For detailed information on Internet Explorer and alternative browser choices such as Mozilla Firefox, Opera, and Netscape, see my Web browsers page under Internet software.
Some corporate Web designers make a practice of opening external hyperlinks in a new browser window, in an attempt to force you to return to their site. This is coercive and unfriendly. If users want to return to your site, they will. If they don't want to return to your site, you're going to force them to? What are we trying to do here, attract customers, or annoy people?
What's worse, opening a new browser window breaks the Back button, the second-most used browser feature (after following hyperlinks). A user whose browser is maximized because they have a small monitor may not even realize a new window has opened, or understand why their Back button suddenly is grayed out.
Opening new windows should be left to the user. From Jakob Nielsen's Alertbox site: "Opening up new browser windows is like a vacuum cleaner sales person who starts a visit by emptying an ash tray on the customer's carpet." (From the article Top Ten New Mistakes of Web Design)
Fortunately, there's a simple way to cope with this misbehavior: just run your browser windowed rather than maximized. (If you have a smallish monitor, you may want to size the browser window so it is nearly full-screen.) If your browser is running windowed, when a site does open a new browser window without your permission, it will be visually obvious: you'll see the new browser window with the old one peeking out from behind it. Then you can just click on the old window and close it, thereby reproducing the behavior the offending Webmaster should have programmed in the first place.
Alternatively, if you're running Firefox, Opera, Netscape, or another browser that supports tabbed browsing, you can close the unwanted new window, go back to the page with the hyperlink, right-click it and pick Open Link in New Tab (or the equivalent in your browser). Then you can close the original page's tab whenever you want.
You can use this technique when you want to make sure you have the opportunity to investigate several links on a page, but you don't particularly want to bookmark it. Just right-click a link and select Open in New Window from the mouse menu. You can then follow subsequent links as far as you like. When you're ready to return to the first page and investigate another link there, just close the second browser window, and you'll find the original page waiting for you in the first window. This technique is especially useful for exploring a series of "hits" from a keyword search engine.
You might also want to open a second window if you want to refer to another site while browsing something, either one with a known URL or something you've bookmarked. In most browsers the keystroke Control-N will open a new browser window, or from the pull-down menus you can do File, New and pick Window. Then you can either type in the desired URL or open a bookmark.
An example: say you use a free Web-mail account for your email, and you want to send an email to someone containing a Web-page address in the message text. You can open your Web-mail in one browser window, and the target Web page in another browser window (or a page containing a link to your target page) Copy the page URL to the Clipboard, and Paste the URL into your email message in the other browser window.
If your browser has tabbed browsing (Firefox and Opera) use tabs for this sort of thing. I find it much simpler and economical of screen space.
Opera has the File, New and Control-N methods, but by default uses a different windowing model. New browser windows, the bookmarks window (called "Hot List") and other windows all open inside the main Opera window, similar to document windows in word processors such as Word. Recent versions of Opera let you choose to open new pages in a separate browser window, as in Internet Explorer or Mozilla, if you prefer.
(This has nothing to do with using search engines sites to search the entire Web; this is more like the Find function in a word processor.)
Once you have a target page displayed, you can do Edit, Find or press Control-F. This enables you to jump to particular text strings on the displayed Web page. Usually there are controls on the Find dialog to let you search backwards and forwards, search for only whole words that match, and choose case-sensitive searching or not.
This function is especially handy when a "hit" from a search engine lands you on a really long page, and you don't see your search keywords anywhere. You can save time by using Find to jump to your keywords on the page.
You can get more out of your Firefox Bookmarks toolbar, if you want. Normally you put a bookmark in the Bookmarks Toolbar folder, and it then appears on the toolbar. You can do this when you first make the bookmark, or later by going to Bookmarks, Organize Bookmarks to open the Bookmarks Manager.
If you put a folder that has bookmarks in it into the special Bookmarks Toolbar folder, it will appear on the toolbar as a button, labeled with a folder icon, that opens a menu of the bookmarks in the folder. This is a lot easier to understand if you try it than if you just read about it here. You can move an existing bookmarks folder into the toolbar folder, or create a new folder there and move bookmarks into it.
You can do the same sort of thing on Opera's Personal toolbar and the Links toolbar in IE6. IE has a Links folder that works pretty much like the Bookmarks Toolbar folder in Firefox. In Opera you use a checkbox found under Properties, for bookmarks or folders, to put them on the Personal toolbar, rather than moving them to a special folder. The key idea in any of these browsers is to put a bookmarks folder on the toolbar, instead of just bookmarks.
Refresh and Reload are functions that both cause your browser to re-plot and re-display a page. The difference is that Refresh will generally use code and graphics from the browser's disk cache; Reload forces the browser to refer to the original files, whether they are local files or out on a Web server somewhere.
If the page has actually changed since the last time you loaded it—either because it's a dynamically updating page of some sort, or because you're rewriting the code yourself—you may need to use the Reload function to see the changes. Refresh can be useful when the browser has made an error in plotting the page, and you want it to try again.
In most browsers, the F5 function key is Refresh, and Control-R triggers a Reload.
Most Web authors are pretty good about providing reasonable contrast between text and backgrounds, either the usual dark on light, or for most Star Trek pages, light on dark. Sometimes you'll find a site that has major readability problems; a busy background image, a loud-colored background, or oddball color choices for text vs. background, any of which can make the text hard to read. I've even seen pages with dozens of links, where the author had perversely forced visited links to be the exact same color as un-visited.
I also note that when I see these kinds of problems, it usually turns out the site wasn't worth reading anyway. But there are exceptions.
Fortunately there are usually overrides for this stuff. I usually have a harder time finding them in MSIE than in Navigator, but they are there. You can force text to black in your browser, backgrounds to white, and even control link colors if you have to. I remember an early Navigator version that had a single set of color/background overrides for both viewing and printing, but generally these days there are separate overrides. This is good: generally you want to see backgrounds on your display, but not print them, for example.
Let's say a user has bookmarked a page about Arctic wolves, that they haven't looked at for a while. They click on that bookmark and get back the familiar 404 Not Found error. Okay, that page isn't there anymore. If this user has been using the Web for a while, they might just do Control-O for Open and try the following URLs in succession, until they get back a page instead of an error:
|
http://www.zoo.org/animals/mammals/canines/wolves/Arctic.html http://www.zoo.org/animals/mammals/canines/wolves/ http://www.zoo.org/animals/mammals/canines/ http://www.zoo.org/animals/mammals/ http://www.zoo.org/animals/ http://www.zoo.org/ |
This is what I mean by URL chopping. Sites should be set up so this works.
You should do another sort of URL trimming when you list a Web address in print. For example, the full URL for the home page of this site is "http://www.icehouse.net/jim_d/", and that's how it needs to be specified in HTML code. Most browsers now in use let you leave off the "http://" part when typing in a URL, and also the trailing slash, so on my bidness cards I cite it as "www.icehouse.net/jim_d".
It should be a little different when you cite a URL in electronic mail. Current email programs (including Web-mail) will automatically display an Internet address as a hyperlink. You can put a URL in an email to somebody, they see it as a link, click on it, and view the site in their browser; no typing and no typos. But at least the way it's done now, email programs only behave this way with full URLs; in other words "http://www.foo.com" works and "www.foo.com" doesn't work. If you copy a URL from a browser Address bar, or by right-clicking a hyperlink and picking "Copy link location" (in Mozilla; "Copy link address" in Opera, "Copy shortcut" in MSIE) what you'll paste in your email will be a full URL.
This is a way you can speed up some of your Web surfing.
Whenever you click on a link to request a Web page, the remote server normally sends the HTML file, plus any graphics the file's tags call for. Web browsers always have a way to turn off image loading. When you do that, the server sends the HTML file by itself, which generally happens a lot faster.
When image loading is off, the browser plots the page using placeholder frames and little image icons wherever the page code calls for an image. If you're lucky, the page author has also provided explanatory ALT text for each image, which displays in tiny letters inside the frame. You've seen these frames: they're also displayed briefly during normal loading, until the actual image file arrives over the network and gets filled in.
The catch is that not all Web pages are designed so they make sense without their images. It's not at all hard to do, and the required elements are all stuff that needs doing anyway for other reasons. See my image tag coding page, in my content on HTML, for more details.
If you use the Firefox browser, you can get the 5K imagepref extension, which gives you a status-bar check box that toggles image loading. Works great for me.
MS Windows, and other graphical operating systems such as the Mac OS, OS/2, and X windows on Unix machines, are collectively described as having a graphical user interface or GUI, sometimes pronounced "gooey." This is in contrast to text-mode systems such as MS-DOS and Unix exclusive of X windows.
One of the more annoying recent Web-design "innovations," to try to get Web users to click on advertising banners, is the inclusion of elements that appear to be standard Windows interface elements, such as pushbuttons, text boxes, radio buttons, check boxes, and combo boxes. I've seen some recently that actually worked, but the early ones were strictly fake, graphics only with no function. Some people refer to this as fake user interface, FUI, or "phooey."
This seems especially pathetic and desperate, since user interface research shows that Web users actually got so massively disinterested in ad banners that they would actually ignore anything on a Web page that was even shaped like an ad banner, regardless of content.
I have no objection to the presence of advertising in Web services when I'm receiving something significant in exchange for tolerating it, such as in free Web-mail services like Gmail, Yahoo! Mail, and Lycos Mail. I think it's pretty clear that users don't welcome banners on ordinary Web pages. Clickthrough rates are low, and the only reason such banners generate anything at all is the enormous number of Web users.
FTP is "file transfer protocol," and is used on the Internet to move files from one computer to another, including posting edited Web-page files on a Web server. There are specialized user-friendly graphical ftp programs, such as WS_ftp and cute-ftp. Windows 95/98/NT come with their own character-mode ftp program, that you can run from a DOS window; this works very much like old-style UNIX character-mode ftp.
You can still do some ftp even if you only have a Web browser. Browsers can even move files both directions, download and upload.
One difference is that browsers can only transfer files one at a time; real ftp clients will let you select a whole series of files and transfer them with one command. Another is that ftp clients may be faster than your Web browser for exploring ftp servers. Some browsers open a new ftp connection every time you open a URL or click a link, whereas a real ftp client will hold the connection open until you explicitly disconnect. Of course, if you know exactly where you're going on the server, this may not matter so much.
"Anonymous ftp" is a customary Internet service whose name confuses new users. It maybe really should have been called "public ftp." These are servers with large collections of downloadable files, that allow any Internet user to access them. The original custom was to do this outside of business hours at the server location.
FTP servers always prompt for username and password; the convention in anonymous ftp is to supply the string "anonymous" as a username, and your full email address in place of a password. Usually your browser will take care of this part for you.
If you do File, Open in your browser, and type a URL like this (or just click on this link):
ftp://rtfm.mit.edu/pub/usenet-by-hierarchy/
...your browser should show you a listing similar to a directory listing, with the items made into handy hyperlinks for you. Some items will be subdirectories and some may be files; the top list item will usually be a link that takes you up one directory level. If you click on a directory item, you'll see a listing of the contents of that directory. If you click on a file item, the ftp server will send you that file.
If you have rights on a certain system as a user, with a real username (login ID) and password, you usually can transfer files to and from that system. For example, your Internet ISP account normally includes a home directory, and a Web-page directory which may or may not be different.
Let's say you're user "johndoe" on system "www.isp.com" and your password is "foo". If you type this URL:
ftp://johndoe:foo@www.isp.com/johndoe/
...you should gain ftp access to your home directory. This method involves typing your password into the address bar for everyone to see. It would also be a really bad idea to create any bookmarks or hyperlinks with this type of URL, because they would record your account password in a form anyone could read.
On most systems you should be able to use this form:
ftp://johndoe@www.isp.com/johndoe/
This method omits the password. The server should prompt you for it, using the normal password box which shows only asterisks rather than displaying the password, which is safer.
Once you're connected, you should be able to download by clicking a file link. Normally the browser will have a command of some sort on the File menu that allows you to upload a single file that you select. You might want to do this to save some bookmarks you found while away from your usual computer, for example, by uploading a copy of the bookmarks file and retrieving it later.
Alternatives that could also meet this kind of need would include learning to use Windows 95 style character ftp, signing up for a free Yahoo Mail login and using their free Yahoo Briefcase file-parking service, or just using Backflip Web bookmarks.
site feedback