Wed, 24 Mar 2010
So today is Ada Lovelace day and we're supposed to "celebrate the achievements of women in technology and science." I don't know many women in science but I do know a few in technology and one in particular seems to go from back breaking task to another with politeness and grace I wish I could muster.
So for my 2010 Lovelace day (and because she'll need all the happy thoughts she can get now she's president of the Perl Foundation) I'm naming Karen Pauley. A long standing member of the perl community who's been involved in getting things done for more years than many people realise. Listing all her achievements would take a LOT of screen space (and annoy the hell out of her) but, to name three, her TPF work, YAPC::EU organisation and involvement in more related FOSS communities than you can shake a stick at are no small matter.
Speaking as someone who's seen her speak over half-a-dozen times, it's easy to see that Karen has a gift when it comes to presenting. Whether it's about technology, business or community its rare to hear her speak and not come out feeling both smarter and entertained, a combination we'd all love to be able to perform.
I've been lucky enough to chat with Karen outside of conferences and I've always come away from our email conversations with a smile and often with an idea of two, it's hard not to when you're speaking with someone who's both intelligent and a remarkable communicator. Karen is an exceptional person who we're lucky to have in the perl world, and I'm very fortunate to be able to call a friend.
Wed, 13 Jan 2010
Many years ago, in the first dotcom boom, I worked for a website performance monitoring company. I was one of the early employees (developer number 3 and sysadmin number 2) and I remember being in a meeting with the company CEO who was telling us about a new pitch we were doing for $SUPERMARKET, they were going to try this new idea of shopping online and then delivering it to your door.
The worst part of it was that they didn't just want monitoring, they wanted a full transaction engine (with some basic OCR), a product I can probably get away with confessing that we didn't have at the time of the sales pitch. We all knew the deal, if we didn't get it life was going to be very hard there for the next six months, so we all knuckled under. The road was long, difficult and uphill in the snow in both directions but eventually we got to the day of the pitch. Which we aced in an astounding display of luck - the new app sometimes got itself in to a little bit of a state if their website had a failure - which it did about 20% of the time. They loved the demo and wanted us to give them full coverage while they did maintenance work. If we pulled it off then we'd pretty much get the deal, none of our competition at the time could match the features, it was just the uptime that was a little worrying.
So we went out and bought a dozen small desktops, monitors and networking kit, installed them all in our spare store room, put some tables and chairs in and had a company meeting. The management were completely open about what was happening, they took questions and then asked how far we'd go to help. We covered the whole weekend from Friday night to Monday morning. Nearly the entire company chipped in, from three letter titles to sales to dev to systems to HR. We had eyes on the machines over the whole period, including when the Solaris admin, the only person to let us down, didn't make his time slot. Out of all the transactions the worst was beans, they had a new version of the code on some of the servers and it'd return very odd results for beans and break the transaction runner in horrible ways. I'll never forget the 4am calls asking what we do when they offer you a lawn-mower instead.
I placed my first ever order online with the $SUPERMARKET yesterday and hopefully it should arrive in the next couple of hours. The interface may have changed and so many of its users take the service for granted that it's a little humbling to realise how much the Internet's changed so very many things. I guess this post's about a combination of things, the best job I ever had (the company was sold in the end to one of it's competitors. I left happy in the knowledge that we ate their lunch until they gave up trying to compete and bought us), how dedicated staff can be in the right environment, why you should push the boundaries of your industry and how sometimes even cans of beans can be exciting.
I had to put a single can in the order to complete the circle. Here's to hoping they don't charge me for a lawn-mower.
Update: They didn't deliver on the night, there was a "problem with the payment" so they took the money out, using the same details and delivered it two nights later. I'll class this one as a draw.
Wed, 30 Sep 2009
It's very easy to become quite blase or even cynical about new technologies but sometimes a project grabs your attention and coaxes out a "that's very cool", the real time augmented Google Earth had that effect on me.
How long will it be before you can roll back an overlay by X weeks and see what happened in that game last Thursday or check the traffic on your new route at 7am on every Friday for a couple of weeks?
Wed, 08 Apr 2009
Is it just me or does everybody seem to go and buy a new laptop just before they leave their current job? Is it the techie version of buying new work shoes?
I've been quiet on the PiP front for a while now. While the day to day stuff has kept me busy it hasn't exactly helped move my career along, I spend most of my time doing things I already know how to do but with a little twist on them.
In an attempt to stop myself from further stagnation I've put a short list of goals below that should be my bare minimum for the next three months.
- Attend one technical conference / training course
- Attend two technical user group meetings
- Read and review 3 books.
- Write and publish a Perl modules.
- Write and package a Python or Ruby module
- Write and publish a technical article.
- Create a personal Debian repo
- Create 3 Debian packages, at least one of which should contain other peoples code.
- Write 30 blog posts - at least 15 of which should be technical.
I'll do a follow up post at the end of the quarter so see how far I got.
Mon, 09 Mar 2009
If someone in the audience asks a question that you plan on answering then please repeat it, with your own wording, before you respond.
This gets us two things - the person asking probably won't have a mic so not everyone will hear what he said, they will when you repeat it. Secondly - by repeating with your own phrasing - you'll get basic confirmation that you understand what's being asked rather than answering the wrong question; which wastes everyones time and leaves the asker frustrated.
Mon, 05 Jan 2009
Today has been one of those death by a thousand cut days. We did a migration first thing in the morning (I'm not supposed to be awake at 6am unless it's from a really late night) and while all the big bits were planned and moved successfully the work list was missing enough little pieces to make the rest of the day very annoying.
What made the work a lot harder was that the changes had to be made through a web front end that abstracted about 20 seconds of vim in to four minutes of clicking buttons that were never in the same place twice. It's been a while since I've had to bulk make production changes using this kind of interface so I was freshly amazed at how awful it was.
First of all was the time it took. The average change was about 8 mouse clicks, most of them on different pages, across a slow application that was working with a very large (for it) dataset. Second was the lack of a safety net. I had to do full copy and pastes to somewhere safe for each thing I wanted to change before changing it. It may not sound like much but if you come from the land of version control and diffing changes then it just feels so risky. And if you don't then I suggest you start learning one. Instead I had to rely on some hastily written post check scripts that confirmed the changes were correct when publicly viewed. We'd normally write these as a double check but without version control they become the single safeguard. Which were only effective after the change was made, which is better than nothing I suppose...
Fri, 02 Jan 2009
Near the middle of December I lost a very dear, and constant, companion - my Sony Vaio 'some model number or other'. After nearly five years the laptop stopped charging and it wasn't worth paying for the repairs. I put off getting a replacement for as long as I could but while I had the work laptop as a standby I needed a machine I could treat as my own. Something outside the company security policy. Something I could install lots of applications and languages I'll only ever look at once on. So I bit the bullet and bought myself a Samsung NC10.
It's not exactly been a long time since I bought it so I've hardly stressed the machine too but first impressions are very favourable. Battery life on wireless is a good four-six hours (depending on what else I'm doing). The keyboard is much nicer than the Asus ePC I used for about ten minutes before cramping my hands up and the screens actually very usable. It'll never replace a dual monitor setup but it's fine for writing little scripts, web browsing and reading my email.
I've got a 1GB memory upgrade on order (it can only take 2GB) and then I'll see if I can make VMWare play nice without killing the battery.
Tue, 02 Sep 2008
Like most of the techy part of the Internet I dutifully downloaded Google Chrome today and had a little play around. And just like all those other people I'm going to write about it. The difference is I'm very ambivalent about the whole thing.
Chrome seems nice enough. It's quick, works with all the websites I've tried so far and does have a killer feature - the task manager. Finally breaking tabs out in to their own sandbox is an idea whos time should have come years ago. Being able to see which sites are doing hugely evil things with my memory is a wonderful thing. I'm also inappropriately happy with the in-page search showing how many matches it found.
Unfortunately that's about it. While the minimal design and streamlined core functionality are lovely, these days I'm used to my extensions - the web developer toolbar, YSlow and the work flow changing Ubiquity are just too useful for me to give up.
It's not just the fact that these extensions
are missing that puts me off, it's the lack of how to write custom
extensions, searches etc. that feels wrong. Firefox is a platform as
much as a web browser. Using Chrome what is the command line for pulling
out the memory usage for the currently opened tabs? Do I need to
screen scrape a running
about:memory? I can't help but think
they'd have three Firefox versions ready for download by now.
So will I be moving over to the new and shiny? Not yet. As useful as the broken out tabs are I need more functionality than Chrome can give me, so while I might use it for some day to day surfing it's no where near ready for me as a developer. Although I;m guessing they never intended for it to be.
Fri, 25 Jul 2008
And not just on the waist line. Not the worst sysadmin appreciation day ever.
I do now have this feeling of dread as I wait for the other shoe to drop though.
Wed, 16 Jan 2008
Ted Neward is the latest person to get linked to in the ongoing campaign to prove that parrot isn't dead, sleeping or pining for the fjords (sorry, couldn't resist). While chromatic rebuffs some of Teds points I can't help but think something is missing - a little outside perspective.
chromatic rightfully points out that the project isn't dead (and has actually been pretty visible in the perl sphere since the start of the year) but look at it from more of an outsiders angle - unless you are already in the perl community it's not obviously moving. A release every month is lovely and all but the announcements are apparently not hitting all the right places if people like Ted (who seems to at least keep an eye on a number of different projects and technologies) haven't noticed. And if he's not seen them then the unwashed masses of other developers won't have either. Is this a problem? Depends on what you want from the project. Mindshare is a wonderful thing but introducing people to a technology before it's ready can destroy its chances of success.
PS I wanted to be glib and just say "motion != progress" but that seems unfair considering the amount of time people are investing.
Tue, 15 Jan 2008
This weeks three things are -
- MySQL 5.1(.20+) can log errors via syslog (finally)
- IBM Blades run quite well despite being very wet. (don't ask)
- Amazon Prime is too helpful. (Wooo individual book orders)
Copyright © 2000-2014 Dean Wilson :: RSS Feed