction
developerauctionhacker.pen.io
developerauctionhacker.pen.io
But for the rest that you say, I agree, DeveloperAuction have an interesting premise, but you cannot draw any conclusions from any profile traffic you get.
I will add one more observation to the discussion, I found the entire DeveloperAuction process completely opaque and thoroughly mystifying. You touched on this in your write-up; What about avatar? What about headline? Does location matter? Who viewed the profile? What did they like or not? More feedback to the people being made offers would be immensely helpful.
I agree, the fakey profile on the front-page was annoying and could be construed misleading.
On the last day of the auction I appended this text to my profile:
Well, it has been an interesting two weeks. Not one single offer. Guess I am not what the start-ups signed up for DeveloperAuction's services are looking for. Can't be that many people around looking for a CTO.
I embedded some basic analytics in this profile, and set up a script to track the general location of the computer being used to view the profile and send an SMS to my phone each time someone viewed the profile, including their general location.
I wrote a script that would cross-reference AngelList's database of companies with the IP addresses used to view my profile. There were a few weird hits, people logging in from home and local coffee shops, but for the most part, I was able to nail down most of the companies that showed up to look. A couple of home addresses, including one of a VC who lives at a winery in Northern California. Very easy to track through LinkedIn and a quick Google search.
This profile was viewed a total of 51 times between 00:01 on 4/22 and 11:59PM on 5/3. The IP addresses (obscured to protect identity) that viewed the profile were:
199.87.84.X,03/May/2013
166.137.191.X,04/May/2013
174.76.140.X,22/Apr/2013
38.104.230.X,22/Apr/2013
199.21.85.X,22/Apr/2013
24.5.70.X,22/Apr/2013
199.47.77.X,22/Apr/2013
199.241.200.X,22/Apr/2013
206.214.33.X,22/Apr/2013
216.231.9.X,22/Apr/2013
216.55.3.X,22/Apr/2013
173.13.174.X,22/Apr/2013
96.229.22.X,22/Apr/2013
96.229.22.X,22/Apr/2013
216.231.9.X,22/Apr/2013
98.155.87.X,22/Apr/2013
99.89.1.X,23/Apr/2013
174.76.140.X,23/Apr/2013
152.179.69.X,23/Apr/2013
12.201.135.X,23/Apr/2013
50.1.62.X,23/Apr/2013
173.228.28.X,24/Apr/2013
50.193.52.X,24/Apr/2013
69.170.175.X,24/Apr/2013
199.47.72.X,24/Apr/2013
67.53.126.X,24/Apr/2013
74.62.176.X,24/Apr/2013
74.73.226.X,24/Apr/2013
64.88.170.X,25/Apr/2013
98.210.198.X,26/Apr/2013
208.90.212.X,26/Apr/2013
208.90.212.X,26/Apr/2013
208.90.212.X,26/Apr/2013
208.90.212.X,26/Apr/2013
208.90.212.X,26/Apr/2013
208.90.212.X,26/Apr/2013
50.1.107.X,26/Apr/2013
24.7.16.X,27/Apr/2013
76.175.77.X,29/Apr/2013
76.110.232.X,29/Apr/2013
166.147.88.X,29/Apr/2013
67.52.152.X,29/Apr/2013
208.87.61.X,29/Apr/2013
209.119.30.X,29/Apr/2013
208.54.4.X,30/Apr/2013
216.55.31.X,30/Apr/2013
207.34.40.5,30/Apr/2013
198.228.216.X,30/Apr/2013
12.177.140.X,01/May/2013
173.247.201.X,01/May/2013
173.228.89.X,02/May/2013
rem first we download the log files from my web server
wget --user=XXXX --password=XXXX -O access.log "ftp://otakunozoku.com/statistics/logs/access_log"
wget --user=XXXX --password=XXXX -O processed.log "ftp://otakunozoku.com/statistics/logs/access_log.processed"
rem process the log files, remove my two known ip addresses, split up the output in to ip address and date
rem should use "tee" here with process substitution but we cannot do that on bash in Windows, so instead, we use a temporary file
rem this is what we should use: grep -h -i "itunes_link" *.log | grep -v -E "98\.154\.169\.XXXX|71\.165\.193\.XXXX|70\.36\.57\.XXXX" | awk -v OFS="," "{print $1, substr($4,2,11), substr($4,5,3),substr($4,2,2)}" | tee >(gsort -t , -k 3,3 -k 4,4 hits.csv > date_order_hits.csv) >(gsort hits.csv > ip_order_hits.csv)
rem instead, we use these three lines:
grep -h -i "itunes_link" *.log | grep -v -E "98\.154\.169\.XXXX|71\.165\.XXXX\.131|70\.36\.57\.XXXX" | awk -v OFS="," "{print $1, substr($4,2,11), substr($4,5,3),substr($4,2,2)}" > hits.csv
gsort -t , -k 3,3 -k 4,4 hits.csv > date_order_hits.csv
gsort hits.csv > ip_order_hits.csv
rem geolocate all those ip addresses
del geolocate.xml
awk -F, "{print \"http://freegeoip.net/xml/\"$1}" ip_order_hits.csv | xargs curl $0 >> geolocate.xml
rem look up those geolocations on google maps and save the addresses
rem damnit, never could get sed to generate the proper formatting, just had to run it through gawk in the end
del addresses.txt
sed -r -n -e "/<Latitude>/ { N; s/<Latitude>//; s/<\/Latitude>/,/; s/<Longitude>//; s/<\/Longitude>//; s/\n//; s/ //g; p}" geolocate.xml | gawk -F, "{ print \"http://maps.googleapis.com/maps/api/geocode/xml?latlng=\" $1 \",\" $2 \"^&sensor=false\"}" | xargs curl >> addresses.txt
rem go to angel list REST interface and pull in all companies
curl ... > all_startups.txt [what? you expect me to supply these commands so you can do the same?]
rem extract the company addresses and geolocate them
awk ... | xargs curl $0 > located_addresses.txt
rem filter companies that are within 1,000 metres of each IP address
rem get all of the jobs for each filtered company that matches an IP address
curl ... | sed ... | grep ... >
rem remove all companies that are not hiring for a technical position
awk ... | grep > hiring.txt
rem cross-reference the geolocation of each angel list company that is hiring for a technical position with each companies' IP geolocation data
rem plot the locations on a google map with the names of each company and a confidence percentage of who they might be, send the image to imgur
comb -i -v located_addresses.txt hiring.txt addresses.txt | r conf.r | googplot ... mapped.png | curl -k "mykey=[my imgur api key]&image=mapped.png" http://api.imgur.com/2/upload.xml
removed street numbers to protect privacy
Cottonwood Court, Hillsborough, CA 94010, USA
Skyfarm, Hillsborough, CA 94010, USA
Pierce Road, Cupertino, CA 95014, USA
East Main Street, Grand Prairie, TX 75050, USA
South Santa Fe Avenue, Vista, CA 92084, USA
Northwest 120th Street, Potwin, KS 67123, USA
Stevens Creek Boulevard, Cupertino, CA 95014, USA
De Anza & Stevens Creek, Cupertino, CA 95014, USA
Cowper Street, Palo Alto, CA 94301, USA
Castro Street, Mountain View, CA 94040, USA
El Camino & Castro, Mountain View, CA 94040, USA
Old Mountain View, Mountain View, CA, USA
Nelson Rising Lane, San Francisco, CA 94158, USA
Mission Bay, San Francisco, CA, USA
South of Market, San Francisco, CA, USA
Anacapa Street, Santa Barbara, CA 93101, USA
3rd Avenue, New York, NY 10003, USA
Natoma Street, San Francisco, CA 94103, USA
Slater Avenue, Fountain Valley, CA 92708, USA
West 17th Street, New York, NY 10011, USA
West 1st Street, Los Angeles, CA 90012, USA
South Van Ness Avenue, San Francisco, CA 94110, USA
Their offering to talented developers seems to be somewhat limited. Basically you sign up, spend a while writing out a huge profile (in addition to your resume and public work on github), then they just dump you into the laps of recruiters as part of their normal hiring pipeline. Hey, if I wanted to sit around doing 15 hours of live algorithm screening, I would have just called those companies myself...
This is probably okay for junior developers, but realistically junior developers just need to have a pulse to get hired (and maybe one or two sample projects). As for senior talent, there's few incentives to participate in these antics. Same antics, different names. I got a bunch of offers, but they require a great deal of effort to follow up on everyone and schedule a billion interviews and screens and other things that I am frankly exhausted by.
I would love something like a 1-hour guarantee: receive an initial offer to see if the range is what you have in mind, spend exactly one hour talking with a company before receiving an actual truly real sign-on-the-dotted-line offer. I am just frustrated about the marketing I think, I'll feel better in an hour.
> You probably got busted because you used getclicky.
I got busted because they found me bitbanging some urls. :) Btw, your self-improvement focus is interesting. I often tell people that the most immediately empowering tool they can pick up is programming. After that, there's other projects worth pursuing (both realistic and hypothetical): http://diyhpl.us/wiki/declaration anyway, might not be up your street..
... wtf? you use rgbds? There's only like 10 of us in existence. I use it here: https://github.com/kanzure/pokecrystal
Regarding the banging on URLs: I don't know if they are becoming more sensitive to that. I did a little digging around via various URLs but never got cautioned on it. You may have gotten a little deeper in to it, or you touched a raw nerve that I didn't stumble upon. :)
I was surprised by how few companies actually viewed my profile. They provide the list at the end of the auction and it was interesting to see how it matched up with my own. Philosophically you have to look at it through a cautionary lens:
How many companies do they have on their list?
How many companies do they have on their list are hiring for a technical position right now?
How many companies do they have on their list are hiring for a technical position right now and looking for the skills I have?
How many companies do they have on their list are hiring for a technical position right now and looking for the skills I have are hiring at the level I am at?
How many companies do they have on their list are hiring for a technical position right now and looking for the skills I have are hiring at the level I am at and making a realistic offer?
I also noted that DeveloperAuction uses Craigslist and LinkedIn to hire their own programmers. Obviously their pipeline is broken or not mature enough and they aren't eating their own dog food.
I sympathise on being dumped in to other companies' generic hiring pipelines once the auction is over. That is a huge disconnect between promises made and the cold, reality of finally getting to a real, sign on the dotted line offer.
Agreed on the Junior developers. There is a fallacy about the tech talent shortage. If you are young, naive, willing to be overworked and underpaid then $100K is a lot of money when you're straight out of college and the only two things you are worried about is whether you will ever have a girlfriend and how much longer can you defer your student loans for.
I think there's only a shortage of tech talent at the lower levels where the machine that has been created by SV and its ilk cannot feed it fast enough with warm, young bodies (that sounds wrong somehow, but you know what I mean.) But this is a discussion for another time and another thread.
The hiring process and pipeline is definitely broken. It is exhausting to go through each company, follow-up, schedule interviews, take time away from work. I am ramping up my “looking for work” efforts right now in anticipation that the start-up I am co-founder of runs out of runway before it can take off. “Dig your well before your thirsty.”
The code on github is never looked at. Not by the “technical recruiters,” not by the hiring managers, not by the engineers and so on. I have sat in an interview and been asked point blank, by the same group of people gathered at once, “do you still code?” “Can you write software?” “Have you ever worked on OpenSource?” and then been told at the end that they felt I “didn’t code enough on a weekly basis to make a significant contribution.” You stand there with your mouth agape at the sheer, unadulterated willful ignorance and lackadaisical interviewing attitude of people some days.
I co-authored slash maintained RGBDS for about 8 years until retiring from active development. During the golden years of GB/GBC/GG development it was the tool to use.
> You're experience went a little further than mine, I got no offers.
Friend of mine (more experience than me, less than you, well known on Hacker News) was in their first round and also got zero offers. Frankly, I think it might be literally random. I don't have enough data to prove this.
> I was surprised by how few companies actually viewed my profile. They provide the list at the end of the auction and it was interesting to see how it matched up with my own.
They didn't notify me about the list at the end, but they did show me a hidden page near the beginning which continued to be populated. I was also able to see about 200 other companies based on logos. I think not every company had logos.
> DeveloperAuction uses Craigslist and LinkedIn to hire their own
also angel.co
> It is exhausting to go through each company, follow-up, schedule interviews
hate this
> The code on github is never looked at.
hate this too; i wish someone would actually read my code for once.
> willful ignorance and lackadaisical interviewing attitude
more hate. The interview process is a terrible experience that I would never wish upon my closest friends or family. I would be absolutely ashamed of admitting I did this to other human beings. There needs to be dramatically more transparency upfront about salary (hell, let me negotiate in advance) and how long the hiring pipeline is. How the company treats people moving through their hiring pipeline is a big indicator for how the hell they treat employees. I want a place that treats me with dignity and respect, and if companies can't offer that, I'll stick with consulting, thanks.
> RGBDS
RGBDS hates me. Thank you for your efforts on RGBDS. It is probably the one tool I regularly use that causes me the most trouble (crashy). I am not sure anyone has run such a large project through it before.
Hey we could be hatred buddies!
re No offers: I was wondering if I was to be the first on DeveloperAuction that would receive no offers. Glad to hear I am not the only one. Agreed on the randomness -- if the stars align just right -- and the opacity does not help. I was contacted by one of the founders of DeveloperAuction to ask if they could use my profile in their marketing, but when I pointed out that I was probably not going to receive any offers, should they really be holding up my profile as an example of what to do. I heard nothing back from them after that.
re List of companies: I got the list at the end. I took a screenshot and saved it off.
re Exhausting: I don't think it will ever change really. "This is how it has always been done" and whilst the system is utterly broken, there is a long chain of component people with only a few of those components willing to change.
re Github: "Github is the new resume" is a lot of bollocks. However, it means I can just direct people to the code rather than having to email it. Though I still hear "We don't know how to use github, can you just email me a .zip?" and the best one "What do you mean all of it? No, that's not right, you must have used someone else's code to base this off of" which usually dovetails in to "Oh, you only work on your own projects? Well that's not good, you really need to learn to work in a collaborative environment." You can do nothing but roll your eyes at statements like these.
Whilst I am not volunteering to patch RGBDS, I am surprised it is "crashy" for you. Like all bad customer service: "I am sorry to hear of your plight but there is fuck all I am willing to do to fix it." Or would you prefer I invoke the OpenSource mantra of "You've got the code, you could fix it yourself!" :)
RGBDS has quite a few quirks (it started as a hack because the GG tools sucked big time). I would guess you're probably pushing it to the limits on the Pokemon games. The tool was used to ship hundreds of games but probably not on the magnitude that you are stretching it to with those particular games.