Welcome

Welcome to our site, temporarily servicing both Stellar Research and our partner company SOSpecialists. We hope you will pardon the mess during construction.

At the moment, only our Information Technology (IT) Services are supported here, but we will gradually be adding support for all of our other services as we go.

If you haven't yet created a free account, you can do so by clicking on Create Account in the navigation bar to the left.

Otherwise, you can log in by clicking Sign In.

Once you have created an account or have signed into the service, you can do the following:

  • Add, edit, or delete Requested Tasks.
  • Approve Estimates.
  • Check the Status on your outstanding Requested Tasks.
  • Review your Invoices.
  • Make Payments on outstanding Invoices.

 

Document Browser
By Daniel E. Patterson - Stellar Research Corp.
Revised: 20080903.0902

Much of the content on the Internet is in the form of structured document information, but it is clear that as of yet, none of the standard Internet Browsers are fit for handling of structured documents such as shown in the screenshot below:

Standard Html Viewing PaneStandard Html Viewing PaneTable of ContentsIndex ListLocal Search FunctionReader BookmarksReader HighlightsReader NotesAddress BarNavigation ButtonsSearch Results

Viewing the above very familiar example, you will notice that a proper structured document viewer has many features in common with standard browsers:

These are all standard features in any generic browser. However, what is painfully missing from all standard browsers is any support whatsoever for the features available on the left and floating panes of the above example, which are described in the following sections.

 

Structured Document Basic Necessities

The reading of a document, though similar in nature to the viewing of a basic web page, has some unique requirements that must be met by the hosting software if a good experience is to be had by the reader. To view and navigate a structured document smoothly and effortlessly, one must have interactive access to the power of context. Namely, the following features are essential:

  • Table of Contents. A tree structure control for navigating directly to Books, Sections, Chapters, and Key Points.
     
  • Index List. A list control capable of locating pages by key word.
     
  • Local Search Function and Navigable Results Pane. An interactive control for locating content within the specifically loaded context.
     
  • Reader Bookmarks / per-Document Favorites. A control that allows the reader to insert his or her own bookmarks on the document.
     
  • Reader Highlights. Highlight formatting of the open pages, on a per-user basis.
     
  • Reader Notes. Control allowing the user to make his or her own annotations to each and every page, as desired.

 

The following applications have some of these features, but also have a number of disadvantages that keep them less than perfect.

 

Adobe Acrobat Reader

Adobe Acrobat Reader is a fine piece of software that generally has a nice presentation and reading interface, having several key features available for reading structured documents that are in a fixed-published format. However, the main disadvantages with PDF documents are the following.

  • The entire document must be downloaded to be read. Many documents of this type are several megabytes in length, and often, the user must wait several minutes just to access a single sentence in a key chapter. Most often the case is that the entire document will not be read during a single access, and because of this, there is a performance and productivity issue that remains unaddressed.
  • Any changes to a single piece of the document require that the entire document be republished to the server.
  • The user of the document is not allowed to save his or her own bookmarks for later, or highlight pieces of information he or she finds interesting while reading the document.
  • More advanced automated features found in help systems, such as tables of contents, indexes, etc., are not available unless the author of the document manually includes them, which is rare.
  • Search is only available if the document was published from a text / vector graphics source. Documents of images have no search results.

 

Microsoft Html Help

Microsoft's Compiled Html Help files are not too difficult to create, and are generally miniature web sites that have been wrapped up into a single package. The Html Help Viewer is available on all operating systems Windows 2000 and after, with free downloads provided to earlier systems such as Windows 95, 98, and ME, making it a suitable choice for backward support. However, the .CHM file also suffers from a number of severe disadvantages.

  • Like the PDF file, the CHM file must be published and downloaded as a complete file. This presents a serious performance issue, especially in the case that the document is several megabytes in size. Even more likely than with a PDF file, the user will typically be interested in reading about a certain subject, instead of justifying the required download time by reading the entire document in a single sitting.
  • In newer browsers, downloading a CHM file represents a security risk, and a warning is presented to the user each time such a file is opened.
  • A CHM file, although capable of supporting hypertext links in the content page, can not load additional content from another site, due to Internet Explorer security settings. These can be overridden, but the typical user can not be expected to make these adjustments for every link on every file, and as a final result, all of the links in a CHM file must lead to pages within that file.

 

Microsoft Document Explorer

Also known as Html Help 2.0 Viewer, the Microsoft Document Explorer has several very advanced formatting / navigation features, and is slightly closer to a generic structured document viewer, but has its own series of drawbacks.

  • The Viewer is only distributed with certain kinds of specialized Microsoft software, and is not installed by default on any operating system.
  • There is no free download of the Viewer, and there are also no plans to distribute it freely, as explicitly announced by Microsoft.
  • For Document Explorer to properly view structured documents, Collections of documents must be Installed on the user's machine, meaning that one is not allowed to simply start opening structured documents from any place on the web. Even though web links are supported, they must be called from within existing Installed Collections on the user's machine.

 

Web Page / JavaScript Implementations

There are a number of sites on the Internet that do host a limited form of structured document browsing, but implementation of this type of feature in Html and JavaScript have perhaps more serious limitations than any of the other approaches.

  • With the danger of Viruses so acute at every turn, many wise surfers have raised their security levels to the highest settings, and have completely disabled JavaScript, which causes a site implementing this type of feature not to work at all.
  • Similarly to the JavaScript issue, many wise surfers have downgraded to older versions of browsers that have no support for ActiveX plug-ins, Java, or JavaScript at all, providing a guarantee that downloading a virus would be close to impossible, unless the user happened to explicitly download it from a link. As in the above example, Version 1.0 browsers do not work with the online versions of structured document viewers.
  • A feature known as cross-frame scripting is required in order to create a decent version of an online structured document viewer, since when the user clicks on a link to a new page in the document, at least the Table of Contents, if not other lists, must be updated on the left side of the viewer. In newer browsers, while the feature is still supported, it is not enabled by default, since this activity presents a security risk.
  • When the user submits a search, that search is not specific to the loaded context, and most generally, results are returned for documents everywhere on the web site, instead of just the context in which it was issued.

 

It is also highly undesirable to host structured documents on the server without the basic navigation and reading tools, because as you might have experienced on a number of sites yourself, reading a loose collection of badly organized 'pages' can be confusing and utterly non-productive.

In summary however, there are currently no real choices for a true Document Browser that provides a complete experience for both the web site administrator and the visitor, even in the case that those choices pertain to out of band (separate session) solutions, such as the Html Help examples above.

In review of these disadvantages, we should also specify that a proper document browser needs to provide the following additional functionality:

  • Standard Html Content. To support all of the hypertext capabilities allowed in modern browsers, the pages in the View Pane should be Html or embedded documents such as Adobe PDF, such as would be found in a single page of a typical web site. In other words, the viewing pane of a good Document Browser shouldn't be any different than the viewing pane of Internet Explorer.
     
  • Support for server-side changes to single pages. The administrator or author should be allowed to edit single pages in a simple Html editor such as Microsoft FrontPage, or even Microsoft Word, and those changes should be immediately available to the visitor within an active session without any further publishing or maintenance.
     
  • Support for Templated Pages. While CSS provides most of the needed formatting for colors, fonts, and other attributes, an additional level of formatting should be provided that allows template clips to be present in every page of the document, such as in the example shown below, where the current chapter number will appear unconditionally at the top of the page, regardless of the user's scroll position.

    The content for these templates is prepared at the server prior to delivery, so the document browser application only needs to render and display the Html. However, this consideration must be addressed as a part of the standard, so authors can code their pages with the proper attributes.

 

The following paragraphs describe some of the function-specific requirements.

 

Table of Contents

The Table of Contents is the most structured portion of the document, and its importance can not be underestimated, especially in the case that the document has been well organized into sections and chapters, or as a well-defined outline. The following image illustrates the usefulness of this control.

In the above example, the Table of Contents displays a number of books, each of which can be opened and explored with a minimum of effort. However, since the information in this control is arranged in a hierarchical structure, any number of assignments can be made, depending upon the author's preference, such as in the following example, which typifies an outline for a specific document.

1. About the CSS2 Specification.

1.1 Reading the Specification.

1.2 How the Specification is organized.

...

2. Introduction to CSS2.

2.1 A brief CSS2 tutorial for HTML.

2.2 A brief CSS2 tutorial for XML.

...

...

When the visitor accesses the document from the Web Server, the Table of Contents for that document is loaded as an independent file. The information from this special TOC file is used to populate the user's tree control. From that point forward, while the user is still navigating within that document or series of documents, the tree control will update according to the user's current location within the document, making it a natural act to also provide navigation to Previous and Next pages (or subjects) within that context.

 

Index List

The Index List is another very helpful feature allowing the reader to get to a specifically desired location in the document.

As with the Table of Contents, the Index List is loaded from the server as a small data file when the document is downloaded. The content of this list is a combination of author-inserted values, and each single word found in all of the pages of the document. The latter of these is found by the web server, and updated each time the source of any of the pages in the document has been changed.

Since any item in the index might relate to multiple pages, an additional control is implemented that shows the topics of all of the pages found. By default, the first matching page is loaded into the viewer, but the user can also click on any of the remaining choices.

 

Local Search

The local search function operates upon the data from the Index List described above. After this list has been downloaded from the server, the user can enter a number of basic or advanced search terms to find pages with esoteric but matching value.

As with the Index List control, a search might yield multiple results. For this reason, the Search Results control is filled with all matching items, and can be clicked upon to explore the eligible targets.

Several levels of logic are used to find the results, and the user has control over some of these, such as whether to match the Case of the letters, whether to match all or some of the words, or whether the term refers to an exact phrase, etc.

 

Favorites

All of the available browsers do have Favorites Lists, but Favorite Sites is not the same as Favorite Books, let alone Favorite Chapters within this Book. Would you want to heap all contexts of favorite things together, then have to determine as you search for them later, which link applies to what kind of thing? Oh wait. That is what everyone does now.

A good document browser should provide the ability to allow pre-categorized storage of favorites, using the active document's hierarchy as a guide. Even though the user can override the location of that favorite, this attention to detail provides a certain structure that can easily be located again in the future.

As with the Table of Contents control, the items in the Favorites list are synchronized to the currently loaded page in the document. If that page appears on the user's favorites list, it will be selected, allowing for the same kind of Previous and Next navigation afforded by the Table of Contents control.

 

Reader Highlights

It is difficult to believe that after all of these years, none of the web or document viewers have allowed users to store their own highlights on a page. Granted, the source of the page might change at some point in the future, but unless the specific sentence is actually deleted from the document (which could be retrieved in the user highlight anyway), finding the location upon the page in which the highlight appears is not a difficult proposition.

The following type of example should be supported for each reader of the page.

One in Five dentists recommendgum for their patients who chew gum.

The highlight directives are stored in the user's local preferences file, and cross-referenced with the globally unique identification of the page upon which they occur. Whenever that page is loaded from the server, the highlighted terms are found, and placed accordingly.

 

Reader Notes

Similarly to the subject of highlights, very few applications have ever promoted support for user annotations to be associated with a given page, and this is a sad situation. Many times, a document is somewhat incomplete, and the reader desires to record some additional information about that page for future reference. In this day and age, while he or she could still write in the margins of a book, or place a small note card in the space between two pages, there aren't any electronic means of support for this.

Something similar to the following example would really help to resolve that requirement.

Note the toolbar, which enables control over richly formatted text, allowing the user to emphasize, organize, and colorize his or her own notes. Linking to other documents is also allowed.

As with the highlights, the Reader Notes are loaded in reference to the page for which they were written. Both Reader Notes and Highlights can also be displayed independently of pages, and searched upon using local search.

 

If all of the above considerations are made, a very smooth and responsive experience is possible for the reader, and the burden upon the network administrator is greatly eased, without any of the disadvantages found in previous structured document systems. Additionally, for web sites that employ a great deal of hierarchical structure, entire sites can be navigated using this approach, and the client can benefit by never getting lost.

 

Project Status

This project is currently underway. When completed, the DocumentBrowser application will be offered as a free download application, and an additional Software Developers Kit (SDK) will be offered (in structured document format, of course) that will allow others to implement structured document support on their own web sites.

 

 

Do you own a service-based business?
We can personalize this very web site for your own company's use.