What makes a hobbyist?

Surely the definition of hobbyist should be who sponsored or paid for their work in the mainline kernel? If they paid for it themselves they're a hobbyist as far as the mainline kernel is concerned? The current definition makes people non-hobbyists even if they work for a company working on an ancient embedded kernel (nowhere near upstream), and the company would never in a million years agree to allow that person to work on the mainline kernel. The definition also traps people who now work on Linux, but did the bulk of their work in the kernel as a unpaid hobbyist.

Squashfs, which I initially wrote in my spare-time while working for an embedded company (on Linux 2.4.2) is a case in point. I would never have been allowed by management to work on it. Later as Squashfs became more popular (but still out of mainline), I discovered most of my improvements to Squashfs only came when I could work on it full time "between jobs", as spare-time working simply didn't give enough time any more.

I failed to mainline it first time in 2005, where the work had been done between jobs. Facing increasing demands to mainline it, but being unable to while working full-time, I left my job in 2008, and worked on it full-time unemployed for a year, finally mainlining it in 2009.

In hindsight it wasn't a good move, because I then found myself looking for a job (after a year of unemployment) in the worst economic downturn for decades. Suffice to say it wasn't an easy time.

I now work for Redhat as a kernel maintainer but only on the RHEL5 kernel (2.6.18).

So, does that make me as Squashfs author, a hobbyist or not? Under your definition it would not, despite most of the work having been while not being paid to work on Linux.

You say "Since most top kernel developers are not hobbyists these days, this is your opportunity to make up for what we're missing.", but you then partially answer your own question next, with "As we recognize most hobbyists don't have the resources to attend conferences ..."

Lack of time and resources and the consequent inability to "network" is what affects hobyists most in my opinion, made worse by attitudes from other kernel developers which have that in abundance.

Again, a case in point. When kernel.org got hacked all logins were revoked, and people wishing to reestablish access under the new system, had to get their PGP key signed by others in the Linux kernel community. To put myself in the position where I could get my key signed by the necessary people, I personally paid to attend LinuxCon Europe in 2011 (Prague). There I discovered no key signing event had been organised out side of the kernel summit itself, so I had to approach people individually, which was generally OK, but a number of people told me they didn't know who I was and then walked straight by without giving me the slightest chance to show documentation or say why I wanted them to sign my key, which was slightly humiliating. It did nothing to dispel the notion of a closed community which know everyone worth knowing by sight.