[Alexandria-list] Just another wishlist for Alexandria

[Pablo Bianchi]

Hi Cathal,

First, sorry for my English, isn't my native language.

On Mon, Mar 23, 2009 at 11:15 PM, Cathal Mc Ginley
<cathal.alexandria at gnostai.org> wrote:
> What version of Alexandria are you currently using? Because some of the
> features you mention are already supported... 0.6.4.1 is the latest
> release.

Sorry I didn't mention. I'm using v0.6.4.1.

>> Same functionalities of Amarok to mp3, to books on Alexandria:
>>    Next and Previews buttons on Properties box, to move...
> I don't really know the Amarok GUI (I mostly use GNOME) but I can
> imagine what this would look like. Sounds like a nice idea actually;
> view the complete properties of a book in the 'properties' dialog, and
> then to get to the next book, just click the Next button (rather than
> closing the dialog and clicking the next book).
Exactly!

>>    Statistics and Labels tabs and functionalities
> You'll have to expand upon this, I don't know what you mean... What
> statistics should be displayed, and in what form? I don't understand
> what you mean by 'Labels' in this context either, unless you're talking
> about 'tags'.

With Statistics for example:

Number of books from this author: n
Author rating: m (average value for all books)
Amazon book rating: (or any other source)

But I think it could have more info..

Labels:
Yes, I think we are talking of the same thing, labels on Amarok are
similar like on GMail. The bibliographic format MARC 21 (the most used
in the world) use descriptors in fields 6xx (fields are numbers, from
000 to 999, with subfields. For example, 20$a is the ISBN. Something
similar but a little bit easier, to group sets of books.
But I don't know the difference (in the behave) from actual tag and
Notes field. I'm talking about a cloud of tags/labels where to pick
someones, related to the book. This information usually comes on 6XX
MARC fields: http://www.loc.gov/marc/bibliographic/lite/elbdsubj.html

> Can you describe what the feature would look like? A pencil-sketch
> would work too - or a screenshot of the Amarok feature you're referring
> to...

Love between a man and a music player is possible:
sudo apt-get install amarok  :)

A screenshot, but with only one label...
http://yfrog.com/08amarokscreenshotlabelsj

>> Export/Import to ISO 2709 Format using MARC 21 bibliographic format,
> Now this *could* be done, since the Z39.50 support does use MARC.
> Unfortunately, I only barely understand the MARC format - it's a part
> of the code I've barely touched.
>
> Would you really have MARC files lying around to use for Import and
> Export? Usually, we just let Ruby-ZOOM deal with accessing Z39.50
> sources and parsing the MARC documents.

Fetching records of books with z39.50 is different from bulk
importing/exporting from/to a ISO2709 (a "MARC file", to be strict,
the file could have any other format, like UNIMARC or CEPAL).

> If importing and exporting to MARC files is useful, it could be added
> quite easily I'm sure.

A little library could become a bigger library, so if is not very
difficult, this option would be interesting. Another standard file
format (modern) is MARC XML: http://www.loc.gov/marc/

>> Using UTF-8, Latin1, or any other coding.
> In the future, Alexandria will use UTF-8 everywhere (where possible).
> At the moment, it's a bit of a mish-mash, and there are occasional
> encoding problems.

Great! Encoding troubles are really annoying.

>> Import also the Synopsis of the book, is the field 520 of MARC 21.
> Thanks! That's something I can add to the "optional book properties" in
> a future version.

;)
>From the record of a well cataloged book you can fetch a lot of
interesting data! Like size of the book, number of pages, translators,
year, edition...
To use z39.50 should be in some place the list of servers. A line
would be the LOC server:

Library of Congress Online Catalog
    Domain name:  z3950.loc.gov    (140.147.249.38)
    Port:  7090
    Database name:  Voyager

>> Import from a TXT with a set of ISBNs
> This feature is already supported: select menu item "Library > Import"
> and select "ISBN List (*.txt)"
OOPS!

>> Import using z39.50 when Amazon, WorldCat... doesn't works (like ILSs
>> like Koha)
> This feature is also supported, but you may not have Ruby-ZOOM
> installed which provides this support. The simple way to install it is:
>
> 1) Install rubygems using whatever tools your platform prefers (for
> example, on Debian or Ubuntu # apt-get install rubygems)
> 2) # gem install marc
> 3) # gem install zoom
>
> Restart Alexandria, and you should be able to add new Z39.50 providers.

Certainly I'm doing something wrong. I apt-get install rubygems (and
of course restarted Alexandria) but I don't see any change.

>> Why if I enter to the link web I can see the cover, but it
>> doesn't download it when fetching data?
> Can you give a specific ISBN of a book for which this happens and the
> provider where you can see the cover image? It may be that the book was
> found at a different provider from the one whose link you followed.

ISBN: 9789871138517 (A brave new world). Perhaps I'm doing something
wrong, but I only see the default blue cover. Right-clicking, Display
online information, Amazon  go here:
http://www.amazon.com/exec/obidos/ASIN/9871138512 where I can see the
cover.

> For example, I can't grab images from WorldCat (the web page uses
> complex JavaScript to display the cover image) but I can from
> Amazon and AdLibris etc. So if your book was found at WorldCat first, it
> won't have found any image.
mmm, I have Amazon on the first place (default order). I'm not
familiarized with Amazon API. Google Books API is an alternative?

> Naturally, once it has found a book, Alexandria stops searching - and
> it can't know that it hasn't found the 'best' data available.
And couldn't add even more data from the other sources?

> One feature that would be nice would be to 'fill-in extra data' from a
> specific provider afterwards.
That what I'm talking about! Z39.50 servers have data amazon don't
give, and vice versa.

>> Clic on cover image (on Properties box) should let:
>>    Save the image
> That's a nice idea, it would be very convenient. Of course, it *is*
> currently possible to view all cover images directly by browsing
> "~/.alexandria/My Library" with Nautilus or another file browser.
Yes, but average user don't know that. The option is more user friendly.

>>    Set url or path of the image (not only path, locally)
> That's a nice idea too: it should be quite easy to implement directly
> (but designing a  simple GUI for this would a bit more complex).
>
> Ideally, for GNOME, the best way would be to use GIO/GVFS... but I
> don't think that's completely stable at the moment, nor is http support
> done yet.
>
>>    If it doesn't found it on Amazon and those: Offer a set of covers
>> using Google Images
> I think that searching a book-specific site like LibraryThing might be
> more helpful here. That *is* a feature I plan to add, eventually.
>
> I would generally be *very* reluctant to display the results of a
> Google image search in Alexandria (even with safe-search on :^)

You are right. Searching the cover in Google Image should be then the
last -manually by the user- alternative.

> (This is why I think in future Alexandria should support plugins; so
> individual users can add a feature like this that *they* might like but
> which others would prefer to stay out of the main body of the
> software.)
Totally agree with that!

>> Possibility to create different notes:
>>    Unknown words (It would be awesome that it search automatically the
>> meaning/translation, detecting automatically (or not) the language,
>> for eg using WorldReference.com).
> You probably meant WordReference.com
OOPS! Yes!

> I think this is too broad a feature for a cataloguing program. I
> personally can't see it being useful for cataloguing your books.
>
> But then, I'm an English-speaker and most of my books are in English;
> if you could give me a Use-Case that describes how you might use this
> feature and why it fits in to Alexandria's overall purpose (keeping
> track of your books) then it might help me understand your wish for the
> feature.

Formally, for a cataloging program this is not a good idea. But, is
Alexandria for a formal library, even a small one, to loan books...? I
think not, only for a personal use (at least firstly).

A use-case (I owe you the UML diagram :P): When I read a book, I often
use the last white pages to make some annotations. If I have not near
a dictionary, I write there the words I don't know to search for them
later. Mainly I have three parts on those pages: "Unknown words",
"Things to search about", "Quotes and page-number of interesting
sections"

>>    Things to search about
>>    Remarkable quotes.
> This would be a nice idea, but where would Alexandria get these from?  I
> don't know of any online book quotes database. There may be copyright
> issues which prevent such sites from existing.
That's because that data it comes from the user, not from any online
database (well... it could be, with some things). For example, after
(or before) reading the Odyssey is interesting to read about Ionian
Islands.

>>    Related URLs, Books, Movies
> Again, how would we generate a list of related items? There aren't
> really any sources that spring to mind that Alexandria could search
> automatically. (Amazon gives 'related purchases' which is not useful in
> the same way at all.)

All this items is in the group "Possibility to create different
notes", personal notes, written, at least in first place, by the user.

>> Link to biography of the author on Wikipedia (look for Context sidebar
>> on Amarok for inspiration) using embedded browser
> Link to author biography would be a nice touch; but probably just
> launching a browser rather than having an embedded browser. I do want
> to keep the user-interface as simple as possible.

OK, but please see Amarok context sidebar. Is not a full browser, it
doesn't look like a browser, just show the html embedded.

> As for how we'd get the URL of the author biography... that's another
> story. We could just construct the URL and hope the page exists, or
> we could search Wikipedia. But it's not trivial to get the canonical
> name of the author (never mind dealing with pen-names), and not every
> author will have their own page of course. GoodReads.com also has Author
> Biography pages, so that might be a useful start.

Good question. I'm not sure how work internally Amarok, but Wikipedia
usually have alias to redirect from a group of names to the exact (or
more common) name of the author.

>> Different color for odd lines on List view
> This is already present. If it's not visible on your system, it's
> something to do with your GTK+ theme; but Alexandria definitely sets
> the "rendering hint" to alternate colours in the List View.
> http://alexandria.rubyforge.org/screenshots.html
OOPS! I use Kubuntu. But.. it could be my monitor or video card,
because I don't see the colours alternate here:
http://alexandria.rubyforge.org/images/gallery-Alexandria%200.6.3/List%20View%20of%20books.jpg

> Other users have pointed out that the user-interface for adding books
> could be improved and simplified. In particular, there are now two
> dialogs for adding books from a barcode scanner; one adds books in the
> foreground, one does it in the background... It's not so easy to figure
> out which is best, or how to combine the two dialogs without losing any
> features that users like.
Yes, is something to think about, but I dont dislike the way it is now.

> Well, that went on for longer than I thought it would. I'll try to work
> some of your wishlist items (as well as lots of earlier input from the
> mailing list, Feature Requests on the Tracker and data from the
> recently-departed Wiki) into a requirements document sometime (soon,
> hopefully). Also, some of your questions could probably go into an
> online FAQ :^)
>
> I outlined my plans for future work on Alexandria on the list recently:
> http://rubyforge.org/pipermail/alexandria-list/2009-February/001598.html
> In particular, I want to get some other people looking at the code I've
> started for Palatina; I hope to set up a corner of the web for that
> project soon, including a list of goals and a requirements document.
> So, watch this space...

Well, a library soft it could be as much complex as you want. I'm
familiar with Koha, an Integrated Library System big and complex,
using MySQL (you can use SQL lite?) over a web server like Apache. You
could think not only about "biblios" a with a list of items associated
(copies of a book), also about a book on multiple volumes. In those
cases you usually have an ISBN for each volume *AND* another ISBN for
the complete group. Also you can add support for serials (magazine)
with ISSN...

Cheers,
Pablo