LinkedIn is flawed for assessing a developer actual skills

Social networks are funny games



I am user of social networks since the beta of orkut the first prototype made by google. I was with fellow free software companions as Sam Hocevar and we have always played social network with the following rules :

- it is fun to have connexions;
- it is fun to play with.

We are instinctively pretty distrusting social networks, thus we were always making fun social groups like eyes with lasers, phô lovers, and stuff like that. If ever you go on linked in, don't hesitate to come in the game and endorse me as either a troll, or any exotic funny skill or expertise that I would be glad to endorse (eyes with lasers for instance).

Distrusting for you means I probably fear that my private data are used ? Well, as a pure hellenist I live according to my moral and value to live in a «glass house». No I just distrust the feedbacks to be genuine. I am who I am, and I voice it gladly.

You can't trust retroaction when they are not gratuitous



Look at linked-in new game: endorsement.

The question is when you endorse people is the endorsement genuine based on professional recognition, or based on your interest that the more you endorse the more you get endorsed?

And by the way did you actually see that person actually doing stuffs right?

Every time I decide to play «endorse people» I have the fear to be misleading, because I actually have not seen the people's real work in depth. In companies, at the opposite of free software your work is protected by the secrecy and the noise of inefficiency that cripples commercial organisation.

And I had a revelation

A fellow free software developers which module I evangelize and use connected to my network. And I discovered I badly wanted to make an endorsement that would stand out as more genuine than the others. So I made a recommendation to Olivier Lausanne based on his module pyQuery that I love. I made sure people could verify it existed, what it is, and what makes this developer special, and that they could check what I say by themselves. And for the first time in my linked in use I was a satisfied user (even though I had to bent its arms so that I could do what I thought fair and genuine).

But, I was pissed off by linked in. First, it was painful to explain how and in which context I used this software. For linked-in it seems we cannot be assessing another peer's work without actually have met him. But I trust more a developer which code I used and never met without the pollution of informal social chit chat. I am more confident to endorse a skill of someone I don't know because I will assess him only on what matters: getting things done the right way. Then I realized Linked-In is missing a massively more important fact than which company did you attend which is : in which you way do you help people and what do you do for the community?

I myself is a proud owner  of a first aid certification (in the context of a company). And I belong to the open source community where I like to help people. I value people volunteering to help the others because in a working context it means a team player someone that can not only produce, but value communication.


Hence I thought: an itch is scratching you then why don't you develop a software to supply these informations that would be on top of linked in helping people belonging to various communities that are actually getting things done (red cross volunteers, benevolent firemen, open source developers and users) endorse each others?

Well, the best solution is sometimes to do nothing


Then I my wits came back: I would distrust this social network. I still had a lot of pleasure giving back my recognition to someone whose work I value, but I value it because I had no selfish interests involved in the process. It came from the bottom of my heart, and I wish no incentive for doing things right.

A social network of recommendations on open source would results in coteries of developer artificially cross recommending themselves to stand out of the crowd (maybe on their  teachers or companies behalf), the same way stackoverflow or HN became crippled by these negative feedback loops of incentive.

 And then I realized I already had this in free software:
- people using my python modules are de facto endorsing my skills;
- people building softwares over my software trust me even more;
- people saying thanks are the endorsements I care about;
- people solving issues I raise value my troubleshooting experience;
- people accepting my patch trust me;
- people reading my doc, value my work.


Why would I need an imperfect version of what I already have?

Can I do something constructive?


I have a trillion ideas :) as usual.

I could write an howto assess a free software contributors that would encompass all the skills (doc writing, social interactions, QA, code, deployment, design...) for instance helping people use the already existing tools (github, bitbucket, free OS/distribution mailing lists, CPAN, pypi, rubygems, read the docs, travis-ci ....).

I could write a gamification tool for developer that would not only grant them achievement for actual code, but also for packaging, documentation, issues solving (both as a coder and a user), successful patch proposition, helping (on stackoverflow for instance) ....

I can actually do nothing because I don't want people doing the right things for an incentive but for the love of doing things right. And I value the benevolent aspect of free software. Still, I realize linked in will remain in my mind just a social game where it is fun to max your connexions, play the endorse game, and bypass the stupid logic of «only companies can be trusted».  Companies lie! Actual software don't.

Conclusion



After numerous interviews, and realizing how linked in is considered a tool of choice by recruiters and Human Resources Department I realized we have a problem software can't fix : recruiting based on recommendation is biased and a software cannot fix broken massive human behaviours. Recruiters are just doing recruiting the wrong way at least in the programming field of competence.

In fact I am liar: my last interviews in Canada were the best I ever had, and actually contradict every aforementioned statements in this post about broken HR process, I am thrilled and impatient to work for my new employer that (like the others) based its interviewed on getting things done and was actually taking my portfolio into account and checking it worked. Since it happened more than once, I endorse the whole Montréal's software industry as doing recruiting right. And as a (probable) result, the «benevolent» developers communities I have met had an outstanding level and were focused on getting things done (and of course while drinking beers :) and having fun). I love Montréal it is a developer's heaven, Tabernacle! And they even value being involved in a community and helping the other.  
  

No comments: