|
Otherwise, page indexing by the spiders of some search engines may be hampered by a few seemingly ordinary characters. The latter are transcribed here as HTML entities, such as ? for the question mark (?). [back] |
|
|
|
<br><br></body></HTML> Side margins are obtained by inserting the page contents into tables defined by a code such as: <center><table width="95%" border=0> [...] </table></center> where [...] stands for the contents, properly distributed between the tags <tr>...</tr> (defining a row within the table) and <td>...</td> (defining a cell within the row). The inter-column space is obtained, if necessary, by creating an empty cell with the code <td width="1%"> </td>. The width of the two cells used by the columns is then adjusted in such a way (here "47%") that the total width of the row matches the width chosen for the table (here "95%", i.e. 95% of the screen width, irrespective of the resolution). Centering the table is of course intended for obtaining equal margins on each side. For more detail, you may examine and/or save the source code of a page in which you are interested. You will perhaps find there the tag <object>...</object> in places where it does not seem of any use. I initially added it where requested by the W3C validation process, but the W3C robot seems to have been corrected in this respect. I have since tried to remove this tag from any place where it was not requested. [back] |
|
|
|
|
|
Javascript error on line 56: http://this_site is not a number You then have to click an OK button to close the window, and things go on more or less as they should do. At the third stage, no link works on the displayed page, not even the browers's back button, and the only escape way available is closing the browser. At the fourth stage, the latter function itself is unavailable, and there is nothing else to do than shutting down and/or rebooting the computer. Some variations with respect to this range of effects may occur, such as, for instance, the freezing of the system while only the Javascript code is displayed, to the exclusion of any other page contents. A particularly frequent occurrence is the case where the browser finds itself confronted with a .js file without knowing how to handle it. A Javascript code set has then been called from an external source file by a code such as: <script src="source_file.js"></script> Some browsers do not implicitly recognise the Javascript code or the .js suffix. For them to behave as expected, the above code should have been written as prescribed by the HTML norm, i.e.: <script type="text/javascript" src="source_file.js"></script> More generally, neither the various browsers nor their script interpretation engines react in the same way to HTML or Javascript programming errors. Netscape 3 for instance tolerates errors in the opening/closing scheme of HTML tags that would be disastrous with other browsers and accepts that non strictly numerical parameters (such as the relative width "95%") be written without quotation marks, which is prohibited by HTML norms. The Javascript engine of the same Netscape 3, by contrast, seems to be much less tolerant of similar errors than the engine of other browsers. In order not to display the above-mentioned alarm window, it would probably have required that http://this_site be written "http://this_site". Testing a Javascript code set with a single browser, especially with the latest version, in which new specific functions may have been implemented, is therefore not enough to ensure that the code is correct. Whereas the full conformity of an HTML code set to the relevant norms can be tested, independently of any browser, through the W3C validation service (or other similar services), I unfortunately do not know of equivalent services for Javascript code. In the absence of such a validation, the wisest choice consists of abstaining from using the code, all the more so because many internauts disable Javascript on their browser as a security measure. [back] |
|
|
|
|
|
|
|
|
|
|
|
|