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

The Update Notification System sounds good - should make the whole updating procedure smoother, just as the last changes in the subbing procedure made subbing a lot easier.

What I would like to propose is a button 'Recommend for Update' in the margin of every Edited Entry (if I got it correctly, the whole community should be able to recommend entries for updating, not just the voluteers), similar in look and feel as the 'Submit for Review' button. If you click this button, you should enter your reasons why that specific entry should be updated, and the posting you've made could automagically pop up on the UN page (just like 'Submit to Review' and the PR page).

The recommender should be obliged to say what is to be done, as a posting saying 'This entry is crap' is of no use, I would think. Reasons for an update are for instance correcting errors, adding links to newer entries, adding specific information, or even major changes stemming from conversations or from another, better entry available at Axxxxxx.

What do you think?

SchrEck Inc.

Absolutely the recommender should be obliged to say what is to be done, otherwise what's a poor Updater to do?

Currently we take suggestions and dig through the Guide for any other nuggets of information - and it is seriously time-consuming work!

Telling the Updater what needs to be fixed and allowing the community to discuss it in a Peer Review-type forum to add even more information or refute one of the recommender's facts is critical and will make this process really functional!



- X

An interesting idea. I know of a number of my earlier attempts that could do with updating and some of the more generic articles could do with being worked on. However, a thought just struck me out of the blue on this one. If you have a button that people push and then give their update, will people continue to have conversations? A lot of conversations often start as someone leaving their nugget of extra information and that may well disappear as they all rush off to the UN (A fine acronym) to get their bit added. Just a thought and I still reckon its a good idea overall.

Good points. My only point on this would be that UN would possibly be *flooded* with Recommend for Updates, with nobody except the poor Updaters doing the actual updates. People might lose confidence in the process if they were not seeing a relatively fast turnaround.

Originally, before Peer Review came into being, the sub-eds were given the job of turning guide entries into official entries. The sub-eds were swamped, and the delay between submitting and release could be up to 6 months. Even poorly crafted entries took ages to be rejected.

I strongly believe that the onus for updating should be on the research community, with the Updaters doing the approval, final editing and release functions, while also perhaps making small changes where appropriate.

Sorry, I am beginning to repeat myself.

Woodpigeon

I think the idea of a 'Recommend for Update' button is a great idea. It should be accessible to all researchers so that new researchers can feel it is possible to contribute to the guide without having to write completely new entries themselves.

You would definitely have to make the user enter a valid reason for Recommending for Update... maybe by insisting then write more than a certain amount of word/characters in the 'further information' box.

If the button 'Recommend for Update' is pressed by someone other than the original author, the author should be informed by a posting to their home space board.

If an entry is updated, then revised entry should go into the original A##### page... to allow pre-existing links both in H2G2 or outside the BBCs website to remain intact.

I agree with all of the above. However, I also think that the "submit for update" button should *only* appear on Edited Entries that have been in the Edited Guide for at least 3 or 6 months. Otherwise, someone will hit the "submit for update" button on *every* new Edited Guide entry as it comes out, claiming that it needs to be updated because there's a word wrong in paragraph three, or because their opinions differ with the author's. These are the kinds of things best raised in conversation threads attached to the entry. If, 3-6 months down the road, people think that the entry needs updating, *then* they could push the button.

Just my 2 cents.

Mikey

ps -- I also agree with everything Jimi X said, per usual.

The button could bring up a list of options, such as "Correct a typo", "Fix a factual error", "Add new information" and "Correct out-of-date information". The last two options just wouldn't appear on articles which were less than three months old...

Also "Suggest a link to another article"...

I agree with the comment posted above. I think in some ways the proposal is taking a bit of the wrong emphasis by placing the onus on updating the entry on people in the UN forum and the volunteers. What I'd have imagined for the update system is that viewers to the page are allowed to make the updates themselves (by somehow being able to submit a version with revisions and updates), which the volunteers can instantly see, and only have to approve/reject, without having to do the hard work themselves.

After all, the person clicking the button to 'update entry' is probably going to have the ideas on how to update it themselves rather than simply suggesting someone else update it. For instance they might see an entry on a specialist subject they know a lot about themselves (having come across it via a BBC/Web search for instance), and want to submit corrections or updates. In this case that person is the one best placed to make the update.

I'm not sure about the time delay either. Many people only come across the entries when they hit the front page (remember new entries are also promoted on other BBC pages), and so the changes should be made straight away. Think about how many suggestions are made in the first 24 hours that an entry becomes edited...

IMO the system should also cater for small and minor updates, such as typos that have slipped through the net and subheaders that should be headers for example. Or the quick addition of a clarifying sentance. If these were included in the '5 a week' limit the system wouldn't be very scaleable at all.

At the moment, the editors have to monitor all the conversations from the new entries to see if there are any changes/typos spotted, and then they have to make these changes themselves. If you could just click on a button, make the suggested change yourself, and then click submit, flagging the update to be quickly checked by a volunteer, it could actually save the in-house staff time...

Any comments?

Right now people *do* post info about typos and factual errors to threads on new Edited Guide entries, and these are fixed pronto by the in-house staff. This is not currently part of the Update process, but a quicker, more streamlined process.

If these issues were to be addressed via the Update process, there would be several problems:

1) The errors would not be fixed nearly as quickly as they are now.

2) The list of entries submitted for updates would quickly become overwhelmed with entries that have only been in the guide for a day or two before being submitted. We would need a volunteer group twice the size of the sub-editors to be able to deal with all of those *plus* the older entries that need more extensive updating.

3) These new entries would then be updated multiple times. Entry hits front page, someone spots an error, hits the "submit for update" button, and it's updated. Someone else spots another error, and the process repeats. And again and again. Having a 3-month time delay would allow these errors to accumulate in the threads -- the urgent ones would be dealt with by the in-house staff ones as they currently are, and the rest could be dealt with en masse, which would be much more efficient for both volunteer and house staff editors. Remember, *every* time a volunteer changes an entry, a house staff person has to re-proof the whole thing.

Mikey

Good points - maybe the current processes are all that's needed for quick fixes.

I often spot things when randomly reading the guide though, and usually fix things via a quick e-mail to the subs list or the editors. This isn't neccessarily something everyone can do.

Also, things that might be spotted on new entries might be factual errors as well as simple typos/GuideML errors. For example a scientist browsing the BBC website might see a science entry advertised on http://www.bbc.co.uk/communicate and instantly spot some things that the writer has mis-understood. There ought to be some way that they can suggest how these things can be fixed.

It's a natural thing that web pages contain errors, I've even spotted stuff on the BBC news site. Ideally you want a system of dealing with this kind of stuff as quickly as possible. Most of the BBC sites do this quite simply with a 'feedback' button that links to a form which sends a quick e-mail to someone (usually generating an automatic reply)...

Hi there,

thank you for your comments; a few further thoughts:

I thought a 'Recommend for Update' button would make the update scheme easier to use, that's the main and only reason for such a thing. You needn't tell people 'go there-and-there, post in the forum, but, mind you, use a specific format'. As with the new PR button, this could lead to more traffic on the UN page, as more people would use (abuse?) this facility.

It has been said before, a further question would be whether small updates (typos and the like) should be handled via the new update scheme or not - these are addressed effectively now by observing the conversations of newly published entries. Should it be left that way?

If not, there is a distinction between small and major updates necessary. For minor things, there has to be a shortcut. I can't think of anything else but having the inhouse editors process them within a day or so and then drop a note in the UN thread that it has been fixed. I don't think that we need a superceding system for these updates, though. Of course, errors could still be reported in the article's forum.

Major updates should be handled by the update volunteers, as proposed by Anna. The only thing is, what is a minor update and what is a major update and who is to decide that?

SchrEck Inc.

But surely that is what the conversations are for currently. If someone comes along and spots an error they leave a note in a forum. This can lead to a conversation, possibly with the editor or author, which may well lead to the need for an update. However others can also turn up and add to that discussion. So I get the feeling it is good to let these things build, as such, to make the eventual update more worthwhile. Better to do one big rewrite than several little changes. Also, what would happen if the person who hits the update button is actually talking through a non standard orifice? Other people who come along to the entry later may not notice, as it is not in a converstion off the entry, and so not be able to rubbish the suggested update, which would then get in (if I'm understanding things correctly).
I reckon we want to keep things to conversations as much as possible. Is it possible to have a UN forum which would appear in UN (allowing the updaters to spot it as well as any lurkers there) while also having the forum appear under the entry, to allow the casual reader to chip in?

Munchkin, I appear to be developing a point of view, how odd.

Hi Munchkin,

you're right, the discussions in the forum are A Good Thing, and to let them settle until an update is worthwhile is also A Good Thing. I think that this wouldn't change significantly by having these conversations in the UN forum rather than in the article's forum.

To have all conversations handy, entry and update thread should be crosslinked, favourably automatically. When the update is done, the UN thread is to be moved back to the article (like the PR thread nowadays). That way, all messages regarding a specific entry are accessible from the entry itself, just the way it is now, and 'suboptimal' update requests will get commented upon pretty soon, I guess.

SchrEck Inc.

True. I just worry that some poor volunteer is going to have to bop back and forth between two threads to glean an update. And then, having updated, someone from the opposite thread will point out the mistake and it will all have to be done again. But then, no system is perfect and Peer Review and what not tend to survive these problems. I'm probably just getting too picky. Its still a good idea which I'm almost tempted to volunteer for.

I agree with the suggestion that in general, update requests might be handled through the normal conversation forum. It's public and people can comment on the validity of the suggestion. I would not be happy with the idea of a feedback email because it is too private, and that's not what H2G2 is about, I think. We should see what people are recommending, and have the opportunity to discuss it.

At the risk of boring people, I take some issue with getting the volunteer Updaters to handle major changes. If someone feels strongly that an entry needs a major rewrite, then let them put their money where their mouth is, and get them to do the rewrite themselves.

Finally, I actually was not aware of the many informal mechanisms present to handle minor updates. It would actually be a good thing to have a forum / page to recommend minor updates to articles. Again, it should be public, so that dissent is allowed if other people disagree that the change is worthwhile.

Woodpigeon

Why all these UN threads? What happened to the Review Forum format? Here's how I envision the system working:

1. Researcher X reads an entry and sees a need for an update
2. X clicks the update button
3. X fills in details qualifying the entry for updating
4. A new thread appears in UN named 'A123456 - Title' with the details as the first post
5. Anyone passing through can read the thread and post comments
6. Researcher Y happens to read the same entry and sees what X saw (or a different problem)
7. Y looks at the Entry Data box and sees that the entry is being updated
8. Y clicks the link and adds contributes the UN thread (if necessary)
9. When an Updater thinks it's time, the update is 'picked' (to borrow a term from the Scouts)
10. 'Picked' updates are allocated to Updaters, who edit the new information into the entry
11. The Updater submits the new entry to the in-house team for final touches
12. The new version recieves the original A-number and the old version is moved
13. The new version has a link in the ED box or at the bottom to a list of earier versions, each with its own credits

I would also suggest that the focus of the UN thread be both to gather information to include and to foster ideas regarding its incorporation.

-Spike A.

It'd be nice if there was an official way of flagging an update as "urgent", such as where an entry suggests you get in touch with someone who has just been killed in a horrific bobsleigh incident...

(Arbitary example - no bobsleighs were harmed in the preparation of this posting... )

Spike, I have a feeling that unless the Updater is an accomplished writer, that the edited entry and subsequent further revisions might lose some of it original character, until the entry morphs into some kind of please everybody hodge podge (look for example at some of the community edited entries where questions are posed on the front page and everybody adds their musings - personally I would not like to see all guide entries begin to take this style). I think that we have to be alert for the "many cooks spoiling the broth" syndrome. Personally I would prefer to see the revised entry also being written by a single committed researcher, rather than by an updater doing their best to stick lots of user comments together into a revised entry.

Woodpigeon