I again approach the wisdom of the monks, not with a brain-twisting code example, but with a design query for assistance:

I have a proposed section of a new website I'm designing that was originally called "Screenshots", bland. I'm going to update this for the new site launch to be more of a "gallery" of images for users to walk through.

My question though, is: What is the best approach to tracking the image filename, alt tag, title tag, and a textual blurb that describes what is shown in the actual image itself?

I could use something like this (vars shortened here for space):

# t = TITLE tag
# p = param passed
# a = ALT tag
# d = Description 
@img_list= 
       ({t=>'palm',        p=>'palm',
         a=>'Plucker running on the Palm',
         d=>'Plucker on the Palm device supports B&W,
             color, VFS, and external storage'},

        # Handera in Scale-to-Fit (StF) mode
        {t=>'handera_stf', p=>'handera_stf',
         a=>'Plucker in Scale-to-Fit mode on the Handera'},

        # HiRes Sony
        {t=>'sony_hr',     p=>'sony_jd',
         a=>'Plucker on the Sony device in HiRes mode'},

        # Sony with JodDial support
        {t=>'sony_jd',     p=>'sony_jd',
         a=>'Plucker on the Sony device in HiRes mode'},

        {...},
        );
[download]

In this example, I could then display each image something like:

$last_img   =       pop @img_list;

for my $item (@img_list) {
        print span({-class=>'mitem'}),  
                a ({-href=>"$script?action=$item->{p}", 
                    -title=>"$item->{a}"},
                img({-src=>"i/$item->{p}.gif",
                     -border=>'0', 
                     -alt=>"($item->{a})"}),
                " $item->{t} " ),
                " | ",
                "\n";
}
[download]

This seems inefficient, and the hash gets uglier and larger as I add more images and more descriptive text to each one. Also, it doesn't allow one to click on one image and expand that into a page of several images (all "Palm" images for example, or all "Sony" images).

I'm wondering if there are other similar/better or different ways of thinking about doing this. I'm already using SQL (mysql) to hold the News articles on the site, so I could probably make a table there to hold the images and image data, but I've never personally done that, and then I have to deal with filesystem-based storage vs. storing the images in the database itself as a BLOB.

Has anyone done this? Any gotchas here? I can implement the perl bits.. I just can't wrap my head around the right design considerations with this approach yet.