Wikisource talk:Annotations/types

Annotations are main namespace only

We should only be defining annotations as for our guidelines as something that applies to our main namespace, that is to say works. All our other namespaces are constructs, and as such they are simply wikilink'd. Accordingly, from our main namespace, annotations (links) will (normally) only be going to (at WS) the namespaces: main, author, portal; (at other WMF wikis) to other main namespace. I do NOT see that we would need to be linking to these namespace: Wikisource, Help, ...

So what falls outside the idea of annotations?

• Links from the {{header}} which takes in other wiki properties, and a link to the talk page through use of {{edition}}.
• Interwiki links to works of the same name

Now some tricky bits that may fit within this that I see that we can move in or out. Links to works in other languages that occur in the text, would we encourage a link to an existing work at another language wiki. Noting that linking to works that do not exist is counter-productive. Similarly, would we link to an author page in another language wiki? I would prefer that we linked to an Author: page locally, and that there is the interwiki link. — billinghurst sDrewth 10:15, 21 June 2011 (UTC)

Hi, the point of this page is simply to define all possible types of annotations, not to debate what types we should or shouldn't be allowed. The idea is to make a list from which we can then design a vote page - after the vote page is agreed on, and after the vote cast, then we start working on a guideline on how the agreed on annotations should be implemented. I'm going to remove the strikeout because in effect its a judgement or opinion. We're not making judgements or opinions yet. I'm sure you'll be right in the end, but leave it to community consensus. Green Cardamom (talk) 16:40, 24 June 2011 (UTC)

I completely disagree, it is not a feasible annotation, hence it is not possible, hence it is out. We simply do not and will not link to Scriptorium from a work, this confuses the page and puts into play something that will not occur. Tell me on one occasion we would link from a work or a header to Scriptorium which is a discussion place?

A vote page, that you so define is simply not how we have been undertaking work at Wikisource, and often ends up making decision making more complex in the discussion of the futile for no express reason. We have been more on having less in the way of policies and defined rules, and setting principles and outcomes. Listing them, grouping them, getting opinions on them is all fine. But coming in and thinking that this is voting territory is way ahead of the game. — billinghurst sDrewth 17:48, 24 June 2011 (UTC)

Additional: things need a scope. The scope for ANNOTATION is something that takes place in the main namespace as we are discussing annotating a work. Getting it broader than that seems pointless.

The problem I'm finding with the ever-increasing implementation of proposed policies or guidelines (or even those demoted to "no community consensus") is the fact that they too often become regarded as accepted policy (which is not to be confused with official policy) even as they remain proposed. What happens when a widely implemented proposed policy (or no policy) is eventually "voted down"? Seems we should stick to internal linking for now at least until official policy is written on the subject. It does seem prudent to create limited works for the purpose of illustrating a proposed policy—so others can at least see whether there's any usefulness in the proposal... But naiive copy/paste Users like me are all too often apt to run with something that appears "accepted" (albeit not necessarily "official") because it is/has been widely used by more experienced Users (if you look at my history and created works here, you are sure to find infringements on official policy—or even liberties taken where no policy exists!)... But thinking this practice through more recently, while well-intended and often innovative, such widely implemented "accepted" practices have merely resulted in confusion, and may perhaps in the future cause a hand to be forced that maybe shouldn't be—all because a thing has already been widely implemented, and the resulting clean-up would be too colossal a task! Seems kind of viral to me. In my mind, priorities should be faithful rendering and author's intent. All other innovations should support those two—not add to them. But that is my opinion. Londonjackbooks (talk) 05:27, 26 July 2011 (UTC)

• The idea of a vote page or a !vote page seems very, very bad. This needs to be discussed, not voted on. All voting will do is show just how divided we are and where there are gray areas will result in long discussions within the vote that someone would have to parse to try to determine consensus. We need to discuss things and come to consensus. I'm not at all sure that this is the best place to do it. The Scriptorium barely links to this sub-page of a blanked draft policy. I'm not sure that we have much chance of getting full participation from a representative group of editors here. We'll only get the very most vocal ones on each side, which will result in a clear lack of consensus being the only likely outcome (and maybe that's good, I don't know). This is a very important policy issue and I would object to any sort of vote. If we discuss here, we should transclude the discussion to the parent talk page and maybe also to the Scriptorium. An actual result in this discussion will affect the definition of our project.
• as for LJB's comments, I generally agree, except I'm not sure I see the point, since the proposed policy generally tolerated link, at least at one point, and a clearly sizable chunk of the community is favorable towards some sort of links, and most importantly, since there is not and never has been a clear prohibition on links, those who wish to add them can. However, as was pointed out on Scriptorium, who would want to take the time if a policy might be enacted to eliminate them all!--Doug.^{(talk • contribs)} 10:51, 26 July 2011 (UTC)

@LJB: I don't see why annotations are inherently seen as an addition separate from the goal of providing the authors intent. My work with wikilinks has always been primarily focused on restoring the context lost through changing times and settings so that readers can comprehend the author's intent. "We've drunk to the [[w:Queen Victoria|Queen]], God bless her!" is not annotation that would not have been useful to contemporary readers, nor is the author's intended reference likely to be inferred by the many of our contemporaries in a cold reading. Another work I remember annotating spoke of Americans moving to settle in "the West". Written shortly after the Louisiana purchase the writer was speaking of lands that is probably now in either the state of Missouri or Illinois. After further Western territories were acquired these lands came to called the "Mid-West". Very, very few of our contemporaries would comprehend the author's intent in that without some initial confusion, and some would certainly give up before they could manage to figure it out on their own.--BirgitteSB 01:35, 5 August 2011 (UTC)

I understand what you're saying... I recently had to explain to a friend that when George H. Earle, Jr. wrote to Justice Harlan: "You have appeared in this respect the incarnation of the spirit and policy that have made our race great and happy..." that Earle was not referring to the "white race", but the human race—just as his sister, the poet, often used the word in her poetry... My point above (I had to reread it) had more to do with "experimenting" with proposed policy ad nauseam (annotations being but one example) before an official policy has been adopted—and that the practice should be limited and "queried" (to use a word removed from the Style guide on 24 July 2011) until then. Hypothetically, what if it is eventually decided that only inter-linking is allowed?—What a clean-up mess that would create! Would the fact that it would be a monumental task to remove all sister links from existing Pages dissuade the community from adopting such an "exclusion" policy?—even if it were the "better" policy? [I am not saying it is, just using it as an example]...

RE: "...some would certainly give up before they could manage to figure it out on their own.": I can only speak from my perspective, and the following words by Robert Hutchins says it best for me... I posted the following on Cygnis' Talk page back in June:

...the above topics prompted me to go back and re-read portions of Hutchins' The Great Conversation. (1952) The following passages seem to apply themselves—whether directly or indirectly—to the current discussion on annotations as well as to the debated purpose/intent of this sister itself:

[Conversation, p. xx] One of the policies upon which the Advisory Board insisted most strongly was that the great writers should be allowed to speak for themselves * * * Since the set was conceived of as a great conversation, it is obvious that the books could not have been chosen with any dogma or even with any point of view in mind. In a conversation that has gone on for twenty-five centuries, all dogmas and points of view appear. Here are the great errors as well as the great truths. The reader has to determine which are the errors and which the truths. The task of interpretation and conclusion is his. This is the machinery and life of the Western tradition in the hands of free men.

[p. 48] Sir Richard [Livingstone] goes on to refer to the remarks of T. S. Eliot: "In my own experience of the appreciation of poetry I have always found that the less I knew about the poet and his work, before I began to read it, the better. An elaborate preparation of historical and biographical knowledge has always been to me a barrier. It is better to be spurred to acquire scholarship because you enjoy the poetry, than to suppose that you enjoy the poetry because you have acquired the scholarship."

"The task of interpretation and conclusion is [the responsibility of the reader]..." It is they who have to "determine which are the errors and which the truths." It might seem like "pointing" the reader to the facts is a helpful act... but is it really? I suppose that is a debatable question that might be worth asking. You might actually be in essence robbing the reader of further research they might have undertaken on their own otherwise.

What I think WS might be lacking is a clear and stated goal or purpose. As far as I could tell, WS fashions itself as an "online library of free content publications," (Main page) and a "project to create a growing free content library of source texts." ("What is Wikisource?") The latter quote goes further, however—where it adds: "as well as translations of source texts in any language"—with the word "translations" pointing to proposed WS policy. Interesting that a stated purpose or goal would point to proposed policy—but that is only part of why I think WS needs to be more clear about it's common goals and policy...and perhaps have a tighter rein on "novelty" until more official policies are cemented. I am one to talk, however—as I have a rather undisciplined track record here!...

I have been giving all this some thought recently, however, and started jotting down some thoughts on annotations, goals, etc. toward July's end in a personal user sandbox. My two cents, however, may not even be worth the space it's written on, but feel free to take a look! Sorry my reply is so long! Londonjackbooks (talk) 03:34, 5 August 2011 (UTC)

Waking up with the following in my head: "Great symphonies require more than one hearing; great poems more than one reading." Words I had transcribed by Florence Earle Coates in her piece about Matthew Arnold... Reminded me what a history lesson I have had just by reading and re-reading (and transcribing) Mrs. Coates' poems—many of which are filled with historical references... What is the "Devil's Isle"? Who is "The Smile of Reims?" And all those Greek references—another area where I am ignorant... But the fun is also in the searching... If something "speaks" to a reader, they will want to pursue the thought further. If they are without understanding of who "the Queen" is, they'll learn more by following a search-engine search of Kipling's line of poetry than they would linking to the Queen's sister-site page... Any reader who would "give up before they...manage to figure it out on their own" is not inherently interested in the subject matter. Are we their teachers that we must then instruct them? Just a random thought, sorry... :) Londonjackbooks (talk) 13:06, 5 August 2011 (UTC)

To be clear, I only imagined readers "giving up" on the much, much more obscure second example where terminology was usurped in the generation following the one in which the essay was written (the writer speaking for himself can hardly imagine the label he uses will refer to an entirely different location after his death than it did while he lived while that location he spoke will be given a alternative label nor will he imganine that four generations after this the fact that this location was ever labeled otherwise will be almost entirely unknown). I certainly don't think we should leave research up to the readers because "the work be good for them" nor delight in keeping people who aren't deeply interested in the subject from comprehending the text. If something wasn't obscure in a contemporary reading, I strongly believe in pointing the modern reader where they might look. I don't like linking were it seems the author was picking out things obscure to contemporary readers to give an impression of the exotic. A good example of this practice by an author themselves, is Kipling. His early poems published in an Indian paper were in plain text and fully understood by the British readers who lived there. When the poems were republished in London for a different audience he added footnotes for the kitchen Hindustani etc. I don't think what we are doing is so unheard of or all that new. But then I started doing it in 2006 exactly because it seemed such an obviously desirable thing to me and was quite surprised when people first questioned it. Regarding experimentation/proposed policies, there is no way to force consensus. I believe heavy experimentation is the best way work out an area of no concensus. Not that I think that is happening here. But still we can't just make a policy firm without any consensus to back it up. I suppose you could blank proposed policies unless people are actively working on the issue. I know some people are driven nuts by "no consensus", but I find the alternative to be much worse. Artificial deadlines just seem to destroy wikis. Although as we have been mainly talking about wikilinks. I must say that that particular method has had consensus for some years and I haven't seen anything to indicate that consensus has been overturned there. The rest is more ambiguous. --BirgitteSB 02:49, 7 August 2011 (UTC)

You have certainly been doing this much longer than me, and I don't know much about wikis, or how "artificial deadlines" (not sure what you mean—deadlines related to voting or reaching a consensus?) affect them... I don't have an administrative bone in my body! :) I do understand your point about the writer perhaps not conceiving at the time of his writing that the "labels" he uses will be defunct at a future time... And pointing to those facts is unarguably a useful thing! Probably the reason why there is more of a reaction surrounding annotations and wikilinking, however, is due to the fact that we are dealing with original source texts here at WS... Public domain, sure—... and WS is certainly within its legal rights to do with the material as it (consensus or no) sees fit; but to some who come here looking for "unadulterated" original source texts (which, one could argue, wikilinking—but perhaps not all forms of annotation—doesn't affect)...

I was just taking the thought one step further... Musing about the annotations debate, and perhaps attributing the lack of consensus about annotations to the question of "What is the purpose of WS?" For I feel that the annotations question is so closely tied together with what this sister's purpose might be (Is it a library? In what sense?, etc.) and where such a purpose might be lacking or vague, that would have an even greater effect on the ability for this community to reach a desired consensus... (I'm starting to talk in circles now, I think, and that is never a good thing!)...

A quick aside on the blanking of proposed policy pages as a result of there being "no consensus" on the matter: seems counter-productive to do so, and that much useful information would be/is contained there, and that removing the information would give the appearance (on the surface at least) that it is not being "actively worked on" (at least by the casual observer). Maybe just don't refer to annotations as "proposed policy,"—yet!... I don't know... But it seems the information should still be available in some sort of structured manner somewhere (as opposed to solely on these Talk pages)... Londonjackbooks (talk) 16:41, 7 August 2011 (UTC)

Noting AdamBMorgan's comments in his View below: "...wikilinking to further information where necessary. Gutenberg, Archive.org, HathiTrust etc just can't do this so we stand out, it's our niche..." I agree with that. And, while I have not yet scrutinized over the "list" of Grade 1 annotations, I am inclined to agree with at least their inclusion based on AdamBMorgan's characterization of them. I am reminded of a comment by Cygnis Insignis on my Talk page, who stated, "The introduction of scans with a near-perfect OCR text layer changed everything, a quantum leap: millions of books could be added and verified! And when these works refer to another—what has been called the "great conversation"—the power and potential of this sister is mind-blowing." It is what makes this sister unique! [Note: Do not assume that my inclusion of Cygnis Insignis' words infer an endorsement on their part of my or anyone else's views of what an appropriate annotation would be. I do not wish or intend to speak for Cygnis, I merely second their sentiment with regard to the "mind-blowing" potential of this sister! :) ]Londonjackbooks (talk) 13:45, 12 August 2011 (UTC)

Added thought on WS vs. WM ("Grade 1") linking: Best advice I've received is to "keep it in the family" (i.e., WS) as much as possible; if a WS page is not available, create one (if the subject matter is appropriate to this site)! Otherwise, link to the best representative WM page available. Londonjackbooks (talk) 14:01, 12 August 2011 (UTC)

Seems Doug was thinking about WS v. WM the same minute I was! To Doug: All Author pages here on WS have/should have links to WP, etc., sites in the notes section that readers can refer to... Granted, some WS Author pages are sparse and uninformative. Key words are if the subject matter is appropriate to WS. If not, such as where you noted Pythagoras and the like (while I am not an authority on their authorship), then "WM it"... As far as AdamBMorgan "describing" the codes, a description is already available on the Project page that is the "parent" of this Talk page. I have printed out a copy for reference, because I don't like switching back and forth either! :) Londonjackbooks (talk) 14:13, 12 August 2011 (UTC)

Another good reason to "WS it" is that an author's work speaks for the author—whereas WP and the like do not. Even though the desire is for WP articles to remain unbiased, there will always at least be traces of bias due to the fact that it is a human endeavor! ;) At WS, we are letting the author represent himself, and our wikilinking should support that wherever/whenever possible... especially when pointing to an outside source within the original source text of the author. Londonjackbooks (talk) 14:25, 12 August 2011 (UTC)

Discussion

View by Doug

• As mentioned above, I see no value to a !vote page, at least at this point. We need to discuss. With that in mind, let me comment on the use of the broad categories:
  • WS - I think we need to be cautious about some of the special namespaces as they can introduce opinions and POV but generally, internal linking like this should be encouraged. One conflict is where a work names individuals, some of whom are authors, if we allow interwiki linking then some names would link to WS:Authorspace and others would link to WP:Mainspace, that seems odd, but at this point it's just an observation. The same is true for links to portals, the example shows a link to Portal:Buddhism rather than to w:Buddhism, I'm not entirely convince that the former is better than the latter. Links to publishers should be limited to the publishing data of the work - and to the header; though I suppose the name of a publisher used in a work could be linked either to publisher space or to en.wp.
  • WM - (leaving aside the non-standard abbreviations) Links to WB and WV within a text are problematic as those projects can contain original research and editorial. Any work that we might find there and want to link to we should also host here in a different form. Someday we may find works that are suitable for hosting here that actually reference WB or WV works, when that happens we'll need a special rule but it's a long way off at this point. Normal interwiki/sister Links to WB from a dab, an author page, or even a header should be encouraged in whatever standard format eventually arises for that, but those aren't annotations. Links to WQ are not particularly useful except for backlinks since the only reasonable place to link in the mainspace would be the actual quote. I can't see the point to a link to News and I'm not sure whether links to wikispecies would be as useful as links to enwp. Pagespace is one big set of links to Commons, that an images should be the only such links generally. Links to en.wp and wiktionary, I think should be encouraged.
  • EL - No business in a text unless the text itself references the link (particularly possible with recent court cases and other recent government documents), in which case the link makes perfect sense. ELs to sources for the works should only be linked from Talk, Author, or meta (index) space or possibly, on a dab to show that the work has yet to be imported into our project but is available.
  • FN - Footnotes are commonly in the source texts and are frequently also PD (must be if the work is on Commons), these should be transcribed (though it would be cool if we could use the same scans to support both footnoted and non-footnoted texts). Footnotes added by our editors should be allowed in those cases where it's really necessary; but clearly marked as such.
  • In summary, I support wikilinks to other works we have and interwikilinks to en.wp and wiktionary where useful. I do not support any user added external links. User added footnotes should be used very, very sparingly. I see a problem though. Many texts that I'm working with are multilingual. If the sister projects insist on "clean text" then we will have a conflict.
  • I don't understand this insistence on a pure text. That could be done without a wiki (and could be done better) and really tosses out a core competency (and raison d'être) of MW projects. And if we wanted to be a pure reference source for the other projects, Commons actually could do a better job at present. We are here to make these works accessible. To a large extent, that means making the dictionary and the encyclopedia entries available at a click.--Doug.^{(talk • contribs)} 11:13, 26 July 2011 (UTC)
    • Something I mentioned on another page that I want to consolidate here: 1) I find no logical basis for preferring en.ws wikilinks to other wikilinks. If I'm reading a work by author x and author y is mentioned I would like a link to the en.wp article on author y, not a list of author y's works. Yet, I've asked about how we can link Plato but not then Pythagoras or Socrates, neither of whom wrote a single word, I was told they should have ws:Author pages - author pages for people who didn't author anything?! I'm confused by this concept. I've also heard it suggested that wiktionary is OK but wikipedia is not. Why? What about the name of a Greek mythological character, they ought to be linked to en.wp oughtn't they?
    • User:AdamBMorgan may have some very good points, I can't tell. I can't switch back and forth to look to see what each of the codes stands for. Please discuss these things by describing them instead of creating codes for each of them.--Doug.^{(talk • contribs)} 14:01, 12 August 2011 (UTC)
      • Re last comment by Londonjackboots, yes I knew the codes were there but I didn't find them very intuitive, re-looking, I found that I was a bit unfair to AdamBMorgan as they are more intuitive than I thought. However, long lists of them are not helpful.
      • Also re comments by LJB, take this scenario: a work cites references both Plato and Socrates (not at all unlikely), I find the inconsistent links to be strange. Part of the problem with author space is that too many pages link to en.wp. I have a report (several months out of date, time to run it again) of pages that have links to non-existent en.wp pages. Most are not pages which were deleted, rather they were added en masse to author space.
      • I'm glad to see there is a lot of agreement between LJB, ABM, BSB, and myself. Although there are certainly differences, there are broad areas of overlap and that's a good sign when we are hoping to find some consensus.--Doug.^{(talk • contribs)} 14:56, 12 August 2011 (UTC)

View by AdamBMorgan

• I do think annotations are within the scope of Wikisource. More than that, they are beneficial for the project. They make full use of the capabilities and possibilites of being a Wikimedia project; and they are one of the potential strengths of Wikisource over other internet libraries (along with the Proofreading extension etc).
• My view of the list of annotations is to grade them by how much they alter the text, and handle each grade differently.
  1. Grade 1: The most basic of annotation, as done by many users without really considering it an annotation. Mostly, this covers wikilinks. Although we could try to ban them, I dont think this is a good idea because: 1) It undermines Wikisource's strength (and unique selling point) as a wikimedia project capable of wikilinking to further information where necessary. Gutenberg, Archive.org, HathiTrust etc just can't do this so we stand out, it's our niche. 2) It is futile; any new user coming to Wikisource "knows" the wiki... means wikilinks. There are a lot works with wikilinks and I'm sure there will be a lot more, trying to chase after them all, removing the links would take up a lot of time for little purpose. Grade 1 annotations should either be made visible or not with javascript or just allowed to remain in the plain version (the latter because the javascript option still involves chasing after people amending wikilinks to use templates; besides the only difference between a clean and wikilinked version is that some words will be in blue).
     • Grade 1 annotations: All wikisource links (WS-A, WS-W, WS-P, WS-PO); All wikimedia links (WM-W, WM-B, WM-D, WM-N, WM-S, WM-C, WM-V, WM-Q); Pop-up messages (O-P); inline translations (O-T2)
     • Note: While Wikipedia and Wiktionary will remain the most common, I don't see any benefit in denying links to the other projects if it is every appropriate. For example, personally, I tend to wikilink latin names for species to Wikispecies rather than Wikipedia.
  2. Grade 2: A level of annotation that actually adds to the text. Where grade 1 annotations would not be picked up by a copy-and-paste or the book tool, these annotations would be carried over to derivatives of Wikisource. Grade 2 annotations should have the option of being removed from the plain version, either by javascript or by a separate instance of the text held on a explicitly marked page.
     • Grade 2 annotations: All footnotes (FN-R1, FN-R2, FN-R3, FN-R4, FN-R5, FN-R6); External links to toolserver (EL-T, EL-TG)
  3. Grade 3: As with grade 2, these add to the text but at a greater degree. Grade 3 annotations should definitely be removed from the plai version in the same way or not be used at all. Even where they are used, it should be sparingly. These come closer to my impression of a Wikibooks-appropriate annotation.
     • Grade 3 annotations: Additional images (O-I1); Text introduction (O-IN)*; External links/further reading (O-X)* (asterix indicates it is in "grade other" too)
  4. Grade n/a: "Annotation" should only refer to the body of the text, anything in the header is therefore not an annotation. Anything designed for navigation and visibily separate from teh body should also not be counted as an annotation. This includes the header itself as well as {{AuxTOC}}.
     • Grade n/a annotations: Aux TOC (O-A); Listen, Media etc (O-L); Wikipediaref (O-W)
  5. Grade other: Special cases that I don't really see as annotations.
     1. External links (EL-A & EL-AF): Can go in the header or the talk page if really necessary but I see no reason for them to be in the body unless it's a very unusual case. Putting them on Author or Portal pages makes more sense.
     2. Additional images (O-I2): Adding a cover image, sample illustration or other image from Commons to a plain text (not DjVu/proofread) work appears to be (or have been) quite common. To grandfather that into any new policy, I would allow an exception for one image (or a small number of images) for this purpose only. This would not allow works to be fully re-illustrated by a user.
     3. Translations (O-T1, O-T3): Translations are related to annotations as derivative works but should be considered separately.
     4. Excessive Grade 3: Text introduction (O-IN) and external links/further reading (O-X) should not be used if too excessive. This is covered in my next point but I see annotations as only notes to assist the reader, not the introduction of new sections (although some of this could go on the talk page).
• I see the difference between Wikisource and Wikibooks as mostly a matter of scale (and, possibly, of intent). Wikibooks' annotations often go beyond simple notes and add new sections and new pages. They can discuss characters, themes etc and add significant extra content to the work. They are primarily intended as textbooks - educational resources useful or required to study a work academically and/or prepare for tests and exams. I see annotations on Wikisource, however, as intended to help a reader simply to read the work. Annotations here provide pointers to further information or some context useful for comprehension. While a reader could be forced to go a do research on their own, that denies the possibilities inherent in a Wikimedia project and many people may not have the time/ability/inclination to do so. For example, I have personally wished that some books I've read might have some sort of translation for "common" French phrases inserted into the text, it is possible but inconvenient and annoying to have to look up something so potentially trivial but potentially important just because the author assumed knowledge in the reader. Wikisource annotations should, therefore, take the form of simple notes to support the work.
  • Note: The most extreme annotation (grade 3) I have made, is an image added here. The text gives names of islands (some of which are no longer used) and some co-ordinates but the image provides more immediate context of the location. Also note that I have included footnotes for the co-ordinates with links to the toolserver based on Wikipedia's co-ordinate templates—this is a moderately extreme annotation (grade 2) that does the same thing, provide context to the reader about the location being described in the text. (I've added this type of annotation to the list as I'm writing.) If agreed during this process, I intend to extend this throughout the work in question (that chapter is the only one that heavily annotated at present).
• I think separate, clearly marked, user-annotated editions have their place on Wikisource just as do user-translations. Having two copies of the same work can be a problem technically, especially if the original is proofread from a DjVu (for example, as the proofread version is edited, it will shift away from a copy-and-paste version even if the copy was made from a fully validated source). Conversely, adding larger or more frequent annotations to the DjVu (and sorting them out by javascript or in some other fashion) risks getting confusing. Having more than one annotated edition could make this worse (I'm not sure why this might occur but it is a possibility worth addressing).
• I doubt there will be many annotated texts (grade 2+) on Wikisource as a proportion of the whole library (as it should be) but, like translations, they should not be banned for this reason.
• I agree with Doug regarding pure texts. While I would want to differentiate between original matter and anything user-added, an absolute stance on any added value (even basic wikilinks) undermines the project. - AdamBMorgan (talk) 12:27, 12 August 2011 (UTC)

A few notes from experience at he.wikisource by Dovi

Moved to the main talk page here.