h2g2 The Hitchhiker's Guide to the Galaxy: Earth Edition

So, the *old* copy should be assigned a new number, with the edited copy being given its original number.

I really like this idea, though I'm checking with the dna team whether or not it's technically feasible. As for what to do with the previous version, I'm not sure we should delete it, as it seems to be within the ethos of the editing side of the guide that we never throw anything away.

In principle I like the idea of storing previous versions in the conversation forums, but I'm not sure how user friendly this would be. Also, if there was a link to previous and/or the original version in the left hand margin, then older versions would only be a click away.

For long entries - I don't think it would be a problem to split entries up. It doesn't happen that often so as to make it a time-consuming process for the inhouse team, and in most cases it would take time for entries to become so big that they needed to be divided.

"So, the *old* copy should be assigned a new number, with the edited copy being given its original number."

There was a definate consensus on this. What to do with the old version is a bit tricky though. Personally I don't think putting it into a forum would work. You'd lose things like lists, bold & italics, pictures, etc. and the forum post would be very long. I think this solution is nicest: "Store it in an archive section of the guide that doesn't show up in searchers (eg 'dna/h2g2/archive')"

The point about splitting up entries was not that the actual splitting up would be difficult, but that if the entry keeps its A number, which part of the new split-up entry does the old A-number point to. I don't think this is a major problem though, pointing to part 1 or the 'general overview' page (like Uni projects) would be most logical.

Well, we currently throw away old copies when we make updates (whatever method), for both Edited and Unedited Entries, so I'm not sure that you can claim ethos as a reason.

My big question is "Why?". I haven't heard anyone give a convincing reason why anyone would ever want to look at the old version of an entry. It's different for Unedited Entries, where it's good to keep the old copy so the original author keeps copyright and control over the original. But in the case of updates, the original author doesn't have copyright and control over the old Edited version, so that motive isn't valid.

Really, where's the benefit in having the old version "only a click away"?

Well, in case something gets edited out that turns out to be important, or a future researcher needs to see another perspective or approach to help with redevelopment.

How about links to past versions and/or the original version submitted by the researcher listed so it is only visible to updaters?

Researcher PSG

Well wherever it gets moved to, it shouldn't come up in searches.

As long as it no longer has its status as a part of the 'Edited Guide' why couldn't it show up in a search? It'd be just like any other regular Guide entry right?

How about this compromise.

Mary writes an entry on Elephants. (call it Unedited-1). She submits it to Peer Review and a copy is taken, edited, and hits the front page (call it Edited-1).

Fred writes an update to the entry (call it Unedited-2). He submits it to the Update Forum and a copy is taken, edited, and so on (call this copy Edited-2). Samantha makes a second update, generating Unedited-3 and Edited-3.

Now, the three Unedited Entries should be retained, and the authors can do what they like to them, just like any other Unedited Entry. But I think it's reasonable to hide Edited-1 and Edited-2. The perspective and information in those entries will be already contained in Unedited-1 and Unedited-2. Plus, the Unedited Entries can be modified or deleted by their authors in whatever way they feel like: specifically they can put back in any bias or personal opinion that they had to remove to get the entries through Peer Review.



A seperate issue is whether Edited-3 version should link to the Unedited versions, and whether the Unedited versions should link to Edited-3. I think yes in both cases, and AFAIK that's already in the <./>Features-Development</.> feature list.

-Martin

Thanks for these responses!

Frankie - in splitting up entries... the old version could point to both new entries. How about that? I absolutely agree with you that not all versions should come up in the search.

Lucinda - linking to old versions. I think there are two reasons for linking to old versions of entries. One comes from traditional publishing (though I'm not sure how well it applies here). In magazine publishing every single proof of a page is kept for seven years in case the publishing company gets sued for copyright issues, defamation, that sort of thing. At least that was the case the last time I worked on magazines. The second reason is this: we will certainly need the old version if we are going to implement some sort of comparison system to help inhouse editors and the volunteers that end up being involved in this scheme.

As for your second post - wouldn't Edited 1 and Edited 2 just turn back into ordinary entries? (That list of copies, just keeps getting longer.)

> "wouldn't Edited 1 and Edited 2 just turn back into ordinary entries?"

I thought the fate of these entries was what we were trying to decide!! I would say, keep Edited-1 while Edited-2 is being subbed and in-house edited, and delete it when Edited-2 actually becomes Edited. Similarly, keep Edited-2 until Edited-3 becomes Edited. You could turn them into ordinary entries, owned by the relevant sub-ed, with new A-numbers, but I don't think that's a good idea.

I doubt there are legal reasons to keep the old versions available - and even if there are, there aren't any reasons to keep the old versions *publically* available.

For the second reason you gave, you obviously need to keep Edited-1 available until Edited-2 has finished being subbed and in-housed. Not least because until Edited-2 has been finished, Edited-1 will represent h2g2's official word on Elephants. Once Edited-2 is in fact Edited, Edited-1 is surplus to requirements, and at *that* point it can be deleted.

Perhaps some kind of timeline would clear this up?

* Unedited-1(UE1) is created by Mary
* UE1 is submitted to PR by Mary
* UE1 is picked by a Scout
* Edited-1(E1) is auto-created with status 'recommended'
* A sub-ed works on E1
* the in-house editors look at E1
* E1 is set to status 'pending'
* E1 is featured on the front page and set to staus 'Edited'
* (time passses)
* UE2 is created by Fred
* Fred posts a thread to E1 pointing at UE2.
* Fred updates UE2 with all relevant info
* Fred submits UE2 to the Update Forum
* UE2 is picked by an Update Scout
* E2 is auto-created with status 'recommended'
* A sub-ed works on E2 (looking at E1 to see what has changed)
* the in-house editors look at E2 (looking at E1 to see what has changed)
* E2 is set to status 'pending'
* E2 is set to status 'edited', E1 is deleted, and E2 gets E1's old entry number.
* (time passes)
* UE3 is created by Samantha
* Samantha posts a thread to E2 pointing at UE3.
* Samantha updates UE3 with all relevant info
* Samantha submits UE3 to the Update Forum
* UE3 is picked by an Update Scout
* E3 is auto-created with status 'recommended'
* A sub-ed works on E3 (looking at E2 to see what has changed)
* the in-house editors look at E3 (looking at E2 to see what has changed)
* E3 is set to status 'pending'
* E3 is set to status 'edited', E2 is deleted, and E3 gets E2's old entry number (which is the same entry number as E1 originally had).



Now, all this would ensure that you never have situations where:
A) There is *no* Edited entry on Elephants, or there is more than one Edited entry on Elephants.
B) Someone is subbing or in-housing an entry without reference to the previous edited version
C) An author loses edit access to their unedited entry

-Martin

Okay - thanks for that - is there any reason to *not* keep all versions? (Even if not publicly visible, with it being put into an archive.)

I think that's a good timeline of the life of an entry Lucinda, but we still have the question over whether or not you delete the old versions from public view.

Providing that the entries don't come up in searches, I can't see the harm in leaving the entries visible (and there might be a few rare cases where it will be useful - say if someone accuses an updater for altering the meaning of the original).

Perhaps the outdated entry could be given a new status of 'Archived', and would no longer show up in searches. Looking at the entry's A number brings up a link to the new entry, plus displays the old version.

I'm not that the edited entry should have links to the archived versions though, at least not in the entry data box. Perhaps the search box could have a new tickbox for 'archived' allowing people to search for archived version.

I agree

If you keep superseded Edited Entries but make them totally invisible, then that's fine by me - the practical impact is identical to deleting them. That's an implementation issue, and I'm sure Jim will pick whatever makes most sense.

Reason not to: clutter. If something has no use, why display it? It'll only get in the way of other entries that actually have a purpose in life. It'll clutter up the search results, clutter up <./>RandomNormalEntry</.>, clutter up people's lists of unedited entries, and so on. Basically they'll get in the way, for no benefit.

In particular, consider when someone searches for 'Elephants', and gets back six different entries, one edited, five unedited, all of which are pretty much similar. And the more times an entry gets updated, the worse the situation.

Also, if the ex-Edited entries are demoted to Unedited, it's unclear who should get ownership (IE, who gets the Edit button). The original author already has a copy each, and there's no benefit for hir to get another. And it seems distinctly odd for the person who originally sub-edited the entry a couple of years ago to suddenly get a fresh new copy of that two-year old entry on hir home space.

-Martin

Yeah, you could do all of that (Archived status, checkbox in searches, etc) - but it all sounds like a lot of coding work for no real benefit. I'd prefer to have Jim's team working on the tools for volunteer translation/moderation, grouping entries, and so on.

> "say if someone accuses an updater for altering the meaning of the original"

*If* that's a valid accusation, then someone will raise it in the Update Forum, or the sub-ed will raise it, or the in-house editors will raise it, all of which will take place while the original edited entry is still visible. The chances of a meaning perversion being missed by all these people but being serious enough to irritate the author... I don't think that's plausible.

However, I'll give way on this one. Y'all seem to know what you want, so go to it, by all means.
-Martin

Technically it would be the same as deleting an unedited entry - which can be undeleted by the author.

Okay. I'll discuss the technical implications with Jim. Thanks everyone.

(I really shouldn't do that - but it's hard to remember which threads you've read otherwise. Bring on the bookmark feature!)