WebWonder reviewBy Martin Hansen. Published: 2nd Sep 2006, 14:58:52 | Permalink | Printable
Creating a simple website in a few hoursDespite having written numerous articles for Drobe, writer and software developer Martin Hansen had, until now, yet to produce his own website. Here he talks us through his first experiences with web publishing and the website design package Web Wonder.
It's fashionable these days to give yourself forty-eight hours to achieve something big; decorate a room, climb three mountains, learn to speak French. A few weeks ago I decided that such was the time that I would give myself to write, from scratch, a website to promote my digital art and get it online. I had no significant previous experience of doing either of these tasks but what I found out was that, with the right tools, it's not actually that difficult. I'm proud of the resulting online presence which is modest, but effective and looks elegant. This article is the back-story of www.mathmagical.co.uk.
Let me begin by remarking that getting old is a peculiar affair. I have been in as good a position as many to embrace the information technology revolution from it's mass market emergence in the early 1980s. In many respects I have ridden the IT wave but a part of me does not like having to relearn new ways of doing an old job when the old way does it fine, for me. So, for example, I've learnt to program very competently in BBC BASIC but opted out of moving on to C. I cut my teeth on RISC OS, but declined to move on for the Microsoft dream. It's not that I don't grapple with new things. More, I'm fussy about what I choose to invest time and effort into, and expect a long term return once I've done so.
|"It's not that I don't grapple with new things. More, I'm fussy about what I choose to invest time and effort into, and expect a long term return once I've done so."|
I've struggled a fair bit to work out what I want from the world wide web. Mostly, when I'm online, it'll be ebay or Drobe that I'm plugged into. Google is fantastic when the need for information on something obscure occurs, and then I'll wander as the search takes me. Increasingly, over the last four years, I've become aware that, as someone seriously involved in RISC OS, I should have my own website. As 2006 dawned I vowed that this would be the year in which I staked out my own small part of cyberspace.
In February I reserved the domain name, mathmagical.co.uk. I'd have also liked to have grabbed the domain than placed an "e" between "math" and "magical" but a theatre company had already purchased that. However, I bought a few other likely "mistakes" such as mathmagical.com and mathmagical.org.uk as I knew that these could all be made to point to my site. In all, I spent 34 quid on domain names.
Website with a purpose
I then did a lot of thinking. I decided that I didn't want the site I was going to create to be big or complicated. I felt that I needed something that would be simple to set up and straightforward to regularly update. Regular updates are a big part of what the 'net is about. Many RISC OS sites are embarrassingly out of date. I didn't want to set my site up to become one of them.
I mulled over the fact that blogging is the latest Internet fashion. I could regularly post news of enhancements to my software, and images of the art that I and others were creating from it. I didn't like the idea of an ever lengthening page of out-of-date news, but a constantly revised current blog entry seemed a neat idea. Many existing RISC OS sites are of no relevance to a non-RISC OS user but I had an idea that I wanted my site to be of some interest to anyone who should happen across it. This, in turn, led me to think that I didn't want my site to simply be an 'in-your-face' commercial for my software. So, emphasising what the software made, the digital artwork, the puzzles and the activities, seemed the right idea to pursue. This mix of ideas simmered for several months.
Three years ago, I wrote a single page experimental website in HTML with the aid of information from an article in Paul Beverley's excellent Archive magazine. I didn't get as far as placing it online. At a RISC OS show shortly afterwards I come across a website creation and editing package written by David Snell called WebWonder. What I liked about it was that it was all RISC OS, came with a printed manual, and seemed targeted at the beginner whilst also being able to do the job properly. In view of the time likely to be saved by not having to hand code every single line of html, I purchased version 1.01 of this software for £50.
|"What I liked about WebWonder was that it was all RISC OS, came with a printed manual, and seemed targeted at the beginner whilst also being able to do the job properly"|
The final necessary purchase was made last month; a web hosting package from 123-reg, valid for two years for £32. Now all I had to do was actually create the site code, using WebWonder, and upload it to the waiting server. The night before web-build day was spent reading the WebWonder manual. The particularly useful part seemed to be the Tutorial tucked away as appendix E, which I read through three times.
The most relevant part for me was the facility to create a table of thumbnails which could be clicked upon with the mouse to take the visitor to individual pages, each of which contained a larger image of the relevant thumbnail. For many potential users, I imagine that this makes it easy to get a collection of personal photographs online. As I had spent a lot of time over the summer using my ArtGraph software to create digital works in a modern art style, it seemed logical to put some samples of these, rather than photographs, onto the first edition of my website.
David Snell's WebWonder
It was clear that WebWonder offered many facilities, but I stayed focused on what would be required to do the minimalist job that I had in mind. I decided that all of my linking would be via image clicking, and that there would be no home page menu, no anchors, few external links, no file downloads, no forms or frames. The bottom line: I'd given myself a weekend to do this job. Any other free time over the next month had to be for developing the artwork itself. There are RISC OS shows looming in October, after all, and one has to arrive with new material on offer.
I've no wish to duplicate the instructional material of the WebWonder tutorial. If you look at my site you will see that I've closely followed what is in the manual. However, I made a fortunate discovery early on: a background tile can be added simply by dragging and dropping a JPEG onto WebWonder's Site Page Colours setup box. From the site menu, pick "Options" and then "Site colours...".
I grabbed the rough leather tile from the WebWonder Resources folder, converted it to a sprite, changed the colours from browns to greys and then saved it as a JPEG. Images for websites cannot be RISC OS sprites. The three formats currently popular on the internet are JPEG, GIF and PNG. On a RiscPC running Adjust, the built-in version of !Paint has options to export RISC OS sprites as JPEGs or PNGs. !Paint on an older RiscPC or Iyonix does not offer this option, but ChangeFSI could be used instead.
Adding text in various sizes is as easy in WebWonder as in any word processor. When developing a website, it's important to periodically look at it using a variety of browsers and in a variety of screen resolutions to get a feel for how it will look in these different situations. I mostly used NetSurf on my Iyonix and Oregano 2 on my RiscPC to check all was going as I thought it should. It's also worth remembering to hit the browser's refresh button each time a change is made to make sure you are not looking at an old cached version of a page. It was through doing this that I decided to centre most of my text and images rather than align them to the left or to the right.
In order to transfer my website from my Iyonix, which is not currently connected to the Internet, to my domain space at 123-reg, I saved it as a zip file on a floppy disk and hopped across to a Microsoft machine. There I used a free Windows program called WSFTP to upload the various images and files that as a whole made up my site. Images are transferred as binary files, HTML pages as ASCII. It is amazing when the transferred site then simply works. And the entire venture had indeed taken less than a weekend. FTPc should work equally well on a RISC OS machine connected to the Internet.
Over the following days I updated my Drobe user profile to include my new website address, edited my software details on Andy Jackson's Nutshells website, emailed Paul Vigay with a request to join the Acorn Webring and ordered the upgraded version 1.03 of WebWonder, now on CD and with an enlarged manual, from David Snell.
Becoming part of the World Wide Web
With my website written and online, my thoughts turned to how best make sure that someone who was trying to find me or my software could do so. There are various ways of moving up the listings tables produced by internet search engines such as Google or Yahoo. I'll discuss, briefly, three options.
Firstly, you can pay. This, I presume, is why a search for "drobe" on Google brings back a response "Buy drobe on ebay" followed by a link to the ebay website. This nonsense response suggests that Google and ebay may have some way to go in getting the arrangement between a search and what is actually available on ebay working sensibly. It's all very well getting your site at the top of the list by brute financial muscle, but if what you have to offer is not even remotely what the searcher is looking for, it's a rather pointless achievement. I decided not to spend any money doing this.
Secondly, you can enter a list of meta tag keywords in your website's
<head> section which search engines will look at in their own time and store in a database - although the more sophisticated web spiders will pay more attention to the actual content on the pages. None the less, when you search the web, this database provides a lot of the thousands upon thousands of results in the amazingly little time that makes surfing the web with broadband such a powerful facility.
WebWonder lets you add these meta tags easily along with another mega tag stating the website author's name. My meta tag list was simply of my software titles; artgraph, turtlechalk, mathbox and sudoku. I added RISC OS, Acorn computers and mathematics as afterthoughts although I want to think about all of these a little more carefully at a later date.
Thirdly, you can link to other websites. The more important the sites you link to the more important your website is deemed to be and, if you can get them to link to you, even better. The iterate algorithm used in determining a website's importance was famously exploited to make a biography of President George W Bush a top Google result. Entertainingly, it is a search on the words "miserable failure" that take you to the President's official biography. Essentially there are a lot of websites around that have the words "miserable failure" as a hyperlink that points to the White House website. I tried this using Oregano2 a few days ago and got 4,260,000 results with the President at the head, although Google seems to have caved in to political pressure to qualify the result with an explanation of what is going on and why.
Andy Jackson runs a RISC OS website that Google and Yahoo see as important. It is called Nutshells. On the Nutshells site any software author can add a description of their work along with - and here is the crucial bit - a link to their own site. It should be high on every RISC OS developer's list as the place to arrange a link from. Andy has been making several enhancements to the site in recent weeks.
After placing my new web address on Nutshells I tried searching for my software on Google. Being on Nutshells really does make a difference. A search for "ArtGraph" brings back "ArtGraph/Nutshells" as the fourth result, ahead of my site itself, thanks to that meta tag, in position six. A search on "TurtleChalk" brings back "TurtleChalk/Nutshells" as the top result. A search on "MathBox" brings back "Mathbox/Nutshells" as the sixth result. Of course, having fairly unique software titles helps. A search on "SuDoku" effectively fails. There are far, far, far too many major sites offer something to do with that word. "SuDoku, RISC OS" gets me back in there, though.
The Acorn Webring
Another RISC OS institution that I thought would be fun to join was the Acorn Webring. This is run by Paul Vigay. The idea of the ring is that a hundred or so RISC OS websites form into a circle and where each links to the site ahead and the site behind in the circle. All sites link to a central site which holds an index of all the sites that comprise the ring. Again, the linking to and fro is mutually beneficial in boosting site importance in the eyes of search engines. More than this, a collection of sites that are in themselves rather small and isolated can form an interesting passage through cyberspace.
Each of the sites in the ring is a potential gateway to catch a random surfer and feed him or her around the ring. There is also a definite community feeling that I got from being a part of this ring. Rather than being annoyed that your neighbour has not trimmed his hedge recently, you frown concerning the lack of regular site updating. I know of some very slick and well organised webrings but the Acorn Webring is currently somewhat frayed around its perimeter. Some of the sites no longer exist or connect to pages that have nothing to do with RISC OS.
None the less, I wanted to join because I like the concept of the ring and felt it should be supported. Also, Paul tells me that he is keen to give the Acorn Webring a makeover to remove dead links and make sure it flows as it should. So, to become a member, I emailed Paul to request the HTML code that needed to be added to my site's home page to make it work as a part of the ring. I got a quick reply that allocated me position number 35, along with twelve lines of HTML.
These gave me, or rather WebWonder, a lot of trouble. If I pasted them in by hand and put the site online it worked fine. If I tried to get WebWonder 1.01 to maintain the code as a part of its system, it dropped the links. The code simply vanished and I then had to add it back in by hand. To cut a weekend's story of frustration and annoyance short, I eventually arrived at the following stripped down and rearranged ten lines of HTML which version 1.03 of WebWonder is happy to maintain. Notice that, rather than fetch the GIF webring graphic from Paul's site, I've stored it locally in my site's images directory, where I edited it to make its background match the background of my home page, and added four rows to its size to make it work properly with Microsoft Internet Explorer.
<IMG SRC="images/ringmap.jpg" ALT="Webring Map" ALIGN="TOP" WIDTH="265"
HEIGHT="67" BORDER="0" USEMAP="#clientmap">
<AREA HREF="http://www.vigay.com/cgi-bin/webring?ring=riscos" SHAPE="RECT"
As you will have gathered by now, although I had succeeded in writing the essentials of my website and getting it online in just one weekend, further dabbling, problem solving and messing around with Google and Yahoo has resulted in the whole project taking a fair bit longer. Isn't that always the way with software? Even so, I am proud of my site which is simple and effective. I could not have created such a site without WebWonder, which is a top class piece of RISC OS software.
I recommend it to anyone wanting to get an uncluttered site out quickly and without bother. It is outstanding value for money. I'm looking forward to exploring more of its capabilities over the coming year. I guess you'll have to keep checking out my site to see how I'm progressing with it.
Search engine statistics from theguardian
Previous: RISC OS 4 caught on Mac OS X
Next: How to write a screen saver
DiscussionViewing threaded comments | View comments unthreaded, listed by date | Skip to the end
Please login before posting a comment. Use the form on the right to do so or create a free account.
Search the archives
Today's featured article
Select hardware profiles uncovered
But it's elementary, my dear Watson
5 comments, latest by caliston2 on 2/1/05 5:59PM. Published: 1 Jan 2005
Nettle, email and spam, ODBC and more
9 comments, latest by AndrewDuffell on 13/12/03 1:31PM. Published: 11 Dec 2003
News and media:
RISCOS Ltd •
RISC OS Open •
MW Software •
Advantage Six •
CJE Micros •
Liquid Silicon •
Chris Why's Acorn/RISC OS collection •
The Register •
The Inquirer •
Apple Insider •
BBC News •
Sky News •
Google News •