Adobe is buying Macromedia. Here’s one possibility for how their services could fit together to create a very interesting end-to-end web workflow. This could possibly be what I’ve been waiting for from Adobe for years - the ability to do web-based structured content editing for print and web publications in the same interface.

More on my regular blog.

RAID (redundant array of inexpensive disks) is a series of disk configurations to get more reliability or performance (or both) out of multiple disks. RAID typically consists of two operations - striping and mirroring. Striping involves writing disk blocks in a pattern across multiple drives. This is very fast, because modern controllers can read and write to multiple disks simultaneously, so you essentially double your disk throughput. But you also double your liability, because a failure on either disk causes data to be lost. Mirroring takes care of that by writing exact copies of the same data to multiple disks. RAID-5 is a combination of the two, and is the most common, balancing space efficiency with speed.

Here are a few common configurations. All of these can be extended to larger configurations of more drives.

Each cell tower is a short range transceiver (sometimes disguised as a tree). Usually, if the density is high enough, you’re within range of multiple towers, and when they move, they hand off your signal to the next one. Each of the towers is individually wired to the cell provider’s internal network, which is bridged to the standard phone system.

(Suggested by Gina.)

Here we have a further evolution, spanning multiple rooms by bridging the network access over powerline ethernet. This can work over much greater distances than wifi (although at somewhat slower speeds given the current speed of wireless - that’s still probably okay for most uses), at roughly the same cost. We’ve also split the network conceptually into “office-side", which is wired with a high speed transfer gigabit subnetwork and “entertainment-side", used for low-bandwidth external facing applications (PVR, game machine) that currently have little comparative need or capability to transfer large files to and from the rest of the home network. This also gives you the ability to easily add another NAT / firewall box in between the two so that when someone finally gets around to putting out a worm that propagates through one of the gaming consoles, the rest of your computers are somewhat isolated.

(See part 1 and part 2)

Slightly more complicated this time, we’ve added another file server, a gigabit unmanaged switch on the backplane for fast LAN communication (internet accesss will still be limited by the connection speed to the ISP), and a VOIP phone. We’re still just talking one room (depending on the wifi range).

(See part 1 and part 3)

A very simple single-room wired network plan for a home network. More complicated examples will come in the next few days.

(See part 2 and part 3)

Managing a large CMS or web development installation requires at least a few different conceptual environments. Depending on individual needs, there may be variations on Development, Staging/Test, and Live (Production). Typically, these environments should be distinct and separate from each other, connected by a central source control system through which all updates happen. Making changes by hand will inevitably foul an environment and let it drift out of sync with the master repository.

Running a multi-developer environment is complicated, and there are a large number of concurrency issues to be dealt with. This is a sample CMS environment to support a few developers with three formal environments, in the case where changes are only made on code in files. It only gets more complex from here.

Next, you might want to add support for automated database schema updates, or more different environments. This also doesn’t say anything about testing environments for content editing - this is development only. I’ll cover content editing workflows in a future post.

Interactive websites offer a large number of opportunities for data collection, personalization, and identity profile correlation with other sources. I think, in most cases, this is an intent-neutral technology that’s branded with the goals of those behind the rules.

On the one hand, non-invasive personalization can make the customer experience better, more natural, and tuned to the needs and desires of the individual user. Highly targeted ads, if they’re accurate (which is a big “if"), tend to remove invasive ads for things the user is actually not interested in, as well as decreasing the number of ads overall, because advertisers will pay more for targeted ads.

On the other hand, there are a number of nefarious purposes to which a highly detailed CRM database can be put. Recently, a wrong arrest was made in an arson case based on Safeway affinity card purchases.

There’s constant debate over whether companies should be collecting all this data on users, and whether those practices should be regulated.

There are, of course, many variations. This is one evolution of how a website will get to know users better over time.

With traditional downloads, the server is a bottleneck. With P2P (ala Bittorrent), downloads get better for most as more peers join. (Clarification: better than without P2P for the same number of downloaders.)