Applying the GDE matrix to cycling?

As a member of the Institute of Advanced Motorists and a regular cyclist I’ve wondering whether the Goals for Driver Education (GDE) matrix can be applied to developing cyclist safety. With my own children learning to ride on the roads it’s more pertinent today than ever.

Briefly the matrix maps four levels to three areas. The levels are:

• Goals for life and skills for living
• Goals and context of driving
• Driving in traffic
• Vehicle Control

The three areas are:

• Knowledge and Skill
• Risk Increasing Aspects
• Self Assessment

It’s a hierarchical model and driver/cyclist training typically focuses on the lower two levels. Various road safety organisations are promoting focussing on the higher levels as a more effective means of developing safer drivers.

For example, you may have great control, be experienced in traffic, but you just missed your turning and you’re late – suddenly you find yourself taking more risks.

The lower levels are much easier to explain and teach. The highest level is about who you are and your attitudes – much harder to fix but surely worthy of the investment.

Why RSS doesn’t work for me – consumers versus producers and one tool to rule them all?

I just realised why I don’t use RSS readers very often – you can’t create content with them.

I doubt there will ever be one tool to rule them all – Facebook is trying very hard in this area – but I want to minimise the number of ways I have of keeping up with and contributing to communities both inside and outside my company.

To successfully follow communities an RSS reader beats visiting each one manually but if I want to contribute I have to step outside the RSS reader.

RSS makes perfect sense for (say) news, eg The Guardian – where I do use RSS – but apart from that I don’t ‘consume’ that much (note to self, reconsider intent to purchase an iPad 2).

When it comes to active participation I find myself being sucked back into communities that run on email – the lowest common denominator in more than one sense. Other communities get less attention, eg Facebook. In some ways the communities that are a priority for me do tend to end up in my INBOX. Others have to wait until I have time to browse.

Are there any RSS readers that will let you post content?

A quick search suggests not and I suspect the problem is not the tools but the lack of a common protocol to post content.

Perhaps another way of putting it is that I tend towards producing and less consuming?

Fixing Thunderbird 3 on Solaris running out of file descriptors

I recently started running Thunderbird 3.0.3 on Solaris with a large 2000+ mail folder hierarchy accessed through IMAP. It rapidly becomes unusable. Problems seen include:

1. In stderr the following sorts of message are seen:

(thunderbird-bin:7792): Gtk-WARNING **:
Error loading icon: Failed to open file
'/usr/share/themes/nimbus/gtk-2.0/stock_dnd.png':
Too many open files


2. The GUI pops up errors like this:

Unable to open the summary file for XXX. Perhaps there
was an error on disk, or the full path is too long.

Where XXX varies.

3. Some icons in the menus appear as a blank page with a
red cross (x) in them.

4. Mouse and keyboard operations silently fail.

Restarting Thunderbird corrects the problem for a few minutes but then problem returns.

Running pfiles(1) against the thunderbird-bin process shows file descriptors being limited to around 256:

$ pfiles $(pgrep -u $LOGNAME thunderbird-bin) | tail

The open files are predominantly the local cached file of the IMAP folder.

Some internal discussion found the following workaround, hurrah!

$ ulimit -n 2048
$ export LD_PRELOAD_32=/usr/lib/extendedFILE.so.1
$ thunderbird

Essentially it’s hitting the 32-Bit stdio 256 File-Descriptors Limitation. See the man pages on extendedFILE(5).

Solaris bug logged (6955102) – though I’m not sure whether the fix is to have the workaround in the Solaris start scripts or work with the Thunderbird developers to have it handle FDs better.

How people actually develop – 70/20/10 rule

Many moons ago Sun introduced SunTOPS (Sun’s Talent Optimization System) for development planning. Included in the notes, sadly I no longer have a copy, was this intriguing diagram:

I’ve since found what I believe to be the underlying research behind this, it’s the 70/20/10 model. There is a reference to it in the Princeton University Learning Process.

I’ve not found a reference to the idea that development plans usually turn this in its head – in other words, we incorrectly assume that development is 70% training, 20% learning from others, 10% job experience. If anyone knows where this came from please let me know.

According to Princeton:

70/20/10 learning concept was developed by Morgan McCall, Robert W. Eichinger, and Michael M. Lombardo at the Center for Creative Leadership and is specifically mentioned in The Career Architect Development Planner 3rd edition by Michael M. Lombardo and Robert W. Eichinger.

The seriously chunky and expensive The Career Architect Development Planner isn’t in my local library or searchable on-line. I’d love to take a peek

NIS+ is now NIS-

As already blogged by Chris,

With the push of this feature into Solaris:

6874309 Remove NIS+ from Solaris
PSARC/2009/530 Removal of NIS+

a bit of Solaris history is made. The namespace that was to replace NIS (YP) has been survived by the system it was to replace.

My first day at Sun Microsystems in 1995 was the day I first touched NIS+ having geekily read and re-read the white papers prior to my joining. Chris was already service’s recognised NIS+ expert world-wide so I was in excellent company as tentatively typed my first niscat.

To witness its removal is eerie to the least. The irony is that I’m now the manager responsible for the team that just removed something that’s been been a golden thread running through my career at Sun for what is almost 15 years. I’ll raise a virtual glass to all the people who’ve worked on it and with it – cheers. As Chris said, “it was fun”.

Working from home? Make sure you’re adding unique value

Interesting article in the Guardian on crowdsourcing – companies using large numbers of distributed people rather than technology to solve problems.

While not directly related to working from home it struck a chord with me. I commented on Rands’s article on “The Pond”. One of the things I wrote was:

“One concern whether remote or not is that if my work is so precisely defined then the company may decide to contract the work elsewhere, possibly off-shore. Human nature means that the unquantifiable work that keeps me valued is so much more visible in the pond.”

Occasionally I entertain the idea of working remotely so that I can live where I want to live and all the other good stuff around home working. The article was a useful reality check and had me thinking about where I and the people I work with add value.

Rules, incentives and Barry Schwartz’s call for “practical wisdom” – and what of goals?

Another extraordinary TED presentation, this time by Barry Schwartz on The real crisis? We stopped being wise. It echoes a theme I’m seeing in many areas around how an over-attention to goals, metrics, rules, incentives, etc not only demonstrates a lack of trust by those setting them, it actually encourages the very behaviour they are seeking to avoid.

I’m a little leery about SMART goals. It’s in all the management books of course and as a manager I’m expected to set them but we tend to over-rely on them. As Barry says in his presentation, we need rules, we need incentives but not more and more of them. They have their place but understand the limitations. I also recommend When Goal Setting Goes Bad which discusses the working paper entitled Goals Gone Wild: The Systematic Side Effects of Over-Prescribing Goal Setting.

Barry talks about how any scheme of incentives can be subverted and calls for wisdom and ethics to be nurtured. There’s a Wired interview with Barry where he says:

“When you rely on incentives, you undermine virtues. Then when you discover that you actually need people who want to do the right thing, those people don’t exist because you’ve crushed anyone’s desire to do the right thing with all these incentives. And if you bring in a new set of people to replace them — virtuous, moral people who want to do the right thing — and they’re subjected to the same set of incentives, they’re going to become just like the people they replaced.

I’m not talking about getting rid of incentives; people have to make a living. But people need to understand that rules and incentives aren’t enough…. The more rules and incentives you have, the less wisdom you will have. There needs to be room left on the one hand to nurture in people the desire to do the right thing and on the other hand to give them the tools so that they’ll know what the right thing is. This incredible pressure to increase payoffs is an obstacle to doing the right thing. You will never be able to create a system of incentives that rewards people for doing the right thing. The system of incentives may start out that way, but very quickly clever people will find ways to … game it.”

Amen to that.

Recommending “Managing Humans” by by Michael Lopp (aka Rands)

I thoroughly recommend the book “Managing Humans” by Rands, aka Michael Lopp. The pearls of wisdom come thick and fast, it’s an easy read and you’ll find plenty to laugh and cry about.

I’m a manager in Solaris sustaining – essentially we fix bugs in our released versions of Solaris – rather than product development which is more Rands territory. Having said that, there’s lots of commonality between the roles.

It was recommended to me by Dave Walker, a colleague of mine here at Sun UK. It was his tip for the engineer-begat-manager – ie me.

All the chapters are available on-line in Rand’s blog, eg Meeting Creatures, but the book neatly groups them and is handy for dipping in and out of when the mood takes you or the panic sets in

On the theme of book reviews – for those with children I also have to recommend the “Mr Gum” books. Full of nonsense words, mad characters and slapstick humour. It certainly amused me and my children (4 and 7) were laughing out loud. Favourite quotes include this description of Padlock the bear:

“He was a proper fat shaggy rumble-me-tumble sort of a roly-poly flip-flap-flopper of a big brown bear”

I had to read that description out loud several nights in a row.

Please read to your children.

The Knowledge Distortion Field

I’m sure there’s a better term for this – the Knowledge Distortion Field is an observable effect created by those who know something that they are unable to share.

I’ve noticed it more as a manager as I’m privy to more confidential information such as personal details and management decisions. It’s particularly noticeable when you yourself are also aware of the confidential information.

Probing the KDF is a little like the old Black Box game. Asking direct questions isn’t allowed, or at least won’t get you anywhere. Asking indirect questions or observing actions taken allows you to build up a picture.

Why do I mention this? It’s simply something to be aware of no matter which side of the fence you sit. If you are in possession of confidential information be wary of any KDF you generate. If you are seeking that confidential information then quiet observation and indirect probing may help.

If anyone knows a better way of describing this I’d love to hear about it.

Why security bugs are different to other bugs

Way back in 2000 I had the following insight on security bugs and had this weird urge nine years later to share it. I can’t believe it’s original…

Security fixes are different from every other kind of fix. As every good troubleshooter knows, when problems occur something almost invariably has changed. For most bugs it is something like load, configuration and so on which can be undone.

With security bugs it is knowledge that has changed and a security weakness can’t be unlearned by the world at large.

I don’t have insights like that very often