OpenBSD 2014 by the numbers | Hacker News Reader

OpenBSD 2014 by the numbers

tedunangst.com

64 pointszdw13 comments

I love OpenBSD and this is in no way a bash, but… ugh

  Who were the most productive developers? Top three in terms of lines added:

I understand this is just a funny page with some metrics, but seriously calling people who added most LOC the most “productive” developers is somewhat disingenious.

  In order to prevent cvs from filling up with all this code, it’s necessary to delete some old code.

Is there any current reason to still use CVS in 2015?

  Special mention to jsing for achieving the most churn and smallest net gain by adding 153802 lines and deleting 152604.

Net gain. Blowing up the codebase size is “net gain”. Maybe OpenBSD team knows something I don't, but I got a bad vibe from this report.

> I understand this is just a funny page with some metrics, but seriously calling people who added most LOC the most “productive” developers is somewhat disingenious.

This sentence contradicts itself.

> Is there any current reason to still use CVS in 2015?

http://www.openbsd.org/opencvs/

The funny thing is that even though they have OpenCVS in their source tree, they do not use it. It never took off and probably had too many bugs and issues.

Instead they are and have been using GNU CVS instead.

OpenCVS was linked to the normal build for a while but has been removed from it four years ago.

http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/usr.bin/Makefil...

I wonder if keeping CVS is not one of the smartest things the OpenBSD project does now. I'm being somewhat facetious but hear me out.

Workflow is as much about people and process as it is about tools. The tools simply serve the people and processes.

This is a project that has consistently hit high quality releases, on a predetermined schedule, for coming up on two decades. That is unprecedented. I can't think of anything even remotely similar.

Version control is a tool for integrating change and managing releases. They are arguably one of the best projects at doing it. See the silliness of the "CVS?!" non sequitur yet?

So for people to drive by, who are statistically more likely part of the problems in the software industry, and critique the OpenBSD development process.. is at best cute and worst delusional.

Inadvertently(?) it functions as a litmus test.. if you care so much about this, you aren't really who we want to work with anyway, similar to the candidate fixated on his title in http://dtrace.org/blogs/eschrock/2012/08/14/engineer-anti-pa...

Makes sense, thanks for the perspective.

Yep. The OpenBSD devs seem to be fully conscious of this too. Like, when they first announced their fork of OpenSSL, and decided to use comic sans for the font. (http://www.reddit.com/r/netsec/comments/23orq0/libressl_open... -- "This page is scientifically designed to annoy web hipsters.")

As expected, that got a bunch of people talking about comic sans, instead of OpenSSL, and acted as a pretty good filter for serious programmers.

...which apparently I've also just fallen on the wrong side of. Darn.

>I wonder if keeping CVS is not one of the smartest things the OpenBSD project does now.

Given that OpenBSD is supposed to be a security-oriented OS, it's slightly weird that they are using a version control system that does not guarantee that what you put in the repo actually stays there, unchanged. Git guarantees that.

> They are arguably one of the best projects at doing it.

What makes you think that?

1) Repo security is a hard proposition. Using git with an ostensibly(SHA-1) immutable history doesn't imply security. git repo integrity is cool, and I'll just point to interesting discussion of attacks and mitigation in http://mikegerwitz.com/papers/git-horror-story.

I'll end my debating with the fact that there is room for improvement.

2) Releasing on time, of high quality, for nearly two decades. Development process.. not that it is some ultimate code or product.

> Is there any current reason to still use CVS in 2015

It works.

> Is there any current reason to still use CVS in 2015?

Is there anything wrong with it?

> Blowing up the codebase size is “net gain”. Maybe OpenBSD team knows something I don't, but I got a bad vibe from this report.

To be fair, I think the "net gain" there just means "net increase" not net gain in the sense of a positive outcome. Plus, the top line deleters are removing way more than the top line adders. If I'm reading it right the total "net gain" was somewhere around negative 2.2 million lines.

The post lacks nuance, but it's really just supposed to be a fun little snapshot with some simple numbers.

> I understand this is just a funny page with some metrics, but seriously calling people who added most LOC the most “productive” developers is somewhat disingenious.

@bottom you can see the real productivity:

  lines added     1192696
  lines deleted   3484520
  files added        2653
  files deleted      9995

As Ken Thompson said: "One of my most productive days was throwing away 1000 lines of code." :)

So it's been a really productive year.

(edit: updated to fix codeblock :) )

"The first commit of 2014 was to bump the copyright date, but then jsing jumped the gun and bumped it again at the end of the year, resulting in a copyright year one day shorter than the calendar year. Last commit, for the curious."

This is a level of pedantry that deserves respect.

Reply on news.ycombinator.com