Several people requested an outlining feature in Zotero. I'm curious, how many people would use it? And also how would you like it to work in terms of interface, export, etc.?
I would love such functionality. Something as simple as Treepad would be great. A smart bib-manager like Zotero that can do minimal outlining would be an enormous resource.
What I would like is to be able to group and order notes, sources, etc. around in an hierarchical format. Notes should be able to move on their own, but without forgetting where they're coming from -- if they have parents.
What might work is if the leftmost pane could toggle between the current library display and an outliner display.
I think Scribe3 really had a hand up on Zotero due to its multiple list functions. I especially miss the chronology function. I know that Zotero is supposedly integrated with SIMILE Timeline, but a chronology by publication date does me very little good and I haven't had the time to invest in figuring out how to really make Timeline work for me.
It would also be wonderful to be able to create a citation from a note in Zotero as is possible in Scribe.
MTBradley--I think if developers add an "event" item type down the road, the current SIMILE function will work for what you want to do.
It is impossible to create a citation from a note unfortunately, because Zotero doesn't recognize page numbers in the note--but this may be supported eventually also.
Bruce's comment at the end of this thread notwithstanding, I like the idea of connecting outlining (or better: dialogue mapping) with Zotero. I would be glad if, once I had my notes (or excerpts) into zotero I could *do something* with them in ways that would help me (1) "see" the existing academic discussion and (2) plan out my own contribution to it.
I've recently been intrigued by two pieces of software which do dialogue mapping, and their potential for bringing together various academic treatments of an issue into a single conversation.
The first is Compendium, a dialogue mapping tool (free, cross-platform, not GPL --yet-- but source-available). It works nicely for identifying the relationships between ideas, but is (a) weak on importing existing data (b) doesn't handle longer notes/excerpts well and (c) has no built-in method to match notes to bibliographic resources.
The second is a web-based second-generation version of a similar concept It's called Cohere (nice screencast available). In this one you don't map the dialogue yourself in a 2D space, but you define relationships between ideas (either from a set list or make your own) and it creates a 3D model of the relations for you, which you can explore and zoom in an interesting spinning visualization (which is hard to imagine being useful in its present form, but is a tantilizing glimpse at what might be possible.)
Now, could Zotero ever interface with a program like these? This is related to the FreeMind thread, since a similar kind of use of zotero data would be involved. I'm even more interested in this however, since so much academic discussion matches the "dialogue" metaphor quite well, seems to me.
In a way the 'guts' of this idea are already in by the plans to establish semantic relations between items. It would just need an ability to easily make those associations, and to visualize the results in 2D or 3D space, ether mechanically, like Cohere or manually (which I like a lot) like Compendium.
This is an old discussion, but outlining will be very very useful to law students and practitioners. All law students use outlines to learn their subjects, which will include cases and statutes to illustrate a particular type of rule, and they'll revise and simplify this as time goes on. When they graduate many will continue to refer to these outlines to refresh their memories, and some will build on them so they can quickly cite to a useful authority in correspondence, pleadings, and conversations, or find a starting point for research. And all attorneys writing an argument for the court or a memorandum for a client will organize his/her reasoning and authorities in an outline first, and the final written product will keep the same organization and headings as the outline.
As for how it should interface, there are outlining software that some people use, which I am not familiar with, that could be looked at. I am new to Zotero but I expect to be putting rules of law that a case or a statute represents under an item's notes (e.g., under the notes for *Harry v. Sally* I might put "Community property principles do not apply to unwed couples."). I think if I could create an outline in Zotero with sentence-length headings and then drag and drop items into the outline (which should be in a pane I can see as I go through my library), and then be able to export that structure, the headings, the citations, and the notes I've written for the item, that ought to do it. All I would have to do to churn out a memorandum then is create an outline, generate the structure, headings, citations, and rules of law that they represent, and then fill in the facts of my case. I'd also want to be able to export just the outline and headings for summaries that may be required for longer arguments, and just the citations for tables of authorities.
Being able to link a subheading in one outline with several subheadings of another outline would be a valuable feature, because areas of law intersect. If you could turn an existing folder structure into an outline that might be helpful too, as that's how I am currently organizing things in lieu of this function.
fearsomeworrier: That's really interesting! But remember that Zotero is (mainly) a citation manager and the core development team has very limited resources.
I wonder though - this sounds like it could just be done with collections, no? The collections could provide both the hierarchical structure and the headings.
The only thing that Zotero can't do then is to print these out as structured data with citations & notes - that doesn't strike me as super difficult. Gracile is right that this is unlikely to be something core devs will attack any time soon, but with the structure already in place, it wouldn't be prohibitively hard to write a plugin that exports the outlines, perhaps drawing on existing create report capabilities?
That's what I think - not terribly difficult, and it would expand Zotero's usefulness quite a bit, for law types and anybody who needs to organize their sources in a meaningful manner, instead of just categories. That's really missing a big chunk of intellectual life. Why have one category of software (citation managers) and another (outlining software) when you can combine the two and use the same information? I'm sure the more interest there is in Zotero the more resources would be available to develop it.
but as I say - this isn't the type of thing that is going to see a lot of time from the core developers anytime soon. The way to go would most likely be to find someone who is interested and has some coding knowledge to do this as a plugin.
Collections don't allow arbitrary ordering, though—wouldn't that be necessary? When we've discussed outlining ability in the past, that's what's been missing (along with the ability to include notes separately from their parents/siblings, but that could possibly be accomplished by using tags).
Can I request an outlining facility in Zotero? It would be tremendously useful and is the logical next step for people looking to make the most of the technology.
I'm not an expert, but it seems that it is almost possible to do so in Zotero presently. Ie, rather than creating child notes on documents, notes are related to the document. Each note is then a one issue / quote / thought 'factoid'. These factoids are then arranged in folders and subfolders. The only thing missing is an ability to arrange the factoids in folders in an arbitrary way, rather than alphabetical etc.
Alternatively it would be really helpful if it was possible to export the notes to another program, while retaining the link to the parent item.
Alternatively it would be really helpful if it was possible to export the notes to another program, while retaining the link to the parent item.
not sure what you mean here - could you explain what you mean by a link to the parent item?
Beyond that: This is certainly not something that anyone is opposed to, but I doubt it's anywhere close to the top of priorities. There is a ton of stuff that's part of Zotero's core functionality as a reference manager that's still not implemented - see here for some examples: http://www.zotero.org/support/requested_features
The issue is that in order to outline, you need to use notes that typically contain only one 'factoid' (idea, reference, or quote). Using the original source is too cumbersome. However, you also need to be able to quickly find the part of the original source where the factoid came from. The existing reports could be adapted, but they presently do not contain sufficient information to identify the original source. They do contain much of the information if they also contain the parent item, but keeping the notes with the parent item makes them cumbersome. If, on the other hand, they are freestanding, then they only carry scant information about the items they have been related to. This means that one has to add the information manually. While this can be done after the note has been exported to another program (though the quick copy function), it cannot be used to paste the reference into a Zotero note.
After a little back-and-forth communication between Paul and myself, I think we've worked out a nice solution for tying Zotero for Firefox to external outlining utilities.
There are two parts to it. The first bit is a couple of small tweaks to the zotero://select/items/[zotero_key] protocol handler, to prevent it from opening an empty tab every time it is called, and to assure that it will not fail if Zotero has not yet been opened in the browser. I've put up a patch so the core team can take a look when time permits.
(Edit: My tweaks "to assure that it will not fail if Zotero has not yet been opened in the browser" do not work, actually, but the empty-tab issue is addressed by the patch.)
The second bit is an extension to the CSL processor to allow individual cites to be wrapped in a link. With a small addition to the Preferences UI, this could be enabled for use with QuickCopy. The code for that was checked in on the citeproc-js archive this evening.
With an external outlining utility that handles HTML (or other) formatted links, we should be able to cut and paste references directly from Zotero into external notes, and retain one-click access back to the item and its associated data in Zotero.
This hasn't yet been used in the field, but the pieces are all in place, and we'll soon be giving it trials in MLZ. I'm pretty confident that it's going to work out quite well.
Rintze's suggestioon of using Docear is great, though it requires Zotero to automatically export a BibTex file after each added item. Would this be hard to implement?
Note that people have reported severe performance problems when using AutoZotBib, presumably with larger libraries. Doing that efficiently would require a different implementation, and possibly some structural changes in Zotero.
Further to Frank's note, I have been using his setup with MLZ Zotero to do outlining. It works extremely effectively and I would highly recommend it. Basically, with a click and drag on your Zotero item, you get a plain text or html link that you can paste anywhere and that takes you back to the item in Zotero. The beauty of it is that you can use this link in any outlining program. I have been using Scrivener and Notational Velocity to do outlining in plain text. Even better than this, a python script Zodfscan will convert the plain text links to 'live' Zotero citations in odf format. These simple steps have added a massive amount of academic functionality to Zotero.
I haven't tried MLZ, but one can move back and forth between an external outline and Zotero references using regular Zotero with some kludging. I use Microsoft Word's outlining feature. My current workaround is to insert an identifier -- {author date} -- in the outline, and then write a macro (in Macro Express) which copies that identifier to the search field in Zotero. This approach offers a great deal of flexibility, though it's not elegant.
@adamsmith: I would strongly support that proposal.
@Heckscher: Have you tried Frank Bennett's Multilingual Zotero? I suspect that would achieve what you're trying to do with much greater reliability. The identifier is created using a click and drag and it is extremely reliable given that it uses a unique identifier. If I understand you correctly (I'm not sure I understand the copying to the search field in Zotero), you are using the RTF scan format style of {author date}. This seems likely to cause confusion when converting to true Zotero references. You could still use word's outlining function.
I just found out about Multilingual Zotero and have been a bit cautious about trying it since the current setup works ok. But I very much like the idea. Is experience with MLZ good -- is it stable enough for serious use?
(Btw: "copying to the search field" just means that I copy the reference in the Word outline and paste it into the Zotero quick search field to find the item. As I say, it works ok, but it's klutzy!)
I have been using MLZ Zotero for about a year and have experienced no problems. I have it set up on a separate Firefox profile, so I can run MLZ Zotero and Vanilla Zotero side-by-side using the same Zotero database.
If you can work out how to install and use the Zodfscan python script (which is not entirely straightforward if, like me, you have no idea how python works), I think you will find that it is far superior to manually cutting and pasting into the Zotero search field.
I've amended the instructions in the README file to include a quick-and-simple way of running the conversion script.
Eventually, it should be possible to recast the code (in JavaScript) as a Firefox plugin; the code for it isn't heavily dependent on Python per se. But for now it seems to do the trick once installed.
This extension to MLZ is very great and I would appreciate it being a feature of Vanilla Zotero! It's very useful for taking notes and writing in other applications than Word/Openoffice.
Upgrade Storage
What I would like is to be able to group and order notes, sources, etc. around in an hierarchical format. Notes should be able to move on their own, but without forgetting where they're coming from -- if they have parents.
What might work is if the leftmost pane could toggle between the current library display and an outliner display.
It would also be wonderful to be able to create a citation from a note in Zotero as is possible in Scribe.
It is impossible to create a citation from a note unfortunately, because Zotero doesn't recognize page numbers in the note--but this may be supported eventually also.
I've recently been intrigued by two pieces of software which do dialogue mapping, and their potential for bringing together various academic treatments of an issue into a single conversation.
The first is Compendium, a dialogue mapping tool (free, cross-platform, not GPL --yet-- but source-available). It works nicely for identifying the relationships between ideas, but is (a) weak on importing existing data (b) doesn't handle longer notes/excerpts well and (c) has no built-in method to match notes to bibliographic resources.
The second is a web-based second-generation version of a similar concept It's called Cohere (nice screencast available). In this one you don't map the dialogue yourself in a 2D space, but you define relationships between ideas (either from a set list or make your own) and it creates a 3D model of the relations for you, which you can explore and zoom in an interesting spinning visualization (which is hard to imagine being useful in its present form, but is a tantilizing glimpse at what might be possible.)
Now, could Zotero ever interface with a program like these? This is related to the FreeMind thread, since a similar kind of use of zotero data would be involved. I'm even more interested in this however, since so much academic discussion matches the "dialogue" metaphor quite well, seems to me.
In a way the 'guts' of this idea are already in by the plans to establish semantic relations between items. It would just need an ability to easily make those associations, and to visualize the results in 2D or 3D space, ether mechanically, like Cohere or manually (which I like a lot) like Compendium.
http://forums.zotero.org/discussion/10994/how-to-install-vue/
It is far from perfect yet, and doesn't work with notes, but it's a good beginning.
As for how it should interface, there are outlining software that some people use, which I am not familiar with, that could be looked at. I am new to Zotero but I expect to be putting rules of law that a case or a statute represents under an item's notes (e.g., under the notes for *Harry v. Sally* I might put "Community property principles do not apply to unwed couples."). I think if I could create an outline in Zotero with sentence-length headings and then drag and drop items into the outline (which should be in a pane I can see as I go through my library), and then be able to export that structure, the headings, the citations, and the notes I've written for the item, that ought to do it. All I would have to do to churn out a memorandum then is create an outline, generate the structure, headings, citations, and rules of law that they represent, and then fill in the facts of my case. I'd also want to be able to export just the outline and headings for summaries that may be required for longer arguments, and just the citations for tables of authorities.
Being able to link a subheading in one outline with several subheadings of another outline would be a valuable feature, because areas of law intersect. If you could turn an existing folder structure into an outline that might be helpful too, as that's how I am currently organizing things in lieu of this function.
The only thing that Zotero can't do then is to print these out as structured data with citations & notes - that doesn't strike me as super difficult.
Gracile is right that this is unlikely to be something core devs will attack any time soon, but with the structure already in place, it wouldn't be prohibitively hard to write a plugin that exports the outlines, perhaps drawing on existing create report capabilities?
That's what I think - not terribly difficult, and it would expand Zotero's usefulness quite a bit, for law types and anybody who needs to organize their sources in a meaningful manner, instead of just categories. That's really missing a big chunk of intellectual life. Why have one category of software (citation managers) and another (outlining software) when you can combine the two and use the same information? I'm sure the more interest there is in Zotero the more resources would be available to develop it.
The way to go would most likely be to find someone who is interested and has some coding knowledge to do this as a plugin.
Any takers reading this? I would buy that person a meal.
I'm not an expert, but it seems that it is almost possible to do so in Zotero presently. Ie, rather than creating child notes on documents, notes are related to the document. Each note is then a one issue / quote / thought 'factoid'. These factoids are then arranged in folders and subfolders. The only thing missing is an ability to arrange the factoids in folders in an arbitrary way, rather than alphabetical etc.
Alternatively it would be really helpful if it was possible to export the notes to another program, while retaining the link to the parent item.
Beyond that: This is certainly not something that anyone is opposed to, but I doubt it's anywhere close to the top of priorities. There is a ton of stuff that's part of Zotero's core functionality as a reference manager that's still not implemented - see here for some examples:
http://www.zotero.org/support/requested_features
The issue is that in order to outline, you need to use notes that typically contain only one 'factoid' (idea, reference, or quote). Using the original source is too cumbersome. However, you also need to be able to quickly find the part of the original source where the factoid came from. The existing reports could be adapted, but they presently do not contain sufficient information to identify the original source. They do contain much of the information if they also contain the parent item, but keeping the notes with the parent item makes them cumbersome. If, on the other hand, they are freestanding, then they only carry scant information about the items they have been related to. This means that one has to add the information manually. While this can be done after the note has been exported to another program (though the quick copy function), it cannot be used to paste the reference into a Zotero note.
There are two parts to it. The first bit is a couple of small tweaks to the zotero://select/items/[zotero_key] protocol handler, to prevent it from opening an empty tab every time it is called, and to assure that it will not fail if Zotero has not yet been opened in the browser. I've put up a patch so the core team can take a look when time permits.
(Edit: My tweaks "to assure that it will not fail if Zotero has not yet been opened in the browser" do not work, actually, but the empty-tab issue is addressed by the patch.)
The second bit is an extension to the CSL processor to allow individual cites to be wrapped in a link. With a small addition to the Preferences UI, this could be enabled for use with QuickCopy. The code for that was checked in on the citeproc-js archive this evening.
With an external outlining utility that handles HTML (or other) formatted links, we should be able to cut and paste references directly from Zotero into external notes, and retain one-click access back to the item and its associated data in Zotero.
This hasn't yet been used in the field, but the pieces are all in place, and we'll soon be giving it trials in MLZ. I'm pretty confident that it's going to work out quite well.
@Heckscher: Have you tried Frank Bennett's Multilingual Zotero? I suspect that would achieve what you're trying to do with much greater reliability. The identifier is created using a click and drag and it is extremely reliable given that it uses a unique identifier. If I understand you correctly (I'm not sure I understand the copying to the search field in Zotero), you are using the RTF scan format style of {author date}. This seems likely to cause confusion when converting to true Zotero references. You could still use word's outlining function.
(Btw: "copying to the search field" just means that I copy the reference in the Word outline and paste it into the Zotero quick search field to find the item. As I say, it works ok, but it's klutzy!)
If you can work out how to install and use the Zodfscan python script (which is not entirely straightforward if, like me, you have no idea how python works), I think you will find that it is far superior to manually cutting and pasting into the Zotero search field.
Eventually, it should be possible to recast the code (in JavaScript) as a Firefox plugin; the code for it isn't heavily dependent on Python per se. But for now it seems to do the trick once installed.