Hi all,

Not sure if this is the right forum for this question, but I'd like to get a sense of how people are using their iLiad to do research reading...

I'm just starting to use an iLiad to try and reduce the vast amounts of paper I use as an academic. In my normal practice, I download academic articles in pdf form, and print them out on paper. I then carry this paper with me everywhere, slowly but surely reading the articles, making lots of notes on them and generally marking them up. They then go in a pile on my desk, sorted by whatever I'm working on. When I'm writing articles, I go through these piles, re-reading marked sections and my notes and incorporating them into my work.

What I've started to do is put articles sorted by topic on my iLiad, scribble on them, and then copy the folder containing the pdf, manifest.xml, and the scribble.irx file to my macbookpro. I use the iLiadPdfScribbleMerger java program to merge the scribbled pdf, placing it in a 'work-in-progress' folder on my desktop. I then use these files when writing.

I have the standard problems of screwing with pdf formatting and using zooming to try and get the text readable but that seems to be mostly working (using the print-to-pdf automator action developed by scotty1024.) I'm currently thinking about trying to find a way to automate the above process a little, maybe by making an automator action than runs when the iLiad mounts, automatically running the java merge, and updating the contents of my works-in-progress folders.

I've gone through the forums to try and get a sense how others are using their iLiad for research reading but I was interested in getting a better overview. If you're willing, (and if you use it this way) could you post how you use your iLiad for reading and marking up texts? I'd be really interested in hearing how people go about this, and what software on and off the iLiad that you use.

best,

Matt

Hi Matt,

I am also using my iliad as a replacement for reading PDFs on screen or printing them out. Here's some of my general approach followed by some of the issues I have worked through to make it work with the iliad, followed by a typical work flow.

A) General approach

i) Referencing:

I use bibtex as my reference manager and until I got my iliad I used to have one large file with all my entries in.

ii) Storage strategy:

I like to name the PDFs that I download using the same keys as I use in my bib files, e.g. blogs03AD would be first author Blogs, published 2003, second author's name starts with 'A', third author with 'D' (to be reasonably sure the key will be unique). I then have a library (directory) with directories for articles, books, reports, and within the articles directory I file the references in alphabetical order. This makes it easy to find an article 6 months down the line.

Important: When I am actually working on something, however, I like to have the articles I am working with in the directory where I am working, and on linux I use symlinks to the stored articles for this.


B) Making this work on the iliad

i) Since getting the iliad, because it uses container directories, I have decided to create a single bib file for each PDF and keep the bib file in the container directory. I then have a script that uses "find" and "cat" (on my linux pc) to get all the individual bib files and makes one big one that I refer to when creating documents on my PC.

ii) I now have all of my references and books on a USB stick so that I can easily access them on my PC or my iliad. I still have a library directory and file articles by author name. I have directories for "A-E" and so on.

I have a script that I run on my PC that takes a PDF and arguments for the description and then creates a container directory, manifest file, an image of the first page of the PDF and puts whole lot in the container directory. The manifest file has a reference to the image so the first page appears in the content lister.

I have another script that I can run from the command line but now have integrated into the gnome file manager that creates something a bit like a windows shortcut to a PDF container. (This is because I can't use symlinks on a USB stick formatted to play nicely with windows and I haven't tried to format my USB stick for ext2/linux). The script creates a new container directory and a manifest file that points to the orginal PDF. I can then move this to my working directory and leave the original PDF in the library. When I open the PDF using this shortcut it refers not only to the original PDF but also to the scribble.irx so I get my previous scribbles and accumulate more as I work.

I do all my reading and annotating on the iliad and use my PC for writing. Reading multi-column articles on the iliad has become much smoother now that I use the version of ipdf with the column-mode patch.


Here's a typical work flow:

i) I do my searching on my PC and save a PDF to a specific directory using my key filename system.

ii) Use pybliographer to get the bib information and create a bib file or download it directly if it is available and save that with the same basename to the same download directory.

iii) Use my script to create the customised container directory

iv) move the container directory to my USB stick

v) create a shortcut and move that to my working directory on the USB stick

vi) Put the USB stick in the iliad, go to my working directory and read and annotate.


This is working well for me. I always have all of my references with me and most of the work is automated.

Here's the script for creating the shortcut. There's one thing that needs to be redefined at the top of the script. This is a variable that says where the USB device is mounted on the host PC. In my case it is:

PCMOUNTDIR="/media/UDISK\ 2.0"

To use this on a machine that uses the gnome desktop, e.g. a standard ubuntu, just unzip this file to the ~/.gnome2/nautilus-scripts/ directory. Then once you have edited the above variable, insert the usb key, open the file manager, find a PDF container directory, right click, select scripts and choose "Create shortcut".

I plan to see if I can get this to work on windows with a copy of bash.exe and sed.exe. Eventually.

For reading papers, there is some unix utilities under the name pdfcrop or similar, or pnmcrop for graphics etc. It can be useful to do some scripts to crop your pdf automatically. Or perhaps to copy the zoom configuration from a paper to other, using bash scripts in the iLiad itself

For reading papers, I rotate to landscape mode and use fullscreen (that's why I wrote it!), then zoom to cut off the side margins. Makes it about 90% of the size of the printed version, and I can read both columns at once.

Of course, wpd's column viewer is really nice and even easier to use.

As for managing files, I just use bibtex, with all my entries in one big file, and I use JabRef to manage them. What's nice is, it can import ISI-exported reference lists. No typing!

For keys I usually use year, followed by the first author's last name, and then a number if that's not unique: 2005Chen, 2005Anderson1, etc.

Then I can put the files in any folder structure under a single directory, and JabRef can find them by name (since the name is the same as the key).

Unfortunately, I haven't figured out a good way of taking notes yet. I don't like writing on the files directly, I like to have all my notes in one place. My current approach is to have a notes file in LyX where I take notes on each article as I read it. This has pluses and minuses, but it's the best method I've found so far...

I do my academic writing in LyX. If you haven't tried it yet, you should!

[...]
As for managing files, I just use bibtex, with all my entries in one big file, and I use JabRef to manage them. What's nice is, it can import ISI-exported reference lists. No typing!

I get most of mine from medline and pybliographer works nicely with that. I've used jabref a bit in the past and I'll have to look at it again.

For keys I usually use year, followed by the first author's last name, and then a number if that's not unique: 2005Chen, 2005Anderson1, etc.

Then I can put the files in any folder structure under a single directory, and JabRef can find them by name (since the name is the same as the key).

That's how I used to have it on my PC, all in one folder. But now that I've moved my library to my iliad (or rather a USB stick that I read on the iliad) I didn't want to have to go through pages trying to find refs. I therefore decided to have a directory structure that makes it relatively quick to locate a particular reference on the iliad. Oh, ... as I typed that I remembered that the iliad can search in single directories. Doh! Perhaps I can get rid of my directory structure.

Unfortunately, I haven't figured out a good way of taking notes yet. I don't like writing on the files directly, I like to have all my notes in one place. My current approach is to have a notes file in LyX where I take notes on each article as I read it. This has pluses and minuses, but it's the best method I've found so far...

I don't feel that I have found an optimal solution either.

I do my academic writing in LyX. If you haven't tried it yet, you should!

I looked at it ages ago, but never really got hooked. I am a latex and emacs guy. With reftex emacs, bibtex and latex work together very nicely. And these days I usually add R into the mix ("Sweave" documents, the ultimate in reproducible analysis). I think it would be too hard to change now.

I wonder if it would be possible to port emacs to the iliad and have gestures for Ctrl and Meta. Then anything would be possible! :)

I haven't gotten my research process fully worked out yet, but I've been getting some really great ideas reading all your comments.

At this point, I read the documents on the iLiad, and scribble. But I need to collect notes about what I'm reading elsewhere, e.g. a laptop, which kind of defeats the purpose. I want to try the new bookmark/comment feature just developed for ipdf and what I'd like to be able to do is somehow extract those comments straight into something like bibtex. Now that we can export the CF card on the iLiad, I thought I might use that for my primary storage, accessed on my laptop via USB as needed (and backed up regularly). But I think it would be tedious to use the keyboard to enter very long comments, and I'm not yet sure how well the handwriting recognition would work for this. I'd be happy enough if I could get a little "post-it" note to write on that would be exported somehow with the reference, so I could read it when I'm getting ready to write. I'd also like to be able to highlight chunks of text to include in my bibliographic database.

All this is really why I justified buying the iLiad instead of sticking with my eBookwise 1150, so I'm really hoping the current trend in innovations for ipdf and related tools will continue, because I do see us headed in the right direction.

I just wish I could use the same annotation tools on HTML and RTF, as well-- I occasionally get reference documents in those formats.

@nekokami

I think we have similar needs in terms of bibliography tools (in fact I've been intrigued by your sig. for a while and was going to ask if you had a spec./wishlist worked out).

IIRC the bookmark feature stores info in the manifest, so if it is stored as a plain text it should not be too hard to have a tool to extract info from the manifest and insert it into a bibtex file (I love plain text files). But as you say writing too much with the keyboard on the iliad will be frustrating. Being able to select a few lines of text and have that saved in the manifest as plain text (with page number maybe) would be very nice. I wish I had more experience with compiled languages so I could try to implement something, but that is not going to happen. Whenever I read about pointers I get a headache and I often have to look 'C' up in the dictionary to make sure I've spelled it correctly. :(

Edit: I convert my html and rtf/doc files to PDF so I can scribble on them.

@nekokami

I think we have similar needs in terms of bibliography tools (in fact I've been intrigued by your sig. for a while and was going to ask if you had a spec./wishlist worked out).
The sig line doesn't seem to show it, but "bibliography tool" is a link to my specs.

Being able to select a few lines of text and have that saved in the manifest as plain text (with page number maybe) would be very nice.
Exactly!

Edit: I convert my html and rtf/doc files to PDF so I can scribble on them.
Yeah, I know I could do that, but I really don't want to have to. They're easier to read in HTML/RTF format, for one thing -- those formats reflow!

Hi all,

great responses - so interesting to read what you all are doing to fit the iLiad to your research practices. I hadn't even considered automating my bibliographic process - I have used Endnote for years (oh why did I ever stop using bibtex!!) but am now exploring using Zotero because of its support for online grabbing of citations.

In my current practice my keeping of references and my note-taking are separate - I've never used the note fields in any bibliographic software, keeping my comments in separate text files in a special "notes" directory. I think this is because what often starts as a short comment becomes a longer piece of text (that I later use in the actual paper.) Keeping notes in small text fields somehow stopped me from doing that kind of extended commentary - and I do everything in my power to keep myself writing! I wonder is this has something to do with disciplinary difference - I'm mainly a humanities scholar that does qualitative research on new technologies.

So are most people here using patched versions of the community iPDF? I see that many of you have created patches and that jharker is working on an all-in-one version - brilliant! I've finally downloaded shell access and ipdf_powersave_bookmarks_for_2.11.zip (the only binary I've found so far for 2.11- I don't have a linux distribution to do patching,) and will eagerly await the full all-in-one version.

Last year, some colleagues and I wrote a grant for the Dutch Science Foundation (NWO) aimed at funding a research project on the scholastic use of ebooks like the iLiad. (I thought it was such a good idea, what with iRex being right there!) We had a bunch of scholars onboard - researchers in typology, librarians, folks in special collections - but couldn't get the science foundation people or iRex interested in supporting the research. One claim we made was that current ebook designs fostered passive reading practices - reading for entertainment - rather than active reading for research (e.g. note-taking, highlighting, etc.) This seemed (and still seems) like a mistake to me - academics and researchers being probably the best fit for ebook early adoption! (Just like the book itself in its print configuration.) I find it so interesting that ebook companies like sony and iRex still haven't taken researchers on board as a market, but also that the researchers themselves (you all) have taken the ball into your own hands with projects like the aforementioned patches and the all-in-one iPdf.

Finally, (and sorry for the long post) has anyone looked at using the iLiad as part of note-taking and annotation software like Pliny (http://pliny.cch.kcl.ac.uk/), AtlasTI (http://www.atlasti.com/), or NVIVo (http://www.qsrinternational.com/)?

Hi Matt,

I agree that researchers are a potentially large market for the iliad. I got mine mainly because reading PDFs on an LCD was killing my eyes, and largely thought of it as an alternative screen. I am now using it much more than I ever thought I would for work, and am also now reading much more for pleasure (books formatted by feedbooks look wonderful!).

Just a quick reply about the versions of ipdf. In the long week or so we have before the first combined version will be released you could try this set of installers (http://davepublic.pbwiki.com/f/ipdf_installers.zip). It includes 3 versions: the standard iRex; fullscreen; and column mode versions.

EDIT: I have just added the bookmark version (that I think also includes the fullscreen patch).

EDIT: Unzip this on your iliad or usb stick or CF disk, go to it on your iliad and then you can choose between versions. You won't get any message telling you which one is currently installed I'm afraid. There's probably no point me making that happen with the new combined version so near.

Last year, some colleagues and I wrote a grant for the Dutch Science Foundation (NWO) aimed at funding a research project on the scholastic use of ebooks like the iLiad. (I thought it was such a good idea, what with iRex being right there!) We had a bunch of scholars onboard - researchers in typology, librarians, folks in special collections - but couldn't get the science foundation people or iRex interested in supporting the research. One claim we made was that current ebook designs fostered passive reading practices - reading for entertainment - rather than active reading for research (e.g. note-taking, highlighting, etc.) This seemed (and still seems) like a mistake to me - academics and researchers being probably the best fit for ebook early adoption! (Just like the book itself in its print configuration.) I find it so interesting that ebook companies like sony and iRex still haven't taken researchers on board as a market, but also that the researchers themselves (you all) have taken the ball into your own hands with projects like the aforementioned patches and the all-in-one iPdf.
You might want to try re-submitting your grant application, citing the posts in this thread and the various threads that describe the patches people are making to ipdf to support active reading and research.

You may also want to look at DotReader, which could be a killer app on the iLiad, as the iLiad supports wireless connectivity as well as pen input. Currently DotReader requires Mozilla, which is too heavy-weight to be practical on the iLiad, but perhaps a grant could fund porting it to a minimo or dillo platform. DotReader could do amazing things for collaborative research and review groups, in my opinion.

Daudi - thanks for the installers. They all work great - I'm currently running in the bookmark with fullscreen version which has added a huge amount of functionality for me.

Nekokami - well, I'm no longer in Holland or working directly with the folks with whom I wrote the grant, but your suggestion is a good one. Definitely the work of people on this forum demonstrates a strong interest in active reading. I'll take a look at dotreader as well. I may be starting a project that relies on the OLPC devices. Supposedly they are fairly proficient ebook readers - might be nice to have some cross-device action going on. You've also reminded me to stay watching dotreader - I haven't been paying attention to it since it hasn't had the ability to read pdfs. However, I read somewhere they they are planning to implement that functionality. If so, that would be a great solution.

Looking forward to the all-in-one release of iPDF!

Thanks for the great thread everyone,
I am also a researcher (medical/neuroscience/philosophy... hell I don't know what to call it anymore!) and am so undecided about whether to get a tabletPC running Ubuntu (Fujitsu T2010) or an Iliad.

The main advantage I see of e-ink is the eye-strain factor but I have a habit of using Lyx/web browser/bibtex referencing software/wiki all at the same time... something I don't expect will be easy with the current Iliad software... although I do have Lyx and Xournal and firefox running on my Zaurus (I guess ports to the iliad should be possible).

I have been a close follower of the development going on in linux for the Zaurus platform (this little device is fantastic!) but useless for reading PDFs... screen too small. Anyway, I also am coordinating some development of online referencing and collaborative editing software so this is an area very close to my heart.

What have been your experiences using e-ink vs modern LCD displays as far as long sessions of reading papers?

What have been your experiences using e-ink vs modern LCD displays as far as long sessions of reading papers?

E-Ink rocks. Much better than LCD for longer reading sessions. And on the iLiad i can read papers very nicely. I tend to read with fullscreen mode and sometimes i use the comlumn mode reader. Though often the order of textboxes gets mixed up. You can read for ours with alot less eyestrain. Well, at least i can. I just would like the E-ink to be faster, so that you can easily flip through pages like you can with paper. But i hope we will have that in a couple of years.

Thanks tribble, good to have some first-hand advice.
I am certainly interested to try linking up my zaurus and an iliad via a peer-to-peer wifi network... which makes me wonder if the iliad can be used as an X-windows client.... off topic I know :)

Hi,

Are there any Zotero users out there? What are your experiences. The combination of Zotero and Iliad looks awesome. But to what extent can they be integrated? Is it possible to run Firefox on the Iliad in some fashion?

Hi,

Are there any Zotero users out there? What are your experiences. The combination of Zotero and Iliad looks awesome. But to what extent can they be integrated? Is it possible to run Firefox on the Iliad in some fashion?

I don't know about running zotero on the Iliad itself, but I do use it on my computer to manage citations and pdfs. It's currently a bit complicated, but basically what I do is download the pdfs online and index them with zotero. Then, I manually copy ones I want to read my iliad, read and notate them, and copy the folders back over to my computer. Then I use iliadPDFScribble merger to merge the scribble and the PDF documents, and copy the resultant single file back into the appropriate location for zotero.

Or I should say that I've done that a couple of times but haven't made it a habit. I need to automate the process.

best,

Matt

Hi Matt,

That sounds pretty straightforward actually. I'm amazed at the capabilities of Zotero -- especially the automatic download. I now have some 500 pdf articles on my computer and I'm not in a mood to print them all out.

A complication in my case is that I run Ubuntu -- you wouldn't have any experiences of that?

yours,

Erik

I just floated this discussion on zotero.org to see if we could get some insight from the other side... theoretically, firefox could run on the iliad... and theoretically, zotero could, right?

A complication in my case is that I run Ubuntu -- you wouldn't have any experiences of that?



I've used Ubuntu but am currently running macosx. Still, all the software mentioned above should run fine on any linux flavor, including the scribbler merger program which is java.

best,

Matt

I just floated this discussion on zotero.org to see if we could get some insight from the other side... theoretically, firefox could run on the iliad... and theoretically, zotero could, right?

seems to me that Firefox is way too resource hungry to run on an Iliad. Personally, I prefer to manage my documents on a computer, and read my documents on the Iliad - sort of the same task division as itunes/ipod.

best,

Matt

Hi Matt,

Thanks. That's very helpful. I should just make the move and buy it!

yours,

Erik

True - but last I checked, getting a song / CD into itunes was a drag and drop - I'm just hoping for a more streamlined approach... just being lazy that way.

With the wireless functionality... that could be an efficient way to "synchronize" with Zotero? or other reference / note manager software... just brainstorming out loud.


seems to me that Firefox is way too resource hungry to run on an Iliad. Personally, I prefer to manage my documents on a computer, and read my documents on the Iliad - sort of the same task division as itunes/ipod.

best,

Matt

http://www.mobileread.com/forums/showthread.php?t=18198&highlight=samba

this thread mentions how to set up samba on the iliad, probably the easiest way to transfer files, as you could mount the iliad on your pc files system and use it just like local storage, this is what I do with my zaurus anyway.

http://www.mobileread.com/forums/showthread.php?t=18198&highlight=samba

this thread mentions how to set up samba on the iliad, probably the easiest way to transfer files, as you could mount the iliad on your pc files system and use it just like local storage, this is what I do with my zaurus anyway.

Actually (and to also address the previous post) the Iliad works in USB host mode - plug it in and it looks like a standard drive. This allows you to "drag and drop" files from your computer to it - which works fine as long as you place them in the appropriate locations.

That being said, samba - along with the Iliad's wireless functionality - does make for a possibly more elegant solution. Run samba on both host computer and Iliad and have untethered drag and drop file transfer.

best,

Matt

[...]
As for managing files, I just use bibtex, with all my entries in one big file, and I use JabRef to manage them. What's nice is, it can import ISI-exported reference lists. No typing!

[...]

Then I can put the files in any folder structure under a single directory, and JabRef can find them by name (since the name is the same as the key).

Unfortunately, I haven't figured out a good way of taking notes yet. I don't like writing on the files directly, I like to have all my notes in one place. My current approach is to have a notes file in LyX where I take notes on each article as I read it. This has pluses and minuses, but it's the best method I've found so far...



I've now have all my bibtex refs in one file and have started using jabref as well and really like it. One nice thing jabref has is an easy way to create export filters. I have created one that creates a manifest.xml from the bib entry. This uses the article/book title as the title, then authors, year, journal name and pages in the description section. I have modified my script for creating container folders to use this manifest.xml, so it is now a very quick and easy process for me to search for articles and import the entries into the bibliography, create a manifest.xml and then create the container folder with the manifest.xml.

I just take notes in the "Review" window in jabref, so they are relatively easy to search, including specifying fields and are all in one place.

Here's the jabref export filter in case anyone is interested. It needs to be named iliad.layout but the file upload seems to want me to call it .txt

daudi, could you show more explicitly how you make the container folders and the files that are supposed to go in there? Examples would be greatly appreciated.

Best,

/ Sp

Here's the bash script that I use. It's pretty simple really. It creates a scribble.irx file (for holding your scribbles), and just creates a directory, moves in the pdf, scribble and manifest files, and renames the directory. I now create the manifest file using the jabref template earlier in this thread.

I have another version of the script that I need to merge with this one that also includes an image, or rather the path to an image depending on the journal.

<image>../_cover-images/$3.png</image>

I make that image line the last thing before the tag that closes the </y-metadata>
This way I only need one copy of each image for each journal.

I have this integrated with my file manager so I just create the manifest file using jabref, and save it in the same directory as the PDF file and then call this script on the PDF file.

Oops, here's the script.

...
I just take notes in the "Review" window in jabref, so they are relatively easy to search, including specifying fields and are all in one place.
...


Daudi, are you running Jabref on the iLiad? I assume from jharker's description of his process that he is running lyx on the actual device. If either is true, how do you switch back and forth between note-taking and PDF reading?

Or do you both (or either) scribble on the pdf and then transpose on the computer?

Hi Matt,

No, I don't run jabref on the iliad I'm afraid (now that would be cool). I'm still refining my work flow, but for now I am reading on the iliad and writing on my PC. I read an article and underline or markup parts that are interesting or important as I go, and then once I have finished the paper I make a quick summary in the review section of jabref using the marks I have made as prompts. Having read the paper, skipping through and taking notes is quite quick.

I am very close to being able to use jabref (running on my PC) to open pdfs on the iliad. I have one little issue to sort out, that I think is a jabref issue that I can work around. I'll try to get time to play with finalising it this weekend.

Thanks for the input Daudi, I will see if I can get something like this to work on my new mac and with my new Iliad.

See, I was thinking if I had the iPDF with comments patch, I could write a script to suck the comments from the XML that creates into a bib utility.

Well, I have this mostly working now. I can use jabref to manage my refs, I have a filter that makes it easy to create a manifest file that contains the author(s) in the title, paper title in the first line of the description and journal name and pages in the second line of the description so they look good and are easy to identify on the iliad.

I also have another export filter that creates symlinks, so I can select a group of refs, export to symlinks and move the symlinks to some work folder on the iliad. That way I can keep all my refs in one folder and have collections for particular projects in another that point to the original articles.

But when you have hundreds of articles and no bibliography manager on the iliad using the iliad search becomes a pain. So, I have installed samba as shown recently (http://www.mobileread.com/forums/showthread.php?t=18198) (thanks ajnorth!) on the iliad and access it from my PCs. I have setup some scripts and registered these as external file handlers in jabref (this is much simpler than it might sound, it's like setting up file associations in Windows 98 or something) and the scripts themselves are quite simple. I then network my PC to my iliad and setup an article entry to do different things depending on which option is chosen. The choices are:

i) open and view the PDF on the PC
ii) Merge the PDF and scribbles and view on the PC
iii) Make a symlink in an easily accessible folder on the iliad so I don't have to search for it
iv) open the PDF directly on the iliad (via ssh)

(The last option works in that it opens the PDF and was what I was after in the first place, but I have trouble closing it afterwards and might have to give up on this approach.)

Here's a screenshot showing the choices. I might even try to make a video and put it on youtube if people are interested and if I can figure out how to do it (never tried youtube before).

Jabref is really nice. I am really getting to like it. At the moment I am liking the static and dynamic groups. And it looks like there will be a plugin for OpenOffice soon (it is already in the dev version).

I have started to work with this and JabRef seems very nice. One thing I do not get is whether I can export several manifest files at the same time. I have a bunch of old articles that I'm starting to sort out and I want to create separate
<bibtexkey>.manifest.xml files in one go and then batch-run daudis container script on all of them. Does anyone know if this is possible?

/ SP

Try registering this attached file as an export filter. Using this to export refs will produce a bash script that will create each manifest file. So, if you make the default extension for this filter to be .sh, and then you select a load of refs and export them using this filter and save it as myrefs.sh you can do:
sh myrefs.sh
to create all the manifest files in the current directory.

Even if jabref would run on the iLiad, you'd have the problem of switching between it and the text you are reading, wouldn't you? We need a window switcher (or a split screen) to be able to do reading and note-taking directly on the iLiad.

Even if jabref would run on the iLiad, you'd have the problem of switching between it and the text you are reading, wouldn't you? We need a window switcher (or a split screen) to be able to do reading and note-taking directly on the iLiad.
Agreed. And it would need to be fast otherwise I think it would interrupt my flow too much. I'm not sure that things refreshes will get too much faster with the current hardware, unless it really is a software issue. I have startd using ssh to login to and poke around the iliad and was surprised how fast it was. I now have a feel for the lag caused by the E-ink.

So, for now, I am focussing on making a PC-based bib manager work *with* the iliad rather than *on* the iliad.

BTW, I made a terrible movie on youtube showing the integrated iliad and jabref-on-PC. It's so bad I'll have to explain what I am doing:

SETUP:
i) PDFs are on the iliad (in fact on the USB in the iliad)
ii) jabref is running on the PC
iii) I have the USB shared using samba (i.e. a windows share)
iv) I have custom file handlers

MOVIE:
i) I select a ref, then right-click to open it on the PC for viewing
ii) I then merge scribbles and open it on the PC for viewing
iii) I then show that there is nothing in the "readnow" directory on the iliad
iv) Next I choose the "make symlink in readnow" menu option in jabref
v) then I show that there is now a link in readnow and open the PDF on the iliad.
vi) I fumble around trying to get the pen out then write "Hi" on the the screen
vii) I go back to the PC and merge and view again and show that "Hi" is now there.

http://www.youtube.com/watch?v=GM-34CA6Cg4

This looks really interesting! When you are done, could you share your scripts with us? Maybe a little howto for the not so experienced users?

Thanks

Sure, will do. I think it should be possible to do this on a windows PC and Mac as well because there's not much to it. Jabref is a java app, so no worries there. I use the scribble merging tool posted elsewhere on mobileread for the merges (java again). jabref makes it easy to specify "associations" and I have just created some very simple scripts that could be rewritten as batch files.

The thing that makes it all so simple now is samba which was contributed recently on mobileread. For (windows?) users who don't know what this is, it setups up linux directories as windows shares. So this should also work fine with a windows PC.

I might not have time for a full howto until next weekend.

Daudi, I tried the batch export from JabRef, but Im stuck. How to do it?

Where do I have to apply the code (sh myrefs.sh) ?? Is this only for linux? (I run Windows).

Thanks
Gogolo

Edit: I dont understand this prodcedure: Try registering this attached file as an export filter. Using this to export refs will produce a bash script that will create each manifest file. So, if you make the default extension for this filter to be .sh, and then you select a load of refs and export them using this filter and save it as myrefs.sh you can do:

Gogolo,

Ah, sorry, for some reason I thought you were using linux. The export layout file creates the shell script when you use jabref export. The shell script is really just a load of "heredocs" but I don't know what the equivalent would be in windows, or as a dos batch file. An alternative would be to write it as a python script (or other language). That would be platform independent (I think) but then would mean you would have to install python.

If you just want to do this once, you could send me the shell file that you create, I could run it and send back the manifest files in a zip file. It might be sensible to try a small number (say 50) refs first before the whole lot to make sure it works.

One other idea: I used to use jython a while back. This is python implemented in java. Presumably you have java already (to run jabref) so perhaps all that is needed is a small change to the export filter to create a python/jython script and then the jython jar file. It's a bit too late for me to get my head around this now. Another thing to add to my list for the weekend.

Thanks for your nice help Daudi. Im still updating my Library - so maybe in a while I will come back to your offer, thanks!

Programming phyton etc... I once programmed games in basic on commodore 64. uuhhhmm seems to be a pretty long time ago :o

Greetings
Gogolo

Doh! It's amazing what a good night's sleep and really big cup of tea can do.
Gogolo, take a look at the batch layout file (posted above), I think you should be able to see what it does. It uses the cat command to write the template to a file name specified at the end of the cat line. You should be able to modify this layout file to produce a DOS bat file that writes the lines out using echo. Something like this:
echo "<?xml version="1.0" encoding="utf-8"?>" > \format{\bibtexkey}.manifest.xml
echo "<package>" >> \format{\bibtexkey}.manifest.xml
echo "<metadata>" >> \format{\bibtexkey}.manifest.xml

etc. Note the single > to redirect output to a new file and subsequent >> to append to that file. This assumes that echo will happily echo everything between the quotes. You might need to escape the quotes in the first line or use single quotes, but hopefully the idea is clear. If echo does agree to play nicely, then once you've done this for each line and then used it to export several refs, you'll end up with a DOS bat file that should create each manifest file.

If there are problems with echoing (e.g. the < or other characters) then it's back to python/jython. I think the layout file would be really simple, something like:

f = open('\format{\bibtextkey.manifest.xml}', 'w')
bib = """
<?xml version="1.0" encoding="utf-8"?>
<package><metadata><dc-metadata>
<Title>\format[Replace("&",and),AuthorLastFirst,AuthorLastFirstAbbreviator,A uthorAndsCommaReplacer]{\author}</Title>
***etc. rest of manifest here***
"""
f.write(bib)
f.close()

So you'd use this layout file to create a python script (e.g. myrefs.py) that you'd run with something like:

java -jar jython.jar myrefs.py
I think I'll play with using jython again a little more it was fun, and it allows access to the java libraries so it could then be possible to create guis and things if necessary.

And for those wondering why I don't just use java in the first place: i) I find it takes too long to write quick code to do simple tasks; and ii) I'm not very good at it.

The example above had a few typos in it. Here (attached) is a layout file for jabref that will create a python script to generate manifest files for multiple references. To run it you either need to install python or you could get jython and create the standalone jar file. To do this, download jython from here (http://www.jython.org/Project/download.html) then run the installer:
java -jar jython_installer-2.2.1.jar
and choose the standalone version. This will ultimately, after a few clicks on the Next button, create jython.jar.

Once you have installed python or created jython.jar and registered the layout file in jabref you can:
i) select some refs
ii) export to a py file, e.g. myrefs.py
iii) run myrefs.py
java -jar jython.jar myrefs.py
or
python myrefs.py
This will create a manifest file for each ref that you had selected. This works on linux (tested) and should work on windows and mac (not tested).

It has just occurred to me that it should not take much more code to have this not only create the manifest file, but assuming that the PDF is named the same as the bibtext key, to:
i) create the PDF container folder
ii) create a scribble file
iii) move the whole lot (manifest, scribble, PDF) into the containder folder.

That way you'd just:
i) dump all your PDFs in one folder
ii) select the refs in jabref
iii) export them to a python script
iv) run the script

And you'd then have all your PDFs prepared for use on the iliad with the appropriate data in the title and description fields.

Here we go. These two layout files now create a python/jython script that when run will create the manifest file, scribble file and put the whole lot (including PDF) into a PDF container directory. You *MUST* have a PDF with exactly the same name as the bibtex key in the directory or it will crash (I have not added nice code for error handling).

The two layout files must be in together. iliad-batch-jython.layout is registered in jabref in the usual way and it will pickup the "begin" layout file when it is run.

So, the steps are now:

i) install python or get jython (as shown above)
ii) register the layout file in jabref (Options->Manage custom exports)
iii) place your (correctly named) PDFs in a directory
iv) select your refs in jabref
v) export them using the new export filter (thereby creating a python script)
vi) run the python script in the directory with the PDFs

I've made some more updates to the layout file but also have some more ideas and think it'll get confusing if I keep posting small updates, so I'll just explain what I have done and intend to do.

I've now changed python script created by the layout file (but not posted it yet) so that it will determine for each selected entry if the PDF is file or a directory. If it is a file it will create the container directory, etc as above. If it is a directory (i.e. it has already been opened on the iliad) it will only create a new manifest file and will leave the existing scribble file. I've added this because I decided I want to be able to use jabref to quickly update how all of my articles appear on the iliad.

If neither a file or directory exists that matches it does nothing (i.e. no longer crashes).

The thing that I am thinking about adding is support for images. I'm thinking that if there is an image (in a user-specified image directory) that matches the journal field then it will add a line to the manifest and copy in that image. So if I have an image called bmj.png and I have a journal entry bmj then it will copy in the image. That way I will be able to have journal covers for each article.

I think this could easily be used to manage non-academic content as well, e.g. PDF books. There's no journal field for book types by default but it is easy to add that and set its value to a name that would map to a cover image.

Great work Daudi and thanks for sharing!

I've made some more updates to the layout file but also have some more ideas and think it'll get confusing if I keep posting small updates, so I'll just explain what I have done and intend to do.
<snip>

I think this could easily be used to manage non-academic content as well, e.g. PDF books. There's no journal field for book types by default but it is easy to add that and set its value to a name that would map to a cover image.

This is great work. Thanks for it and for your detailed descriptions and notes!

A slightly off topic question (however probably useful for researchers):

Tribble, in post #15 of this thread you say that you are using sometimes the column mode and sometimes the fullscreen mode. Is it possible to have both these patches to iPdf installed at once? And how do you go about installing them both?

/ Sp

A slightly off topic question (however probably useful for researchers):

Tribble, in post #15 of this thread you say that you are using sometimes the column mode and sometimes the fullscreen mode. Is it possible to have both these patches to iPdf installed at once? And how do you go about installing them both?

/ Sp

I simply install the ipdf that i need. I have severall installers, with wich i can switch quite fast. I wish there was a package with all features included. Like fullscreen, gestures, bookmarks and columnmode.

Daudi,

It would be very nice to have a summary of how you get JabRef to work with Samba and everything else you have been playing around with. I know that time is scarce for everyone but at least I would find a concise guide extremely useful. So if you have time to write a "Beginners Guide", I would be most grateful.

Best,

/ SP

I've been meaning to get around to doing this, so thanks for prompting me . OK, here's what I've done. I'm doing this with a linux box, but as this makes use of windows shares (that's what samba is) it should be easy for people using windows as well, except that I use bash scripts and these would need to be re-written in something for windows (although there is bash.exe so perhaps this could work as is). And I guess it should work fine for Macs (but I don't have one to test it).

1. Set samba up using ajnorth's samba for the iliad

2. I currently keep my refs on a usb key (but plan to move to a CF card) so I have set up samba to share the usb key.

3. I mount the shared usb key (which is plugged into the iliad) on the PC. I happen to mount it at /home/dave/work/refs/iliad. I use a script for this:

sudo mount -t cifs //192.168.11.3/usb -o uid=dave,gid=dave,credentials=iliad_credentials iliad

There are other ways of doing this and it will depend on your OS and flavour of OS. I presume it should be simple enough for windows users to connect to the shared USB key (that's connect as in networking, not USB connection even though it is a USB key).

4. At this point the usb key with the references is plugged into the iliad and connected to the PC via windows/samba networking, so it can be accessed just like a normal share/drive.

5. I then set up the file part of the jabref entries to have multiple file entries for each reference.

file = {View on PC:/home/dave/work/refs/iliad/albala02.pdf/albala02.pdf:PDF;
Merge scribbles, view on PC:/home/dave/work/refs/albala02.pdf:PDF merge scribbles, view on PC;
Make symlink in read-and-review:/home/dave/work/refs/iliad/albala02.pdf:PDF make symlink in readnow}

I plan on writing a script to write all these entries for me. jabref can auto-discover refs for normal PDF viewing and I'll just take those and automate the others. Eventually.

The first entry (View on PC) just points to the PDF file (not the container) so that clicking on it opens the PDF in the normal way.

6. The next two entries use customised file handlers that I have added. If you go to jabref and choose "Options" you will see "Manage external file types". This allows you to add new file types. All I did was add the path to some scripts that I wrote (see below) for handling PDFs.

7. The script that merges the scribbles and opens the result on the PC is:
#!/bin/sh
# Merge a PDF file on the iliad with its scribbles
# then display it on the PC
# Assume that jabref points to container folder not PDF file.
OUTFILE=$(basename "$1")
java -jar "$HOME/bin/merge/ILiadPdfScribbleMerger.jar" -i:"$1" -o:$OUTFILE -m:n
evince $OUTFILE
rm $OUTFILE

This uses bash and I was planning to re-write this using python/jython to make it usable on all platforms. That's partly why I had not posted anything about this sooner.

8. The other script I created is one that creates a shortcut to a reference and places it in a "read and review" directory. This means that I can quickly select some refs to read and dump shortcuts to them in one place and not have to move references around. It simply creates a directory, a manifest file and fixes the path from that seen when the share is mounted on the PC to the path that of the file as seen by the iliad. This is also a bash script:

#!/bin/sh
# Copy a symlink to read-and-review

ILIADMOUNT="/home/dave/work/refs/iliad"
DEST="read-and-review"

SYMLINK=$(basename $1)
NEWDIR=$ILIADMOUNT/$DEST/$SYMLINK
mkdir $NEWDIR
sleep 1 # USB is slow.

PDF=$(echo $1 | sed -e "s|$ILIADMOUNT|/mnt/usb|")

cat <<EOF > $NEWDIR/manifest.xml
<package>
<symlink>$PDF</symlink>
</package>
EOF

Thanks Daudi, this will help alot.

I made a Firefox plugin to export files simply to the iLiad. A more detailed description can be found in a related thread (http://www.mobileread.com/forums/showpost.php?p=177774&postcount=1)

hi guys i've been doing a lot of research on this feature but i just can't find it anywhere:

more specifically, i'd like my scribbles/text-selections (whatever, be it a selection on text based PDF file, be it a scribbled circle around some region on an image) to be automatically collected/put together on a separate file, say "textmarks" or "highlights".

for example i wished i could go on reading and selecting text/scribbling circles around paragraphs. when finished, there would be an option to extract these selections/circled regions and gather/sum up these things on a separate file all at once by processing the hole document(or a given page range).

thanks for reading all this and apologies for my poor english; it's soo important to me... so does anyone has any clue on whether this feature already exists or where else should i be asking this question?

i'm new to all this and this feature would justify by itself my iliad purchase as this feature plays an important role on my research process. :thanks:

If you are using linux then I think this should be what you are after (http://www.mobileread.com/forums/showthread.php?t=22804). If you are on windows it might still work but you might have to put in a little work to get it working smoothly (I have not tried it on windows). At the end of the thread it is also points you to a search tool that allows you to search the selections on the iliad.

If you are using linux then I think this should be what you are after (http://www.mobileread.com/forums/showthread.php?t=22804). If you are on windows it might still work but you might have to put in a little work to get it working smoothly (I have not tried it on windows). At the end of the thread it is also points you to a search tool that allows you to search the selections on the iliad.

THANK YOU MAN
thats precisely what i was looking for, as it allows the automatic extraction of snippets(bits of text, image areas). i'm using windows so i'll be dealing with some difficulties, but it's no problem i suppose...

Great. Let me know if you need help. It would be great to have this worked out and tested for windows. First you'll need to have pdftohtml working for windows. I saw a couple of references to a windows implementation in passing, but did not look at these too carefully. Assuming that is able to run from the command line and can create an xml representation of PDFs then the python script should run fine on windows.