Just to add my little contribution to this sudden flurry of activity, I'd like to mention PiggyDB (piggydb.net), which I've been playing with for the past couple of days. It's described as a knowledge creation tool, but actually it's a kind of textual concept/mind-mapping engine - very powerful, very entertaining. I'm not absolutely sure if it's something with long-term legs, but as a conceptual exploration it's fascinating, and very clever. Kind of reminds me of Luminotes, but it's (much) more flexible. The whole thing runs on a Java server, and doesn't take up many resources (you can even use the standalone version as a server - I've had it running on my (very slow) Asus 1101HA netbook, and it's perfectly responsive. Better still, I can access it from my desktop (using a different user ID, even!), and it's even more responsive!

Nice piece of programming, and functions as a rather cool outliner as well (with all entries naturally folded). Filters, tags, it's got all the toys...

Cheers,
Bill

It seems brilliant... thanks for the heads up!

Given the flexibility of its structural approach, it is probably more useful to me and my team than Kerio Workspace which we were looking into. I'm a bit cautious about the non-existent business model, but the fact that it has reached v.4 is quite encouraging.

Spoke too soon; it's a single user application. Ah well, you can't have it all in this mortal coil apparently.

Ah! No, it's not - you can create users very easily, although everybody has the same access to every part of the database, which does restrict the functionality somewhat. You simply create a fragment entitled with a user name (e.g. "Bill") and tagged with #user, and ping! you have created a new user. Their password is initially the same as their username, but can be changed easily from the System menu. Once you've created a few users, you'll find that each tag is labelled with the author's name, date of creation etc.

I'm actually quite impressed, I have to say!

Best of luck with the ongoing search for a collaborative solution (I'm assuming you've already looked at Plone 4.0? And MindTouch? Have you checked out ocPortal?).

Cheers,
Bill

It's terrific. Similar to TiddlyWiki, which was, for me, a productive system. But where TW hides and exposes a single document, Piggydb drives a database, so it won't get bogged down when entries number in the thousands. By using a browser interface, these applications are jump-started with features that too many standalone applications neglect: zooming, spell checker, control of the background color. Try a split window with the Firefox add-on Fox Splitter, aka Split Browser. All the engineering that has gone into browsers and add-ons accrues to Piggydb and its kin.

Piggy's primary gaps are in screen clutter management and import capability. With a little more control of the front end, I'd remove the sidebar (Calendar, Classified by, Tag Palette, Recently Viewed), and the "Edit this fragment" bar. These should be available as text-optional icons on the top Piggy bar. And an import of some kind will be a must. Though there's bound to be a way to access an H2 table independently of Piggydb.

Of course each user will have to determine whether he's prepared to standardize on plain text markups rather than full-featured editing. I'd be feeling that loss profoundly.

Jerome

Never mind the digression on screen clutter. To customize the layout of a Piggydb screen, you can simply edit the corresponding HTML file, then shutdown and restart Piggydb. The script calls are in nested tables. Removing the sidebar from the fragment view window was just a matter of opening fragment.htm and removing the cell in which the sidebar widgets reside. Pretty neat.

Jerome

Yes, I have to say I think it's pretty amazing, and the developer (who is, I assume, Japanese) is also very responsive and has outstanding English. He's got some interesting entries on his blog, too, discussing e.g. the difference between concept and mind mapping, and his frustrations in attempting to create a knowledge builder rather than a knowledge finder. I like the idea of using it with Firefox - being able to view multiple fragments side by side is, for me, a sine qua non. Take a look at Barnes & Noble's NOOKstudy, incidentally, for an amazing example of side-by-side document viewing built into an application from scratch (the software is free, but you need to create a B&N account, which is a mildly annoying). The link is:

http://www.barnesandnoble.com/nookstudy/index.asp

MadaboutDana wrote:

you can create users very easily, although everybody has the same
access to every part of the database, which does restrict the functionality
somewhat.

Many thanks for this guidance! Common access is not a problem for me, as there are no secrets in my team (that I am aware of... :-)

Is there any indication of how many concurrent users the system can support? We would not require more than 7-10 tops.

MadaboutDana wrote:

Best of luck with the
ongoing search for a collaborative solution (I'm assuming you've already looked at
Plone 4.0? And MindTouch? Have you checked out ocPortal?).

Plone and MindTouch yes, and I find them more complex than needed. I need something that would require minimal training and administration. In this regard Kerio looks ideal, but I'm still waiting for the price announcement. Thanks for the heads up on ocPortal, I'll be checking it out.

It's a shame that many of the applications that we discuss here are not client-server applications (even though internally they may use such an infrastructure via SQLite) allowing for easy deployment in a joint SOHO environment; there are exceptions of course. I'm looking forward to test Citavi Team as soon as it is out.

As hosted solutions go, especially for multinational teams, the most powerful and value-added solutions I have found are Huddle and Glasscubes.

Piggydb v6 is out and includes some interesting goodies including a more streamlined interface. There's now a supporter's edition ($10) which adds .CSS customisation.
http://piggydb.net/2012/08/19/piggydb-v6-0-released/
http://piggydb.net/2012/08/20/piggydb-supporters-edition-launched/

Is PiggyDB quite similar to ConnectedText?
What do you see as the major differences?

Can they be used for largely the same type of project?

I've just purchased a QDA program to finish organizing my research data.
Later, once I've got more free time, I will re-work the data using another program like ConnectedText or PiggyDB.
I would prefer to use an open-source application, but if CT is really much better than everything else for organizing Rich Text data, then I'll just get a license.

Thanks for any advice.

To copy from the accolade that I read of one of my favourite CDs, that is a "very excellent" question :-)

I hadn't considered Piggydb as a wiki, but that has most to do with its aesthetics than anything else. As in any wiki including Connected Text, you can relate any 'fragment' (topic in CT) to any other. However, such relationships are not made via internal links ('wikiwords') which are part of the topic text, but by dragging a connector from one fragment to the other. Relationships can be uni- or omni-directional and are shown underneath the fragment.

Pdb also employs hierarchical tags as a second organisational approach. Tags and relationships are used complementarily which I find a very fexible approach. You can read more on Pdb's organisational philosophy here http://piggydb.net/2012/06/20/the-piggydb-way-1-tag-as-concept-over-tag-as-index/ and in related posts.

CT from its part can provide structure within topics themselves. In Pdb, fragments are just that, with the possible addition of attachments.

Aside from other CT advantages described here in the past, by Dr Andus in particular, CT is a much more mature product. For example, unless I am missing something, there is no way to export Pdb content, other than opening each fragment one by one as a document (it opens in a separate browser tab) and saving it. But then, all the interlinkages are lost.

In short, for the moment I do not believe that the two products are comparable, but in the future who knows? The Pdb developer is clearly dedicated.

Carrot wrote:

Is PiggyDB quite similar to ConnectedText?
What do you see as the major
differences?

Can they be used for largely the same type of project?

I've just
purchased a QDA program to finish organizing my research data.
Later, once I've got
more free time, I will re-work the data using another program like ConnectedText or
PiggyDB.
I would prefer to use an open-source application, but if CT is really much
better than everything else for organizing Rich Text data, then I'll just get a
license.

Thanks for any advice.

I posted here http://www.outlinersoftware.com/topics/viewt/4445/0/anybody-met-this-outliner
the following statement

By the way, Piggydb is evolving nicely into a paradigm much more solid than wikis.

and Dr Andus, most undesrtandable, asked me to support it.

That’s a bold claim ;) Would you care to elaborate? (I’m intrigued by Piggydb but haven’t had time to test drive it).

It makes sense to do so in the Piggydb thread. I've already posted just above the main characteristics of knowledge structring in Piggydb. I will now explain why I believe that this approach is more 'solid' than a wiki's.

First of all, what do I mean by 'solid' in knowledge management? I refer to the extent that a tool supports an organisation of concepts conducive to higher levels of conceptualisation.

The way I see it, concepts are represented by 'entities' in the sense discussed here http://www.outlinersoftware.com/topics/viewt/4266 i.e. different tools will use different kinds of entities as basic building blocks.

In a wiki, the entity is the wiki page. However, because the main connectors between pages are the hyperlinked words, it becomes easy to create pages with several concepts. Not so in Piggydb where, since the main connectors (relationships) are set for each entity ('fragment') as a whole, it supports the user to maintain only one concept per entity. I believe that this is conducive to mental discipline and to clearer descriptions of concepts.

Also, in a wiki, tags can be used to label pages, but they cannot be organised themselves --they are just part of a long flat list- so their contribution ends there. Piggydb allows the hierarchical organisation of tags, supporting conceptual development at the meta-level too. Assume that you are making a database of living organisms: you can have entities under hierarchical tags Kingdom / Phylum / Class / Order / Family / Genus / Species, as well as a multitude of other possible sets of terms.

By the way, here's part of what the developer himself has written; if you follow his blog, it seems that he's got a pretty clear vision:
http://piggydb.net/2012/09/25/lets-play-piggydb-knowledge-creation-1-on-writing-by-stephen-king/

Alexander Deliyannis wrote:

Also, in a wiki, tags can be used to label pages, but they cannot be
organised themselves --they are just part of a long flat list- so their
contribution ends there. Piggydb allows the hierarchical organisation of
tags, supporting conceptual development at the meta-level too. Assume
that you are making a database of living organisms: you can have
entities under hierarchical tags Kingdom / Phylum / Class / Order /
Family / Genus / Species, as well as a multitude of other possible sets
of terms.

At least in ConnectedText, categories (which are the same as tags), can be placed into a hierarchy. Any tag can be in multiple tag hierarchies. And you don't need to know in advance what your tag hierarchies are going to be, because you can tag the tags along the way (by placing any category page into one or more other categories). Finally, you can also limit a search to items contained within a particular tag tree.

I'm not sure if there are any online, collaborative wikis that have the same functionality, though.

Alexander Deliyannis wrote:

First of all, what do I mean by 'solid' in knowledge management? I refer
to the extent that a tool supports an organisation of concepts conducive
to higher levels of conceptualisation.

In a wiki, the entity is the wiki page. However, because the main
connectors between pages are the hyperlinked words, it becomes easy to
create pages with several concepts. Not so in Piggydb where, since the
main connectors (relationships) are set for each entity ('fragment') as
a whole, it supports the user to maintain only one concept per entity. I
believe that this is conducive to mental discipline and to clearer
descriptions of concepts.

Also, in a wiki, tags can be used to label pages, but they cannot be
organised themselves --they are just part of a long flat list- so their
contribution ends there. Piggydb allows the hierarchical organisation of
tags, supporting conceptual development at the meta-level too. Assume
that you are making a database of living organisms: you can have
entities under hierarchical tags Kingdom / Phylum / Class / Order /
Family / Genus / Species, as well as a multitude of other possible sets
of terms.

Thanks very much for that Alexander, this is very interesting.

As Chris has just pointed out, you can implement that behaviour with CT, provided you know how to do it (so it's down to a personal decision and also sufficient knowledge of the tools) and you stay disciplined in doing it. I think Manfred Kuehn seems to use CT like such a knowledge base as well (built out of fragments).

I, on the other hand, use CT very differently, for my current project at least, to some extent violating the wiki principle of staying with short fragments, partly because it works for me that way but also because CT doesn't discipline/restrict me that way.

I can see the appeal of Piggydb, at least the way you describe it, if it indeed forces one to implement such a bottom-up fragment-building approach and makes it easy to do so. It sounds like it's trying to implement a bottom-up, "reverse-cascade," "bubbling up" type abstraction process from the detailed data to its abstractions, which is something I'm trying it emulate in CT as well.

How stable is Piggydb at this stage of its development? I looked into it a while ago but for some reason it didn't feel like it was mature enough yet for mission-critical projects. Also, I'd want some easy import-export options, to switch stuff back and forth.

Chris Murtland wrote:

At least in ConnectedText, categories (which are the same as tags), can
be placed into a hierarchy.

CT is a very advanced and flexible tool by wiki standards. For example, the ability to create customised fields and organise/filter the model accordingly is surely not common.

As far as categories go, in the wikimatrix.org it is listed as a single feature with the question "can pages be added to (multiple) categories?", thereby differentiating categories from inclusive folders, but that's about it. For 88 of the listed wikis the answer is "yes" without more information http://www.wikimatrix.org/search.php?sid=846 From my (limited) experience with wikis, I'd say most use a flat list of categories.

In any case, I am discussing here mostly the principles of organisation. In principle, for example, a wiki is very different from an outliner. The fact that there are information managers combining both approaches into one, doesn't make the two approaches any more similar. I would say that it is a good thing when multiple approaches are available to the user, but it also goes to highlight the limitations of each.

Dr Andus wrote:

How stable is Piggydb at this stage of its development? I looked into it
a while ago but for some reason it didn't feel like it was mature enough
yet for mission-critical projects. Also, I'd want some easy
import-export options, to switch stuff back and forth.

In my (non demanding) use of it I have found it very stable. I would expect as such from a tool that is server based: it is not so difficult to render foolproof as the browser interface itself limits the kind of stuff you can pass to the actual program.

Your choice of software is wise, especially considering the importance of your data. As noted earlier, Piggydb is still fairly new and lacking in several areas, import-export being a major one. But it does feel well-built, and moving in a consistent direction.

Is there anyone here who has used Piggydb over a long time as a main knowledge base? I'm just wondering how it scales up when you have hundreds or thousands of fragments. Any performance issues? Any issues with large number of tags and dealing with the complexity of the amassed information?

I went to the piggydb.net website, thinking there might be a forum there where users might have brought up the issue of scalability. Didn't find any forum. Daisuke seems to be working steadily on piggydb - he's up to version of 6.13. You might try asking him about scalability .. perhaps by posting a comment to one of his blog posts. In his "about" post, "scalable" is one of the first words he uses to describe piggydb. But in a comment, he says he doesn't plan to implement substring searching because of scalability concerns:

Hi,
this piece of software is quite uselful to me … thank you. It should be fine to be able to search substring not just exact match of the searched word. Do you plan to implement this ?
br
milan
Reply
Daisuke Morita
January 10, 2013 at 4:02 am
Hi Milan,
Thank you for your feedback.
I don’t have a plan to implement searching substring because I think it won’t scale (maintain certain performance) when your database grows larger.

jimspoon wrote:

You might try asking him about scalability .. perhaps
by posting a comment to one of his blog posts.

Thanks for looking into it and for the suggestion. I'd still be interested in users' experiences though, rather than just the developer's opinion on the matter.