The online community for annotating James Joyce currently seems pretty fragmented

amanda, will iu 1.0 support images?

I can’t speak for Amanda, but I can say that image annotation is a bit of a holy grail at the moment, particularly with the Annotator.js library she’s using.

@mwidner have you seen Annotorious before? https://github.com/annotorious/annotorious

it’s a plugin for annotator.js ; also I believe Mirador has image annotation baked in at some level http://iiif.github.io/mirador/

I have, but it doesn’t work with the latest 1.x branch of Annotator. You’d have to run an older version to use annotorius. the original idea was that they’d pick up development again once 2.0 was released, which it has been, but I haven’t heard of any progress on that front.

Mirador does not use annotator.js for image annotation, because the 1.x branch of Annotator is OA compliant, so the Stanford Libraries decided to develop their own image annotator.

huh, ok — yeah, last time i looked at it was a few years ago — just didn’t think it was that much of a holy grail

I’m hoping that they’ll work with http://Hypothes.is to update the image annotator in Mirador so that it’s an Annotator.js plugin.

but even then, the version of Annotator that Amanda’s site uses is Annotator.js 1.x, so that would need updating. Probably not too much of a challenge for her site, but for Lacuna Stories (which is partially what hers is based on) has a lot of custom Annotator.js plugins that would need updating.

Getting a community together focused on image annotation working with Annotator.js is a goal of mine that I haven’t had much time to pursue lately.

i see

Yeah… :disappointed:

you have a lot invested in annotator.js i guess

seems like since you are at stanford getting on the mirador train might be a good idea?

Well, I developed Lacuna Stories (http://www.lacunastories.com) and it uses Annotator.js. Plus, I like settling on a widely-used, open source library like that. The other options are not widely adopted. For example, Columbia’s MediaThread has image annotation in it, but afaik, that’s a home-grown solution.

Mirador is very focused on images (not surprising, considering that it’s a demo for IIIF). I don’t know that they have any plans for text annotation. I should ask again, though.

oh, text too ; last i heard they were working on image based transcription

but that’s sounds different from what you’d want

yeah, that sounds more like they’re doing a T-PEN-style approach. Not sure, though.

I haven’t talked to the developers about it in a while.

i guess even stanford libraries is a big place :simple_smile:

It’s very big, especially the team that Drew is on: Digital Libraries Systems and Services (DLSS). I’m in a different part and actually don’t spend much time in the libraries proper because I’m embedded in a literature department.

thanks for bringing it up. I just wrote to Drew to see if there’s been any movement there. My guess is no. :simple_smile:

there’s been movement, perhaps not the movement you are looking for

:simple_smile:

yeah, Mirador is great stuff.

@timfinnegan: do you mean annotating the text with images?

@literature_geek: I don’t know if you’ve seen, but Lacuna Stories allows you to embed media in your annotations now: images, video, gifs

I’m going to play around with https://github.com/danielcebrian/richText-annotator for adding rich text and images to annotations

But I’m not sure yet how well it works or will integrate with Drupal. @mwidner and I have emailed a bit about getting that plugin to work with existing Drupal WYSIWYG

@mwidner: Oh wow!

just grab the plugin I wrote. it’ll save you time. as soon as I get the install profile working properly, I’m going to open source everything.

just don’t want to release software people can’t install! :simple_smile:

@mwidner: Thanks, that’s fantastic! Maybe after the release we could Skype? I’m waiting to hear back on two grant proposals to support more annotation coding, and it would be neat to compare wishlists and avoid redundancies.

sure thing!

I’ve added so many new features this past summer that when it’s finally publicly released, I’m calling it Lacuna Stories 2.0

has anyone seen a classification of annotation types, eg: link to sourcetext, bio-of-real-person, word-definition, illustrative-image, meta-commentary-image, etcetcetc?

In what context? I’ve seen a few examples of categorization of annotations. Lacuna Stories allows for that. We have “Comment”, “Question”, “Analyze”, and “Compare”, but it sounds like you’re looking for something more nuanced. It’s customizable to allow for that, but it’d probably get rather messy in the interface with too many categories.

dream feature: highlight a name and type ‘bio’ and have it link the wikipedia bio, or generate a map, or translate a passage, or fetch a synonym

(belated realisation: ‘image annotation’ can mean either adding annotations to images, or adding images to annotations)

is everyone open to interlinear placement of annotations, rather than popups? (for finnegans wake it’s the only efficient approach, i think: http://fwannotated.blogspot.com/2014/09/p25h.html )

has anyone else been burned by http://genius.com , where seniority is based on a gamified point system that promotes obsessives over insight?

With storage so cheap, a site like Genius or Infinite could avoid edit-wars by letting every annotator create their own forked view

@timfinnegan: Something I’d like to do is let any reader generate a custom URL that you can visit to see just the annotations that reader curated (their own annotations and/or ones they favorited).

@timfinnegan: I know Jonathan Reeve and Alex Gil are thinking about letting people fork texts on GitHub (https://twitter.com/j0_0n/status/643401337334468608) and then maybe make custom editions (https://twitter.com/elotroalex/status/646831403963969536)

ASCIIDOC looks too simple for Ulysses, we need: newspaper-headlines, dialog-speaker, stage-directions, song-lyrics, quote-dash (so css can restyle it), various other odd indents…

@timfinnegan: I like your bio/map/etc. dream feature. I could see this returning the first result of a search of Wikipedia or Google Translate, and then letting the annotator click a button to confirm adding the result or to say “that isn’t right, show me a different result”

@timfinnegan: I should get a master list of desired features up in a public place so people can add to it and let me know which are priorities for different types of reading…

Just joining and looking around, but wanted to respond to @literature_geek’s comment: https://digitalhumanities.slack.com/archives/annotation/p1445024939000043

We’ve been working on adding annotation functionality to http://readux.library.emory.edu based on annotator.js (the release is almost done and should go out soon); that rich text annotator plugin didn’t work so well for us, so we created a different one using markdown and meltdown: https://github.com/emory-lits-labs/annotator-meltdown

I also have an annotator.js plugin for selecting and annotating portions of images using imgAreaSelect http://odyniec.net/projects/imgareaselect/ but I see now I haven’t yet pulled that code into its own git repo. (I’m not sure how generalizable it is yet, but would be interesting to see if it worked for someone else.)

@suttonkoeser: Cool! Thanks, that’s really helpful to know.

ooooo… that’s really exciting. @suttonkoeser Is it off the 1.x branch of Annotator.js?

If so, I could probably integrate it into Drupal pretty quickly.

@mwidner: no, we saw there was a 2.0 alpha release of annotator and didn’t want to target a version that was about to be out of date, so it’s compatible with the 2.0 alpha.

gotcha. That makes sense. Unfortunately, I don’t have time to update all my plugins for 2.0 right now. :disappointed:

If you post the plugin on Github somewhere, though, it might still be useful to take a look.

Yeah, I hope to do that soon. I’ll drop a link in this channel when I get it up on github

Also: I saw in the intros thread you have a PhD in English. Me, too! What was your speciality?

20th c poetry (although that was a while back now!); dissertation on Eliot’s Four Quartets, Susan Howe’s Nonconformist’s Memorial, and ideas of reading those texts nonlinearly / out of order / as exploring a textual space. I see projects like Infinite Ulysses now and wonder what I could have done had that kind of work been farther along at the time! (if you want to discuss more maybe we should move to dm, but thanks for the interest)

Just curious! :simple_smile: Sounds like Amanda’s work would’ve been great for you, for sure.

in annotating ulysses, there are three distinct levels of ‘consciousness’: 1) no access to gifford or equivalent, 2) access and agreement, 3) access and disagreement. (gifford compiled the standard in-print collection of annotations.) it might get messy/ impolite to assign these levels to others, but could we ask contributors to classify themselves?

underwhelming annotated bartleby: http://www.slate.com/articles/arts/culturebox/2015/10/herman_melville_s_bartleby_the_scrivener_an_interactive_annotated_text.html

in theory, the annotations at any point in a text can be expanded to mirror the whole universe as reflected in that point… in practice, choosing any point to begin from is near impossible… unless it’s joyce

the corollary here is that every annotation is potentially infinite, so if you’re soliciting reader input you want to offer/ allow for infinite whitespace

(so interlinear trumps hyperlinks)

eg in finnegans wake when HCE meets the cad, joyce is pretty unambiguously inviting us to map all meetings ever onto it, back to primal carbon atoms meeting hydrogen.

and the mythic ai Singularity must perfect this universal mapping

has anyone else noticed a conflict between crowdsourced online annotation projects, and academic cv-enhancement? how do you get credit for creativity you give away?

@timfinnegan: This is an interesting question—and really important given that the crowdsourcing/community-building projects I know of are all really concerned with crediting annotators as co-authors. Some related thoughts:

1) Is there an issue with citing on your CV work done under a username not affiliated with your real name? I.e. How do other people know the username is you? (Not something I personally much worry about—I think there’s a culture of trusting what’s on a CV, as one incorrect statement crushes your whole reputation.) Still, interesting in terms of how you design a site to give people anonymity when they want it, and still also give credit. Maybe a site admin could create a “real name” page per username, which lists the person’s real name and # of annotations authored (but isn’t linked to their username); CV can point to that, and even though there are ways of figuring out from annotation # which real name matches username, at least that’s some plausible deniability. Not sure whether that’s something users would want or not…

2) Maybe we can look at other places where people do unpaid academic work, such as academic journals. Dino Felluga, the general editor of the BRANCH site (http://www.branchcollective.org/; open 18thc literary criticism/history) has discussed “fixing” a set of annotations at some predetermined deadline date, having the annotations available then undergo peer review, and then having those reviewed annotations available as a fixed set (for example at a specific URL) in the future. That would give users a fixed URL to point to as well as peer review.

There could be “calls for annotations” just like journals have calls for articles that match an issue’s theme, perhaps with a focus on annotating for a certain mission (the post-colonialism annotation mission, the Finnegans Wake overlaps mission).

These ideas work better with shorter works like on the BRANCH site (in my case, I’d probably want to focus on one Ulysses chapter per mission).

could IU offer a page with all my annotations numbered chronologically so i could cite any or all?

@timfinnegan: yes, i’ll put that on my to-do list.

As I mentioned the other day, here is an annotator 2.x plugin for image annotation: http://emory-lits-labs.github.io/annotator-imgselect/
It seems to have a display issue with the main annotator ui viewer, which I didn’t discover until I pulled it out separately (it was developed with our marginalia plugin) and I haven’t had a chance to really look into yet. (And possibly some other quirks specific to our current use)

And our other annotator 2.x plugins: http://emory-lits-labs.github.io/annotator-meltdown/

http://emory-lits-labs.github.io/annotator-marginalia/

Each site includes a little demo page that will let you play with it and try it out.

@suttonkoeser: Cool! Looking forward to checking these out

Thanks!

re @literature_geek ‘s first point on Oct 26, I don’t think a real name is necessary, unless an organisation wants to check a reference by asking a project about a participant. A lot of voluntary work (my ‘go to’ analogy for online participatory work) doesn’t come with formal recognition.

fyi if you want to cite an earlier comment in slack you can copy the link from the timestamp and paste that in like this: https://digitalhumanities.slack.com/archives/annotation/p1445863889000014

twitter has lots of alleged joycefans, but precious little interest in closereading via online annotation projects. might it help to gamify the process of annotation so people start to see it as a shared responsibility?

(i’ve heard a dismissive term ‘party joyceans’ for people who’d rather drink than read)

we might distinguish joyce democrats (sharers) from joyce republicans (status seeking)

even rough estimates of how many/where annotations, like “infiniteulysses: 100, ulyssespages: 5000”

@suttonkoeser: thanks! I thought there must be something like that but it’s not obvious

if i wanted to write a script that visits every online ulysses project and counts all the annotations and who did them… is perl still state-of-the-art? would a site like http://genius.com automatically block it?

dug out my old python book (o’reilly rat 2nd ed)

@timfinnegan: if you’re doing screen scraping in python, beautifulsoup is pretty nice - http://www.crummy.com/software/BeautifulSoup/

@timfinnegan: are you planning to scrape http://genius.com then?

@timfinnegan: or other places too?

i’m thinking genius and amanda’s and mine and a couple others, if possible noting dates and authors and ratings so i can acknowledge their consistent contributors

Given the licensing for annotations on @literature_geek ‘s site, I bet she might just provide you the data. A lot simpler than trying to scrape it all.

okay, after lots of bleeding, i’m getting some screenscraping progress via bs4 with Requests (is Requests a good choice?) i’m on windows 10 so nothing quite works the way the docs promise

will screenscraping turn out to be too slow or blocked once i figure out bs4?

i want joyceans to see online annotation as a group responsibility

joyce built his prose out of ambiguities that can only be addressed via annotation

@timfinnegan: Requests is a good choice - I’ve got no idea what the current situation is with Python on Win 10 - certainly remember it being painful in the past on earlier versions of windows

Infinite Ulysses currently has an RSS feed of annotations, and I’m interested in offering other ways to reuse the content (for Twitter bots, other Ulysses sites, etc). The 1.0 site will probably offer JSON. @timfinnegan and others, what would your ideal annotation/annotator info export look like (for any site letting you annotate)? Do you have a preference as to file type (e.g. CSV)? I’m thinking both about reuse and not getting in the way of preservation (e.g. how can I let users’ annotations be best transferable to other/future platforms like Hypothesis?)

an export-format wd be awesome– page, line, highlightedtext, author, date, annotation, tags? gifford-or-not status, ratings?

a thriving online database for finnegans wake annotations: http://fweet.org/pages/fw_srch.php (this started by transcribing mchugh’s book, which collated many others by page and line)

What is IU using for annotation implementation, and how are annotations stored? If you were using annotator.js I would wonder about using the annotator store api to make them available, but perhaps not all of the information you want included is part of the annotation exactly (although the data model is flexible and can include any extra data fields you want)

just view-selection-sourced a random one: “<span data-annotation-id=”2891” class=”annotator-hl” href=”“>Universally </span>” http://www.infiniteulysses.com/ulysses/366

cf http://genius.com: “<a data-id=”4143452” href=”/4143452” data-editorial-state=”accepted” data-classification=”accepted”>princely presence</a>” http://genius.com/James-joyce-ulysses-chap-2-nestor-annotated

Reeve embeds the annotation as a title-popup in a (linkless) anchor: “Fleshpots of Egy<a href=”#name0” title=”refers to Exodus, in Bible; Jews complain to Moses they were better off in Egypt” onmouseover=”self.status=’‘;return true” onmouseout=”window.status=’’“><u>pt</u></a>” http://jonreeve.com/ulysses/ch05.html

Hunt: “<a style=”color: rgb(245, 150, 39);” id=”050005walkedsoberly” class=”box-images” href=”notes/050005walkedsoberly.htm”>Mr Bloom walked soberly</a>” http://hs.umt.edu/joyce/index.php?chapter=lotus

Wikibooks credits annotation-authors but doesn’t use embedded hyperlinks: https://en.wikibooks.org/wiki/Annotations_to_James_Joyce's_Ulysses/Telemachus/003 Only Genius and Amanda are crediting annotation-authors from inside the text.

counting/comparing number-of-notes-per-episode will be my relatively easy first challenge

@suttonkoeser: It’s weird because I’m using Annotator.js, but with annotation stored as nodes in the Drupal db (thanks to @mwidner!) so I can easily tie them into Views etc. But I bet I could expose them that way too, and just have my site continue using what’s stored as nodes—that would be a good longterm plan.

@timfinnegan: Attributing authors next to the annotations is really important to me—otherwise it isn’t clear to the casual reader who’s doing the work. Maybe the biggest problem in my survey of other Ulysses digital prototypes was unclear attribution (usually meaning wholesale using Gifford/Blamires without citing either). I’m thinking about putting major annotators on the masthead of the 1.0 site, but I’ll contact everyone first to make sure they’re okay with it and if they want their username attributed or something else (e.g. link to personal site or other project)

could the masthead be like a leaderboard with tallies? genius does this but they also reward people with editor powers which i found dangerous because quantity isn’t quality

Hm, maybe! I’ll probably also have a leaderboard with tallies for the top ten all-time annotators and/or annotators in the last month or week on the front page, so either that will be separate and a little lower, or I could play around with masthead designs like “Infinite Ulysses [logo] General editor Amanda Visconti, Key annotators: timefinnegan (470), amandavisconti (249), bbogle (159)”. Especially if the leaderboard will be by week/month (since some use it to encourage current reading/annotating) and not all-time, there should be recognition for all-time prolific annotators on the front. I’ll need to think about the cutoff of # annotations to get listed on the masthead, since there’s other helpful site activity not captured by # annotations authored (e.g. marking typos for me to fix, tagging).

in the longer run, building a database of paraphrases of gifford/thornton/slote/johnson that everyone could freely add or not (maybe resembling http://fweet.org)

(is 470 for real?)

@literature_geek: The Drupal version also uses the Annotator API, so it’s already available as JSON if you want to expose it that way. We’ve also created some custom Views for Lacuna Stories that export annotations with more human-friendly information, like title of text, excerpt, annotation, etc.

@timfinnegan: (470 is what I’m seeing—very exciting. I have to check, but that probably isn’t counting typos you’ve noted and I’ve fixed—I need a way to include that kind of activity but separate from regular annotations. Sorry, it’s on my task list to get this and other stats like what pages your annotations are the greatest % of the annotations on that page, etc. up on your user profile so you can see these every day yourself! My happiness to export data for you and display things like this is only, for better or worse, limited by trying to make time for Infinite Ulysses work… )

@mwidner: Nice! I think I did have it as JSON through Views JSON briefly but it was fiddly, so I’ll check that out. Your annotations = nodes magic makes so many cool things possible with Views!

That’s why I did it! :simple_smile:

Here’s the Annotator API output for your site: http://www.infiniteulysses.com/annotation/api

Actually, this is it:

http://www.infiniteulysses.com/annotation/api/annotations

if you’re not logged in, you won’t see much except the node ids

But, that’s enough for someone to grab the text of every annotation just by visiting the urls for each node.

@mwidner: Awesome! Thanks, I totally forgot that was a possibility

well, i haven’t figured out how to use ‘data-*’ attributes in bs4, so for genius i just used the python string count. by episode: 200, 76, 244, 52, 36, 103, 113, 81, 284, 6, 109, 38, 24, 70, 13, 32, 11, 106 = 1598 total

the 284 in *scylla are almost all from michael bible, who also did 2/3 of the 244 for proteus, and 3/4 of the 200 for telemachus

running the code for all 18 pages took less than a minute

Reeve: 304, 91, 272, 117, 186, 227, 334, 472, 737, 355, 254, 235, 140, 32, 335, 0, 0, 349 = 4440

Hunt: 200, 108, 207, 22, 23, 60, 39, 35, 46, 29, 19, 46, 23, 69, 128, 26, 50, 37 = 1167

IU: 103, 62, 34, 45, 41, 39, 14, 6, 16, 141, 15, 38, 20, 31, 11, 6, 54, 10 = 686? (my loop-indexes may be off)

@timfinnegan: requests is really good for python ; i’d be curious to know what isn’t working on windows

mostly ‘tar.gz’ and ‘PATH’ (and resistance to RTFM)

oh, getting it installed?

where it = requests ?

i did beautiful soup first so requests went smoother

@timfinnegan: i think python on windows comes with pip right?

pip install requests ?

that should download and install everything

i never got pip or easyinstall to work at all

@timfinnegan: was pip available to you at the command line when you installed on windows?

@edsu @timfinnegan : It doesn’t seem like pip is available if it was python 2 that was installed

looking at : http://www.anthonydebarros.com/2015/08/16/setting-up-python-in-windows-10/

i thought python >= Python 2.7.9 came with pip

I thought so too, but I’m not familiar with the windows installer so…

oh hmm, that kind of terrible - i wonder if it’s some variation of http://stackoverflow.com/questions/28621436/pip-not-working-on-windows-python-2-7-9

python2.exe -m pip install requests

maybe?

No, you’re right about 2.7.9 onwards

https://docs.python.org/2/whatsnew/2.7.html?highlight=pip#bootstrapping-pip-by-default

it was python 3.5 …and 2to3.py for bs4 was also torturous

i never use commandline in windows so as a beginner i didn’t know when to use python’s vs windows’ commandline

@timfinnegan: because you installed it manually rather than using pip?

bs4 will automatically convert it’s codebase to python3 when you install it with pip

there’s no need to run 2to3 for bs4

http://www.crummy.com/software/BeautifulSoup/bs4/doc/#problems-after-installation

oh, i take that back

that does look convoluted

(I’m surprised to read it as well - always just pip installed it and never thought about it)

maybe i’ll just go back to my windows free universe and keep my mouth shut :smile:

I’ve never tried to use Python on windows - I only recall trying to help someone with Postgres setup on a windows machine and ducking out on it once it started to get to these ported psycopg installers

http://www.stickpeople.com/projects/python/win-psycopg/

dragons all the way down

win10 hides python at: C:\Users\tim\AppData\Local\Programs\Python\Python35

right, but that should just be the exe or alias that your account can run

but modules need to be installed there (no?) and trying to find where to run 2to3 from took a million tries

no - you need pip installed and working and shouldn’t have to run 2to3 - you’re stuck doing it the hard way at the moment

So, one route that might be more reliable is to get VirtualBox, set up something like an Ubuntu virtual machine to play around with, then use all the package management tools to install python and all the libraries. It’d be command-line, but you’d probably find a lot more tutorials for that online than Windows and things would just work.

it’s working now, mostly, for my purposes

this was how i scraped infinite: import requests start = [3, 24, 37, 53, 68, 84, 112, 144, 176, 210, 245, 280, 331, 366, 408, 569, 619, 690] end = [24, 37, 51, 68, 84, 112, 144, 176, 210, 245, 280, 331, 366, 408, 567, 619, 690, 733] for episode in range(0, 18): temp = 0 for page in range( start[episode], end[episode]): url = “http://www.infiniteulysses.com/ulysses/” + str(page) r = requests.get(url) temp = temp + r.text.count(‘field-text’) print(temp)

next is my own site, which i’ll try to use bs4 on because nothing is wellbehaved there (messy blogger urls, messy blogger markup)

i should be able to trim away everything that’s not annotations, and then count linebreaks, ignoring multiples, for a first approximation

@timfinnegan: Not sure if this helps, but the feed of Infinite Ulysses annotations (updates automatically as they’re authored) is at http://www.infiniteulysses.com/annotations.xml (clunky table view that will be improved on 1.0 site: http://www.infiniteulysses.com/all-annotations-feed). I put the latest (11/5/2015) export of raw annotation data up at https://raw.githubusercontent.com/amandavisconti/infinite-ulysses-public/master/Annotations/node_export.txt.

@timfinnegan: I also made an RSS feed for just your annotations (http://www.infiniteulysses.com/timfinnegan-annotations.xml) and a page you can visit to see a table of all your annotations (sortable by Ulysses page number, date authored, quote, text): http://www.infiniteulysses.com/timfinnegan-annotations

@timfinnegan: I’ll try to add some more stats/cleaner export options to that soon

looks stunning! (for some reason the default by-page sort puts a couple of 34s before the 28s?) and wouldn’t the most useful default view be newest-first?

i just ran my ‘nestor’ pages and got: notes 214 links 126 images 22 maps 0

…so. How would one annotate a video game? Video of a playthrough w/ youtube’s tools?

(any good examples of youtube’s tools?) here’s a vr ulysses project that’s supposed to get annotations eventually: https://inulysses.wordpress.com/

columnar comparison of preliminary ulysses notecounts: https://files.slack.com/files-pri/T0CELH6PR-F0E0P3FKR/preliminary_note_counts.txt

one surprise is that 86% of my notes (last 4 columns) have a link or an image (3605+947 = 4552 = 5312-760)

i guess this is because images and links are the most fun places to start, so that ratio will drop as more speculative interpretations get added

i forgot to count links to youtubes of mentioned songs, which should be about 300

is there any consensus on what user-comment system works best? (since in the end every annotation is a potential flamewar)

@shawngraham: Interesting question! Wondering whether there’s a way to caption (as annotation) screenshots taken with hot key while in a Steam game (while in-game=don’t think so; while on Steam profile=maybe?) I can test when Fallout 4 drops next week : D

@literature_geek: the open source release of Lacuna Stories 2.0 will probably drop next week, too.

the common form of video annotation is a commentary track; an alternate would freeze playback and maybe shrink it to a corner while the screen is taken over by a commentator

(i’m imagining a hyperlinked set of wandering-rocks animations on youtube where you can click to jump to any parallel timeline…?)

((start with just the audiotracks?))

@mwidner: I am excited for the release! Congratulations!

@timfinnegan: by user-comment system, do you mean platforms like disqus, or like best practices for moderation?

@timfinnegan: That comparison file is really cool. I’ve currently got Infinite Ulysses’ annotation at 1,043 total, though, rather than 686—could you tell me how you got that count so I could investigate? I’ve got a table of annotations per page here (http://www.infiniteulysses.com/most-annotated-book-pages), but I’ll work on getting the counts per chapter (not sure of an easy way to tell Drupal to group by ranges of pages in specific chapters but I’ll figure it out…). (Changed your annotation view page to default-sort by newest first, by the way: http://www.infiniteulysses.com/timfinnegan-annotations. Will add a total annotations count next)

@timfinnegan: Though I just added up the annotation counts by page on http://www.infiniteulysses.com/most-annotated-book-pages and got 1137… that might be including the small number of annotations marked as private? (e.g. annotations of typos to fix)

i counted “field-text” https://digitalhumanities.slack.com/archives/annotation/p1446657224000050

my count for sirens is 15, yours is 25– could it be that my ‘field-text’ search is skipping the typos?

imagine a ulysses e-reader that pulls in all the notes from all these sites and lets you like/dislike them individually, learning which sources you prefer but also compiling your own custom annotated edition

with ulysses and even more with finnegans wake, publishing annotations in a paper book has become impractical– there’s too many possible insights, and links and images and maps and songs. so progress is effectively stalled until some such community e-reader gets off the ground

http://fweet.org is one individual’s annotated edition, so far without links or images or genetic layering (manuscript variants)

(is it possible to agree on a set of formats for tagging notes, so they can be shared, no matter if the edition embeds them in the text, or embeds a link to them on the same or another page, or presents them interlinearly?)

((and imagine if your bots could detect updates on every site, like a smart rss for annotations?))

(((has there ever been a commenting system that lets each commentor host their own comments, but tracks others’ replies?)))

I remember seeing something where you self-hosted content and that was linked out to other social platforms (e.g. Twitter) if you wanted. It might have been Diaspora? https://diasporafoundation.org (I remember the “decentralized web” was a thing it discussed.)

@timfinnegan: Section 2.5 (“Ulysses under hypertextualization and hyperannotation”) compares various digital Ulysses prototypes for pros/cons and what they seem to value (you mentioned Genius and Infinite Ulysses as being the only sites with inline attribution to annotators, and that’s one of the things I thought about there)

http://www.theguardian.com/books/2015/nov/13/the-poems-of-ts-eliot-annotated-text-christopher-ricks-jim-mccue-review