* The changing face of your newsletter
Over the next few weeks, you’ll be seeing some changes in this newsletter. Joanie has decided to hand over the co-authorship of this newsletter to another WAN expert, someone with whom we have worked closely in the past and referred to several times in this column. But before we introduce Joanie’s replacement, we want to spend the next few newsletters with both a look back and a look forward.
We published our first newsletter – then called Network World Frame Relay Newsletter – on May 29, 1998, almost seven years ago. The topic was whether frame relay was still being used primarily as a replacement for leased lines, with the clear implication being that the full capabilities of the technology to build fully meshed networks was not being exploited. Instead, the primary reason to move to frame relay was simply to save money.
And while we would never claim that “price doesn’t matter,” this does indeed signal perhaps the most significant change that continues to occur for networks in general. Price erosion, which we keep thinking has bottomed out, continues. Seven years ago, telecommuters were really lucky if they could get ISDN BRI service, and most corporate branch offices were connected at 56K bit/sec or 64K bit/sec. Now, literally as we were writing this issue, a notice arrived from Steve’s cable provider that his cable modem (download) speed has been increased from 3M bit/sec to 5M bit/sec – for the same price.
Interestingly, this still raises a constant question and challenge for networking professionals. On the one hand, there’s the “old school” that remembers when a T-1 line across the country cost tens of thousands of dollars per month. This mentality says that compression and closely monitored traffic management are still of prime importance in terms of spending your money wisely.
On the other hand, a new school of thought has emerged, which believes that since bandwidth is so cheap it’s easier to throw bandwidth at the problem than to try to micromanage its usage.
In our opinion, the answer lies somewhere between the two schools of thought. Neither is entirely right, and neither is entirely wrong. Throwing bandwidth at the problem works in many cases, but, at the same time, the applications tend to expand to use the available bandwidth.
Just as we need more and more powerful PCs to support overly fat applications, we also need the bandwidth to continue to grow to support those fat applications. At the same time, the PC analogy works in that there’s a limit to the needed bandwidth, and the price for that bandwidth is constantly eroding.
