Netsight Blog

So I bought my daughter a cheap little netbook a few months back, and downloaded an image of Ubuntu Netbook Remix (9.04) and copied it to a USB stick (dead simple, just using 'dd' on my Macbook Air to copy the image to the USB stick), plugged it into the Netbook and booted it up.

Wow! It was fantastic, a really great little OS, and amazingly everything seemed to 'just work' including the built in webcam, the wifi, audio etc. Linux has certainly come a long way in out-of-the-box- usability. However after a while noticed a very odd problem with the mouse pointer seeming to stop working. Well the pointer move about, but when you hovered over something it didn't notice and clicking didn't work. You could click on an app to launch it, but then you couldn't click anything in that app. Killing X so that gdm or whatever it is restarted X and logging back in seemed to fix it. A bit annoying.

So I thought with the release of the latest and greatest Ubuntu (9.10) I thought I'd try that. Now, I don't want to upgrade the current installation via the software update tool, as I want to check everything still works. I want to book a USB stick and run it in 'live' mode and just make sure all is OK.

Should be easy right? Go to the main site to get the latest Ubuntu Netbook Remix and selected a mirror and downloaded the image. Thinking I knew what I was doing (same as last time right?) I just copied the file to the USB stick using dd and stuck it into the netbook to boot it. No joy. The USB stick was ignored and it booted from hard disk. I double checked the boot options and it seemed all OK with the USB stick listed first. Still no joy.

After a bit of ferreting around online I realised that the file I'd downloaded was an .iso not an .img file. ISOs are images specifically designed for CDs. IMG files are designed to burn to USB stick. It was 15 years since I cut my teeth on BSD Unix and can reboot and rebuilt a server on the other side of the world with nothing more than a serial cable and some chewing gum. Yet I can't tell you the difference between an ISO and an IMG file. Probably something to do with boot sectors.

So I went back to the site above and saw the great big step 2. 'Create your UNR flash drive'. Ahh.. so I should have read the instructions in the first place (I'm a geek, what do you expect?!). Ahh problem. There are instructions for Linux and those for Windows. But nothing for OSX. Bummer. So I followed a link to the detailed manual instructions and they seemed a bit incomplete and contradictory. I wasn't sure if they were applicable to this version of Ubuntu or not as they kept referring to past versions in the notes. The comments all then had about a dozen different other ways to try and skin the same cat.

The basic process seems to be:

1. Download the ISO file
2. Mount the ISO file
3. Format a USB stick to FAT
4. Download some other files from online and copy to the stick
5. Use 'syslinux' to install a boot record on the stick
6. Make the stick partition bootable in fdisk
7. Copy over all of the files from the mounted ISO to the USB stick (different comments gave different instructions as to exactly which and where).
8. Rename some of the files on the USB stick.
9. Unmount USB stick, put in netbook, boot and install.

But could I get it to work?! No. A few times I got the Ubuntu installer screen at which I could select my language and keyboard layout, but then after that - Nothing. Just hung. Probably looking for some files I didn't have.

So... an open question to Canonical and the Ubuntu people in general:

Why the hell do you distribute the image for Ubuntu Netbook Remix as an ISO?! 99.9% of Netbooks don't have CD-ROM drives. 99.9% of people who try to download it to use it will need to create a bootable USB stick using either the programs you point to for Windows or Linux, or manually for other OSes. WHY NOT JUST DISTRIBUTE A .IMG FILE IN THE FIRST PLACE?!

THAT IS WHAT YOU DID WITH THE LAST RELEASE AND IT WORKED GREAT. WHY ARE YOU MAKING ME THINK?!

Like many digital agencies, Netsight has long held the view that speculative design work is damaging to us, to the client, and to the industry as a whole. But not so long ago we were talked into producing a creative pitch, our first in several years, and the process turned out to be of questionable value for all involved. So I'm taking the opportunity to remind myself (and anyone else who might be reading) why speculative design work is just so counter-productive.

A piece of speculative design work is an uninformed sales pitch, not an informed response to real requirements.
The sole point of a speculative design pitch is to sell your services as a designer. Therefore, to be successful, that piece of design is most likely going to show the client what they want to see, not what they need to see. So begins the process of a designer second guessing what the client will want to see. No matter how good the brief is, the pitch design will be deliberately indulgent (with that ever-elusive "visual impact" applied in spades) and will not necessarily serve the client's users in anything other than a superficial way. In the normal design process, a good designer will ask difficult questions of the brief and make suggestions which may challenge convention. He or she will do this as part of an iterative, investigative process. But these things rarely go down well in a speculative pitch as there is a high probability that, to begin with, they will land wide of the mark - and the creative pitch is a one-shot deal. Often the worst outcome of the speculative design pitch is that the piece of pitch creative is actually used, straight off the bat, as the final production design - without ever going through that process of discovery, collaboration and refinement which is essential to great design. This is something that many clients don't consider.

Speculative design is wasteful and ends up costing everybody money.
If the worst outcome is that a design born out of a sales pitch is adopted as the final design, then the only alternative is that the speculative pitch concept is largely discarded and a "proper" IA and design process is then followed to arrive at the final design. Clearly then a significant portion of work done to produce the sales pitch design has just been wasted, which will inevitably be paid for out of the overall project budget. And of course, agencies which regularly engage in speculative design work have to pad out their project rates overall to pay for the time they lose in pitches for projects they don't win. This is bad for everyone.

I do understand why it is so tempting for a client to request creative pitches. Hold a beauty parade, and you'll get 5 or more completely different responses to a brief which you can consider and choose between without having to pay a penny. You might even be tempted to take your favourite bits and pieces from the different pitches and instruct the appointed agency to combine them into one (frequently horrific) chimera design...

So what is the alternative? Of course the obvious answer is to look at your potential agency's past work and case studies, and where appropriate talk to their previous clients to find out how well they fulfilled the brief.

But there is another option, and it's one that we're using very effectively with more and more clients. We call it the "Pre-production Phase", and it's simply a small package of requirements analysis, IA and design prototyping which stands alone from the main project. This approach is not uncommon but is often mistakenly thought of as a "paid pitch" - although to think of it as this is to miss the point. The pre-production phase aims to understand the client's values, their users, their competition, their business processes, their objectives and criteria for success - and to distil these down into a meaningful proposition which is independent of the technology that would be used in production (in our case, Plone). Both the process and the deliverables of the pre-production phase have intrinsic value to the client, and can be considered in isolation and in advance of the main bulk of the project.

The pre-production phase might typically cost something in the region of 10% of the total project budget, but this is in no way wasted investment. Virtually all of the work done in the pre-production phase would have been needed in the main build phase - it is simply being moved into its own stand-alone work package - and the clarity that the pre-production phase brings to the main project is invaluable, frequently leading to savings elsewhere. This then is surely the ideal solution: reduced risks and lower costs for all parties.

Well, a pretty good flight to the Plone Planning Summit. A long, but so far good journey. The taxi picked me up at 4:30am, and at 10pm I'm finally in San Francisco waiting for the train into town. Every city has its own nuances with their public transport system, but San Francisco has a new one on me: With the BART train you buy a ticket from a machine... sounds simple enough, but when you put in a note, eg. $20 it defaults to a $20 ticket, and you have to hit buttons to *remove* value from the ticket. I needed a $5.30 ticket, so put a $20 note in, and hit 'remove $1' several times, until it got to $15 then stopped and wouldn't go an further. Confused I ejected my $20 bill and looked around and spotted a change machine.

Sticking the 20 into that, I got four $5 bills back. Back to the ticket machine, I put two $5 bills in, and then had to hit 'remove $1' four times, then 'remove 5c' button fourteen times until I got to my fare. Wouldn't a numeric keypad be easier?!

Speaking of which, getting cash out the ATM was also an odd one, as it asks you to key in the amount you want to withdrawal, but with two decimal places, ie for $100 you hit one then four zeros. Yet above the entry box it says it only dispenses $20 bills. So why the two decimal places?!

You do have to wonder sometimes if the moon landing conspiracy theorists were on to something!