lashtal.com
LAShTAL.com Problems/Suggestions - Bibliography improvements
ianrons - Feb 08, 2008 - 01:49 PM
Post subject: Bibliography improvements
93 All,
Just a quick note to let people know that, in honour of one of our most recently distinguished members, I've now added a search-by-username facility to the search page of the bibliography. Whilst doing so, I noticed that when I wrote the search feature, for some reason I didn't include historical entries, so a search will only show entries that currently have the title/author/etc. that you're looking for, even if previous versions of that entry are different. This is particularly annoying for the username search, but I don't have time to fix that now.
Anyway, what I'm asking is that if anyone has any suggestions for improvements to the bibliography, please let me know here & I'll try to include them when I re-code the system for our development site.
93 93/93
Ian
the_real_simon_iff - Feb 08, 2008 - 02:21 PM
Post subject: RE: Bibliography improvements
93!
I found it pretty easy on the Admin page of the Bibliography to sort by User, but searching ability might be an improvement.
Love=Law
Lutz
ianrons - Feb 08, 2008 - 02:34 PM
Post subject: RE: Bibliography improvements
Ah yes, the edit log. But only a few people -- the original beta testers, basically -- can access that screen.
OKontrair - Feb 08, 2008 - 03:20 PM
Post subject: RE: Bibliography improvements
Quite some while ago, in the galleries , there was a section called 'Those who followed'. Before too long some entirely relevant picture appeared although not one of a person who could fairly be called a 'follower'. Someone raised the point and the situation was resolved by changeing the gallery name to 'Personalities'. As a result we can now enjoy pictures of all sorts of bit-players, villains as well as heroes and, so far as I know, no one gets overly peeved.
Is there not perhaps a way to restore focus to the Bibliography by division into categories, or renaming something? If there is a storm in a teacup why not use a coffee mug.
I have been meaning for some time to send a few contributions but since I was tricked into having Microsoft Vista my scanner won't work.
OK
lashtal - Feb 08, 2008 - 03:39 PM
Post subject: Re: RE: Bibliography improvements
OKontrair wrote: › I have been meaning for some time to send a few contributions but since I was tricked into having Microsoft Vista my scanner won't work.
Hah! I think you're onto something there... I've had a lovely time on the site since getting the Mac - just binned that and gone back to Vista and now the legions arrive!
ianrons - Feb 08, 2008 - 04:12 PM
Post subject: RE: Bibliography improvements
That's an interesting suggestion about splitting it into sections. When I first designed the schema, I did want to include sections of some sort, but decided that there would be no taxonomically-satisfactory way of dividing books/media entries into sections except keyword tagging. This came on the back of my experience with an earlier online library project that never was, and also on the back of experience at the Open Directory Project (where taxonomy is hell).
I fundamentally disagree with categorization used quite often (esp. by local libraries) that divides things into mutually exclusive areas such as "fiction" and "non-fiction", esp. where those categories can be very blurred. In fact, to the extent that one can express ideas about the world in fiction, it's not really fiction is it?
Anyway, I wondered what would be the point even of dividing things up into relatively sane categories such as poetry, music, etc., when those categories can and do overlap. Sorry if it sounds all a bit obvious at this stage, but keyword tagging is the thing, because then an item can exist in several different categories at once. This doesn't require a hierarchical structure and is inclusive rather than exclusive in nature (and it's also easy to program).
(There are valid reasons to use hierarchical structure for taxonomy, of course: where hierarchy is built in, such as in zoological classification; but I just don't think it's an efficient model for library classification, since we are dealing with "ideas", "expressions", etc. that have no true hierarchy. It'll be easier to find things if we don't split things into too-specific compartments.)
Now, I did include a sort of proto-keyword in the form of the "Subjects" field, but never got around to making an interface to query it. This is because I didn't just want to have a "tag cloud". The trouble is, with keyword-tagging one just gets a big mess of keywords which bear no relationship to each other. By getting rid of hierarchical structure, one ends up with protoplasmic soup.
So, I thought that we'd need something to relate the various taxa together. This is where Drupal (which is what I'm building the new site with) could come in, since it has "free tagging" and "related terms".
So yes, I agree. Something like that would be excellent, and I plan to do it.
Incidentally, the back-end structure of the bibliography may run purely on the built-in Drupal architecture when it gets transferred. This is because my idea of how to set up a library system is very close to their idea of how to set up a CMS, so we'll probably have bibliography cards as "nodes" and I'll use Content Construction Kit to create the various inputs like "Author", "Title", etc.; then we'll have the Revisions module to run what's now called the "History" section, and the Taxonomy module to do the tagging, whilst the permissions system is flexible enough to cope with all of this. Hopefully I can just put in some validation & styling routines (with a front page and so on) and bob's your uncle.
lashtal - Feb 08, 2008 - 05:03 PM
Post subject: RE: Bibliography improvements
It's at times like this that I realise just how lucky I am that Ian works voluntarily and tirelessly as the Tech Admin here!
OKontrair - Feb 08, 2008 - 05:33 PM
Post subject: RE: Bibliography improvements
I think I agree. No tag clouds for Uncle Bob.
Inclusive rather than exclusive has a rosy feel to it but if there are no limits then there are no limits.
Collectively, the folk here probably have every Crowley book there is and they are kind enough to share information on them. These good people undoubtedly have much else which they like to present, share, recommend or even just plain boast of. That's wonderful but it leads to protoplasmic soup and not everyone likes carrot.
The division I had in mind was something like core/periphery or Crowley/everything else. This would exempt anyone from making value judgements and leave people free to post pinups of their other prizes.
Nomenclature is probably going to be a bigger issue than content. Dog/Fleas might not go down too well.
OK
ianrons - Feb 08, 2008 - 06:05 PM
Post subject: RE: Bibliography improvements
I hadn't really considered taxonomy as a way of circling the camp, as it were. I think it was quite early on that Paul & I decided that the best way to approach the subject of inclusion into the bibliography as a whole ("appropriateness of content") was to avoid laying down formal rules, and to look at everything on its own merits. The thing about taxonomy is that it's so inflexible, and what is core and what is periphery is moot; so immediately one is setting up an arguing ground. Rules always have to be interpreted. The recent "Satanic Bible" silliness illustrates this perfectly, and shows I think that the best taxonomy is one guided purely by human judgment (benevolent despotism) rather than formal rules that can be argued interminably.
Example problems with a "Crowley/everything else" division: The Treasurehouse of Images, Perdurabo.
*However*, I would agree in principle with a sort of overall "relevance" score, with a view to eventually promoting higher-ranked content over the other stuff (similar to what Google does); but I would want to make it a sliding scale rather than a choice of core/periphery. In practical terms, this seems like quite a difficult proposition to manage. There are 2 options:
a) relevance ranking by human input (i.e., each user gives each entry a score from 0 to 10).
b) relevance measurement based on keyword analysis of each entry versus a keyword analysis of the whole lashtal.com site.
The trouble with (a) is that our "sample size" is probably too small.
The trouble with (b) is that it would be difficult to construct an appropriate algorithm. e.g., there may be several editions of the same book, each with its own entry, some of which are detailed entries and some not. They would almost certainly rank quite differently by any feasible algorithm, but would actually be the same in terms of "objective" relevance. Other problems would be that keywords like "thelema" and "crowley" would probably swamp the top half of the rankings (although this may be what you intend with the "core/periphery" division).
Coming back to that "core/periphery" division, I think in practice it would be a lot of work and very little to show for it. I mean, would one really want to browse the collection based on that sort of criterion? Or how would one present it? That's not meant to be a rhetorical question, by the way!
ianrons - Feb 08, 2008 - 06:23 PM
Post subject: RE: Bibliography improvements
P.S. Keyword tagging has its problems too. Just look at some of the rubbish on slashdot.org. However, it is extremely easy, organic and useful, even if obvious things don't get tagged "properly" sometimes.
MichaelStaley - Feb 08, 2008 - 07:33 PM
Post subject: Re: RE: Bibliography improvements
lashtal wrote: › [ I've had a lovely time on the site since getting the Mac - just binned that and gone back to Vista . . .
As a matter of interest,
Why have you gone from OSX to Vista?
lashtal - Feb 08, 2008 - 08:57 PM
Post subject: RE: Re: RE: Bibliography improvements
"Needs must when the devil drives."
I'll be taking some Microsoft training courses and, to be frank, Microsoft Office on the Mac really doesn't cut the mustard. I also fear that OSX is a triumph of style over substance...