Effective Mac - Using Time Maschine

On Saturday at 08:00am it happend to me. I move/dragged an email into a folder on my Mac and the machine froze. Diagnosis: Catastrophic harddisk failure. 

But hey ... look at the bright side: I got to replace my 160GB disk with a bigger 250GB disk and got to test, if my Time Machine backup really works.

And there is where the problem starts. In an ill-conceived effort to minimize the disk space and the time I need for my backup, I excluded all directories besides my "User" directory. This will allow you to recover single files (and/or mails, photos, etc.), but will make it very hard to recover a machine, since Time Machine (when you run the app) cannot restore all data from a given date.

The better/only way to do it, is to backup everything (including all apps, system folders, libraries) and just exclude a folder (e.g. the download folder), which has recoverable (fast changing, big, volatile) data in it (e.g. my iTunes Library). Basically reverse my backup strategy. This will create Time Machine backups, which include a snapshot of the entire system (including the operating system).

These backups can then be used with the migration assistant to restore the machine with the last backup or even better you boot from the install DVD and use the restore from Time Machine utility to restore to a backup of your choice.

I can tell you, restoring a system file by file is tedious :).

Effective Mac - Open Source Mac

Was looking for an opensource search engine for the Mac and found this interesting website. I am already using half of the apps, but will take a look at the other half too :).

Effective Mac - Growl back to work

Stumbled over James's blog entry about useful software for the Mac and added Growl to my list of software that you need have/use to be effective, but was not able to make it work with X-Chat (AquaChat). Confronted with the option to figure out what is wrong or to try something new, I took the easy way out: I tried Colloquy. This worked out of the box. I know need to find out how to make my Twitter tweeds show up in Growl.

Open Source Business Models - Open-Core Licensing

OH NO, please!!! YABOOSBM!!! Yet another blog entry on open-source business models :).

Ok, let's make it as short and as sweet as possible :).

This morning I was reading Matt Aslett's comment on Andrew Lampitt's blog entry on "Open-Core" licensing (which actually contains a link to a good presentation by MySQL chief Marten Mickos).

What I like about the approach is that it tries to take the emotion out of the discussion. We obviously all know in our head that open-source is about free speech and not free beer, but the heart then still reacts (at least sometimes) disappointed, when companies show that they are not pursuing an open-source strategy for purely philosophical and/or altruistic reasons. In that context some of the hybrid models (e.g. selling closed-source features on top of a open-source core) got a bad reputation and got dismissed as "bait-and-switch" models.

The value of open-source is NOT in the absence of a desire to make money with it. This is actually healthy and will make sure the approach/model with survive in the long-run.

For me an interesting value of open-source is that it redefines the customer-vendor relationship. It creates a very open, honest pay-as-you-go relationship. Customers are not forced to give large sums of money to a vendor up-front and trust the vendor that it will do the right things with this money. Open-source is the ultimate customer empowerment tool (within limits as discussed by Matt Asay).

A very interesting side effect of this is that it allows the CFO to shift money from his capital expense budget to his operational expense budget, which in general makes his business more manageable.

These are good reasons for open-source. And yes, to keep it alive, we have to find ways to make money with it and share the ROI over the entire delivery/value-chain.

Shortening the list of open-source business models in Marten's presentation, I right now see the following main models in the market:

• Open-Core Model - have an open-source core and sell closed-source features on top of it (e.g. SugarCRM)
  
• Dual Licensing Model- one product/project that gets licensed with a viral, GPL-style license and a commercial closed-source license (e.g. MySQL)
  
• Services Models - where you get to download a productized version of an open-source project and pay a fee for the support you get on it

All of these (and variations/combinations of these) are good valid models, which can deliver a lot of value to customers.

It remains to be seen, which one will proof to be the most innovative one.