A quick link to a small post I wrote over on LinkedIn on growing agility into an organisation instead of scaling. Includes a link to a much better read than something that I could have written. Go have a look: https://www.linkedin.com/pulse/grow-agile-dont-scale-ramin-miraftabi
A day for joy! Rationality won and the EU rejected software patents. However, EFFI has already stated in an interview that some sort of patent legislation should be in place EU-wide to prevent the problems with different patent laws currently in place in EU countries. Count one victory for the small businesses and people.
I recently started playing around with F-Spot, an application for personal photo management for Gnome. The functionality of the program looks promising and might end up being the program we use to manage our photos. After all, we do have enough of them… But don’t get me wrong, F-Spot still needs quite a bit of work before it’ll be usable for the masses.
What interested me and got me to start hacking with the F-Spot code was the ability to export to Original galleries. During the export I noticed one interesting problem with the EXIF data that is saved in the image. Most applications use the EXIF field
DateTime as the time when the picture was taken when the EXIF specification clearly states that it is the
file change date and time. What should instead be used is the attribute
DateTimeOriginal. The attribute
DateTimeDigitized could also be used for images from digital cameras, but with scanned images it won’t work (as it should state when the image was digitized).
F-spot correctly changes the
DateTime of the images output to Original galleries (as the filenames are changed), but I wonder why no other image manipulation program that I’ve used changes it. I do know that I’ll have to revisit the code that creates the Life of Jalo so that the correct attribute is displayed just in case image manipulation programs start working like they’re supposed to.
As a recent victim of backup failure (although not due to tapes failing) I loved the John Cleese video on backup failure (via). Even if it is an ad for LiveVault, go take a look and have a laugh. Why, oh why, can’t we have better systems available for backups? I know that I’m simply too lazy to set up anything complicated for backups. And this is even after I’ve suffered the consequences…
The recent testing of mobile phone viruses and car systems (see F-Secure’s notes on their testing) got me thinking that sometimes limiting the amount of options and choices is a good thing.
The current trend in most devices with any kind of semi-permanent memory (hard-drives, flash memory) and connections to other devices (think iPods, phones, etc.) is that you can store anything on them. With phones I like the idea that I can upload any file and carry it with me. It’s at least better than carrying it on a disk with me since disks aren’t always on me (my work phone is always with me — no I’m not addicted to work ;). But however much I’d like to have a car with a display that has custom backgrounds and such, I’d be more than a little worried if that were possible.
The approach taken by Toyota in their Prius (based on the testing by F-Secure) appears to be that only certain types of files (vCard) is allowed to be uploaded/downloaded through the generally available interfaces. This makes it easy to secure the file transfer and the handling of the file. Allowing generic file transfers would only open up a can of worms in an environment in which software bugs or malicious code can be fatal.
If a car’s systems would allow generic files to be uploaded there is always the risk of bugs in the systems causing some serious problems. While I have the utmost respect for those who program and test such systems, they are only human and such systems are too complex to be given 100% reliability. There is always the possibility of infinite loops, race conditions, or buffer overflows to mention a few problems. Even if the system that allows the uploading of generic files would be sandboxed from the systems that control the cars essential functions there is some communication that takes place between the systems. And thus there is always the possibility of a bug causing havoc.
Do I sound paranoid? Probably, but I’ve learned enough of quality assurance and software to be very skeptical of any larger system being totally bug free. I’ve also experienced first hand some of the problems that errant software glitches and problems between hardware connections can cause in cars. Problems that don’t even show up in the cars logs… (This was a problem in a Golf that I had that caused the car to stall for fractions of a second at high speeds.)
To wrap this ramble up, there is a reason why some electronic devices should not allow just about any kind of file to be stored on them. We do need different devices to take care of different tasks even if many tasks could be handled by a single device. I’d love a Nokia N91, but don’t let it talk to my car.
And as a final disclaimer: I do drive a Toyota, but it’s *ahem* — well, let’s just say that it’s of an older vintage.