Tuesday, May 06, 2003
The principle of "good enough" -POGE
- Blogger lost this post, originally written april 20. Google cached it. If it weren't for google's cache, and a random netizen copying my work into his private web page - I'd have lost my System Administration, The Movie too. Thanks for the service, netizens!
Sometimes a meme resonates in our supersaturated web solution of ideas... becomes part of a standing wave... then - wham!! it crystalizes in a short phrase that contains all the meaning you need to hang a further discussion on.
Doc wrote a few weeks ago about POGE - the principle of good enough.
Heh. I wrote about that, about a year ago. I called it the rule of good enough. My technical choices as both a programmer and as a manager have been guided by it for nearly two decades. It's worked more often than not. POGE beats back PHB based disasters. POGE beats software sold by powerpoint presentations, press releases and sexy marketing girls hands down. POGE means that after actually looking at the code and running the software, you can determine if it's good enough, improving, or abandonedware. If it's good enough now - it's going to get better. Plausible Promise is where engineers get involved in a project, POGE is where users do.
You can't smell a conventional company's software to see if it is decaying or not. Huge marketing resources are expended to ensure that you don't.
POGE means that if a piece of software is good enough - it will find a niche for itself.
Doc wrote:
- One advantage of POGE is gradual improvement. From their beginnings, all members of the LAMP suite (Linux, Apache, MySQL, PHP, Python, Perl, PostgreSQL) have been triumphantly adequate efforts. Their adequacy keeps growing, of course, because so many who apply them contribute back to their steady improvement. But their appeal remains a matter of uncomplicated utility, not does-it-all polypurposefulness. And that seems to be the main difference between XMPP and SIMPLE.
POGE means that if a piece of software is "good enough" - it will eventually take over the niche dominated by a pricey, more feature-full competitor. POGE-ware finds new markets and creates new ones. Borland's 99 dollar Turbo-C pushed most competing compilers off the market - GNU's gcc compiler pushed out the rest, and eventually Turbo-C itself.
I like the acronym "POGE" over "ROGE", (although I often felt like a rogue when I dragged open source into an organization, or had to defend it from those PHBs) - so whenever I get the dozen pages I've written on the topic over the past year blogable, I'll do that. I've got some examples that doc didn't use - 802.11b vs bluetooth is one and I have some data showing the growth and change of open source projects over the past 5 years that ties into it. OK, some text, some graphs, a workable acronym, hot dang I've got a theme! Maybe there will be some net-wide discussion on this rule, because it explains so much about the growth of open technologies vs proprietary ones.
¶ 10:56 PM
