Aside, sorry for the top posting--I was going to just put the intro (the first
two paragraphs) as a top post, but I think the rest of my response works here
as well (i.e., basically just responding to your subject line).
I debated whether to reply privately or publicly (sp?)--still haven't
decided--I guess I'll know when I press send. I decided to post it to the
list, so it seems less like I'm sneaking around promoting TWiki.
I like your idea--I'm slowly working on something with some similarities (but
perhaps a much broader scope) (something I loosely call an askSam / (T)Wiki
clone or workalike). Anyway:
I use TWiki (keep reading for clarification) and recommend that you look into
it, for the following reasons.
* The back end storage is almost plain text files, one per TWiki page.
("Almost" because they are marked up with TWiki markup which is a fairly
typical wiki markup (but like most wikis, not standard across multiple
wikis). With almost plain text files, your interface from Firefox could
start out pretty simple--all it has to do is find the right file and, for
example, append the TWiki marked up text for the link and any text you might
want to add. For example, to show a link in a bulleted list item, this is
the basic markup (the angle brackets are "metadata" (??)--you don't need (or
want) those in the actual markup of a link):
* [[<URL>][<webpage "title">]] <Whatever else you want to say.>
* TWiki has lots of other features that I'm not sure Instiki has, among
them automatic tables of content based on headings.
* TWiki has:
* a lot of plugins already (so you could create another if that was
required or helped for your application)
* a community of developers (many wikis, when I did a fairly thorough
search several years ago, had only a single developer)--thus, you might
interest one of the developers in helping you with whatever you need done for
an API
* I just think the TWiki people (partly for the reasons above) might be
more receptive to having an API--I'm not sure whether they do at this point
or not
* written in Perl (which is one of the reasons I looked at Instiki--I
didn't want to learn Perl, and at some point in time decided I did want to
learn Ruby)--obviously, you don't need to know Perl to use TWiki, but if you
want to add an API, you probably need knowledge of Perl.
* (aside) TWiki allows what they call multiple webs (I'd call them
subdirectories) but their terminology (afaic) is a little strange--I forget
what they call a page (maybe a "node"?), and then they call a subdirectory a
web--there is other strangeness which I can't recall atm.
I should point out a few other things:
* my "application" (I guess nowadays I could call it a mashup) doesn't
actively use TWiki at this point, although the data is generally being
maintained (i.e., with TWiki markup) so it can be moved back to the TWiki at
any point in the future. A slight problem is that I've extended the markup
language slightly, so some of the things I do now will not currently show up
properly on a TWiki. (ATM, I can't recall what those extensions are ;-)
* TWiki is about to make a new release, 4.2. With luck it will be out
tomorrow, but depending on the timing of any contact you make with the
developers, many of them may be busy with other things. (I used a much older
version of TWiki when I used it--in general, my opinion is that TWiki is
somewhat bloated since then.)
* I did attempt to use Instiki at one point in time. I don't remember the
little things that might have been difficult at the time (they might have
been solved since then), but the back-end data storage was my biggest
stumbling block. At the time, iirc, they used I think what is called
marshalling to serialize the objects into a file. IIRC, there was some talk
about a different approach, maybe that's been done by now.
Sorry for posting this to an Instiki list--I still am subscribed to this list
and "lurk" (don't interpret that as "laying in ambush"--its to attempt to
keep up with Instiki and maybe learn some things).
Randy Kramer
PS: You can find TWiki at twiki.org. They have a TWiki which you will find
rather quickly, which includes a sandbox web where you can experiment.
On Tuesday 22 January 2008 02:53 am, Spencer Collyer wrote:
> I've got a slightly-left-field usage of a wiki planned, and I'm wondering if
Instiki would be a good one to use to implement it.
>
> Basically, I want to store my web browser bookmarks in a wiki. This makes
more sense to me than having them on a menu, for a couple of reasons.
>
> Firstly, I use several computers around the house - one in the office, and a
couple of laptops - so having one central repository for my bookmarks makes
sense. I don't see any point using one of the online bookmark repositories,
as I've no interest in sharing them.
>
> Secondly, I have a *lot* of bookmarks, and have them heavily categorised on
my bookmark menu, so sometimes I have to go down three or four menu levels to
get to the category I want. This is fine if it's just one bookmark from that
submenu, but gets annoying if (as often happens) I'm working my way through a
load of pages on the same or related menus, so having them all on a tab I
keep permanently open would make a lot of sense.
>
> Thirdly, if they are on a web page, I can easily add comments alongside the
actual links, just as a quick reminder if I need one of what the page is
about and so on.
>
> In order to make this easier, I have a couple of requirements.
>
> 1) I want to be able to easily convert my bookmarks into a set of pages (one
per sub-menu) that I can easily load into the wiki. To do this I'll probably
write a Perl script to parse the bookmarks.html file and generate a set of
wiki pages ready to be used. How easy is it to get Instiki to accept
externally generated content like this?
>
> 2) I want to be able to write a Firefox extension that will allow me to
emulate the current 'bookmark page' functionality, but store the bookmark in
the appropriate wiki page. Is there an API external programs can use, or some
other way to get externally-generated changes into Instiki? I know I can
cut'n'paste the bookmark, but it would be nicer if it could be done by
emulating current bookmarking.
>
> Thanks in advance for your attention.
>
> Spencer Collyer
>
> --
> <<< Eagles may soar, but weasels don't get sucked into jet engines >>>
> 7:39am up 25 days 0:45, 16 users, load average: 0.44, 0.54, 0.42
> Registered Linux User #232457 | LFS ID 11703
> _______________________________________________
> Instiki-users mailing list
> [email protected]
> http://rubyforge.org/mailman/listinfo/instiki-users
>
_______________________________________________
Instiki-users mailing list
[email protected]
http://rubyforge.org/mailman/listinfo/instiki-users
