Proof versions of articles

[Chris Day]

Question to the biologists who have tracked many draft pages: in your estimate, what proportion of edits made on a draft page actually make the article worse?  Surely it can't be a high number: why else would people be making the edits, if not to improve the article?

If a high enough proportion of edits actually do improve an approved article, and if the edits are rarely if ever so bad that you would want to revoke approval, then perhaps we should be allowing people to edit the approved versions of articles directly.

Worse is a relative term here. I mean with resepct to major rearrangements.  If sections get moved and reworded there can be major discontinuity in the flow of the article.  For example, terms or concepts defined to late. this has not happened yet, but it will in the future.

Maybe i exagerate this potential problem, it is too early to say how these approved articles will evolve.  Possibly authors will shy away from major changes to 'approved' articles.  At present there are not enough authors chiming in to see how much of a problem this will be.

With regard to the new approved test template being tested here:
http://en.citizendium.org/wiki/User_talk:Larry_Sanger/Approved_in_pagehist

I think the approved version needs to have a distinct template compared to the current page, reader know they are on an approved version.  This would entail adding the 'approved:historical' template and immediately replacing it with the the 'approved:current' template. Then linking back to the version with the most recent 'approved:historical' template.  Does this make sense?  if not i can set up an example for you to check out.

[Larry Sanger]

I think the approved version needs to have a distinct template compared to the current page, reader know they are on an approved version.  This would entail adding the 'approved:historical' template and immediately replacing it with the the 'approved:current' template. Then linking back to the version with the most recent 'approved:historical' template.  Does this make sense?  if not i can set up an example for you to check out.

Makes total sense.  Maybe you could edit the example I already have up there?

[Nancy Sculerati]

The best way for experienced readers to know whether or not they are reading approved or unapproved versions is not with a template but with a style sheet. If approved articles had a light yellow background, or VERY light green background  instead of white and a dark navy blue or very dark purple type color instead of black, the reader would know immediately, even at parts of the page distant from the template, that this page was in an approved article, and not the draft. I'm not saying it has to be this way, but a different font would also do it - and the change of all- page color, font color, and font, would nail it down. The colors have to be previewed in different browsers but it can be done- easily.The code can be written into the articles manually to start, just to save the tech crew from overload. Besides, there will be a period where the "best look" will require establishment.

PS- my style suggestions are not instead of the template, but in addition to the template. With such a distinctive style, the template can then be less intrusive.

[Chris Day]

The best way for experienced readers to know whether or not they are reading approved or unapproved versions is not with a template but with a style sheet

Sounds good.  We can play with the details.

[tkjazzer]

I think the graphics can be modified.  Font is too big.  I think that if this is supposed to be different than the approval template on the approved article then it should be different color. 

[David Tribe]

Question to the biologists who have tracked many draft pages: in your estimate, what proportion of edits made on a draft page actually make the article worse?  Surely it can't be a high number: why else would people be making the edits, if not to improve the article?

If a high enough proportion of edits actually do improve an approved article, and if the edits are rarely if ever so bad that you would want to revoke approval, then perhaps we should be allowing people to edit the approved versions of articles directly.

A number of useful edits does not compensate for one awful edit that ruins a whole theme, or the blister of a bad idea that grows like a cancer. Article quality it not a simple metric where small bricks of improvement add to the pile of goodness of the article. If that concept were true, Wikipedia would be full of brilliant articles   I'd say there not much wrong with the approval process except the need for streamlining the last copyediting phase and organising the authors to only do copyedits. Unfortunately we cannot currently revert someone who breaks this pattern

In Biology we have taken some articles through numerous editions and can get approvals done quite frequently. The main reason we have several approved articles is to do with the intentions of the editors to get a job finished. It is not connected much with the approval software structure in my view, and will probably continue whatever the structure as long as the editors are enthusiastic and like working with each other. For me that getting better with time as I learn more about the ways of the other editors (eg I now understand Gareth's wicked sense of humour).

Can I repeat at the risk of boring people, that the problems instantly recognised by Catherine as a newbie is one familiar to the editors in biology, as shown by several of them seeming  to support features of  Carolines idea as a path to improvement (and me saying Deja vu  yet again).

[Chris Day]

Maybe you could edit the example I already have up there?

Now you can see how it looks with a separate style for the historical 'approved version' so it can be more easily distinguished from the current version.  Note that in the printable version the historical approved template only appears as the disclaimer for brevity. Lastly, i added an option to see the changes between the approved version and the most current version.  This should make it more simple than having to hunt down the last apporved version in the history.

[Joe Quick]

I don't want to step on anyone's toes here, but I dislike most of the ideas suggested so far:

1) A proof version that coexists with a draft version will necessarily diverge from it unless all changes are also made to the draft version.  And all changes would need to be made to the draft version - otherwise all corrections made to the draft would be lost upon re-approval of the draft.  This means that the implementation of such an approach would only serve to create extra work.

2) A proof version that is simple a stage in the approval process works fine, but the concerns raised by others that this locks authors out of content changes for that period are very strong, I believe.  So are the predictions that authors will not conform to the rules about what they are allowed to change during the proof stage.  Do we really want to revert a good addition to content on the grounds that it was made during a period that was supposed to be for copyediting?

3) On Larry's suggestion about direct editing: I think Steve's brief argument against this idea here is solid.  An I side with Nancy here.

4) On Larry's suggested rules here: these simply put too much distance between contributors and articles.  A draft version as we have now allows authors to jump in, "be bold," and make some good contributions, but these rules are a step in a direction that I don't think we want to go: they require authors to make suggestions and then sit around and wait.

5) Chris's point here, about Larry's further ideas on having the approved version as a page in the archive is good, I think.  We need to make sure that the approved version is copyedited.  I don't like the idea of protecting the article while this happens, however; we don't want to block authors from contributing to the article if they don't have another place (the draft page) to go to do this.

By now, you're asking what my solution is.  I think we need to simply have the (proposed) copyediting workgroup involved in approval.  If we ask the chief copyeditor to sign off on an article before it is approved, then we can be reasonably sure that the most egregious spelling, punctuation and other such errors have been set right.  The system is obviously not perfect, and I'm aware that there have been a number of arguments against the idea of requiring this, but I believe it would be the simplest way to address the current problem and it doesn't need to interfere with any of the other edits that take place during the approval process - it is simply the last of the edits before the article is officially approved.

[David Tribe]

2) A proof version that is simple a stage in the approval process works fine, but the concerns raised by others that this locks authors out of content changes for that period are very strong, I believe.  So are the predictions that authors will not conform to the rules about what they are allowed to change during the proof stage.  Do we really want to revert a good addition to content on the grounds that it was made during a period that was supposed to be for copyediting?

Yes I agree with a proof version as a stage, and there indeed is no problem if a good edit appears. The problem is with bad edits. They prevent the copy edits going though because the bad edit cannot legally be reversed. The proof stage concept does not ask people not to make edits, just hold-ff for two days maximum till all the copy editing is  finished

[Joe Quick]

2) A proof version that is simple a stage in the approval process works fine, but the concerns raised by others that this locks authors out of content changes for that period are very strong, I believe.  So are the predictions that authors will not conform to the rules about what they are allowed to change during the proof stage.  Do we really want to revert a good addition to content on the grounds that it was made during a period that was supposed to be for copyediting?

Yes I agree with a proof version as a stage, and there indeed is no problem if a good edit appears. The problem is with bad edits. They prevent the copy edits going though because the bad edit cannot legally be reversed. The proof stage concept does not ask people not to make edits, just hold-ff for two days maximum till all the copy editing is  finished

Perhaps what I have in mind is closer to what is being proposed than I first thought.  I always feel silly after making strong comments.

Just as in the current system, in which editors mark a specific version in the history for approval, the editors' role in the approval process would end with them specifying their "final" version for the copyeditors, who would clean it up and make the process complete.  All the while, others would be free to make good and bad edits to the draft.

I would be willing to put a temporary moratorium on content edits if it is for a very short time, but I think the article must then be locked to all but the copyeditors during that time.  If not, we'll just end up spending too much time reverting edits by people who don't know or don't understand the convention.  The article would also need to be marked very prominently during this period to explain why it is locked and where potential contributors may make suggestions (the talk page, of course).

Or maybe we need to create the draft page as a part of the initial "pre-approval" process.  Editors presumably nominate articles for approval because they feel those articles are worthy of approval.  This doesn't mean that improvements can't be made.  Those improvements could continue to be made on the draft page while copyeditors touch up the nominated version.  If any egregious problems are found in the nominated article before the deadline, then it loses its nominated status.  If not, then it goes up.  Changes made on the draft page wait for the next nomination.

[Catherine Woodgold]

Larry, I'm not sure you understood my suggestion.  There may be reasons not to to it this way, but my suggestion was to have a /Proof copy which can undergo copyediting changes, and at the same time have a /Draft page which can undergo content changes.  This is not much different from your suggestion, that a class of copyeditors could edit the main page -- presumably content changes could still continue on /Draft while that was happening.  The main difference is that there would not be a period of time during which the "approved" version contains a significant number of spelling errors.

In answer to some other suggestions:  I believe that an editor should approve the final, copyedited version before it is made official.  I don't feel it's the role of constables to check that only minor changes have been made.

Example:  A mathematician wrote an article that used the word "homeomorphism" many times.  (I may have some details of this story wrong but I believe it's based on a true story.)  While the mathematician was away, the mathematician's secretary looked up the word in the dictionary and "corrected" the spelling to "homomorphism" everywhere it appeared, and then submitted it for publication and it was published like that.  Apparently the two words had significantly different meanings in mathematics and the mathematician suffered considerable embarassment over it.  Similarly, someone might think they were improving an article by "correcting" "discrete" to "discreet".  There are many changes that might look like corrections to a non-expert.  A constable might compare two versions, see that only spelling "corrections" had been made, and think the article was ready for official status.  Besides, I thought constables were not supposed to be involved in content.

I think it would be OK for mere authors to begin an approval process and start a copyediting phase, and have an editor involved only after the copyediting, at which time the editor could either approve the article, or not.  Perhaps authors could be limited to doing this only a maximum of once per month per article (or once per six months or whatever).  Where there is a shortage of editors, ideally at any time there would be a most recent copyedited version ready for any editor who comes along to read and immediately approve;  in general it would not contain all of the most recent content of the /Draft.   I would vote against any system which does not have an editor reading a final version before it's marked as the approved version.

If there is a copyediting phase on the /Draft page, it makes sense to me that there should be a prominent template at the top of the talk page and maybe also at the top of the /Draft page, notifying people that it's a copyediting phase, and then if someone makes a content change, someone else should revert it, and move the change to the talk page, and politely tell them that the change can perhaps be put back after the copyediting is finished.

[Joe Quick]

I think it would be OK for mere authors to begin an approval process and start a copyediting phase, and have an editor involved only after the copyediting, at which time the editor could either approve the article, or not.  Perhaps authors could be limited to doing this only a maximum of once per month per article (or once per six months or whatever).  Where there is a shortage of editors, ideally at any time there would be a most recent copyedited version ready for any editor who comes along to read and immediately approve;  in general it would not contain all of the most recent content of the /Draft.   I would vote against any system which does not have an editor reading a final version before it's marked as the approved version.

Now this I like!  It also addresses the suggestion I brought on the [http://en.citizendium.org/wiki/CZ_Talk:Approval_Process#A_suggestion talk page for CZ:Approval Process], which I think is one we'll need to deal with sooner rather than later.  I'll let you read my argument there.

[Anthony.Sebastian]

Anthony - of course it's a wiki. That's the whole idea behind a proof version- NOT a time period, let an article be approved then RIGHT AWAY there is a proof version, for copyediting and a DRAFT version for continued creative work. Could we just try it and see? Isn't that the spirit of a wiki?

Nancy, I agree we need a 'proof' version.  I presume by 'proof' version you mean the current working version with an approval-nomination tag stating a future date when that version will receive final approval.  Until that date contributors will only copy-edit.  Correct me if I have that wrong. 

I hope you don't mean to have open simultaneously for contributions both a 'proof' version and a 'draft' version.  In that scenario, if copy-edits get made on the 'proof' version and content-changes get made on the 'draft' version, with both versions open, how do we reconcile the former with the latter?  I see a nightmare there.

Perhaps we could have a very short 'proof' ''period'', two days, at most, say, on the current working version, whether 'draft' or 'never-before-approved', strictly for copy-editing.  That would minimize loss of momentum for content-contributors, who would just wait out the two days for the copy-editing to conclude.

As for fact-checking, I argue for not tagging an article as nominated-for-approval (i.e., turning it into a 'proof' version) until fact-checking over.  The first editor asking for support to put up approval-nomination tag should have checked facts.

I recognize I may not understand your point.  For myself, personally, I will happily go along with your proposal however that finalizes, as I will adapt.

Thanks

[Anthony.Sebastian]

Larry, I'm not sure you understood my suggestion.  There may be reasons not to to it this way, but my suggestion was to have a /Proof copy which can undergo copyediting changes, and at the same time have a /Draft page which can undergo content changes.  This is not much different from your suggestion, that a class of copyeditors could edit the main page -- presumably content changes could still continue on /Draft while that was happening.  The main difference is that there would not be a period of time during which the "approved" version contains a significant number of spelling errors.

Catherine, with both a '/Proof' and a '/Draft' version open for contributions concurrently, how do you get the changes made in the '/Proof' version into a content-modified '/Draft' version?

[Larry Sanger]

I think it would be OK for mere authors to begin an approval process and start a copyediting phase, and have an editor involved only after the copyediting, at which time the editor could either approve the article, or not.  Perhaps authors could be limited to doing this only a maximum of once per month per article (or once per six months or whatever).  Where there is a shortage of editors, ideally at any time there would be a most recent copyedited version ready for any editor who comes along to read and immediately approve;  in general it would not contain all of the most recent content of the /Draft.   I would vote against any system which does not have an editor reading a final version before it's marked as the approved version.

Now this I like!  It also addresses the suggestion I brought on the [http://en.citizendium.org/wiki/CZ_Talk:Approval_Process#A_suggestion talk page for CZ:Approval Process], which I think is one we'll need to deal with sooner rather than later.  I'll let you read my argument there.

I agree, if we're going to have any copyediting phase, there's no need to make a special "proof" page for it.  Good job Catherine.  :-)