Intkey can access its data and image files over the Internet. To do this, the data-set index file (see ‘Introduction’ in the Intkey online help) or a link in a Web page must point to a special startup file, with extension (type) ‘.ink’ or ‘.intkey’. The startup file tells Intkey where the data set and its associated images are to be found, and the name of the intialization file to use.
When Intkey is installed, it configures itself as a ‘helper application’ for Web browsers. When a person using a browser clicks on a link to an Intkey startup file, the browser activates Intkey and passes it a copy of the startup file. Intkey itself then retrieves the actual data set from the Web, extracts its contents, and begins the identification.
Note that Intkey runs only under Windows 95/NT or later. Even with Web-based data sets, users will not be able to use Intkey on Macintosh or Unix computers (unless they use a Windows emulator). However, only the computer in which the Web browser is running needs to be running under Windows; the Web server that provides the data set can be any sort of platform.
Here is the procedure to follow to make an Intkey data set available over the Web.
1. Generate and test an Intkey data set on a Windows-based PC. Be sure that it does what you want on your PC before you consider putting it on the Web.
2. Assemble together all the files that Intkey requires to use your data set. This will typically include the following:
(a) the data files ‘ichars’ and ‘iitems’ (not your original ‘chars’ and ‘items’ files);
(b) the initialization file, usually named ‘intkey.ink’;
(c) any Intkey ‘input’ files, e.g. the file ‘toolbar.inp’ defining a custom toolbar;
(d) any special .bmp files used in the custom toolbar (those already provided with the Intkey distribution need not be included);
(e) any ‘Contents’ files (usually ‘contents.ind’) and the files they reference (usually ‘*.rtf’).
Do not include image files for characters, taxa, keywords, or startup at this stage (you actually may do so if you wish, but a more efficient mechanism is provided for the retrieval of images on an ‘as-needed’ basis).
You can find the required files by examining your initialization file (‘intkey.ink’) and contents file (‘contents.ind’). Determine what files are referenced either directly or indirectly from these files, and include them all.
3. Put the files from step 2 into a single zip file.
4. Determine where you will later place the data set (and any associated images) on a Web server. You will need to know the full URL that the data set will have once it is on the server. You may need to consult with the manager of your Web server about this.
Create a startup file specifying five different values, each having the general form ‘keyword=value’. Use the extension ‘.ink’ (or ‘.intkey’) when naming this file, and save it as a plain text file. Comments are preceded by semicolons. Here is an example.
; To run the Intkey identification package from this file,
; you need Intkey5 and Windows 95/NT or later.
; Name of this file
; Name of the compressed data file
; Name of the Intkey initialization file within the compressed data file
; Image path
; Image path
The value for ‘InkFile’ specifies a fully qualified URL at which the startup file itself can be found. This enables Intkey to re-locate the data set in a later session without the help of a Web browser. This entry is mandatory.
The value for ‘DataFile’ specifies a fully qualified URL pointing to the zip file created in step 3. This entry is mandatory.
The value for ‘InitializationFile’ tells Intkey the name of the file within the data set archive to use as its initialization file (usually ‘intkey.ink’). This entry is mandatory.
The entry for ‘ImagePath’ is optional, but should be used if your data set contains images that have been associated with the data via the Confor directives CHARACTER IMAGES, TAXON IMAGES, CHARACTER KEYWORD IMAGES, TAXON KEYWORD IMAGES, and STARTUP IMAGES. It should indicate the URL of a directory from which the various image files can be retrieved. This should normally be a subdirectory ‘images’ of the directory in which the ‘DataFile’ is stored on the server.
The entry for ‘InfoPath’ is optional, but should be used if your data set contains external ‘information’ files that have been associated with taxa via the Confor directives CHARACTER FOR OUTPUT FILES, ITEM OUTPUT FILES, and TAXON LINKS. It should indicate the URL of a directory from which the various information files can be retrieved. This should normally be a subdirectory ‘info’ of the directory in which the ‘DataFile’ is stored on the server.
5. Place the files on your Web server. The startup file should go into the location specified by InkFile; the zip file containing the data set should go into the location specified by DataFile; and any image and ‘information’ files should go into the directories specified by ImagePath and InfoPath.
6. Both the HTTP server and the PC must be able to recognize that the startup file references an Intkey data set. The HTML protocol uses ‘MIME types’ to distinguish how different files are to be used. Your startup file should be associated with a MIME type of ‘application/x-intkey’. This association needs to be set up by the manager of your Web server. The exact method for doing so will depend on the server software in use, but in most cases it will simply involve defining an association between the file extension ‘.ink’ (or ‘.intkey’) and the MIME type ‘application/x-intkey’. A similar association on the PC is generated automatically by the Intkey installation procedure.
7. You will probably want to create an HTML document pointing to your Intkey data set (see sample file ‘ident.htm’).
8. Test to see whether it all works. N.B. The Internet Explorer browser does not need the MIME type to be set on the server, so it is advisable to test with another browser as well, e.g. Mozilla Firefox or Opera.
There may be circumstances in which you may wish to restrict access to your data set. There are two basic ways this can be done. The first approach is to have the manager of your Web server deal with this via the server’s security features. The second is to encrypt your zip file when you create it (step 3) and protect it with a password. Intkey will then require that the user enter the correct password before the data set can be used. Your needs will determine which of these approaches is most appropriate. The first approach could be used to restrict access only to PC’s located on a university’s campus. The second approach could be used to restrict access to students within a single class, or to world-wide collaborators.
|DELTA Home Page|