Monday - 24 July 2000
Nessus Security Scanner. A big mahalo to Shawn Wallbridge (and Dan Bowman for pointing me in his direction). Shawn spent part of his busy Saturday afternoon testing my Coyote Linux firewall by running the Nessus Security Scanner against it. I guess Coyote must be doing a good job because Shawn reported no problems. Thanks again, Shawn.
Speaking of Security Tools. This week's Security Watch column over at InfoWorld lists the various scanners and other detritus that the columnists (and you better believe script kiddies and/or hackers) use to test security. For those of you who have not thought much about security, you should be afraid. Very afraid. These tools can show you why.
Office Sweet. For those of you who want MS Office compatibility but don't want to spring for the big bucks can check out 602Software. I have not tried their software myself but they say they are compatible with Word and Excel. And the cost is right. Free. YMMV.
City and County of the Valley. What? The San Fernando Valley in Los Angeles wants to secede from LA and become its own incorporated city? Strange as this idea may seem to most people familiar with the Southland, the movement is at a point that they are actually doing an inventory of various assets. Right down to the last stop sign. See the Los Angeles Times story here.
While you're over at the Times, take a look at this article on Web Phones (see it here). Is the Palm the wireless internet appliance of the future? Or is it the cell phone? Will either or both exceed the number of online PCs? No one knows for sure but you can read about what some people think will happen.
Outrage of the Day. So you are a megabucks drug company raking in the money with one of your new, cures everything including the common cold drugs. Then, you find out a rival company will be coming out with a generic drug exactly like your cash cow. But selling at a much lower price. So what do you do? Well, you have lunch with your rival and you agree to pay them millions of dollars per month to keep their generic off of the market. Which they accept because they would be making more that way then by actually marketing the drug. Ain't life grand? See the story from the New York Times here.
Aloha!
Top / Home / Monday / Tuesday / Wednesday / Thursday / Friday
Tuesday - 25 July 2000
Let's See What Develops. Polaroid announced today an agreement with Olympus camera to jointly produce a digital camera, called the C-211 Zoom (see a picture of it here) that prints images on Polaroid film. The price? About $800 USD.
I would guess Polaroid needs this camera more than Olympus. Yet, if you search Polaroid's site for the C-211 you won't find anything. On the other hand, Olympus has a whole section (see it here) on the camera. Including, at the time of the announcement, a live feed of the event.
I don't know if the world needs this camera although it is being marketed towards those who need a print of something that you can say has not been altered. That would be the insurance industry and law enforcement, for two examples. But, you can do that now with a $20 Polaroid. Why spend $800 to do the same thing (other than to have a backup digital copy)?
AirFrance. My condolences to the families and friends of the passengers and crew of AirFrance flight AF4590. It's times like these which can help to focus our attention on the important things in life.
Speaking of which, there's lots of things going on at work and at home so I need to get out of here. Lord willing, and creek don't rise, I'll see you later.
Top / Home / Monday / Tuesday / Wednesday / Thursday / Friday
Hump Day Wednesday - 26 July 2000
Flame On! Me thinks a flame war is about to erupt between JHR and Brian. Actually, I think Brian is too civil a person to do that. But who knows? It's interesting reading anyway it plays out. Just remember to wear your flame proof suit when you handle it.
As for me, I'm a late convert to Linux. When I first played with Caldera Linux ver 1.0 things were still very much in an infancy. But under its latest version (OpenLinux 2.4 eDesktop), the install, which is half the battle, is at a point where most people should be successful in getting it running. The next battle is to get comfortable working with the applications that run under Linux. This is the other half of the battle. The part that is still being "won" by Linux. Not its users. But that part will come. When, I can't say. Ask me the next time X-Windows does something totally irrational (which is how I feel sometimes working with this stuff).
Speaking of Linux, Caldera announced yesterday the shipping of its developer preview of the Linux kernel 2.4, KDE 2.0 w/Conqueror, and Java HotSpot VM. See the press release here. It should hit the shelves on 31 July and may also be purchased online for $19.99 USD. Don't ask me if it's worth $19.99 when I think you can get most of this stuff for the cost of a download elsewhere. But I guess it's helpful to have everything in one place. And perhaps, as they say, this is the beginning of the end of not ready for prime-time releases of Linux applications. YMMV.
There's tons of stuff to do today at work so I am out of here.
Aloha!
Top / Home / Monday / Tuesday / Wednesday / Thursday / Friday
Thursday - 27 July 2000
Search Me. I see that fellow Daynoters Tom Syroid and Jonathan Hassell have installed the Atomz search engine. Thus, satisfying Daynote Requirement 2.013.294, Mark IX. And this seems to be the engine of choice for some of the other 'Noters also. To each his own.
On the other hand, I've never felt comfortable using an external service for searches. Perhaps on philosophical grounds more than anything. It just seemed to me that if you are serious about security. If you are serious about controlling what is going on with your site. Then perhaps it is not a Good Thing to blow holes in that security by adding an external service for something that you can add yourself.
Yes, if you don't have CGI access, and you want a search engine, you will probably need to use an external one. But otherwise, other than it's easier to install, why use an external service (not to say that easy to install is not a consideration in our busy lives)?
And yes, did I mention it's easier to install (as I digress for a moment)? Some of you may have noticed that I've switched to ht://Dig. The Perl script I was using earlier, Search_Engine.cgi is still there and works just fine (you can access it by using http://www.seto.org/search.html). But at some point, it will begin to slow down since it does not create an index. In other words, it does a live search of all html documents every time you search for something. This can be a sloow process as you exceed the practical limit of about 500 pages. This is opposed to how most other search engines work. That is, most others create an index of non-common words in your html documents (actually, any files you tell it to index). And when you do your search, what is actually being searched is the index.
While index searchers solve the performance problem by not having to do a real-time search of all documents, they introduce some other problems.
First, the index needs to be updated periodically. Preferably, just after you've done an update. Otherwise, if someone were looking for something, the index would not have any pointers to the new information.
And secondly, the index typically does not include every word in the original document (otherwise, what you would be doing is a pseudo real-time search, something we've seen can cause a performance degradation). And further, the index typically indexes only the first X-bytes of information of the original document. This is a further aide in reducing the size of the index. Obviously, if what you are looking for is at the end of a large document, it is possible that the indexer will miss it. The alternative (i.e., tell the indexer to scan a larger portion of the document), creates its own problem in the sense that the index file will become larger, thus taking up even more of your precious disk space, and also slows down the search process itself. But I digress.
So what are the problems with using an external site? Well, the external site will be logging all of your readers, their IP addresses, what search terms they used, and perhaps the results of those searches. Is this a concern to you? If not, it may be to your readers. Do you know what the privacy policy is of the external site? Do they follow it? And have you informed your readers what that policy is? Hmmm? Thought not.
The service may also be indexing everything on your site (and the data held within). Every file. Every password. Every Quicken folder. Every spreadsheet that you used to figure your taxes. Every mp3 file snagged through Naptser. Every jpeg file of you taken when you were a baby. Everything. That's its job. And it does it well. While indexers are not the only ones trawling through your site (bots do it all the time), why invite them in?
Third, if the connection to the external site goes down, or the site itself goes down, you will have no search capability. Even more frustrating to me is when the service and connection is up, but due to heavy traffic along the way, or at the site, performance slows to a crawl. Normally, if it's dead, you just time out and move on. But if it's up, but running like molasses on a winter morning, then you just sit there. And wait.
And fourth, most of these external sites will be using an indexer. Thus, how often they update that index is important (for the reason given earlier). And something you may not have any control over.
And so, we come to the bottom line. Control. Is it important to you to control what, when, and how your site is indexed? Or is it more important to save some (okay, perhaps a lot) of time by adding some bits of html to your page as opposed to adding and configuring tons of Perl scripts? The answer for each individual will be different. But when you decide, be aware of the costs and benefits. YMMV.
Surly Bastard Chronicles. For those people who just have too much time on your hands, go check out the SB Chronicles here. WARNING! This is not for the weak of heart or mind.
Top / Home / Monday / Tuesday / Wednesday / Thursday / Friday
Aloha Friday - 28 July 2000
It's Friday!
Server Side Includes. The look of these pages will change a little starting next week. In line with what several of the other Daynoters are doing, I'm creating a header and footer. By that I mean, what you will see is the combination of three different files. A header page (header.txt), the body (j20000804.shtml), and the footer (footer.txt). Through the magic of Server Side Includes (SSI), all will be combined at the server and then displayed as a whole by your browser. This is a little different from how others may be doing it but this way seemed the cleanest and most universal way to me (as opposed to using ASP or a Perl script).
First, lets take a brief look at what SSI is. After that we can go into why anyone would want to use them. For those who want to get a little deeper into SSI, feel free to take a look at Chapter 5 of CGI Programming on the World Wide Web By Shishir Gundavaram (see it here from O'Reilly).
What is SSI? SSI has been described as a way of creating an html document by allowing the server to cut and paste it together. In one sense this is true. But in a broader sense, SSI dynamically creates the page based on the commands you embed in it. This means not only can you cut and paste from different existing documents (using the <!--#include file="foofile.txt" --> command) but also dynamically display the date and time, the link you used to get to the page, what your IP number is, the name of the file you are reading, and what browser and operating system you are using. There are other commands, but you get the idea.
So, why would you want to use it? What I am using SSI for is to allow me to include boiler plate type information at the top and bottom of the page. While at the same time, giving me the option of updating this information, for all pages that call it. All at once! No more manually going through a years worth of posts to update something (if this sounds a lot like using Cascading Style Sheets, another feature I use on these pages, you are absolutely right).
So, for example, say I need to update my email address because I've moved my mail account. No problem. Edit the header.txt page. And henceforth, whenever anyone reads a page that includes the header, voila!, the new email address will be displayed.
Okay, but surely there must be some downsides? Very perceptive of you. First, your server must support the use of SSI. Not all do because of possible security risks (you can arbitrarily execute programs from html pages). And second, even if enabled, server performance is degraded since it now has to read the commands in, execute the commands, and then assemble the page for your browser to display.
As always, only you can decide which path is the best for you. But now, you at least have one more tool than perhaps you had before. Class dismissed.
Ill Winds. Hurricane Daniel (no, not Dan Bowman's son) is headed in our general direction. Current sustained winds are 90 kt (103mph or 191kph) with gusts to 110 kt (127mph/235kph). The current predicted storm track (see image here) has it moving just north of the islands on Monday. It is still about a thousand miles away at this time and who knows what will happen between now and then. But it would be very unusual for a hurricane generated so far from here actually making it to our neck of the woods. What usually happens is that it looses strength as it gets into cooler waters. But then, who knows. If you don't see a post on Monday...
Have a Great Weekend Everyone! - Aloha!
© 2000 Daniel K. Seto. All rights reserved.
