EnigmA Amiga Run 1996 February

home *** CD-ROM | disk | FTP | other *** search

/ EnigmA Amiga Run 1996 February / EnigmA AMIGA RUN 04 (1996)(G.R. Edizioni)(IT)[!][issue 1996-02][Skylink CD III].iso / earcd / faq / kjvcbibl.lha / cbible / ThinkerDemo < prev next >

Wrap

Text File | 1993-09-28 | 57KB | 1,167 lines

Press the Help key for info on this text reader, then press Help again to escape from Help. Introduction This file is a very abbreviated version of the Thinker manual. It is intended only to help in using the demo version of Thinker. The Demo version of Thinker cannot save any changes to files, does not import or export ASCII, nor does it print. Creating a new file will write on your disk and will in fact create a new Thinker document with a single statement (The copyright notice). ****************************************************************** ****************************************************************** This disk has been updated for Thinker Release 2.2. See the Thinker project "UPDATES" to see what has changed since the last release of this demo. Lots of exciting new stuff!!! Documentation for The ARexx interface is not included in this demo package. ****************************************************************** Thinker is available for $80 from: Poor Person Software 3721 Starr King Circle Palo Alto, CA 94306 (415)-493-7234 Thinker is available at many Amiga stores and mailorder houses. Source for Tinydraw is available on request at no charge. ******************************************************************* Introduction to Hypertext Thinker can trace its heritage all the way back to Doug Englebart's work at SRI's Augmentation Research Laboratory. By combining elements of text processing, outline processing, and hypertext, Thinker becomes an Idea Processor that embodies the basics of Doug's ideas for augmenting the human intellect. What is HyperText? Footnotes are a form of HyperText. Within the body of the text is a directive to look somewhere else in the text for more information. With footnotes the directive is a footnote number while in HyperText the directive is in the form of a character string that "names" another section of the text. In order for HyperText to work the text is divided into sections and each section has a label. One often encounters this form of textual link when a phrase like "See Chapter 2, section 1" appears in the text. This type of link is good but can be carried further. Imagine that each paragraph in a document had a label that described the topic discussed in the paragraph. Now one could encounter the phrase "See the section <correct use of the passive voice>" which would take you directly to the paragraph on passive voice. However, where is that paragraph? What Chapter? What Page? When the document is online and being accessed via a HyperText processor, one simply points the mouse at the phrase <correct use of the passive voice> and asks to see that paragraph. Whether it is in chapter 10 or 2 is of no concern to the reader, the HyperText processor will find it and present it to the reader. HyperText documents do not have to be read linearly like a book. One can begin in the section with the most interest and follow the links around the document to explore the subject. The creator of the HyperText document had only to attach descriptive labels to the various sections of the document. What is Hierarchical Text? Hierarchical Text is an extension of outlines. Outlines are the arrangement of topics in subordinate relationships. Rain and snow are two examples of precipitation and they are properly subtopics of precipitation in an outline of a paper on weather. If the idea of an outline is extended to an arrangement of ideas in subordinate relationships it becomes Hierarchical Text. Ideas are paragraphs or groups of paragraphs and a paragraph on precipitation might have subordinate paragraphs on the different kinds of precipitation that one might expect to find. How Thinker combines HyperText and Hierarchical Text Thinker is a Hierarchical Text processor with HyperText. A Thinker document is made up of a collection of paragraphs (referred to as "statements" or "branches" in this manual) arranged hierarchically and containing textual links to each other. Each statement in a Thinker document can have a number of labels and can be referenced by a textual string (a link) that "names" one of the labels. A statement that begins with "(" is labeled, and the labels are strings of non-punctuation characters separated by commas (",") between the first "(" and the closing ")" at the beginning of the statement. A link in the text consists of the non-punctuation characters between "<" and ">" characters anywhere in the text. When a label is a single word, the link to it can be expressed without the "<" and ">" characters. Thus, each word in a Thinker document is a potential HyperText link to a statement somewhere in the hierarchy of statements that make up the document. To expand the flexibility of Thinker, a link may contain the name of a Thinker document other than the one in which the link is found. The document name is separated from the label by a comma. Thus, <Thinker:example,link> is a link to the section labeled "link" in the Thinker document called "example" on the disk named "Thinker". When a statement has subordinate statements it is called a branch. Thinker allows the manipulation of a document by moving, copying, and deleting statements and branches. When a branch is moved, all the subordinate statements are moved as well. Display parameters control the format of the text on the screen. Thinker can be instructed to display only the first line of each statement and give the appearance of a simple outline. By controlling the depth of the outline visible, the user can actually "see" very large portions of the document at one time. When the document is manipulated while viewing it in outline form, all of the text that is part of the document is moved when the visible portions are moved. What can Thinker "link" to? A HyperText link can link to one of four things: 1) A Thinker document, 2) An IFF picture file which can be displayed in a new window (the picture is thus compressed in size and colors) or a separate screen, 3) any Workbench application, in which case the application is launched in its own window and runs as a separate task (If a project is to be launched, there MUST be a project icon. If an application is to be launched, it MUST have a Tool icon), or 4) an ARexx port, in which case a message is sent to the port. The ability to link to other Workbench applications makes Thinker into a HyperMedia product. Drawings, sound applications, database applications, word processing applications are all possible links in a single Thinker document. How does one interact with Thinker? All interaction with Thinker is via the Mouse and Menu system of Intuition. There are no keyboard commands other than keyboard activation of menus. Gadgets are available for common commands and menus are available for the complete command set. Requesters are used extensively to provide a highly responsive user interface. These requesters always appear directly under the mouse cursor with the cursor positioned over the most common choice. Text on the screen is modified as in many word processors. The cursor is positioned in front of text that is to be modified and typing is always in "insert" mode. Text is "selected" by sweeping the cursor over a block of text while holding the left mouse button down. Selected text can be "cut" from the statement and "pasted" elsewhere or simply replaced by typing. Thinker has "search" and "replace" functions to complete the editing capabilities. Thinker will print a document or prepare a text file for manipulation by a word processor or typesetting program. This manual was produced using Thinker and HPPrint (a Poor Person Software typesetting program for the HP LaserJet Printer). Thinker can also import text prepared by a word processor into a Thinker document. Thinker appears deceptively simple by design. The fact that labels and links are merely parts of the text eliminates much of the complexity associated with large numbers of obscure commands. However, one should not be fooled by the appearance of simplicity. The combination of Hierarchical text, Hypertext, and Word Processing functions makes Thinker a powerful Idea processor. Take the time to learn Thinker by working with the tutorial project (EXAMPLE) found on the Thinker distribution disk. NOTE: Thinker uses a lot of memory. Lack of memory on CP/M and PC DOS systems is what limited Thinker on these machines to a mere toy. One of the uses of memory is to cache blocks of documents in memory to avoid disk access. Poor Person Software recommends the use of a Floppy Accelerator such as Facc II by ASDG. The cache available in Thinker requires a good deal more memory than Facc II to be as effective in reducing disk access. Do not attempt to use Thinker without either enabling Thinker's read cache (see menus explained under Options) or using a program like Facc II. Definitions The terms used in describing a Thinker document have other meanings in other environments. In order to clarify the following discussion of Thinker, the most common Thinker terms are defined here. Anchor The statement that appears at the top of the window is sometimes referred to as the anchor statement of the window. The user navigates around in a document by changing the anchor statement. The anchor statement is changed with Jump commands. Branch A Branch is a statement with subordinate statements. When a statement is inserted into a document, it may be inserted at any level relative to the other statements in the document. A branch is a statement at any level that has statements below it. A branch may have a branch subordinate to it. A Statement with no subordinate statements is sometimes referred to as a Branch. Clipping Level Clipping level refers to the number of document levels that are displayed in a window. A window with a clipping level of 2 will show only the two highest level statements. Clipping can be used to view a hierarchical document at different levels of detail. The document can be manipulated with clipping levels that exclude much of the document from the screen. All of the invisible text is moved with the visible text. The clipping level can be chosen for the entire document or for single branches. The Clip gadget at the top of the screen or the Clip menu sets the clipping level that applies to the document as a whole. "+" or "-" gadgets in front of statements manipulate the clipping level applied to the single branch. Group A group is a selected sequence of branches at the same level. Jump When the anchor statement of a window is changed, a different portion of the document becomes visible in the window. This is referred to as a jump. Jumps can be made relative to a statement visible on the screen or relative to a statement named in a link. It is also possible to jump to picture files and Workbench applications. Label Any statement in a Thinker document may have labels. A label is a name by which the statement is known and is like a key in an indexed file. Thinker recognizes a statement that begins with "(" as its first character as being a labeled statement. The labels are substrings of all of the non-punctuation characters between the initial "(" and a trailing ")" separated from each other by commas. "(These,are,labels)" is a label string with 3 labels. The case of letters in a label is not significant and only the first 19 non-punctuation characters are significant. Labels in a Thinker document need not be unique. Any request to jump to a label that is not unique will cause a requester to appear which displays the first characters of all statements with the non-unique label. The user points at and clicks on the desired statement to select the target of the jump. Level Statements are arranged in a hierarchy. The most important statements are at level 1. Level 2 statements are subordinate to level 1 statements as in an outline. Thinker does not limit the number of levels in a document but the display on the screen will be less than satisfactory for more than about 30 levels. Link A "link" is the appearance of a label in the text. A link is either a single word or a string of characters delimited by "<" and ">". A link consists of two parts. Part one is optional and names a project, Thinker document or file. Part two is separated from part one with a "," and names a branch. Picture file links consist of only the file name and the comma as in <Thinker:art/mirage,>. Links may by typed into requester strings as the address of statements when certain operations invite the user to designate a statement. Links typed into requesters do not have the surrounding "<" and ">" characters. Links can be activated by putting the cursor over the link in the text and double clicking the left mouse button or by using the Jump Link command. Links may designate labeled statements in the same or different Thinker document, IFF picture files, any Workbench application such as a music player, paint package, or CAD application, or an ARexx port. Use the file name or program name in the file portion of a link followed by a comma to specify a link to an application. Mark Thinker sometimes invites the user to designate a statement as the target of an operation such as "delete" or "move". The user may elect to mark the statement on the screen or type in a link that designates the statement. A statement is marked by moving the Pointing Cursor (a hand) such that the hand points to the statement and pressing the left mouse button. Origin The first statement in a document is referred to as the origin statement. The origin statement may be moved or deleted in which case some other statement becomes the origin statement. Plex A complete group of statements. All of the subordinate statements in a branch constitute a plex. All the first level statements in a document constitute a plex. A plex is designated by pointing to any member of the plex. See Through Link Links can be used to extract statements from one document into another. The statement named in a "See Through Link" can be displayed on the screen in place of the link. This allows a Thinker document to be a composite of other Thinker documents. Details about the nature of "See Through Links" are found in the section of the manual describing the "Enable See-Thru" menu selection (under "Options"). Statement A statement is a block of text. Statements are like paragraphs in many word processing programs. Thinker does "word-wrap" within a single statement but not across statements. Text selection (by sweeping the cursor over text with the left mouse button held down) is limited to text within a single statement. Statements may contain up to 2000 characters of text including standard and extended characters from the Amiga keyboard. The <ESC> and carriage return characters are also allowed. View Specifications The appearance of the text in the window is controlled by three parameters that together are called the view specifications. These parameters are "Clipping level", "Spacing", "Number of Lines" and are each explained in the Options Menu section. Window Only a small portion of a Thinker document can be displayed on the screen at any time. Thinker uses Intuition windows to view portions of a document. Up to 8 windows may be open at any time. Each window may display a portion of a different document or different portions of the same document. Thinker Documents Thinker documents are not always meant to be read linearly. While this manual was created using Thinker, many uses of Thinker take more advantage of the Hypertext and Hierarchical text features. A skilled Thinker user creates documents with a definite flavor. These documents do not always read well when flattened and printed on paper and are meant to be read online. Outline nature of Thinker documents If a Thinker document is viewed with a clipping level of 1, only the first level statements (often these statements are topic names or section headers) will be shown. It is rare but not impossible to have blocks of text at this level. An executive summary is presented with the clipping level set at 2. As the clipping level is increased, more and more details are revealed. If Thinker is used to design a program one would expect to find code fragments or pseudo code at the deepest levels of the document. Hypertext features Information is only presented once. A reader should be able to start reading a Thinker document at the topic of most interest. All references to terms that require a definition in the context of the document should always have a link to the section of the document where the term is defined. The fact the Thinker will treat any word as a link means that a Glossary section could define all terms and contain links to the sections that provide further clarification. When a new term is encountered, the reader double clicks on the word to open a window over the Glossary section defining that term. If there is more information available then the Glossary entry will contain a link to the section of the document where the term is introduced and described in detail. Double clicking on the link can move the Glossary window to the detailed information. These links make Thinker ideal for online help where the reader can go directly to the section of interest and still be able to locate quickly all the pertinent information. See Through Links make it possible to automatically extract information from one document into another without double clicking on the link. A new Thinker document might summarize the information in many other documents by using See Through Links. When the information in the original documents is updated, the new summary document will display the changes. Planning a book One possible use of Thinker involves coordinating all the details of a novel. One technique that might be of value is to have several major sections of a Thinker document for various aspects of the planning process. One section would have a labeled branch for each character in the book. Other sections would describe historical events of interests and descriptions of each location. Each of these sections would have labeled statements for each detail. The section of the document that describes the story line (plot) of the book would contain links to the sections containing detailed information. The plot can be reviewed without having to wade through details that might confuse the issues. As needed, each reference in the plot to details about characters, scenes, events, etc. could be checked by jumping to the section of the document where this information is found. Multiple windows help organize the thought process. Small Example: (Characters) (Anastasia) - The heroine Lots of stuff about Anastasia (Scrully) - The hero Lots of stuff about Scrully (Places) (San Francisco) - Home of Anastasia Lots of stuff about SF (San Jose) - Home of Scrully Lots of stuff about SJ (New York) - Place of story Lots of stuff about New York (Plot) Anastasia meets Scrully in <New York> They marry and raise a family The End Organizing picture files Since links may be to IFF picture files, one use of Thinker might be to build a catalog of picture files. Each section of the document would have a first level statement with a label indicating the type of pictures indexed. The subordinate statements would contain descriptions of the picture and a link to the picture file. Pictures are located by jumping to the section of the document with the descriptions for the correct type of picture and then double clicking on the link to the correct picture. Small Example: (People) (Lauren) <Thinker1:art/lauren,> (Bob) <Thinker1:art/bob,> (Birds) (Bluebird) <Thinker1:art/bluebird,> ..... Thinker documents as databases Labeled statements in Thinker documents are much like indexed records in a database. There is a great deal of flexibility above the typical database, however. Records are not a fixed format. Records can contain a variable number of pointers to other records. Records can be part of a document meant to be read. Using Thinker to write is like writing inside a database as references can be checked with the click of a mouse button. Using Thinker is like having a completely free form database with references to applications, pictures, and documents. A Small Example: (Distributors) (American Software) - Address (Dealers) (Winner's Circle) - Address (HT Electronics) - Address (Unregistered buyers) (Nymous-A) - Address Orderdate: Shipdate: (Registered buyers) (Huxtable-P) - Address Orderdate: Shipdate: Amount: Registration: Thinker as a desktop organizer A desktop document can be used as an alternative to the Workbench. Within the document are labeled statements covering various working task areas such as Letter Writing, Programming, Finance, etc. Subordinate to these statements are labeled statements for each task in that area. These statements describe the task and may include sentences describing the progress of the task. Finally, there are links within the statements that name the Project or Tool used to work on the task. A description of a program might contain the name of the source for the program so that you can link directly to the source and work on it within your favorite editor. Other statements can link to other Projects and Tools, any of which can be launched in multitasking mode. A Small Example: (new projects) (Thinker) - Learn thinker <Thinker:example,> (old projects) (whoopydo) - Work on favorite program <Source:whoopy.c,> (appointments) (Monday) - Meet Alice at Restaurant Viewing a Document A Thinker document is a hierarchy of statements. There are three view specifications that control what portions of this hierarchy is visible on the screen. Clipping level The clipping level controls the depth of the hierarchy that is displayed. A clipping level of 1 displays only the first level of the hierarchy (the major topics) and as the clipping level is increased more details are displayed. There is a global clipping level controlled by menu options and the Clip gadgets at the top of the screen. Each branch has a local clipping level controlled by gadgets in front of the statements. Local clipping levels are always higher than the global levels. A document can be searched rather quickly by starting with a global clipping level of 1 and using the local clipping gadgets to expand the areas of interest. Number of lines shown in each statement The document can be viewed as an outline by setting the view specifications to display only the first one or two lines of each statement or the full document can be viewed by setting the view specifications to display all lines in each statement. Spacing between statements The spacing can be adjusted to maximize the amount of material on the screen or to maximize the readability of the material. One views a linear document by scrolling over the document. One views a non-linear Thinker document by moving the window to particular places in the document with the Jump command. Jumps can either be relative to the document structure or can use links. Structure relative jumps include moving the window up and down in the hierarchy or forward and back at the same level. Links provide the most flexible way to navigate around a set of Thinker documents. A link in the text ties the displayed text directly to text elsewhere in the document or in another document. Links may be typed into a string gadget or found displayed on the screen. If a link is displayed on the screen merely double clicking the cursor over the link executes the jump. Any single word in a Thinker document can be a link! For those times when it is desired to scroll linearly through a document several scrolling gadgets are provided. Scroll forward can be by single line or single statement. Backward scrolling is by statement only. There are Jump commands that jump forward by a full screen or back by a statement. It is worth remembering that Thinker documents can be much larger than the Amiga Memory and only the active portions of the document are in memory. This and the fact that Thinker documents are randomly accessed files makes it impossible to provide scroll bars. Editing Text within a statement is edited much like text in a standard word processor. The cursor is positioned by using the mouse or cursor keys. The cursor keys move the cursor within a statement, between statements, and will cause line by line scrolling forward or statement by statement scrolling backward. When the mouse is used to position the cursor, the left mouse button is used to mark the place in the text. Subsequent typing inserts characters at the cursor location. "Backspace" deletes a character to the left of the cursor and moves the cursor to the left. "Delete" deletes a character to the right of the cursor and does not move the cursor. All printable keyboard characters are allowed in text (including alternate characters). In addition the <ESC> character and <CR> character are allowed. The <ESC> character is useful for imbedding printer commands in text. The <CR> (carriage return) character is used in formatting tables as a single statement rather than as a group of statements. The <CR> character is visible as "~" and causes the cursor to move to the next line of the statement in the first column of the statement. A <New Line> character is put into the text. A block of text may be selected by holding down the left mouse button while moving the cursor over the text on the screen. Only text that is totally within a single statement can be selected in this fashion. Once selected, the text can be "Cut" from the statement and "Pasted" at some other location. Selected text is replaced by any typed characters and the selected text is forever lost. Undo is supported with the "Revert" menu selection. All modifications including structural changes are deleted. The user may select "undo points" by using the "Save" menu selection. In fact no changes are recorded until Save is selected. There is a reminder warning presented if the user attempts to leave Thinker without a Save. The requester gives the opportunity to save or discard the changes. Search and Replace operations are selected from the Edit menu. Manipulating the Structure Statements in a Thinker document may be moved within the hierarchy, moved to another document, copied to other places in the same document, or copied to other documents or deleted. The movement, copying, or deletion of statements does not require that the entire structure to be moved, copied or deleted be visible on the screen as parts of the structure may be obscured by the selection of viewing specifications. It is the ability to manipulate whole structures of the document that make the hierarchical text such a powerful tool. The viewing specifications can be set to view the document in its outline form (set the number of lines for each displayed statement to 1) and the entire document can be restructured. Move and Copy and Delete options are available for individual statements, branches, and groups of statements or branches. In all cases a sequence of requesters walks the user through the operation. Some of these requesters have a "Previous" gadget which will recall the previously typed string gadget. The structure manipulation commands can be initiated from the menus, from the gadgets at the top of the window, and from the "power tools" requester. The "power tools" requester is initiated by double clicking on the right mouse button. From this requester it is possible to directly insert a statement or initiate a Move, Copy, Delete, or Jump command. A Tutorial Example The first part of the tutorial will walk the user through the process of building a Thinker document. The second part of the tutorial makes use of an existing Thinker document called "example". In the following examples do not type the '"' characters. The Thinker document "Example" is a tutorial that leads the user through a short example of using Thinker to view and manipulate a document. Duplicate the EXAMPLE document using the "Duplicate" option of the "Workbench" menu before trying the tutorial. If you want to recover the original EXAMPLE document you can discard the EXAMPLE icon and "Rename" the "copy of example" icon. Start the first part of the tutorial by double clicking on the Thinker icon. This will start the Thinker program and bring up a window with the basic gadgets and menus but will have no document displayed. At this point we might jump into some existing document or create a new document. One would jump into an existing document with the Jump Link command and create a new document with the Create New menu selection (in the Project menu). As an example of jumping into an existing document, there is a section in the "example" project that explains the format of a link. The label on this section is "link". To jump to that section, begin by selecting the Jump Link command (either from the menu bar or by selecting the "Jump" gadget at the top of the window and then selecting the "Link" gadget in the Jump Requester.) Next, type in the string "example,link" (without the " marks) followed by typing the "RETURN" key. Finally, select the "This Window" gadget in the Jump Confirmation Requester. Typing the string "example," (note the comma but no label typed after the comma) would jump to the beginning (origin) of the document "example". Whether or not you take the diversion suggested by the previous paragraphs, you create a new Thinker document by selecting the "New" menu option from the "Project" menu. Do so now. Note that a requester pops up inviting you to enter a project name. The first part of the requester is filled in with the name of the current Drawer. Since you double clicked the Thinker icon directly, this Drawer name will be the Drawer in which the Thinker program was found. The Requester is already activated so simply type in the name of your new project. In this example we use the name "tutorial". If you entered the name "example" you will be greeted by a requester asking if you want to replace the "example" project that already exists. You should respond by selecting the CANCEL gadget and begin again typing the name "tutorial". After some disk activity the window will display the new document (note the change in the window title) and the document will contain a single statement (a Copyright notice). The first thing to do is to replace the Copyright notice with the first statement of our new document. You might try using the Delete (Branch) command but you will find that you cannot delete the only branch in a document. Move the pointer over the Copyright notice and observe that while the pointer is within the statement it is shaped like a cursor. Move the cursor shape to the very beginning of the statement and press the left mouse button. While holding down the left mouse button, move the cursor to the end of the statement. Notice that the selected text becomes highlighted as you move the cursor. When you have the entire statement selected, release the left mouse button. Type in the string "What I did last Summer" (do not type the RETURN key). Notice that the selected text is removed and replaced with the typed string. Now we are going to add some statements. This initial document will look a great deal like an outline. Remember that any statement you type could be as long as 2000 characters and fill the entire window. In order to keep this first example short, we will only type a few words in each statement. Select the "Insert" gadget at the top of the window. Note that the pointer changes into a hand with a pointing finger. Move the hand so that the finger points at the first statement. Press the left mouse button and notice that an Insert Confirmation Requester pops up with the pointer pointing to the word "After". If you wanted this new statement to be directly after the first statement, you would simply press the left mouse button again. Instead, move the pointer over the word "Down" in the requester and press the left mouse button. Notice that the text insert cursor appears on the screen just below the first statement and indented by two characters. Type the string "(school)Went to Summer touch typing classes.". Select the Insert gadget again and move the hand to point to the statement you just typed. Press the left mouse button twice (the first time selects the statement and the second time selects the word "After" in the confirmation requester.) Type the string "(beach)Went to the beach with Bob and met his little girl Lauren." (Who is Bob?) Bob is someone that attended the same typing class so you will want to put any information about Bob in statements that are subordinate to the statement labeled "school". Position the pointer over the first statement (the one labeled "school") and double click the right mouse button. The Insert Confirmation Requester should pop up and you should select the "Down" gadget. Type the string "(bob)Met Bob, a computer science student who wanted to learn typing skills." Move the pointer back to the last statement "(beach)Went...." and position the cursor over the word "Bob" and double click the left mouse button. Notice that the Jump Confirmation Requester pops up with the pointer positioned over the words "This Window". Move the pointer over the word "New Window" and press the left mouse button. Notice that the new window opens with the statement "(bob)Met Bob,..." as the anchor of the window. This is an example of a hypertext link. "Bob" is the label on the statement "(bob)Met..." and the double click of the left mouse button with the cursor over the word "Bob" was a Jump Link command. The same effect could be caused by selecting the "Jump Link" command from the "Commands" menu, or by selecting the Jump (Link) gadget at the top of the window and typing the characters "bob" followed by RETURN. Who is Lauren? Position the cursor right after the word "Lauren" and press the left mouse button once. The text input cursor should appear between the "n" and the ".". Type the string " (See <Thinker:art/lauren,>)". This is a link to a picture file. Note that the comma must be present to indicate that the link is to some other file and not the document shown in the window ("tutorial"). You have just added a hypertext link to an IFF picture file. This link had to be surrounded by "<" and ">" because it contained punctuation characters (":" and ","). Move the cursor between the "<" and the ">" and double click the left mouse button. When the Picture Confirmation Requester pops up select the "Full Screen" gadget. After a pause, a picture of Lauren appears (From the NewTek "Digi Paint" package.) Typing any character or pressing the left mouse button will remove the picture and return the Thinker screen. Repeat the selection of the picture link and select the "New Window" gadget. This time an approximation of the same picture will be drawn in a new workbench window that you may move around on the screen. This approximate picture is drawn with the 4 colors available on your workbench. The window has an invisible close gadget in the upper left corner that will close the window when selected. If you want to save the changes you have made to this new project, select the "Save" menu selection from the "Project" menu. "Revert" will restore the project to its initial state (with the Copyright notice). This completes the first part of the tutorial. The second part of the tutorial uses an existing document called "example". You can begin by closing the Thinker window (select the close gadget in the upper left corner) and starting all over again by double clicking on the EXAMPLE icon, or by doing a Jump Link to "example," (don't forget the comma). In either case you should read the next paragraph before going on with the second part of the tutorial. Note: When a file name is selected in the file requester, the comma is optional unless you wish to add a label. Start the second part of the tutorial by double clicking on the EXAMPLE icon or by selecting the icon and use the "Open" option of the "Workbench" menu. Read carefully and follow the instructions. It is helpful if you read ahead a little before executing the instructions. The tutorial is not very long or complicated but is intended solely to give the reader a little familiarity with the behavior of Thinker and build a little confidence for the first time user of Thinker. Not So Obvious Features Most of the operation of Thinker can be discovered by exploring the menus and selecting the gadgets. There are, unfortunately, some behaviors that are not manifest by this exploration. There are also some curious restrictions. Poor Person Software tries hard to produce software that is useful and complete but without all the frills that would make it expensive. Restrictions When the Mark option is used for both the beginning and end of a group, both ends of the group must be in the same window. Only 19 characters of a label are significant. Only the first 32 statements with duplicate labels can be reached via a Jump Link. Features Window Memory Buffer When statements are displayed in a window, the text information is in a memory buffer where it is easy to update the data based on the keyboard input. Any time that a Jump command is executed, "Save" is requested, or another project is opened, the changes in the memory buffer are moved into the Update Pool. When there are changes in the Update Pool a "*" character appears in the title line in front of the "Thinker n.nn" string. Since the Window Memory Buffer is not moved into the Update Pool at each keystroke, the "*" will not appear immediately upon making a change in the window. Selecting the window sizing gadget will write the Window Memory Buffer into the Update pool. The presence of the Window Memory Buffer is only important when there are multiple windows open over the same document. Changes made in one window are not visible in another window until the Window Memory Buffer for the changed window is written into the Update Pool. The previous paragraph contains a discussion of when the Window Memory Buffer is written into the Update pool. The Update Pool is where modifications to a Thinker document are stored until the user selects either "Save" or "Revert". After one of these selections the Update Pool is emptied. Read Cache Thinker keeps a cache of the all modified portions of a Thinker document in its memory (the Update pool) but only a small portion of the document that has been read from the disk is kept in memory (a read cache). Since Thinker documents are randomly accessed AmigaDOS files, updating the screen could be a painfully slow process without this read cache. The read cache is small and can be turned off if some Floppy Accelerator like Facc II from ASDG is used. ASDG did such a good job that Poor Person Software did not feel that it was worth the time to include much of a read cache in Thinker. The read cache is large enough to insure that screen updates are reasonably fast. As an alternative, small documents can be copied into the RAM: device and manipulated there. When you are finished and have saved all your changes, copy the document back to a disk. Undo Once modifications have been applied to a Thinker document they cannot be "undone". However, Thinker will accumulate all the updates to a document in its memory (the Update pool) until "Save" is selected in the Project Menu. The user should learn to schedule updates to allow for backout of recent errors. Remembering Local Clip adjustments When the clipping level for a branch is adjusted using the "+" gadgets in front of statements, this action is recorded in a small in-memory data base. This database has room for 32 selections of "+" gadgets in each project. If the 33rd entry is made into this database, the oldest entry is purged. Power Tools Double clicking the right mouse button with the cursor positioned over a statement will cause Thinker to enter "power tool" mode. An insert confirmation requester and a set of tool gadgets will appear and invite the user to select a relative level to insert a new statement (as in the last step of the insert command) or to select a Jump, Move, Copy or Delete tool. If the pointer is within a statement (showing the text pointer) and the user selects one of the levels, a statement will be inserted. If a tool gadget is selected Thinker behaves as if the tool gadget at the top of the window was selected. The number of open documents is limited to 16 When Thinker jumps to a new document, it must open a new AmigaDOS file. Since there may be outstanding updates on the document that was jumped from and you may soon return to that document, the document is not closed. After 32 such jumps you will have to use the "Close" menu item under the "Project" menu to close some project. Close closes the document visible in the active window. Fast Jump Link Double clicking the left mouse button with the cursor positioned over a link (or word) is equivalent to doing a Jump Link command with the Mark option and positioning the hand over the link (or word). Keyboard Confirmation of Jumps In all cases when the confirmation requester is brought up to complete a jump. The confirmation can be effected by typing the RETURN key on the keyboard. Summary of Release 2.1 features With the features added in release 2.1, Thinker has become a general purpose word processor and database as well as a powerful Hypertext idea processor. The major features are summarized below. The section that follows this one describes the rest of the enhancements, and following that are the details of all of the changes are organized in a fashion that parallels the Thinker manual. Images as Statements Images that are displayed in workbench windows can be incorporated into as statements in a document. To make this option more attractive, the picture display options have been expanded to include a new smaller size and extended color selection using a color dithering technique that allows the display of 10 colors on the Workbench screen. The new smaller size is 160 x 50 pixels on a non-interlaced screen or 160 x 100 on an interlaced screen. This size has been dubbed the "Xfer" window size as it is anticipated that this is the most common size to transfer into a document. Pictures are first displayed in one of the workbench windows and then the most recently displayed or touched (click the mouse over the window to touch it) picture window can be inserted into the document using the "Insert Image" command. No provision is made for printing pictures. Printed or exported documents will display a small box for each included image. Mixed Text Styles and Colors It is possible to display portions of statements in any of the standard text styles (Italic, Bold, Underlined) and any combination of these styles. Simply select the text by sweeping the cursor over the text with the left mouse button held down and then use the Style menu to select the style. A typing style may be selected for future typing by selecting from the Style menu without selecting text. Foreground and background text colors are handled similarly. You may wish to select an alternate color for text that represents a link but the color selection is NOT considered when isolating or selecting links. Saved Environments A complete multi-window, multi-file environment can be saved and restored later. The Save Options menu selection saves the entire environment in the icon of the file displayed in the window that is active when the menu option is selected. When this file icon is used to launch Thinker, all the windows that were active when Save Options was selected will be opened displaying the same files. If any window was positioned at a label in a file when Save Options was selected then the window will be positioned at that label when Thinker is launched. Options in the Thinker Icon The default options for Thinker can set in the Thinker icon. Any time that Thinker is launched the Options and Behavior settings are first read from the Thinker icon and then from the file icon. The options strings in the Thinker icon can be inserted and edited using INFO from the Workbench. Improved display speed while typing The display speed has been improved and now keeps up with moderate speeds of auto-repeat. New Behaviors This section describes the remainder of the Release 2.1 features. Be sure to check to sections that follow for the details of menus, requesters, and gadgets. CLI Interface The special link <CLI,command> uses ICONX to execute the "command" command in the CLI environment. "Command" can be any legal CLI command or set of commands separated by line feed characters. Line feed characters are entered at the keyboard while typing into a statement using the "Return" key and show up on the screen as ~. When typing a CLI command string into the Jump Link string gadget, type the "~" character to indicate Carriage Returns. The ICONX program must be in the C: directory for this special link to work. File Name Prefixes For some applications it is awkward to have the full pathname of files as part of links. If the file is moved from one floppy disk to another or from one directory to another, then all the documents which contain the link must be changed. The use of file name prefixes alleviates this problem. The CLI environment has a mechanism called ASSIGNS that solves this problem for CLI programs. If fact ASSIGNS can be used for Thinker links. A link of the form <ABC:filename,label> would look on the floppy disk with the label ABC: or if there were an active ASSIGN of the form "ASSIGN ABC: HD0:xyz" would look on the hardisk in directory "xyz". Thinker has a similar mechanism. You can use INFO to add file name prefixes to the Thinker icon. These prefixes act similarly to ASSIGNS. If there is a tooltype of the form "ABC:=HD0:xyz/" then links of the form <ABC:filename,label> would be resolved as "HD0:xyz/filename,label". The string on the right side of the tooltype is substituted for the string on the left side when found at the beginning of a link file name. Cursor Positioning The cursor can be moved quickly to the beginning or end of statements or lines using combinations of the cursor keys and the Alt or Shift keys. Shift-LeftCursor moves the cursor to the beginning of the line. Alt-LeftCursor moves the cursor to the beginning of the statement. Shift-RightCursor moves the cursor to the end of the line. Note that the cursor always appears at the beginning of the next line when it is logically at the end of a line. Alt-RightCursor moves the cursor to the end of the statement. Keyboard scrolling Scrolling can be accomplished using the Up/Down cursor keys in combination with the Alt/Shift keys. Alt-Up and Alt-Down scroll the window one statement backward and one statement forward respectively. Shift-Down scrolls the window forward a single line. Null Replacement Strings When replace is selected from the Search/Replace requester the replacement string is now always used. In previous releases a NULL replacement string was treated as a search only. In this release the NULL replacement string is treated properly. The located string is removed from the text. Minimizing Jump Return Points Scrolling the window no longer fills up the Jump Return list. However, if one jumps to a statement and then scrolls forward followed by a jump, the last scroll position is put into the Jump Return list. Thus the sequence Jump to "A", scroll to "B" and jump to "C" would leave "A", "B", and "C" in the Jump Return list. Jump return from "C" will go to "B" and then to "A". The positions between "A" and "B" that were visited as part of the scroll operations will be skipped. Wider Requester for Duplicate Labels When duplicate labels are presented for selection on a Jump to a label that is duplicated, the entire width of the screen is used to display the text of each of the statements. New Definition of Up For the Insert, Copy, and Move commands the definition of "Up" has been changed. Selecting "Up" from the confirmation requester puts the statement(s) immediately above the marked statement. This is equivalent to splitting the plex at the insertion point and putting the second half of the plex subordinate to the inserted statements. Search behavior with small windows In previous releases a search that located a string in the second half of a large statement might not show the located string on the screen if the entire statement did not fit in the window. This situation was particularly evident with small windows displays only a small portion of a statement. This error is corrected in this release so that the located string is always displayed even if the statement must be displayed starting in the middle. In effect the statement is scrolled until the located string is the first thing displayed. If you need to do a CHANGEALL operation, you will find that it proceeds much faster in a small window. Clock alternates with Free Memory The clock display in the upper left corner of the screen alternates with a free memory display. Automatic Close of Unneeded Files With this option selected, files are closed as soon as they are not displayed in any window and have no updates. The advantage is that if you jump from file to file on several floppy disks while browsing, you will not be asked to insert all the floppies when you select Quit. Case Independent Sorts An option on the Sort command allows a choice between honoring or ignoring the case of characters during the sort. If the case is honored, then the lower case alphabet follows the uppercase alphabet. Link Extensions Labels in links can be followed by a series of directives that extend or refine the link. Commands include movement relative to the label (Down, Preceding, etc) and highlighting a phrase within a statement. Line Length Selection The length of the line used for PRINT and for EXPORT with PREFIX=! can be selected.