HTML 'Alt' tags

Many websites do not use the html alt tag with images and as a result, people who either cannot download images or have no use for them are offered little useful information when browsing the site thus making navigation almost, if not actually, impossible.

The Alt tag has been supported since Netscape 2 and Internet Explorer 2 and since Netscape 4 and IE 3, moving the mouse over an image displays the alt text in the same way that tool-tips are; for example in the screenshot snip on the right. Alt text is now considered 'required' if your pages are going to be useable.

The alt text is put into the html page either as alt text in an image properties dialogue box on your WYSIWYG html editor or in the html code itself as follows...

<img src="images/sheep.jpg" 
alt="flock of sheep" align ="right">

...which would present itself as the picture on the right. Just hover the mouse over the image if your browser does this.
If you have your images turned off, it will look like this...
If your alt text is particularly long, Internet Explorer will break it up for you when displaying it like a tool-tip but you can choose where to break it yourself by taking the final html file, loading into notepad or kwrite and breaking the lines yourself thus...

<img src="images/sheep.jpg" alt="This flock of sheep is one that we saw when we went on a camping and walking holiday on the Tissington Trail in Derbyshire in August 2003" align ="right">	unformatted
	formatted

You have to be careful with tags though because you can put together images with descriptions that, whilst sounding innocent in their own right, can produce curious effects. For example...

The two images on the right could (conceivably) be found next to each other on a website catering for outdoor activities such as walking and rock climbing. On the left, we have a flock of sheep and on the right, a carrabina with an Italian hitch for abseiling (a super 8 gives a better ride and if you double the rope ... well, that's another story).
To a person experiencing just the alt tags, we are treating them to a flock of sheep abseiling down a cliff face. (Clearly, just one sheep at a time is safest but should she be wearing a chest harness?)

So, if your site is intended to be a useful and useable source of information that is easy to navigate, as opposed to just some pointless graphics, you need to use the alt tags and also be careful so that their content is not ambiguous or consists entirely of "blue bullet, red bullet, red bullet, yellow bullet..."

.png images - graceful degradation

One other thing to remember about images is the format of them. Many people have started to use Portable Network Graphics (.png) files and like .gif s, they carry alpha channel (transparency) information. However, where a pixel in a gif is just on or off, in a png, it can be set to any level therefore drop shadows or other images can be used so that detailed backgrounds can be made use of.

One of the principles of the Internet is that the browsing experience should be made as good as possible and as inclusive as possible. Unfortunately, some people make sure that their site is aimed at only IE6 or whatever so everybody else has to put up with a degraded experience. The same is true with png files - not all browsers are advanced enough to support alpha channel levels between the extremes.

The graphic below is a png file on a blue background and makes use of a drop shadow.

Actual png image	IE6 on Windows	Konqueror: Internet ready

As you can see, under GNU/Linux (Konqueror), it is reproduced properly but under Microsoft Internet Explorer 6 it is not - IE6 is not as advanced at this stage so you should really use png files without transparency if you are to make your site available for people with Microsoft systems that do not yet understand how to use png files properly.

If the image of the png file on the left appears properly, your browser is ready for this new (20th Century) innovation.

Spam sifting

The Problem and Existing Solutions

If you need to have your email address on a website, you have a problem with spiders crawling for mail addresses and this then being passed on to spammers as; "buy our list of 10 million genuine email addresses for only $xx"

One solution to this was to put "NOSPAM" into the address so it might read "joeNOSPAMbloggs@foobar.com" with the intention of the user removing it before they sent you the mail thus leaving "joebloggs@foobar.com". However, there are programs that will remove this automatically.

Other tactics include presidents names and so on but you are assuming that the genuine sender has the requisite knowledge.

Interestingly, I receive spam email to addresses that have never existed - they are either guesses (find a mail domain and prefix it with "info@", "sales@" "admin@" and so on) or completely made-up "o8hllln@", "f64gopp98@" and so on. Clearly, even the people selling the millions of email addresses are padding out the thousands they have harvested with others that they have just made up to fill the gaps.

Spam is effectively one of a group of email activities that you can call "unauthorised" and therefore use an authentication system to block unauthorised email. Instead of trying to guess if it is unauthorised, 'know' that it is authorised.

Body Content Filtering Systems

Normal mail filtering systems rely on trying to guess whether or not a message is spam by looking at the type of wording of the mail and whether there are certain clues such has heavily disguised URLs and so on - using escape characters instead of normal typeable characters such as http://. If you look at the source for a lot of spams, the 'forbidden words' are divided up with html tags that are just made up - html that is not understood is just ignored - so when the words are displayed, they appear as though they were meant to by the spammer so, a text containing the 'forbidden word'; "finance" would appear in the email source as...

fina<lkasjhdflk>nce

...and appear on the recipient's mail program reconstructed as...

finance

All the anti-spam engine needs to do is strip out the html tags to see the message. In response to this, the spammers have started doing the following...

Enla<font size="1">i</font>rgement

so that when the anti-spam engine strips out the html, it is left with

Enlairgement

and is passed through. However, when the end user sees

Enlairgement

the word is corrected by their brain.

Tackling spam this way is basically just a guessing game with the spammers hitting the email address target every time so, why not look at it from the other side and make it a double bullseye to hit?

Upping the Cost of Spamming

This particular solution relies on the fact that most (if not all) spammers will have obtained your email address by buying it off somebody else and have never visited your website in the first place. Simply by visiting your website, they can authenticate themselves by reading the string but this will up the cost of emailing 10 million people if they have to visit all of their websites first. If they are prepared to put this amount of effort into contacting you, you can at least read through it (or at least the first bit of the email) before you delete it.

Note that this will stop email from people that have not visited your site but the scammers (Nigerian bank scams and so on) might well visit your site first as there is so much more for them to gain.

Subject Line Content Requirement

The solution is to have on your website a string that must be included in the mail - the subject line being the easiest to see - whose inclusion is the key to going somewhere instead of straight to the deleted folder.

If, on your website, you say...

my email address is foo@bar.com and you must include "JD3Y" in the subject line

... then any mail that does not include this can just be deleted. If everybody did this then spammers would have to visit all "10 million email" addresses' sites to find the string - they would have to know where the sites were to start with :-).

If you are going to implement this, you need to make sure that the known genuine senders that will fall outside the net can get their email through so one logical filtering procedure to set up on your mail client is...

1. Sift out the emails from granny and others who will forget (generate rules that look at the sender's eamil address);
2. Sift out the emails from any email newsgroups (majordomo lists and so on - generate rules similar to 1. above);
3. Sift out the emails with the string in the subject line and put them where you want them (generate a rule that looks for the string in the Subject line);
4. Delete everything else.

By doing this, you can get your machine to sort out the half dozen or so genuine emails from the 200 spams you get each day.

Images - negatives and colours.

This is our original photograph of the Big Ben tower in London. It has good highlights and shadows and poesesses red, green and blue so we can see what is going on when we start messing around with it.

If we just make a straight negative, we see that blacks have become whites, reds have become cyans (bus) the blue-cyan sky has become brown (our eyes are better able to distinguish between reds, oranges, yellows and greens than cyans and blues - this is why there are so many words that describe them and so many people think that cyan is blue) and so on. To return the colours to their individual values (ie, red back to red and so on), we have to rotate the hues by 180 degrees.

Instead of doing that, we could just separate the image in to HSL channels (Hue, Saturation and Luminosity) and then, make a negative out of the luminosity before recombining them back into the composite - ie turning the HSL back into an RGB image.

Both of these methods produce the image on the right which still has a blue sky, red bus and yellowy-green leaves on the trees but instead of a near-white sky at the horizon, it has turned close to black.

Separating the Luminosity out and processing it on its own (thus saving having to reprocess the hue later on in order to recover the colours of the picture) has the advantage that we can mess around with it in any way we like. On the right, I have solarized (or more correctly when talking about image processing on the computer, pseudo-solarized) it so that the highlights become inverted and the shadows stay the same - it is just as easy to do the opposite, either by inverting the pseudo-solarization transformation graph or by inverting the luminosity image after it has been pseudo-solarized.

On the right, we have the final composite image with the highlights, mid-tones and shadows pseudo-solarized as you would expect but the hue information in the colours in both the normal and inverted portions of the image has been retained.

Pseudo-solarization best works where there are bright skies and intricate structures that protrude into them. Whilst a lamp post is not particularly intricate, it does demonstrate that this can work even when the intrusion is only a pixel or two across On the right is the MI6 building.

So, we make a copy of this and then turn that into a greyscale image so that the luminosity information makes sense (remember that greens are brighter than blues and so on).

Next, use the threshold function so that we create a black and white mask. Here, it is done at 50 per cent but you can use other values if you are prepared to sort them out later on (you could adjust the luminosity mid-point of the image to start with so that the 50 per cent threshold separates the image the way you want it and then change it back later).

Next, go back to your original colour image and select the alpha channel (this should be all black as you haven't got anything masked off yet). Then, paste the threshold image you generated into the alpha channel and, if necessary for your image processor, combine it with the alpha base. With the mask in place, you can get out of the alpha channel - you will see that the mask outline matches your threshold image - and then invert the masked part of the image so that the brighter part of the image is now effectively pseudo-solarized.

Next, and with the mask still there, rotate the hues by 180 degrees. In the picture on the right, the sky has become blue again.

Next, remove the mask and then adjust the tone balance so that it covers the whole range, brightening up your image.

If you wanted the sky the other way around, either invert your threshold mask or take the image you have just created, invert it and rotate the hues - you get the same result.

The above transformations were performed using the HSL colour model in Micrografx Picture Publisher. However, there are many other colour models and the one in the GIMP is HSV (Hue, Saturation and Value). Whilst it is possible to invert an image and rotate its hue or even solarize by using masks and rotate the inverted hues to get the images above, just taking the Value (which very tenuously replaces the Luminosity) and inverting that in the same way that you would with the Luminosity in the HSL model simply does not work. The images on the right are the result of inverting one or some of the HSV images and then recombining and whilst interesting in their own right (and not easily reproducible using HSL), demonstrate that you need the HSL model to do the above properly.

Running KNOPPIX

System requirements

Based on Debian, KNOPPIX is a GNU Linux distribution that runs entirely off a CD ROM. It only needs you to have a computer that will boot off the CD ROM (or a floppy disk and then the CD ROM) and have enough RAM (I have run it on a 200MHz Pentium MMX with 96MB of RAM with no problems). In fact, it doesn't matter whether or not your machine has a hard drive, it will work just as well without it as long as it has at least 82MB of RAM or, if you want to use it in text only mode, 20MB of RAM.

Getting a copy on CDR

First of all, you need to have a copy on CD. On the SuperDVD this month, we have the ISO image for KNOPPIX 3.3 which you should burn to a CDR. Instructions of how to do this are elsewhere on this disc but for those who cannot wait, you just copy the ISO file to a suitably large and fast drive on your computer and then burn it to a CDR. In Nero, get rid of the Wizard and click on File/ Burn Image, select the image from where you stored it, click on 'Write' and let it get on with it.

Running KNOPPIX

KNOPPIX will load up and present a desktop environment that is very similar in many ways to Windows. On the desktop, you will see the trash can and any disks and other media that you might have on that machine.

You will note that there are some - or at least one - cryptic looking labels for your hard disks. Instead of drive C:\, D:\, E:\, F:\ and so on, where, say C:\ and D:\ are the same physical drive with two partitions and E:\ and F:\ are another drive with two partitions, you will see something else. Physical drives are given letters and the partitions are given numbers. So, in this case, C:\ will be hda1, D:\ will be hda2, E:\ will be hdb1 and F:\ will be hdb2. If you have a usb drive - whether it is a flash drive or a large external drive - it should be called sda1 (and so on).

Along the bottom is the taskbar and if you click on the 'K' on the far left, you get a menu like the Start menu in Windows.

USB FlashDisk or LAN

For backing up or rather, recovering data, the big question is: "Ethernet or sneakernet?" If you have a LAN that you can plug the computer into, everything is a lot easier to do even if it is just one other machine and a crossover link. If you have not, you will have to use some other form of storage device to act either to hold the lot or as a bucket brigade to ship the data chunk by chunk to another machine.

A bit about the GNU/Linux file system compared to Windows

Windows uses the drives as a root to the directory tree and more recently, has the desktop as the tree root with My Computer and Network Neighbourhood running off that with the drives under My Computer and other network shares under Network Neighbourhood. (This is ignoring all of the other stuff in there like recycle bins and so on). Thus...

GNU/Linux has a similar directory structure with the root directory (not to be confused with 'root') at the base of everything with various directories off that. Within this semi-virtual/ semi-real file structure, various disks are 'mounted' in various positions as are network shares. It is irrelevant what file system each uses as they are considered (broadly) equivalent by the file system. Extra drives (those that are not part of the GNU/Linux system) are usually mounted in a subdirectory called /mnt (the leading slash meaning that the directory path is absolute).

The devices are to be found in a directory called /dev so this is what you will find when you look at the properties box - something like /dev/hda1 which will be mounted at /mnt/hda1. If you were going to look at the contents, you would look at where it was mounted because there, it is translated into the file structure. If you look at it in the /dev directory, you will see the raw data on the disk.

If we created a directory called /lan and mounted the network shares on that (remembering to create the subdirectories first) we could end up with a structure that is very easy to navigate. Thus...

Windows likes to inhabit the entire and only partition on the first hard drive so if it is Windows that will not boot up, it is likely that you will find any data that you need on hda1 as the other drives should be okay (unless you have Windows spread out and it is one of these that has become corrupted).

FlashDisk or USB Drive

If you have a flaskdisk or another USB-based mass storage device, it will probably appear on the desktop as sda1. Before you click on it, you want to make sure that it is mounted into the operating system as writeable rather than read-only.

Right-click on the icon (sda1) and then, in the Device tab, uncheck the Read Only checkbox. Click OK and then click on the desktop icon to mount it and display it in the file browser

Setting up the LAN

If your LAN uses DHCP (Dynamic Host Configuration Protocol), all you have to do is click on the KNOPPIX penguin on the taskbar - the second from the left - and select Network/Internet | Network card configuration and when asked if you want to use DHCP broadcast, click 'yes' - the computer will do the rest.

If your LAN does not use DHCP or you want to specify the settings yourself, click on the KNOPPIX penguin on the taskbar, select Network/Internet | Network card configuration and when asked if you want to use DHCP broadcast, click 'no'. Next, enter the IP address - eth0 is the ethernet card (it might be that you are using something different or that you have two cards (if it is the latter and you don't know which is eth0, set up eth0 and if it does not work, you can go back to this dialogue as many times as you like until it works the way you want it)). Next, enter the network mask - it will guess at this (see previous PC Plus HelpDesks for how) - then the broadcast address, the default gateway and the name servers and then it will set up the card for you.

Just to test it out, load up a copy of Konqueror and point it at a webserver

Pushing or pulling data on the LAN

GNU/Linux was built from the bottom up with networking and compatibility in mind therefore it has a lot of options of how to transfer data. You can either pull it across the network using FTP from another machine or you can use Server Message Block (SAMBA) to push data to a Windows machine. If you have your network neighbourhood configured, you already have SAMBA. Otherwise you need to set up network neighbourhood on Windows.

smb:/ as an alternative

Everyone is familiar with typing http:// at the address bar of Internet Explorer or their favourite browser but in GNU/Linux, there are more options. In issue 214, we mentioned KIOSlaves. Here, we are going to use one. If, on the KNOPPIX machine, we type smb:/, we should be able to navigate our way to the target machine by selecting workgroup, machine, share and so on - eventually leaving us with a window that we can see the contents of - files and directories.

If you encounter the problem of having navigated part of the way to your destination, only to be told that you cannot get there for some reason (not found and so on), try deleting the workgroup name in the address thus:

smb:/workgroup/john/webpages/ becomes

smb://john/webpages/

WARNING: if you are copying the files to a drive on the machine you are running KNOPPIX on, do not copy to a partition that is formatted as NTFS unless you know for sure (have checked yourself) that the version of GNU/Linux you are using will write safely to NTFS. As an aide-mémoire, think of NTFS as Not That File System.

Just drag and drop

If you have two browser windows open, one at your source directory and one at your destination (whether it is an smb connection across your LAN or directly onto an sda or another hda) you can now simply highlight the files you want, drag them from one browser window and drop them into the other.

With your files now safe, you can locate SCANDISK, if you have not already (use the find option in KDE under the Tools menu) and then you are ready to start recovering your Windows functionality.

One word of warning - whilst you are in KNOPPIX, your hard drive is protected so that you cannot damage it (unless you have made it writeable) so you have an unfamiliar, professional operating system that is free to explore. In the games menu, under 'Tetris-like', you will see a game called 'Frozen Bubble'. By all means, have a play with it but when you realise that you can hear the dawn chorus outside, remember that this is highly addictive and you should limit your time spent playing with it (strictly). For those who are morbidly curious, the screenshot on the right is at level 70.

You might toying with the idea that you prefer GNU/Linux over Windows any way and decide to install it instead - with KNOPPIX, you have an opportunity to play around and get a feal for it without damaging your existing installation (such that it is).