On Fri, Apr 22, 2011 at 7:37 PM, Adam Porter <a...@alphapapa.net> wrote:
... 8< ...

> I'm not sure how to reconcile this other than simply saying that Zim
> doesn't seem to handle this use case very well.  Perhaps some
> different UI widgets would be handy, like a page-list that could
> occupy the top half of the window, above the note-editing/viewing
> widget, next to the index pane.  It would resemble a mail client in
> this configuration, but it could display some metadata and a summary
> for each page.  But I still wonder if, after storing several thousand
> short snippets of text, if this would not be very efficient in memory
> and on disk, and if it would cause Zim to load slowly.

Sounds to me like you want either some more functions in the search dialog
(like showing the paragraph closest to the tag) or some better way to sort
trough sub pages in a "Quotes" namespace.

In general I would say it would be most fitting with the design of zim to
store individual quotes in separate pages, that allows to link to them,
index them separately in a search or tag interface etc. But zim is mostly
optimized for medium size pages, not for huge amounts of 1 paragraph pages.

Best trade off would probably be to organize quotes over multiple pages,
e.g. by author, by source or whatever and propose some improvements for the
search interface to find one quote out of these pages. If the improvements
are very specific for this usage it might be better suited as a plugin, if
they are more generic they could go in the main application.

> I also wonder about some of the ways Emacs Org-mode works, how it
> parses text into individual items with separate metadata and presents
> different views of the items.  It would be very interesting if Zim had
> some of this functionality, although I wonder if, in this use case
> again, parsing a long page of text into several thousand small items
> would be very inefficient.  I wonder if a SQLite database would be
> more apt to this situation (not that I'm necessarily suggesting that
> Zim use SQLite instead of files).  How does Zim's index handle this
> right now?

With an SQLite database :)

However in zim the text files are the data, the database is used only as a
cache to index what is in the text files. You can throw away the database
and rebuild it from the text files.

In fact the side pane showing the page index is directly wired into the SQL
table. Also it speeds up e.g. the tasklist plugin and the sorting by tags.
Without the database zim would be very slow even for a small notebook.

> I'd love it if Zim could do some of the things Evernote does, but it
> seems like it would require some fairly fundamental changes.

That is a bit broad statement. Probably should discuss feature by feature,
somethings may be easy to do, others more difficult.

> Thanks for your thoughts.  Keep up the good work, everyone!  Zim has a
> bright future.


Mailing list: https://launchpad.net/~zim-wiki
Post to     : zim-wiki@lists.launchpad.net
Unsubscribe : https://launchpad.net/~zim-wiki
More help   : https://help.launchpad.net/ListHelp

Reply via email to