Forums | Email a Colleague | FAQ

Dr. Website: Serving the Web Developer Community

Forum Categories

This is just a partial listing of our discussion forums, including listings for:


Search for a Host

Frequently Asked Questions (FAQ)

Please peruse our frequently asked questions list below, check out the archive of past columns, read up on the categories that interest you most.

HTML

Forms and Interactive Design Graphics, sound, and video Java and JavaScript Frames Site management Statistics and access checkers


HTML

Dear Dr. Website®: Someone told me that there are sites that check your HTML for coding errors. Where can I find one of these sites?
Because even the most eagle-eyed Web designers can forget closing tags from time to time, the best way to ensure that your HTML is up to code is to run your files through a validator.

You should note that these HTML checkers do only that; they're no substitute for testing how your pages display in several browsers.

A comprehensive list such as Yahoo's list of HTML validation checkers would be a good starting point for you. One well-known HTML validation tool is Weblint (which recently became subscription-based).


Dear Dr. Website®: What could cause the occasional stray text characters and other weird symbols that show up on my Web pages when there's nothing in the source code to indicate anything's there?
You may be inadvertently entering hidden control characters into your Web documents as you type in your word processor or HTML editor. Some HTML editors offer a menu command for stripping out these troublesome characters, or "gremlins," as they're sometimes called. If you're a Mac user, you can strip them out with Bare Bones Software's BBEdit or World Wide Web Weaver, among others. On the Windows side, the Textpad text editor from Helios Software Solutions will help you strip out hidden control characters.

You should note that these HTML checkers do only that; they're no substitute for testing how your pages display in several browsers.

A comprehensive list such as Yahoo's list of HTML validation checkers would be a good starting point for you. Two well-known HTML validation tools are Weblint (which recently became subscription-based) and WebTechs (formerly HAL's HTML Validator).


Dear Dr. Website®: Is there any way to set up table data to retain some semblance of structure when accessed by users whose browsers don't support tables?
One approach you could take would be to produce two sets of Web pages, one optimized for users with browsers that can support tables and a set that will be viewable by text browsers and other browsers that do not support tables. In the long run, this approach -- while successful -- will prove time-consuming and can lead to confusing and intricate page updates and maintenance.

The other solution is to design your tables with standard header and paragraph formatting tags -- such as <H3>, <P>, or <BR> -- built into each table cell. Even if a user's browser fails to recognize table tags, it can maintain link breaks and a certain amount of text layout. This may be a difficult process for more complicated tables, but this will often prove a successful compromise.


Dear Dr. Website®: How do I merge cells in my Netscape tables? I have several mini-headers below my table header that I want to spread across two columns instead of one.
In order to spread a header across two columns, you'll need to specify the COLSPAN attribute in your table for the appropriate cells. A similar attribute is ROWSPAN, which spreads the contents of one cell over several rows. COLSPAN and ROWSPAN can appear in any table cell (<TH> or <TD>) tag. With COLSPAN, you can specify how many columns of the table this cell should span. The default COLSPAN for any cell is 1. The example below spreads Item 2 across two columns:

Item 1 Item 2
Item 3 Item 4 Item 5

<TABLE BORDER=2>
<TR>
<TD>Item 1</TD>
<TD COLSPAN=2>Item Item 2</TD>
</TR>
<TR>
<TD>Item Item 3</TD> <TD>Item Item 4</TD> <TD>Item Item 5</TD>
</TR>
</TABLE>


Dear Dr. Website®:Are there any HTML editors that let you spell-check?
HTML editors that act as add-ons to word processing programs will, of course, let you perform spell-checks--but you're liable to be hampered by the spell-checker hiccuping on every HTML tag. Others, such as Quarterdeck's WebAuthor, an add-in to Word6 for Windows, come with custom dictionaries for recognizing HTML tags. Another solution for this is to add the various HTML terms (such as HREF or EM) to a custom dictionary of correctly spelled words.

A couple of HTML editors that will let you spell-check are Bare Bones Software's BBEdit (for Mac users) and SoftQuad Inc.'s HoTMetaL Pro (for Windows users).
Dear Dr. Website®: I want to include some technical articles on my Web site, but I'm having difficulty replicating the Greek letters that appear in equations. What can I do to create these characters?
The HTML 3.2 specifications include support for the <MATH> tag for enclosing mathematical equations. Until the leading browsers include more support for this tag, though, you're pretty much on your own. The ISO-Latin-1 (ISO 8859-1) character set supported by most browsers includes hardly any Greek letters, so don't expect much from its list of codes for special characters. You'll find character codes for a Greek beta (ß) and mu (µ), but that's about it. A completely HTML-based solution will require support for multiple character sets, and that may be a ways off for some platforms.

There are a couple of workarounds using small inline graphics you could try. If you used LaTeX to write your technical articles, there is a latex2html converter that will convert most LaTeX documents to HTML documents and will convert any math equations or tables into individual inline figures. I also recommend you look for online resources for transparent GIFs of the Greek alphabet, such as the excellent collection the UMass faculty Web site makes available for educational or non-profit organizations. (Commercial groups should contact the site's author before using these images.)


Dear Dr. Website®: How can I redirect visitors to my new page or domain using META tags?
Simply set up your old page something like this (and be careful with the quotes):

<HTML>
<head>
<meta HTTP-EQUIV="Refresh" CONTENT="0;
URL=http://www.newdomain.com/newhomepage.html">
</head>
</HTML>


Forms and interactive design

Dear Dr. Website®: Do you have any suggestions for troubleshooting when a CGI script doesn't work?
After you debug your script, making sure you can execute it from your shell account's command line is the first--but only the first--step toward ensuring that it will execute properly for your Web visitors.

You need to make sure that the file permissions are set up properly for your directory. Improperly set file permissions are among the most common reasons why scripts don't work. You should make sure that the script file itself is readable and executable by your server, as well as other files called by your script. If your image generator uses a database of graphics, make sure that database can be accessed properly. Improperly set file permissions are among the most common reasons why scripts don't work.

There still are a number of other factors that could be responsible for a script failing to execute within your Web page. You might get a better idea of what's going on by examining your server's error log for messages related to your script. For further troubleshooting, it's also always a good idea to call on your server's system administrator for help.

If you obtained your script from an online script resource, you may want to check the site for further detailed information or a FAQ list about how to install your script properly. You can find extensive CGI troubleshooting help online at Selena Sol's Public Domain CGI Script Archive and Resources Library, Matt's Script Archive Frequently Asked Questions page, and The Common Gateway Interface Frequently Asked Questions List.


Dear Dr. Website®: On my site I have several interactive forms where the message is sent to me with a mailto URL instead of being passed to a script. The problem is that when I get the form's data, it's coded as one giant text block with all kinds of extra characters. There are "+" signs instead of spaces, and brackets like "(" and ")" come through as "%28" and "%29". Is there an easy way to strip these characters to make the input easier to read?
The garbled form output you describe has been one drawback for forms with mailto ACTIONs, which are not decoded by the browser before being sent on; another drawback is that some older browsers do not support forms with mailto ACTIONs. Still, this method of returning form data has become prevalent among users who either don't have access to their server's cgi-bin directory or don't have the resources to develop and implement CGI scripts for their forms.

If you use forms with a mailto URL, there are several online shareware utilities (and many online scripts) for stripping these extra characters from the returned data. A Windows program that creates a tab-delimited version of your form data is the Mailto Formatter by Robert Fries. For Macintosh users, there is a Hypercard utility called the InFORMer by Doug Jacobson that strips out extraneous characters. One DOS utility with the same functionality is Convert by Cyber Quail.


Dear Dr. Website®: I need to set up a Web site where people can e-mail information to our company and have it appear on a Web page automatically -- for example, as a link to their message. I've seen this done but can't find an appropriate script.
The situation you describe calls for what is commonly known as threading software. Threading software programs process a file of mail messages into a collection of cross-referenced HTML documents. Each of these HTML documents represents a separate message in the mail archive and each in turn contains links to other documents, so that all the mail messages can be read one after the next or browsed by following links. Threading software also organizes posts according to topic, depending on the contents of the message's subject field.

There are a number of threading packages available both commercially and freely on the Web. One commercial package is net.Thread from net.Genesis Corp. in Cambridge, MA. A popular freeware product that will likely meet your needs is Hypermail.


Dear Dr. Website®: I am building a form with a number of input fields, but unfortunately none of the fields seem to line up with each other. The <ALIGN=CENTER> and other positioning tags don't seem to work with the <INPUT> tags and their attributes. As a result, my HTML documents have input fields scattered all over the page. Is there something I can do to make these forms look less cluttered?
It sounds like you're making the common mistake of using the ALIGN tags for items other than wrapping text. In order to correctly position your form's input fields, you'll have to use tags such as <CENTER> </CENTER> to correctly position them. Another suggestion is to keep most of the field lengths the same. The uniformity of the input field lengths will give your pages a cleaner look, which will help make them look more professional and readable.


Dear Dr. Website®: I am interested in learning how to code a script that allows the same HTML document to be accessed many times, but with different graphics each time it loads. I've seen this used by many of the search engines on the Internet. How can I use this technique on my pages?
Search engines such as InfoSeek and Lycos have used this process to place a different graphical advertising banner on the search results page each time a user conducts a text search; whenever possible, an advertiser related to the subject of the search will have its banner displayed. Although this is a fairly commercial use of this technique, there are many other possible uses. While the context-sensitive approach that the searching sites are employing will require some database integration so that the search keyword can be matched to the appropriate graphic, you should be able to implement a more simple version of this process by either pulling from a random pool of graphics or by rotating through a number of selections.

In either case, what you will need to do is write a script that will grab the next graphic and add it to the page that is being loaded. This will require some familiarity with the CGI scripting and the use of server-side includes. You can consult the NCSA repository of CGI resources for more information.


Dear Dr. Website®: Our site has grown to the point where we're soliciting major advertisers, and we want to be prepared to implement banner advertisements, rotate them and keep track of the responses. Where can I find a program to streamline displaying and tracking banner ads on our site?
There are several commercial products and services for managing advertising for a site, with widely varying pricing scales. Web Elite has a product called Ad Master designed to help Web sites place, rotate and track banner advertisements, including generating content dynamically. The company also offers on its Web site a free 45-day demo of its ad management software, which runs on most Unix platforms.

At the high-priced end, one big-league vendor is NetGravity, whose AdServer software includes automated scheduling techniques based on impressions to maximize hits for advertisers and ad targeting based on keyword entries. AdServer runs on Windows NT and most Unix platforms. A 30-minute online demo of NetGravity's AdServer is available on the company's site.

While both of these services enable your site to retain control over your advertisers' content and ads, you should note that another approach to Web ad campaign management is used by centralized Web advertising services like Double Click and Focalink's SmartBanner, which give advertisers site-by-site comparisons of ad responses.


Dear Dr. Website®: I'm looking for an easy way to include a graphical index on my company's site. How do client-side imagemaps work?
Unlike imagemaps that depend on the server for their processing, client-side imagemaps require no CGI scripting in order to work. With client-side imagemaps, the "hot spots" that users click to jump to new locations are specified in your HTML coding instead. You would use the <MAP> tag, which is supported in the HTML 3.2 draft specifications, to embed a map file, and <AREA> tags to define each hot spot's shape and attributes.

Client-side imagemaps, therefore, provide a good solution for content developers who don't have the time or readily available access to resources for implementing imagemaps. The client-side approach also frees designers from needing access to their site's cgi-bin directory, where interactive scripts are usually stored; depending on your company's security concerns, those access privileges may be granted to only a few users or just to the site's administrators.

The catch is that not all browsers support client-side imagemaps, so they won't work if your visitors are using, for example, a version of Netscape Navigator earlier than 2.0. But if you can ensure that your user audience is standardized on a browser that supports them--for example, on a corporate intranet--then client-side imagemaps are the easy way for you to create a graphical index for your site.

A couple of useful online resources for further information about client-side imagemaps and the syntax to use in your HTML documents can be found at Netscape's Extensions to HTML 3.0 page and Microsoft's HTML Support: Client Side Images page.


Dear Dr. Website®: In which cases do you recommend using imagemaps rather than real links?
As a graphical index to your site, an imagemap can offer the benefits of nonverbal communication. Even if your imagemap contains some labeled text, the graphical portion can help to convey information more quickly. This is similar to the way that the icons used by airport signs and some traffic signals are designed to communicate to a broad, international audience. From a strict design standpoint, the two main caveats to bear in mind are the importance of providing a text-based alternative to allow access for all users, and to make sure that the amount of time it takes your imagemap to display isn't so excessive it drives away readers


Dear Dr. Website®: From time to time I've seen Web sites that offer to e-mail you when they've added updated information on their pages. How can I add this feature to my Web page?
One service that provides this kind of e-mail notification free of charge is NetMind's URL-minder. Users of the service register the URLs they want to be notified of changes on and receive an automatically generated e-mail message when URL-minder detects that the page has changed. Since the service currently searches for updates on a weekly basis only, it may not prove useful for pages that change daily -- but then again, would you really want to bombard your visitors with so many e-mail messages if your pages have more frequently updated content?

NetMind actively pitches its free service to Webmasters, encouraging them to embed a short form on their sites with the URL-minder script for registering a URL. You can use NetMind's sample form code and change the default value in the form to your own URL; all that users need to submit is an e-mail address.

There are a number of additional notification services available for your personal use; a short list--although by no means an exhaustive one--is given here.

Smart BookMarks from First Floor Software (which is basically the same product as Netscape's SmartMarks) is a Web site list manager that automatically creates a clickable icon for each updated page linked to the site. Smart BookMarks runs under Windows and lists for $24.95; a 30-day evaluation copy is available from First Floor's Web site. It's also included in Netscape's $55 Power Pack.

NetBuddy is a freeware Web page notification package that runs under Windows. You can manually enter the names of the Web sites to track or import a list from a bookmarks file; scans for updates can be scheduled as frequently as you like. Updated pages appear on a list of changed pages, and you can configure your browser to load the revised page automatically.

CE Software's WebArranger includes a notification agent for telling you when Web pages are updated. You can drag and drop URLs into a monitoring database and set how frequently the agent should scan for updates. The WebArranger collection of Internet utilities for Macintosh users retails for $100.

Graphics, sound, and video

Dear Dr. Website®: What is the PNG format, and how does it differ from GIF or JPEG?
The Portable Network Graphics (PNG) format, pronounced "ping," emerged after a 1994 patent controversy over the GIF format appeared to threaten GIF's widespread use.

Although GIF is still with us, PNG format development is continuing. PNG has a lot of the same features as GIF, including indexed color palettes, interlacing and transparency (in GIF89a), and lossless compression (unlike JPEG, which sacrifices some image data in reducing file size).

It also has some features that GIF doesn't have, such as support for 48-bit color and an alpha channel for masks--and, as you might expect, its compression scheme is not patented. PNG also automatically adjusts graphics for gamma correction to compensate for cross-platform differences in brightness and contrast, which is a major bonus for designers of Web graphics.

It still remains to be seen whether PNG will become a popular Web graphics standard, given the widespread use of and browser support enjoyed by both GIF and JPEG. Not all image editors or Web browsers support the PNG format, though you might be surprised by how many do. The PNG Home Page has links to lists of image editors and Web browsers that can display PNG images.


Dear Dr. Website®: How can I get rid of the small gap between a left-aligned image and the edge of the browser window?
As you've discovered by now, most browsers will automatically offset the contents of a Web page by several pixels along the window's top and left margins. This can result in a lot of headaches for a designer who tries to position a GIF precisely within the browser window or a background image. Interestingly, the number of pixels your image is offset when viewed by your readers will differ, depending on which browser is used; this can even vary among several versions the same browser. Designer David Siegel offers real-world examples of how these variations in browser versions affect your page's display at Browser Offsets: The Good, The Bad, and the Ugly.

Microsoft's Internet Explorer has two proprietary tags, LEFTMARGIN and TOPMARGIN, that give you control over these margin settings. You can change the default settings to any desired value, or eliminate the margin altogether by setting LEFTMARGIN and TOPMARGIN both to 0.

Netscape's Navigator currently has no comparable proprietary tags, although using frames--with their accompanying margin attributes--is one way to control margins to some extent in pages viewed in browsers that support frames. Bear in mind, though, that workarounds and tags that optimize your site's appearance for a particular browser comes at the expense of designing your pages for universal access.


Dear Dr. Website®: What are the most popular video formats, and what can I do to reduce their file size as much as possible?
Web video files can take a very long time to download and display because the amount of information they must contain is enormous. By now, QuickTime is the most commonly used video format on the Web. MPEG (which, like QuickTime, is fully cross-platform) and AVI are other frequently used video formats.

All three use a compression technique called frame differencing, which stores only the information that changes from frame to frame. For example, if most of the frames have an unchanging, static background, that information will likely be stored only once while the key frames (the action in the foreground) are stored in their entirety.

Additionally, MPEG uses lossy compression, which results in higher compression rates at the expense of losing some detail. However, with these high-resolution images, this should not pose a noticeable loss of detail.

If you plan to use computer animation tools that are supported by Netscape's browser via plug-ins, such as Macromedia's Shockwave plug-in for Director movies, you can serve up inline video that uses proprietary compression methods for even smaller file sizes. Note, however, that such videos will be viewable only to visitors using both Netscape and the appropriate video plug-in.

In general, keeping the reins on memory hogging involves using a smaller video capture window, reducing frame rate, and/or increasing the amount of compression. Ideally, you should capture your video files at the highest quality and frame your system is capable of, and then later make decisions about reducing quality and frame rate in order to reduce file size.
Dear Dr. Website®: I need help figuring out how to create a good-looking drop shadow for an image on my page. The images I'm using are color GIFs in Photoshop. I use Photoshop to add feathered drop shadows or blurred edges to the graphics, then I open the file in Transparency to drop out the background color. Instantly, the drop shadow or the edges look like they have a whitish border around it, which is not the effect I'm going for at all. How can I get around this problem? Should I just admit defeat and give up my favorite effects?
One of the reasons drop shadows look so nice in Photoshop is that the edges are often anti-aliased slightly against the background color. The same goes for the effects of the various blur filters. If your image is placed against a white background. that meansthere will be a lot of very light grays and off-whites in your image. Since Transparency will knock out only one color in your image, these slightly off-white, anti-aliased pixels will suddenly stick out like a sore thumb. You may have to go into your Photoshop file and correct some of the glaring colors by hand, pixel by pixel.

Is there a better way, you ask? Well, there are a couple of solutions you could try. When you're creating your original effects in Photoshop, try placing the drop shadow in a separate layer and applying a dissolve effect instead of a blur. This will dither the pixels so that at least you won't have a halo effect around the drop shadow. This technique works best with larger images. If you absolutely must use a blur effect, reduce the number of pixels that are included in the blur; examine and edit the edge pixels carefully to ensure a smooth, solid edge.

Another workaround you could try is creating your drop shadow against a color similar to the background gray used by many viewers. This is far from a perfect solution, of course, since you'll find that different browsers do use different background colors. Some of your users may have changed their background preferences anyway, so your carefully applied subtle gray color will stick out like a sore thumb.

In short, creating clean-looking graphics with blurred edges is a nit-picky process, but it's something you can accomplish with some tweaking and patience.

Java and JavaScript

Dear Dr. Website®: Recently I visited a page with a very dramatic background effect. When the page loaded, the background seemed to fade very quickly from black to white. How did they do this, and how can I add this kind of effect to my site?
The effect you describe is one that can be achieved easily through JavaScript. Kouichiro Eto's fade script is a publicly available JavaScript script that can generate this kind of gradual background color fading. One bonus is that you'll have a good deal of flexibility for modifying the colors used and for varying the rate at which the color change takes place.


Dear Dr. Website®: I've seen a bunch of sites that have custom messages that scroll across the bottom border of the browser window. How can I create this kind of scrolling text?
While there are a number of ways to put scrolling text and graphics on a Web page--from Microsoft Internet Explorer's MARQUEE tag to Java applets--the common way to incorporate scrolling text that appears in the status bar at the bottom of the browser window is via JavaScript.

Implementing these JavaScript tickers is a straightforward process. You can cut and paste the applet's source code from a page that uses such a ticker, making sure to initialize it inside the <HEAD> tag and call the function in the <BODY> tag.

But before you add this capability to your pages, you should note its potential design problems. Many end users complain about the way these scrolling status bar tickers interfere with other status bar messages. For example, the scrolling ticker can interfere with the display of URLs that should appear when users pass a mouse over hyperlinks. In addition, these tickers may flicker jerkingly at times, and continual scrolling may be annoying to users who stay on a single page for any length of time. Webreference.com is one site that lists pros and cons for using JavaScript tickers in your browser window's status bar.


Dear Dr. Website®: How hard is it to learn JavaScript without already having learned Java or C first?
While it's not really a prerequisite, it's very helpful to understand the basics of Java or C before learning JavaScript--or as much as you can about programming in general. If you can spare the time, I'd recommend picking up a beginner's book on C to gain a crash course in its command and syntax, since JavaScript uses some of the same conventions. If you ultimately want to become well-versed in JavaScript, it will be worth it to get a good grounding from the start.

That said, if you need to supply JavaScript programs in a hurry you'll find there are certainly many resources available to you to make that job easier. There are already a number of comprehensive JavaScript computer books in print, like Gordon McComb's "JavaScript Sourcebook" from John Wiley & Sons; some include sample JavaScript applications and routines on enclosed CD-ROMs.

JavaScript Index is one Web resource comprised on links to online tutorials, book publishers, mailing lists, and sites that make good use of JavaScript. Gamelan's list of JavaScript links offers another index for finding JavaScript resource. One of the most useful ways to gain programming skills is to study the work of other programmers, and viewing the source code of JavaScript-enhanced pages will afford you a wonderful opportunity to do so. You may also find it useful to subscribe to the Usenet newsgroup comp.lang.javascript for ongoing news and discussion of JavaScript.

Frames

Dear Dr. Website®: Is it worth the trouble to convert my Web site to using frames?
Debating a large-scale change to your site's design interface, such as frame-enabling all content, should be driven more by considering how the redesign would better serve your site's function than by simply wanting to use a new or popular technology.

Frames will alter the way visitors to your site will see your information organized and how they will navigate the site's contents. For example, if you determine you would have a use for a toolbar or other navigational aid to remain in a fixed position on all your pages, putting that fixed content within a frame--while using other frames or framesets to display independent content--would serve as a very effective interactive aid.

If you become adept at designing frames, you may want to create even more sophisticated framesets--for example, refreshing several frames at once to load several new documents, or adding distinct headers and footers to your pages. This could have great potential for making clear to your readers the connections between various documents and the relative importance of the different kinds of information on your site.

On the other hand, if your site's organization is very linear, it won't gain much in appeal by your utilization of frames technology. For example, if you make your frames too small and compel visitors to scroll excessively--potentially leading to missed resources and unsatisfied visitors--this information structure will not be successful. It's a helpful exercise to articulate what you see as the trade-offs involved in incorporating frames on your site.

You should also keep user considerations in mind, since framed pages can take longer to load than their plain-Jane HTML equivalents. With the passing of time, users have become more familiar with how to navigate frames--and Navigator and Internet Explorer have both refined their frames implementations--but you should be certain that the interface won't be a burden to at least the large majority of your users.

There are a number of excellent online tutorials for developing expertise in designing sites with frames; many of these will offer additional suggestions for judging whether frame-enabling your site will be worth the investment of your time.

Here's just a small sampling: Webspinners' Introduction to Frames includes a layout guide, as well as examples and templates for creating frames quickly. NewbieNet hosts a guide to learning how to design frames on its site called The Netscape Frames Tutorial. Netscape offers a thorough guide on its Frames Syntax page, as does Microsoft, on its page devoted to HTML Authoring Features
Dear Dr. Website®: Is there any way to get rid of extra space at the top of a frame?
There are a couple of attributes available for the <FRAME> tag that might help you do that. MARGINHEIGHT is an optional attribute that can control the upper and lower margins of a frame. Similarly, MARGINWIDTH takes care of left and right margins within a frame. Your <FRAME> tag would then read something like this: <FRAME NAME="url" SRC= "somefile.html" MARGINWIDTH=0 MARGINHEIGHT=0> The value you enter for both margin-related attributes is measured in pixels.

Site management

Dear Dr. Website®: How can I improve my site's placement in search engine results?
The appropriate way to include information about your Web pages for indexing or cataloging programs is to use the <META> tag (for describing "meta information") within the <HEAD> part of your Web documents.

You would use the NAME attribute to describe the type of information and the CONTENT attribute to list the keywords themselves, as in the following example using HTML code:

<META NAME="keywords" CONTENT="word1 word2">

The descriptive terms you use in the CONTENT attribute should add to, not repeat, the terms used in the <TITLE> tag.

Dear Dr. Website®: I want to make a number of application files available for downloading from my Web site, but users aren't able to decipher the files after they download them. How can I solve this problem?
When a Web server responds to a request for a file in a URL, it returns a type of information--called a MIME type--that specifies what type of data will be returned. MIME types can describe text, graphic, and all sorts of other application formats. But if your Web server isn't properly configured with a MIME type that specifies the kind of data in your Persuasion files, then it won't recognize and know how to handle these specific file formats.

In addition to configuring your server, you may have to tell your users explicitly how to set their browser preferences to download such files instead of searching for a helper application to display them.

The easiest solution is to serve up these files in a common compression format--such as a .ZIP file--whose MIME type is already recognized by servers. If you still prefer to serve up the application files themselves, consult with your system administrator on how to configure your server to recognize these additional MIME types.



Dear Dr. Website®: I have a restricted area on my site that's only accessible via a common password. Right now, I use a support script to implement password checking. I'm discovering that this form-based password system isn't very secure. Is there a better way to control access to my site?
It's probably best to put these documents in their own directory protected via your server's built-in security features. With this general access-control scheme, you can make your own lists of users with access privileges to certain directories.

The NCSA HTTPD Mosaic User Authentication Tutorial is a straightforward guide to configuring your server's security.


Dear Dr. Website®: Some users on my company's Web site have a tilde character (~) that appears in their URLs. Is it completely necessary for them to include the tilde when circulating their URL? One user has said he'd like to omit it if possible.
The tilde character tells the HTTP server to access the user's public_html directory, rather than the main document tree, for the file in question. That way, visitors can pull documents from your users' directory directly without accessing your site's main document directories. You or your system's administrator can, of course, reconfigure your site's directory structure to let users access whatever subdirectories you'd like. It's most important, though, to consider the access restrictions you want to impose on your site's directory structure instead of just accommodating one user's desire for a simpler URL.


Dear Dr. Website®: Can you suggest the best mechanisms for maintaining a solid firewall while still providing Internet access to people behind the firewall and to telecommuters? In my corporate communications I have had a lot of trouble, particularly with attachments.
It can be difficult to achieve that balance between maintaining corporate network security and providing functionality to remote users who are telecommuting or are clients. A great resource for this issue can be found at the Security Mailing List FAQ (frequently asked questions) list. Dealing with attachments is also a very common problem that is encountered when working with several disparate systems and protocols. The solution to this problem is not easy, as you may already have certain networking protocols and practices in place. Standardize on MIME (Multipurpose Internet Mail Extension) for handling attachments. Part 2 of the comp.mail. mime FAQ offers an extensive listing of MIME products.


Dear Dr. Website®: Is there an easy way to change the default filename for a Web directory?
Once you've established the filename you want to use for your default page, ask your system administrator to make the necessary change to your server's configuration file for you. For many servers, this file will be called srm.conf. Your Web administrator will simply need to specify in the server's configuration file what filename you plan to use as the default page to be served up.


Dear Dr. Website®: I've heard that some Internet service providers don't let their customers implement forms or imagemaps because they say giving users access to the cgi-bin directory is a security risk. My provider has made accessing the cgi-bin directory convenient for me by putting one under my directory. When I ftp to the site to upload changes, the imagemap or other scripts go into the subdirectory under the directory of my home page, instead of a cgi-bin directory that encompasses the entire server. Why isn't this option used more? Is it more work for the provider? It seems like a great solution to me because the cgi-bin can be accessed at any time without having to bother the overworked provider, and the cgi-bin is safe from hackers who might sabotage my site. I would like your opinion on this method.
While this option may make it easier for you to maintain your Web site on your provider's servers, it also makes it easier for you to violate system security and overuse resources. Such outright access to a cgi-bin directory without screening of the applications by your provider, means that you can run whatever kind of application you want from your cgi-bin directory. While this may be a convenience for you, it is a risk to the provider in terms of both security and resources; there is nothing to stop you from running applications that take advantage of the server in ways that many hackers dream of daily.

Statistics and access checkers

Dear Dr. Website®: Can you help me interpret my site's traffic reports? What are unique URLs, unique hosts, and "304s"?
The total number of unique URLs is the number of separate pages requested by all users, including any nonexistent URLs requested in error. The total of unique hosts is the number of separate machines that request files; this is more useful for counting unique visits rather than unique users, because there are some machines--such as proxy gateways for the large online services--that request files for many users.

A 304 occurs when a file has already been served to the user and is cached--that is, it is resourced from the user's machine when it needs to display again instead of being transmitted once more from the host machine.

The number refers to the HTTP status code that describes this situation. A couple of status codes that may be more familiar to Web users include 403 (access is forbidden) and 404 (URL is not found). If you actively try to reduce your server load by reusing graphics instead of storing many unique ones, for example, you'll see an increase in the number of 304s.


Dear Dr. Website®: Is there an easy way for me to put an access checker on our home page? I want people to see a little announcement that says something like "Greetings! You are the 512th visitor today!" Everyone I've asked has said I need permission to access our server's cgi-bin directory, but what does that mean?
While putting in an access checker on one's home page is not overly complicated, it does entail the need for permission to access your server's cgi-bin directory. The cgi-bin directory is where the script (usually written in a programming language called Perl) to execute this process will live. CGI stands for Common Gateway Interface. Using this interface lets a standard Web page incorporate extended features such as forms, on-the-fly page generation, and access statistics.

You'll probably want to contact your system administrator (or support staff if you're partnering with an outside firm) for assistance in gaining access to your server's cgi-bin directory and installing any scripts. Don't hesitate to ask your service provider for this kind of technical advice. Your provider may even have already set up a program that you can use to perform this function.

There are numerous Perl archives on the Web that contain scripts that compile server statistics, with names like WAMP (Who Accessed My Pages), or RefStats. You may also want to check out the relevant Usenet newsgroups. One exciting Web development in the making is functionality for monitoring your site's traffic on a broad scale, instead of merely counting hits. For example, I/PRO requires new visitors to the sites it monitors to supply a registration code; the company then tracks subsequent visits by number.


Dear Dr. Website®: I run a small Web site in Albuquerque, and recently we have had inquiries from advertisers who would like to place sponsorships on our site as to how many "hits" and "users" we have been getting. While we have no trouble measuring the number of hits to the site and the pages contained therein, we have not been able to tally the number of users that come to our Web site. This number is difficult to arrive at specifically because of ISPs that allocate IP addresses dynamically as well as by the commercial online services, which tend to cache frequently accessed Web pages on their server. Have you found any way to get around this roadblock?
The issue of the number of unique users that come to an individual Web site is a complicated, yet important, number. This number is most apt to represent the "readership" of a Web site as defined by traditional media terms.

Your problem in measuring this number is confounded by two factors: (1) your inability to account for users who browse your Web site using a different IP address each time, and (2) shielding from hits and any unique user IPs from the commercial online services.

The first problem is nearly impossible to solve. Users that have their IP addresses allocated dynamically will always appear to be one among a common group of IPs, thus blocking your ability to keep track of their accesses via reverse-mapping their IP address.

The second problem has come about as a result of the implementations the commercial online services and other organizations operating behind firewalls have chosen for their users. While there may be some speed-up for the end users of these services, it complicates the ability of a Web designer to keep track of these users. For example, you may notice occurrences such as:
[a-z]+[0-9]+-[0-9]+.compuserve.com
piweba[0-9]y.prodigy.com
in your log files, as well as numerous hosts with "proxy" in their names. These servers are acting as a gateway for numerous users, and unfortunately at this point there is no way to determine how many unique users the hits from these servers represent.

You should be open and frank with your prospective advertisers about this subject, and offer them info such as visits to a certain .html file they might like to sponsor, as well as the number of users who have taken links to advertisers' sites once that sponsorship is placed on a page. Trying to use elaborate formulas and estimates, as has been attempted by various Web sites and publications, will only serve to complicate the matter further. There are a number of software tools that are currently available to assist you in this analysis.


Dear Dr. Website®: Is there any way to determine what other sites have links to my site and what the URLs of those sites are?
A standard way to check which sites contain links to yours is by analyzing your server's referer_log file. A more convenient--and less technical--method is provided by the Alta Vista search engine.

With Alta Vista, enter a search query like this one in the form: link:http://www.my company.com/. That will return a list of Web pages that reference the URL http://www.mycompany.com/.

Several browsers, although not Netscape's, pass a "referer" header field to the Web server.

This field tells the server the URL of the last document the client displayed immediately before coming to yours. In many cases, users follow hyperlinks from one page to another, so it's likely that the "referer" page contains a link to yours.

You may find some false returns in the referer_log. If users access your site by typing in your URL, you'll find that the previous page they viewed probably has no connection to your site. Ask your system administrator for help parsing your server's referer_log.

jobs.webdeveloper.com

internet.commerce

Partner With Us














internet.com

IT
Developer
Internet News
Small Business
Personal Technology
International

Search internet.com
Advertise
Corporate Info
Newsletters
Tech Jobs
E-mail Offers

hm-v6-139x61.gif

Database Journal: The Knowledge Center for Database Professionals
find a web host with:
CGI Access
DB Support
NT Servers
UNIX Servers
Telnet Access

advanced search

jobs

Get FREE Development Info via your Email!

The Webdeveloper channel
FlashKit
GIF.com
HiermenusCentral
webdeveloper Jobs
Java Boutique
JavaScript.com
JavaScript Source
ScriptSearch
StreamingMedia World
WDJ
WDVL
WebDeveloper.com
WebReference.com
XMLFiles.com

ForumsEmail a Colleague · FAQ



JupiterOnlineMedia

internet.comearthweb.comDevx.commediabistro.comGraphics.com

Search:

Jupitermedia Corporation has two divisions: Jupiterimages and JupiterOnlineMedia

Jupitermedia Corporate Info


Legal Notices, Licensing, Reprints, & Permissions, Privacy Policy.

Advertise | Newsletters | Tech Jobs | Shopping | E-mail Offers
Solutions
Whitepapers and eBooks
Intel PDF: Virtualization Delivers Data Center Efficiency
Intel eBook: Managing the Evolving Data Center
Microsoft Article: BitLocker Brings Encryption to Windows Server 2008
Symantec eBook: The Guide to E-Mail Archiving and Management
Microsoft Article: RODCs Transform Branch Office Security
Go Parallel Article: James Reinders on the Intel Parallel Studio Beta Program
 Avaya Article: Advancing the State of the Art in Customer Service
Adobe Acrobat Connect Pro: Web Conferencing and eLearning Whitepapers
Avaya Article: Avaya AE Services Provide Rapid Telephony Integration with Facebook
Go Parallel Article: Getting Started with TBB on Windows
HP eBook: Storage Networking , Part 1
MORE WHITEPAPERS, EBOOKS, AND ARTICLES
Webcasts
Intel Seminar: Efficiencies in Hardware/Software Virtualization
HP Webcast: Disaster Recovery Planning
Go Parallel Video: Performance and Threading Tools for Game Developers
 HP Video: StorageWorks EVA4400 and Oracle
HP Webcast: Storage Is Changing Fast - Be Ready or Be Left Behind
MORE WEBCASTS, PODCASTS, AND VIDEOS
Downloads and eKits
IBM TCO eKIT: Your IT Budget is Under Attack, Get in Control
IBM Energy Efficiency eKIT: Learn How to Reduce Costs
30-Day Trial: SPAMfighter Exchange Module
 Red Gate Download: SQL Toolbelt and free High-Performance SQL Code eBook
Iron Speed Designer Application Generator
MORE DOWNLOADS, EKITS, AND FREE TRIALS
Tutorials and Demos
Microsoft Article: Silverlight Streaming--Free Video Hosting for All
Featured Algorithm: Intel Threading Building Blocks - parallel_reduce
 HP Demo: StorageWorks EVA4400
MORE TUTORIALS, DEMOS AND STEP-BY-STEP GUIDES