Daniel’s Posterous

"Ultimately somebody should design a filesystem explicitly for use as an interchange format and license it for free, then everyone can deal with their internal FSes and do what makes the most sense for their OSes and markets."

Louis Gerbarg makes so much sense here. What we need to recognise is that not all disks are equal.

Internal drives are only ever really used by your OS. If the disk is in a machine stick a layer of software abstraction on top such as Samba and boom - we've got interoperability. (Remember - we live in the future now - clock cycles and RAM are cheap as chips, if you excuse the pun). The whizz bang is only needed by the OS isn't it?

It's the plethora of little flash drives that have superceded the floppy that need to be truly interoperable, and the simplicity of One Filesystem to Rule Them All (all being portable drives in this instance, not magic rings) really appeals to my inner geek. As Louis points out, it will free up everyone to do what makes most sense with their internal FSes without worry and this can only be a good thing. It just makes sense.

Now think of the other benefits. I was really interested to discover Apple had the same number of engineers working on HFS, HFS+, UFS, NFS, WebDAV, FAT, and NTFS combined as a University team working on developing a new FS. Now this admittedly says something about the number of engineers at Apple (what special productivity sauce is Steve feeding them?) - but what would these undoubtedly talented engineers have been able to accomplish if freed up from what is effectively donkey work? How many equally talented engineers are doing similar donkey work at SUN, Microsoft and for open source libraries? Isn't this sort of problem one of Open Source's biggest selling points? ZFS is open source, and licensing issues scuppered it for Apple.

We strive hard to stay DRY in development. This a reflection of working with small teams of one or more and needing to target the hours we have wisely. So we strive to cut out donkey work. Is this a concern the Big Boys don't have to worry about? I think not.

In other news, a universal mobile phone charger has been mandated this week. This has been reported everywhere and it's a good thing. What is really interesting about this though is the numbers on what this actually means. We're talking Big Numbers. For example, according to the GSMA 51,000 tonnes of redundant chargers are manufactured every year. How much carbon and oil does that represent? How much human endeavour has been spent on repeatedly designing mobile phone chargers?

The thing here is that it's taken a 3rd party body to mandate this. Most manufacturers are jumping on the bandwagon of what is ultimately just a recommendation. Someone said "Why not all do it this way?" and lo! they all will. Definitely a good thing. But why couldn't these manufacturers do this on their own?

Moving away from file systems and phone chargers, how many other examples of massive redundancy are occurring in industry? How many talented engineers are spending their time on donkey work? There's an oft cited example that the entire Wikipedia project represents the same amount of human endeavour that Americans spend watching commercials in a single weekend. That is NOT a good thing.

Then again, commercial-watching Americans are not talented software engineers. Not all of them anyway.

How many good things are we replicating away for the sake of corporate fear, licensing issues and ultimately ego?