ScanGraphics FAQ

The program looks different

It will, it's a new version

I can't see anything in the main window

All reports, lists, and graphics in the program use an embedded copy of Internet Explorer to display HTML files. For example, when you ask for a list of links, it writes the list to disk in HTML format, and then displays it.

The List Gfx or List Links button doesn't work

Make sure that you use the full website address. For example, try

• http://www.blibbleblobble.co.uk/

(note the http:// at the front, and the / at the end)

The program can't find any pictures/links

Valid HTML

The program works only on sites which use well-formed html. For example, using the program on yahoo.com will not work, because they don't use quotes around attribute names, hence their page fails an HTML validator.

To check whether a site has valid HTML, try the W3C validator (click the Valid-HTML! button at the bottom of this page to see what the validator does)

Frames

If the site uses frames, try entering the URL of the document itself, rather than the overall "frames" page. To find out what this is, right-click somewhere on the page, and click "properties" or "open frame in new window" (or the equivalent in your favorite browser)

Dynamic URLs

If the site uses URLs of the form http://download.cnet.com/downloads/0-3282264-100-5550892.html?tag=st.dl.10001-103-1.lst-7-1.5550892 then the program should still work - although the URL looks like noise, it is actually valid.

Redirection

If links are of the form "/ReferTo.php?URL=nextpage.htm" then the program won't see them

Scan Site takes ages

Scan Site will take ages - it has to download every single HTML document on the web site. For example, blibbleblobble is about 50Mb big, about a third of which is HTML pages. Try dividing that by 5k/sec for a modem connection.

Also, watch out that you've not found a ridiculously big HTML file. I was having problems running ScanGraphics on this site, until I realised that it was parsing my entire 3rd-year project (all 200 pages of it) for links and graphics.

Scan Pictures takes even longer

Scan pictures is the same as scan-site, except that it downloads each document twice (once to look for links, once to look for pictures) so it will take twice as long.

I plan to fix this by caching the most-recently-loaded web page, which should double the speed of this function

"Scan Site" and "Scan Pics" also work, but they will take quite a while to run, especially on a large site, or a slow internet connection.

The Report button doesn't work

It won't, it hasn't been programmed yet. You're welcome to write the code yourself and mail it to me for inclusion in the next release.

The Save Graphics button doesn't work

It won't, it hasn't been programmed yet. You're welcome to write the code yourself and mail it to me for inclusion in the next release.

Try using the "save page as" function on internet explorer

The program is too slow

That's Visual Basic for you. Try the perl version.

The pictures aren't displayed - they all display as broken links

This is just a bug in Internet Explorer (well, I assume it's a bug, it could be a 'feature' )

Given a long list of image files on a page, internet explorer will often give-up half way through, and not display the rest. Try using images.google.com in a low-memory situation to see examples of this.

• You may have some success by refreshing the page (right-click on it, and click refresh)
• If it's only a few pictures, you can right-click on them, and click Show Picture (the fact that this works proves that it's IE's fault and not my HTML!)
• You may do better with more memory available (i.e. shut down other programs) or with fewer other programs accessing the internet.
• If all else fails, try loading up the page in Netscape or Opera. (To find ScanGraphics' temporary file, search for scangraphics.htm - it should be in c:\temp or c:\windows\temp)
• I plan to fix this by allowing you to download the graphics directly, or by listing the graphics in a table, and having check-boxes to select the ones you want to download. You can then use GraphicExplorer to browse them.

The program doesn't run on [operating system]

Try the perl version.

I don't want people running ScanGraphics on my site

• It's easy to create a site that will break ScanGraphics. Try some of the following
• Use links of the form "../thisdirectory/anotherfile.htm" - it doesn't understand the .. parent directory
• Remove the quotes from around your links - this will make your site non-HTML-compatible, and will stop ScanGraphics from seeing the links.
• Route all links through one page, of the form "/ReferTo.php?URL=nextpage.htm"
• Don't bother with circular references - it won't work
• Don't bother linking to massive external HTML files - it won't work
• Don't bother with a robots.txt file - it won't work.

ScanGraphics takes no notice of my robots.txt file

• This feature will be added sometime (with the option of turning it off, probably) - 2 notes:
• The program works for the user, not the website. If a website owner wants to protect their pictures, and the user wants to take those pictures, then my program sides with it's users.
• If you ask most sites (e.g. mine) for a robots.txt file, it will return the 404-not found page, which may itself be a large, valid HTML file. Currently, it's difficult to tell the difference, and it will take me a while to get a parser working.

ScanGraphics is crap

• Try images.google.com, and use their picture-search engine instead
• Write some code yourself to improve it
• Try the perl version.

 

(Valid HTML 4.01)