Definitiion: InTRAnet versus inTERnet
Types of documentation to put on an InTRAnet
Overview of publishing strategy
Many people use the InTERnet for marketing and other purposes. You can use the same tools to work with InTRAnets, also. InTRAnets are internal networks of a company, such as a Novell network, or company divisions in different locations across the Internet. This session focuses on building that internal company publishing strategy. Issues discussed will include: the necessity of one data set, levels of training for different personnel, incorporating graphics from your favorite publishing tool, and HTML forms and their output.
Many of us have heard nothing but InTERnet for the last three to four years. We are just starting to hear about InTRAnets. What is the difference? Internets are open to the world while InTRAnets are private.
"Think of the Internet and World Wide Web as a vast sea of information, punctuated by a few sharks in the form of hackers who would love to steal or confuse proprietary data. Fortunately, some companies have discovered they can avoid such sharks by using Web technology to create internal Internets, called "InTRAnets."
"These miniature Webs are like private company beaches, sheltered behind firewalls that keep unwanted intruders at bay. Resembling a combination of groupware and the Internet, InTRAnets let users do the same kinds of things they can do on the Web -- post documents, send electronic mail, chat with other users and keep employees up to date -- but on a more intimate scale.
"InTRAnets also permit real-time collaboration on documents and drawings, video conferencing and continuously updated access to sales and service databases. People can use them within an office, in multiple-company offices or remotely from home or while traveling. And InTRAnets are faster and cheaper than many other methods of accomplishing these tasks, such as groupware and that old standby, interoffice mail [1]."
Why would your company need an InTRAnet? Here are five reasons:
- InTRAnets solve the problem of information overload.
- InTRAnets are inexpensive.
- InTRAnets are cross-platform.
- InTRAnets are robust.
- InTRAnets are fast [2].
Since the documents are on an internal "safe from competitors" network, any type of communication can be put on an InTRAnet. Listed below are some ideas of documents to put on your network:
Company:
- Company Announcements
- Company Phone Book
- Employee Handbook
- Company Forms
- Company Policy and Procedure Book
- Company Newsletter
- Résumés or Résumé Database
- Inside Job Listings
- Library Listings
- Web Site Listings
- Capability Profile Power Points
- Employee Message Board
- Proposal Sections
- 401(k) Distributions
Program/Department:
- Program Schedules
- Action Items
- Minutes of Meetings
- Work in Progress
- Reports
- Chat Sessions
Anything that you need to share with another person can be put out on an InTRAnet. This makes use of the CALS philosophy, "Create once, share it many times."
You have decided to have an InTRAnet, but what are the rules of display? Who gets to see what, and how can you tell if you are using "old" data? What kind of bells and whistles are on your page and do they all have the same feel and touch? Who approves what is on the pages? All these items must be discussed up front so that you do not run into problems later on.
On an internal InTRAnet most display rules depend on trust. The best display rule is access security on your InTRAnet Web page. Managers would have high level access and workers would be limited to their departments.
Dates of creation and update must be carefully maintained so that the user knows that he/she has the latest schedule, report, etc. The best way to do this is to have a self generating change date on your document. This will be discussed later.
Web developers must be careful to use only items that are free of copyrights if they add GIFs, animated GIFs, and other images taken from the Internet. Internet pages are considered copyrighted and to use something without permission is illegal.
Another policy and procedure rule that needs to be dealt with is "Who owns the Net?" Are individual departments to keep up their information on it, or are there "Web gurus" who perform upkeep? Is there a corporate policy on design, color, etc.? Who pays for the additional bandwidth that may be needed? Management needs to approve the initial design, set up policy on the approval cycle for Web pages, and set up a color/display scheme [3].
The final rule of display concerns browsers. After having seen different groups go through multiple browsers on the desktops, free browsers versus non-free browsers, the non-free browsers win. The reason is that most free browsers end up costing money or do not support the latest technology. There are also bugs in these browsers that can cost your implementors time. (Have you ever searched for an * on the page when there is not an * in the whole document?)
Why use the same browser? Because regardless of vendor propaganda, the same coded page will look different on different browsers. Has anyone ever tried to figure out why there is a space between <P></P>’s in Netscape and none in Mosaic? With the same browser, everyone views the same documents and graphics in the same format. All format junkies are satisfied, or can be satisfied easily.
Firewalls are the best security for InTRAnets that are on an in-house LAN. They protect the InTRAnet from external Internet users. For InTRAnets that span over the Internet, it is best to use encrypted software. A lot of free encryption software is available for US sites, but not for the International sites. Public key signatures should be added if items are so important that thousands of dollars would be lost if they were tampered with.
This section gets into the down-and-dirty of Web publishing. It is mostly for the Web novice that wants to know how to set up a Web publishing system.
Many people start with one-man publishing teams, usually consisting of the most aggressive person in the computer department. He/she creates a Web page, shows it to management, management is overjoyed, and it gets put into the system. Then, it either works badly or costs too much for this person (usually also the highest paid) to maintain it. Identifying the data flow procedures starts with creating a publishing team effort. For example, at Uwohali, most technicians create their own documents. The documents are passed to a publications department for wordsmithing and creation of visual masterpieces for our customers. At Uwohali, we made five data flow process decisions, listed below:
1. Identifying who would be doing most of the HTML markup. For us, it is easier to have publications do it. Publications personnel make about half the salary of our technical personnel and have a more detailed eye for mistakes.
2. Identify how each piece will convert to HTML. For example, will your graphics go straight into a Web picture? What are the standards in font size, borders, etc.? Is your image publishing system capable of outputting to an HTML format? The conversion of the document into HTML will be discussed later in this section.
3. Identify your linking requirements. Are you linking to a database? Is there any free coding available for this? Are you linking to forms? Image maps? Other sites?
4. Identify the management approval cycle. Will some data be "approved finalized" data and some data be "unapproved work in process?" One of the best questions to ask during this phase is, "how easy is it to change the document?"
5. Identify how the pages, forms, images, image maps, etc. will get put on the server. Who is the responsible person(s) for this task? Should they also get the Web mail?
Now that you have addressed the items of data flow, it is easy to flow the data from creator to production on the server.
Take the data you have decided will be on your Net and ask yourself these questions: (1) Will anyone want to view it in a browser; (2) Do changes need to be made while viewing? and (3) How long is it?
For example, a document you would not want to view in a browser is the phone book. Think of the time it would take to load one page containing the entire book. Or having to click to get to the page that has "M"s on it, then doing a find. The best idea for row and column data is a linked database. Create a form that allows the user to enter the data needed, then have the server search the data and come back with the appropriate answer.
Some documents you want people to change while they review; for example, a proposal/report that your group is working on. Having everyone pull up the document in the browser, printing it or saving it to a file to pull it up in their word processor is not the best way to handle this type of document. The best way is to identify what the item is on an HTML page and make download features available. Or, to make it even easier, identify the document’s name and location.
The size of a document also determines how you display it. Even though the above method of display is easier than HTML’ing the whole document, browsers do a better job of allowing the person to instantly jump to where they want to go. So if the Accounting personnel send out a Cost Report, the person in Logistics can jump to the Logistics page and bypass other people’s data that does not involve them.
And what about HTML’ing? Uwohali’s first effort in HTML’ing was to take a Word document and save it using Word’s Internet Assistant. The generated table of contents jumped to the marked heading paragraphs. Paragraphs were marked. Pictures, tables, bulleted lists, etc. did not make it. Neither did "see paragraph..." jumps. But we were thrilled with this markup and proceeded to markup the rest of the document. Unfortunately, we now had two documents to maintain: the Word document, (because some people had to have it in Word) and the HTML version. We could not resave using the Internet Assistant because that would have negated two days of work and caused us to re-mark those areas.
On our next document, we used a Web markup tool called "Hot Dog" which you can get free for a specific period to test it. We liked Hot Dog because, unlike the second round of coding in the first document that had to be input by hand, Hot Dog allowed us to select text and through input boxes fill in destinations, etc., and Hot Dog generated the code around these selections. However, you could only print from the browser, and somehow, it just does not look as good when printed from a browser. We also did not have the helpers we were used to, such as spell check, thesaurus, etc.
After working with these and other tools, we came back to our old favorite, Word. We compiled a list of items we wanted Word to do for us and started trying them. The first thing we did was look at hidden coding. The hidden facility in most word processors allows the user to tell the document what to do. For example, those paragraph marks at the end of each paragraph in Word are hidden codes -- they disappear if you deselect hidden and they do not print out. What about hiding the HTML code? This would leave us the pretty printing with the ability to save in text format and have an HTML document. There is only one document to update, which resaves as an HTML document at the same time. The HTML document is never updated, but only used to view on the browser.
We developed macros that allow the user to push a button and the HTML code appears. Select a target, copy the target name, push the button, copy the target name into the macro box, and the code is instantly generated around the target. Then we started centering and bolding. What if the user could hit the bold button after selecting text and not only have the HTML code generated around it but also have the text bolded for Word purposes? We ended up with two button bars that can be instantly added to any document by adding an HTML template.
There are many different types of graphics for your pages. The most essential decisions concerning graphics are:
1. Does it enhance the page -- useless graphics take lots of time to download and ultimately make the viewer angry;
2. Can the graphic be broken up into pieces -- most people think of a graphic as a whole piece and design their Web picture as such. If you split up your graphic so that you have many small graphics, you get better response time; and
3. What type of graphics do I put on the page?
The most prevalent graphics today are Graphics Interchange Format (GIFs). (Note: GIFs are falling out of favor due to copyright problems.) GIFs are the Internet standard for graphics, even though they result in a large number of bytes to download. One of the main reasons a GIF may be large is because of the 256 colors or True Colors available. One of the tricks of making a GIF smaller is to concentrate on using only 4 colors, then save as a 4 bit GIF [4].
Joint Picture Expert Group (JPEG), is another option. Although smaller than GIF, the quality of the resolution is not as good.
If you are a good CALS person, you always think, "create once, use it many times." Your first response is that you can use all your previously created PowerPoint presentations and other graphics. Instant Net graphics! Unfortunately, graphics that look good displayed on a large screen do not always look good displayed on small screen. The font size may be readable on the wall, but not suitable for the Web page. Also, the minimum of color used may need to be fortified with more color. For general purposes, most graphics can be reused but must be retailored to fit the Web.
The main problem with using GIFs and such is that there is no zoom associated with these images. If you want to put a CCITT Group 4 C size drawing on the Net, the text would be too small to read and the details would not be available. That is why some people are using plug ins that tell their browsers to display all PDF images, for example, in Adobe Acrobat. Although a little more time consuming to load, if you need zoom features, this may be what you want to consider purchasing. (Note: all viewers that you anticipate using this will need this software.)
Graphics allow the use of mapping. Mapping is when you are able to click on a graphic and the browser jumps to the target of that graphic point. Mapping is very easy. Get a free mapping software such as MapThis, pull in your GIF or JPEG, use your mouse to select the area that you want the user to click in, then save the map.
Now that you have an image map, what do you do with it? There are two ways you can use the map:
1. Place the map on the server and point the page to the server. When the user clicks, the server receives the query and jumps the user to the appropriate place.
2. Place the map within the HTML document by recoding in HTML and using the map coordinates generated by MapThis. When the user clicks, the user’s browser determines where the next jump is and jumps to the appropriate place.
Although option two is the best for your server, and also saves time going over the Net if the jump is already in the preloaded page, not all browsers support the HTML mapping. Some people use both techniques to make sure that it works.
Forms are the Internet’s way of letting you enter or search data. To use forms, you must have a server for your Web page. Forms allow the user to input the data needed, then send that data to the server. The server processes the data and sends back information to the user or other individuals.
Proofing HTML documents is bothersomeare a bear to proof manually. If you have a 60 page document, you may have over 300 links that must be tested. Format testing alone is exasperatinga bear because the WYSIWYG browser does not allow you to change documents as you view them. A proofer must toggle between the viewer and the creation tool. What you see may look fine but the actual code is not in standard format. For example, the sentence <P>The lazy brown dog <B>jumped. </P></B> is not standard in that the hierarchical relationship is not followed.
For quality documents, use an HTML proofer, whether one you purchase or one available on the Net. (Note: you must put your page on the Net to have the Net proofers review it.) Use the HTML jump tools that automatically test each jump for you.
Creating an InTRAnet homepage can be fun and rewarding to the company if you follow the rules. Savings made by InTRAnet Web pages can be seen by the following example: a five page change for an employee handbook cost Uwohali about $500.00 per change. This includes the paper costs, copy costs, the salary of the people making copies, getting the copies to the employees, and the time of the employees inserting the new pages into the handbook. A Web page change for the same information costs less than $50.00.
1. I. Meyer, "Seeking Shelter in the 'Intranet'," Communications Week, April 3, 1996.
2. M. Elgan, "Love the Web? Now You Can Build Your Own -- Here's why your company should have its very own Intranet," Windows World, June 1, 1996.
3. P. Schnaidt, "Ask Hard Questions Before You Intranet," Network Computing, April 15, 1996.
4. "Picture This," Windows World, July 1, 1995.