[#bnf]: “Backus Normal Form” (“BNF”), a.k.a. Backus-Naur Form

BNF is a method of being able to define terms and then describe how those terms may interact. This is often called a “meta-language”. (A language would often have actual definitions, whereas the BNF specification is focused more about allowing terms to be defined and handled.)

As an example of how BNF differs from an implementation of BNF: BNF would commonly be used to specify that a program accepts an option as a command line parameter. BNF would commonly be used to describe the syntax of such a command line parameter, such as stating that single-letter command line parameters may start with a hyphen or a slash (which is common in MS-DOS), while multi-character parameters may start with two hyphens before other letters (which is common in Unix). Other portions of documentation may focus more on describing the effect of each parameter, while the BNF portion describes more details about how the parameters are described.

In addition to command line parameters, communications protocols are another example where BNF (or variations of BNF) may often be seen. The effect of certain commands will be described outside of BNF, but BNF describes the syntax of the commands.

For more details about how this is more commonly used, see Extended BNF and/or Augmented BNF.

[#baseline]: baseline

The term “baseline” refers to a record that is made, and can be preserved so that the record can later be compared with newer situations. For instance, a baseline recording network performance can include some numbers that indicated how quickly equipment responded. Later, if the network seems slow, people can re-measure the network and then compare to an older baseline. Such measurements can help determine how much the network has really slowed down (and may help to distinguish from some other causes of apparent slowness, such as perception when compared to newer technology).

In Microsoft Windows: there may be a program called “Performance Monitor” (which may be in the “Administrative Tools” Control Panel applet). Creating a new “Data Collector Set” may help to effectively create a baseline.

[#baud]: baud

This appears to actually be a highly technical term, related to how signals are delivered. The following is based on a rough understanding, based on some material that has been taught.

In a nutshell, baud relates to the fluctuations of a signal. So, if a signal is represented by a sine wave, the baud can refer to how many times the signal changes direction. When the sine wave stops growing and starts decreasing, that is one baud. When the sine wave stops shrinking and starts increasing, that is one baud. So when a sine wave is growing, and then that sine wave switches to shrinking, and then that sine wave switches to growing again, then that is two baud.

With the creation of the first dial-up modems (which were devices that allowed computers to communicate using a telephone line), the computers used baud to communicate. Each bit was represented by either keeping the sine wave at its present location, or switching the sine wave to the other location. In other words, something like this:

  • If the sine wave was at a high location, and remained high: zero change. Bit is zero.
  • If the sign wave was at a high location, but then dropped: change. Bit is one.
  • If the sine wave was at a low location, and then rose to a high location: Change occured. Bit is one.
  • If the sign wave was at a low location, and remained low: zero chang. Bit is zero.

Using that old method, modems would transmit one bit per potential change in voltage level.

Baud measures how many changes do occur, or could potentially occur. This is measured in changes (actual or potential) per second. So a 300 baud communication involved 300 potential changes every seconad.

Eventually, modems began to support a faster baud rate, meaning that the sine wave could potentially change more frequently. The modems still communicated at the rate of one bit for every potential change in the direction of the sine wave. Since there were more potential changes in the sine wave each second, there was more bits per second. Modem manufacturers marketed/promoted the new equipment supporting higher baud rates.

Up until this point in time, many people thought that baud rate was the same thing as the number of bits per second. Well, up until this time, those numbers were always the same, so it was a sensible thing to think.

However, later modems supported more advanced ways of transmitting bits. The modems would not only pay attention to when the sine wave changed direction (either growing in voltage or shrinking in voltage), but would also pay attention to how quickly the signal was growing or shrinking. A modem could cause some delays in how quickly the signal was growing or shrinking, and those delays could be recognized by the remote modem. This became another way to transmit bits.

The result is that there could now be multiple bits sent within a single growth or reduction of the sine wave, so bits per second were no longer restricted to the baud rate. Knowledgable technicians began to understand that there was a difference between “bits per second” and “baud rate”.

Some related discussion: Appendix C: “baud” vs. “bps”

Eventually, higher speed communications methods led to people using “bps” as a standard way to communicate speed. By the time people started using broadband connections instead of “dial-up” modems, the word “baud” stopped being commonly used. As a result, some technicians probably never received any training that updated their old knowledge from when the word “baud” was treated as meaning the same thing as “bps”. The words are, technically, not identical.

[#bleem]: “Bleem!”

If a company is innocent, then it is innocent, right? How comforting it is to know that it is protected by the law. Err... right?

Wikipedia's article for “Bleem!” summarizes the sad fate of the company known as “Bleem!”. (Like “Yahoo!”, the company name included an exclamation point at the end of the name.)

“Bleem!” is not the only company that made some popular software, only to end up closing its doors after it was officially the “winner” in a lawsuit. Other examples include Connectix and Stac Electronics. However, a key difference is that the lawsuit involving “Bleem!” was filed against the company two “days after Bleem! started taking preorders for their” first product (quoting Wikipedia's article for “Bleem!”). As the quote does say “preorders”, this indicates that the product hadn't even shipped yet. So, while the other companies mentioned had one or more products that sold for years before the lawsuit, there's no indication that “Bleem!” had any decent period of time to obtain wealth from the investment that was made when creating the product.

BNF
BNF is an abbreviation for “Backus-Naur Form”.
[#brdcstad]: broadcast address
No broadcast addresses in IPv6
For IPv6, RFC 4291: “IPv6 Adressing Architecture”: page 2 states, “There are no broadcast addresses in IPv6, their function being superseded by multicast addresses.” For equivilent functionality, see Glossary Entry: multicast, RFC 4291: “IPv6 Adressing Architecture”, section 2.7.1: “Pre-Defined Multicast Addresses” which specifies FF02::1 and FF01::1. These addresses perform the same task, but are not called “broadcast addresses” because, unlike IPv4, they don't require special support that differs from multicast support.
broadcast address in IPv4

Each subnet and network has a broadcast address. For modern networks, this broadcast address is (overwhelmingly) generally the last address. Therefore, the last address of the 0.0.0.0/0 address, which is the address 255.255.255.255, may be viewed as the broadcast address for the entire Internet.

For every (sub)network, there is a single address (which on modern networks is most commonly the last address) which is treated as a “broadcast address”. Information send to that network will then be “broadcast traffic” which, in theory, gets received and processed by every device on the entire (sub)network. However, routers tend not to forward such “broadcast” traffic.

The most famous IPv4 broadcast address is the last address of the entire IPv4 Internet.

(Some information has been moved from this glossary entry to the section on broadcast addresses.)

Ethernet Broadcast

See: Broadcast traffic.

[#brdcstdm]: broadcast domain

A broadcast domain refers to all of the devices that will receive a “broadcast” communication. If an IPv4 packet is sent to the universal “broadcast” address, or to a subnet's “broadcast” address, then the traffic is received by all devices (on the whole network, or on the specific subnet). A router typically does not forward on broadcast traffic, while a switch does. So the “broadcast domain” ends up being all of the equipment connected (directly, or more likely, indirectly through switches) to a single port of a router. Devices plugged into a different port of the router will not receive the broadcast traffic, and so they are part of a different broadcast domain. Since all routers (typically) perform this function, each router ends up creating separating devices into different broadcast domains.

In normal/working/optimal conditions, separating traffic into multiple broadcast domains will reduce how many devices end up seeing an IPv4 DHCP broadcast. In the unfortunate condition of a broadcast storm, having different broadcast domains can help by localizing the problem, so that less equipment is being overwhelmed by the extra broadcast traffic. (However, since the router itself is part of the affected broadcast domain, this could potentially affect devices in other broadcast domains that rely on a fully functional router.)

[#brdcstrm]: broadcast storm

If there is a switching loop, then a broadcast may be sent out multiple ports and end up taking both paths on a switching loop. However, even more condemningly, if the received information then gets re-broadcast again (perhaps each time the traffic is received, on both ports that receive a copy of the traffic), then another copy of the traffic may be generated. The result can be that the traffic gets multiplied very quickly. The result is a “storm” of information. (The term “storm” is used similar to the idea of a “flood” of information.)

This may cause hardware failure. Hopefully this is simply a failure in the hardware's ability to keep up with delivering traffic, and the situation is remedied when the hardware is power cycled. However, such a fix will generally be temporary, as the situation may be prone to re-occur until the actual cause is remedied.