Building an academic site

1. Every significant file must have a full file name and location

An academic site has requirements that other sites do not. The first priority is for a clear system of page referencing. Almost every page must have on it the link, so that if someone saves or prints the page, they can find it again. One of the current most frustrating problems academics have is remembering where the saved page or saved file comes from. Until some automatic stamping system is developed then the academic community must take this into their own hands and make sure that everything savable or printable has the absolute url on it.

The way I have solved that problem is to build vital referencing information into the headers of my pdf files.  In addition, the first page has the copyright notice, title, and when the document was last updated. Again, the ‘last update’ information is crucial - especially when minor tweaks are made, and it is important for someone to reference the exact version of a document.

It is important to think through carefully the headings: changes are possible later, but are a lot of work.

2. First buy your domain names - all three net org and com

I had to first buy my domain name, years before I had the site ready. But that was the easy bit.

3. Forget instinctive ways of thinking

What I wanted to do was to set up a master page - within Word - as an html file, then link from that file to the subdirectories and file names. This is extremely easy to do. Then I wanted a program which would load my files, allow me to add navigation etc, then upload the ‘site’ I had created on my hard drive, and change the root to the www. website name. I even considered as a pilot attempt asking someone to set up a small partition on my own computer as a ‘server’ then I could put in the full website name as my root from the word go. Later when satisfied I could transfer the site  to a commercial hoster.

You see, when I view a website, I view it as a glorified tree system of folders and files. The trouble is, the syntax of a website is not the same as the syntax of local files.

on the computer


becomes on the web


Now, programs exist which download whole sites or parts of sites, then  export them onto the Hard Disk and so doing change the syntax of the page names. Maybe I could have got something like this to work, but, I lacked the ability to test it - my own local server.

4. Purchase a website maker

Unfortunately an easy perfect program does not exist. I chose Serif X2 though X6 is now available as a stable release.

How to construct an academic site using Serif X2.

Critique of Serif X2. Review of Serif X2.

1. To my horror, I found that Serif did not take kindly to html files. You cannot create an html file then import it or link it. It is easy enough to turn a Word Processor text into html, but Serif does not like it.

2. Importing from Word Perfect was a disaster. I am using version 8 - an old version admittedly, but the Word Perfect standard for basic files has stayed more or less the same since versions 6. I finally, after many hours and many attempts realised that Serif needs to import files from WordPad. Therefore I have to copy and paste into WordPad, and repeat the copy and paste from WordPad into Serif.

3. Page length is a problem in Serif. If you make it too long you end up with a lot of white space. If you make it too short you lose material, or you are forced to make new pages - and that is strictly against the academic site requirements, where every single page has a unique logical filename and location that fits into a tree just like navigating a hard drive. So, what I do with every new page is to base the page on a copy of a previous page - in this way I get the navigation bars etc that I have set up. Then I delete what I do not want, and add text. When it overflows I go into page properties and alter the ‘height’. Then I extend the html frame, and move the bottom navigation box.

4. Unlike Serif, my priorities are printing and specifying file names. Serif want to break a long piece down into a series of linked files. I think asking someone to scroll an html page is normal. Far more important is this. It is vital that when someone prints an html page it prints with page breaks in reasonable places. I need to see page breaks as I am building a page, and adjust content arrangement accordingly. The editor in Serif does not show the page breaks. For that you have to go to page preview in your browser then print preview within the browser, and go back and make your adjustments. Phew!

I had to set up the page structure, and choose to move files to subdirectories within Serif. I wanted to use absolute file location but was dissuaded from that. Serif Help files are not very helpful on this point.

5. Fortunately, most academics use pdf files, therefore I have got into the habit of turning my work into that format. Linking to a pdf file within Serif is easy. But it is very important to go into the Export Options, and there to set up an actual tree structure, representing the original tree design, and place the file into the required directory.

Note: not all these directories are yet visible, or yet included in the navigation. There are some extra directories added by Serif.

6. Use a clean page layout.

One of my pet hates is the way html files print with junk on them. Therefore I decided, except for the first page, to NOT have a sidebar. In addition, the sidebar and topbar should be text rather than image.

7. Poor help files within Serif

Now, with the exception of linking to pdf files, most of this information I had to discover for myself. The helpfiles in Serif are incredibly brief, and not at all context sensitive. The tutorials seem to teach what I do not want to do and leave you stranded with the basics I have covered above. For instance, I recently looked in the Help file for “Site index”. Now this is one of the menu choices, but the Help files have no information on it whatsoever.

But Serif Forums are quite good - though well used, so you have to work hard to sort out the information you want. Several people have commented that the easiest way to have a navigation bar is to do it with a text box and hyperlinks. Sound advice, but NOT in the introductory wizards and help files.

8. Sound files. A friend hearing me argue my case encouraged me to record some of my pieces. I use Audacity, but have found the tagging of files to be quite misleading. I had to really hunt to find this information, but eventually I found a free program called Mp3tag, from which enabled me to correctly add tag marks to the files and to export a playlist which was readable by VLC and Windows Media Player. I experimented with Mp4 tagged files, but could not get them to work, so stuck to the simple solution as above, and provide a collection of files as a zipped package.

9. Serif’s excellent Site Checker Tool.

The time comes to check the website before uploading. This is easy to do via File > publish site > publish to folder. At this point you get the chance to do a full site check using the Site Checker Tool

Make sure you choose full scan and all site problems.

Also, let your firewall give access to Serif, because it will check the links in the html files though not in the pdf files. When it finds a mistake in the file it will offer to correct it and it did so in this site when I had the link visible and typed in the page, by making it unclickable. How it corrects a hyperlink where the URL is not put in full, I do not know.

When finished checking, let it save to a folder,

Then open the folder in Windows Explorer.

10. Handling files that are not in the right place

Remember, one of the chief requirements for an academic site is accurate file name locations. After using the Site Checker tool you find a list of the files that are not in the right place.

All the pdf files in this case are in the wrong place. To take one example. The file adult-l1-acquisition.pdf should be in the adultl1a directory.

To correct this, go back to the original hyperlink. Mark the link, Right-click >hyperlink

Notice I have checked the box, “Embed file into site”. I unchecked that, but that is not the problem. Click on ‘Export Options’ and make sure the file is exported to the directory structure which directly matches the planned tree structure.

That should fix it. But you can only see that you have fixed it by visual inspection. A NEW copy will be put in the folder, and the old one left behind. You can see this by inspecting file-date and times saved. Check to make sure you do not have more than one link to the same file. Then save as usual, but this time to a new folder. Only in this way are the files actually deleted from the root. Now I have:

As you can see, there are still filenames in the wrong place. reviews.html should be in the directory reviews, to to put it there, go into page properties.

The problem is that the file name does not include the folder. So, click on the box ‘change’ then click on the folder you want it to be part of.

In this way I put it in reviews, where it belongs.

Similarly, contact.html is in the wrong place. The other pages are either not important or can be left alone.