The LangaList

But wait; there's more! Al Brown, another reader, uncovered yet another glitch with Norton AV:

Fred - Here's a hot tip for your readers. I was going nuts trying to figure why it was taking a whole minute for my Quicken 2000 checking program to open and then most of the operations took what seemed to be eternity. My machine uses an AMD K6 - 2 cpu with 126 meg of ram running at 450 MHz.

Well, it seems that Norton antivirus 5.0 (maybe 4) with the latest definitions gets in the way. All you have to do is disable the Auto protect while you are doing your banking and the problem goes away. I did update my files - that was one of the first things I did but it didn't work for me. My 17 Yr old son told me "Why don't you just disable the auto protect?" (Smarty pants :) Don't you hate it when your kids point the obvious solutions?

Check this out: http://www.intuit.com/support/quicken/faqs/win3/5199.html 

I have noticed similar problems with my FrontPage 98 editor - loading the images on a page that I know is ok. Turn off the Auto protect and bingo - back in business.

Enjoy your News letter. Thanks,  Al!

All this amounts to a bad time for Symantec/Norton users; if you have any Symantec/Norton product, it might be worthwhile checking the support sites and message boards to ensure you're up to date.

Click to email this item to a friend
http://www.langa.com/sendit.htm

return to top of page

4) Bits and Bytes Redux

Several issues ago, I wrote about the right and wrong ways various applications calculate online throughput speed. I did this in response to questions from many readers who were confused when, say, their browser reported one speed, but BrowserTune or another speed-measuring app reported a different speed, sometimes at the same time. (See  http://www.langa.com/newsletters/1999/dec-23-99.htm#3 )

To recap: The correct method for reporting online speed is in BITS per second. For example, a 56K modem gets its name because it can operate at maximum speed of 56,000 bits per second--- that's bits, not bytes. But many browsers and speed-measuring software improperly report speeds in BYTES per second. Because there are 8 bits in a byte, a 56Kbit modem can operate at a theoretical maximum of 7K Bytes per second---7KBytes and 56Kbits *are the same speed.* 

Some readers thought this explanation was wrong because of the start, stop, and parity bits that modems may add to the data stream: If a single 8-bit byte also requires a start bit, a stop bit, and a parity bit, then sending that 8-bit byte actually requires 11 bits to flow through the wire--- think of it as three extra bits of baggage that must accompany each byte. 

It gets more complicated because there are actually several different modem data-encoding schemes, including raw data compression, all of which affect how many bits it takes to transmit a byte. Plus, not everyone uses modems--- network-based Cable, DSL, and ISDN connections use *different* schemes.

But here's a surprise: None of that really matters. Here's why:

None of these encoding or compression schemes have anything to do with a connection's rated speed. What they affect is how much of the total bandwidth is available for transmitting actual data. That is, how fast you can move data (words, pictures, whatever) from the net into your system. 

Measuring the final, net throughput automatically takes into account the varying efficiency (or inefficiency) of various encoding and compression schemes. It tells you, *regardless of the type of connection you have,* how long it takes to move the actual data you're trying to get to. That's the only thing that really matters. That's the only measure you really need.

And that's the kind of measure BrowserTune makes. BrowserTune actually takes nine separate measurements: It twice measures the internet's  "latency" or lag between you and each of three different servers, and then measures how long it takes you to download a 35K block of  compression-resistant text from each of the servers. By dividing the exact number of transmitted data bytes by the download time, and subtracting the network lag, BT2K can determine what your actual average delivered throughout is, regardless of your connection type, encoding scheme, or anything else.

In short: It doesn't matter whether you're on a modem, a network, or whatnot--- BT2K will always tell you your actual, delivered throughput.

Other web sites use simpler and less complete versions of the BT2K method. For example, the Microsoft site at http://computingcentral.msn.com/topics/bandwidth/speedtest50.asp uses the exact same basic technique, but runs the test only once, from one server, and makes no allowance for network latency. It's a much rougher test than what BT2K provides, although the basic concepts are the same.

Plus BT2K not only does nine basic tests on three servers; the full-test version of BT2K also lets you test your throughput with 1 MB or even 10MB test files. These longer tests give you a better idea of your connections *sustained* throughput.

I believe BrowserTune is actually the most-accurate web-based speed test there is. I know of no other site, anywhere, that lets you test so many ways, on different servers, with different files sizes, and with built-in latency corrections. And, BT2K is free. 8-)

If you know of a better throughput test, please le me know. But until we know of anything better, check out http://www.browsertune.com/bt2kfast/ for the most accurate throughput tests money *can't* buy!

Click to email this item to a friend
http://www.langa.com/sendit.htm

return to top of page

5) Millennial Heat

In the first issue this year, I  talked about how I'd gotten a ton of mail after I colloquially referred to the last issue of 1999 as the "last of the millennium." In the Jan 3rd issue (see http://www.langa.com/newsletters/jan-3-00.htm#1) I made the point that the common calendars we use are all arbitrary anyway, and the nit-picking about "the millennium doesn't start until 2001" is, well, a little silly.

Although my original use of "millennium" got a lot of mail, my explanation of calendrical silliness got even more, but from the opposite camp!  Here are two sample letters:

Fred: 

A response to your comments in the LangaLetter regarding those who claim the millennium doesn't end until 12-31-00. I've been dying to get this off my chest, and you seem like you might be a sympathetic ear, judging from your comments. 

All those pinhead, know-it-all whiners who claim the millennium didn't just end don't know what they're talking about! The word millennium simply means 1000 years, but that 1000-year period could start or end ANY time. One particular millennium ended the year I was born, in 1956. It started in 957. So if you REALLY wanted to get technical, there is no such thing as THE millennium!

Obviously, the millennium that all of us are talking about is the one that ended in 1999. Evidently, it started in 1000. The millennium before that ended in 999, and evidently IT started in 1 BC. This is where those pinheads would pitch a fit. But I'd like to respond to their screaming by calmly pointing out that nowhere does it say when "the" millennium has to start, so it's up to us to say. And "they" might want it to have started in 1 AD, but there are FAR more of "us" that want it to end in 999, and 1999, etc., so that the millennium change matches the digital rollover from 1000's to 2000's.... so I say We Win, A New Millennium Just Started, and Stop Whining!!!!

And as far as I'm concerned, your comments about the uncertainty of the calendar fit right in here perfectly, too. Thanks for letting me get that off my chest! --- John Ratzlaff

-----

Fred-- I have enjoyed receiving the Langalist, and especially needed the info about Nortion Antivirus 2000, which I just installed. But, for the first time, I'm e-mailing you.... I'm sick to death of folks who've read a few articles and insist that the third millennium won't begin until January 1, 2001, because "there was no Year Zero." The issue isn't really mathematical, but religious. As it happens, the monk Dinoysius Exiguus (Dennis the Little, or Little Denny) was given the calendar reform job in 532 AD (At least, it was 532 AD when he got done!). He's the one who was off by a few years. Denny's motive for reckoning years from the Incarnation of Christ was to ditch the old "Era of Diocletian," which had been in effect for some two and a half centuries. Little Denny didn't like the Emperor Diocletian, who had persecuted Christians. 

A close look at Denny's work shows that he considered the Year One to have started on the January 1st that *followed* the birth of Christ (by Denny's time, that was believed to be December 25). He reckoned the Era of the Incarnation to have begun with the Annunciation to Mary, traditionally March 25. Those nine months leading up to the birth of Christ were part of the "Years of Our Lord," as far as Denny was concerned. So while the concept of a "Year Zero" hadn't been invented yet, the First Millenium did not *begin* with the Year One, but well *before* it! --- David Finster, Senior Minister, Niles Congregational Church