Not perfect, but perfecting feedback loops

by Michael S. Kaplan, published on 2005/04/30 12:00 -04:00, original URI: http://blogs.msdn.com/b/michkap/archive/2005/04/30/413720.aspx

I was reading Paul Vick's discussion about Imperfect Feedback Loops and it got me thinking a bit about the various community efforts over the years.

A few years ago I was visting a friend in Orem, Utah and had an opportunity to see Jeffrey R. Holland speak. I did not agree with everything he was saying, but one particular bit seemed to resonate with me. He said something like "I am not perfect and cannot be perfect, but when I am at my best I are perfecting; we all are." As he started to say this many people around me seemed quite shocked (they clearly did think of him as perfect!), but when he continued they understood what he was trying to say to them. He went on talk about perfecting as a process, a way that one can better oneself.

So I was thinking about those words. Perfect and the state of being perfect (perfection) are absolutes -- one cannot really be "more perfect" after all. But they are also absolutes that we as software people are all unlikely to reach, be it in a product or a process or in anything else. But the word perfecting has a different impliction (imputation? I sometimes mix them up!) -- it suggests to me striving to try to reach that absolute, as a process. The fact that one cannot reach it does not make the goal any less good of an idea to try, does it?

Though words are not my strong suit here. Perhaps my intuitive definitions do not match what the actual definitions are. :-)

As I stated in the end of How whining a whole bunch got a feature added, I and all of the people I know on my team and the BCL team are taking bugs and feature requests posted to the MSDN Product Feedback Center quite seriously. The triaging of the issues is done as quickly as we can manage and there is a sincere effort to look at the issue being raised. As Paul states we cannot really address everything (especially this late in the product cycle) but we do work to address everything we can, and what we cannot is put in the hopper for the next version.

And as I also stated in the article, I take feedback to posts and articles here quite seriously, even if it annoys me (which does happen sometimes). And in several cases others have taken feedback posted here seriously as well. I admit that I do not always report on those issues; perhaps that is something I should work on as a part of my own perfecting process.

I have found and fixed bugs based directly and indirectly on both the blog and the feedback center, and thus as imperfect as things may be, I really do feel that things are getting better here, not worse. And both processes have a great deal of transparency since both involve talking much more directly with customers and their thoughts, their ideas, their concerns, and their bug reports.

So, things are not perfect. But were they ever really going to be? The main thing is that they are getting better all the time.

I think one of the big things behind open source is that it usually does it's own triaging, in this sense.

You "scratch your own itch" because you have a really strong need to change some piece of software, strong enough to dig into the code and do it.

The problem is Most Important to you, so therefore you fix it. You have the need and urgency that another developer will not have.

Or heck, they might never be in the situation that you're in. For instance, I mis-installed CUPS (a UNIX printing solution) and didn't get *any* debugging info. It turned out that I'd not installed any fonts as part of installing the Postscript engine.

No one would have ever been able to help, as there were no error messages, it just didn't work, and any tech support would have had to improperly install it the same way I did.

Also, I submitted the patches to pass the PS engine error output to the logs anyway, in case someone else ended up in that situation. Which means I got a feature added w/o any whining at all.

(I'm not knocking MS, or even proprietary software, just making an observation from another angle.)

Keep up the excellent writing.

Hi Gene -- thanks for the thoughts. I do not see a problem with open source in this context, though of course it can be harder sometimes if there is not enough of a formal framework from within which ornery people like me are whining, and working. :-)

For many open source projects that limitation does not apply though -- and in that case good work can also happen, consistently.

Glad you are liking the blog! :-)