User talk:Wrh2/Why Wikivoyage was not my primary travel planning tool

New Regional template suggestions

Hi, Ryan. I really like these suggestions. I think they'll greatly improve a lot of regional articles that now have a lot of empty fields. Ikan Kekek (talk) 04:24, 7 November 2014 (UTC)

Thanks for the feedback Ikan. Please feel free to make any edits to this essay that might help make things clearer, or to include additional suggestions. I'll solicit additional feedback in the Pub eventually, but I'd like to hold off until either people are likely to be more engaged or when I have more free time to follow-up should there be interest in actually making some changes. -- Ryan • (talk) • 04:40, 7 November 2014 (UTC)

I think most users won't notice this draft while it's in your user space, unless you solicit contributions and comments at least in Requests for comment if not also the Pub. Ikan Kekek (talk) 05:25, 10 November 2014 (UTC)

Understood. I'll solicit wider feedback eventually, but don't want to do so until I have the time to fully follow-up on the discussion, otherwise I'd worry that it will end up as yet another Wikivoyage discussion that elicits lots of feedback, rambles in lots of directions, and ends up changing nothing. -- Ryan • (talk) • 06:00, 10 November 2014 (UTC)

There's always that risk, anyway. I'll think about your ideas, but I surely support you on eliminating a lot of the required subtitles from regional articles and substituting summaries. That'll make us more like a paper guidebook, in a good way. Ikan Kekek (talk) 07:20, 10 November 2014 (UTC)

Just found this draft. Ryan, well done. You brought-up a very important issue. This is the same thing I wanted to bring up last year while writing Sindh but I was afraid to start a discussion because I didn't had any solid solution. --Saqib (talk) 10:04, 10 November 2014 (UTC)

When will we able to see this proposal on the table Ryan? You seems active these days so why delaying this one? --Saqib (talk) 14:26, 5 January 2015 (UTC)

See #Update below. -- Ryan • (talk) • 15:14, 5 January 2015 (UTC)

Sub-regions

Ryan and IK, since soon we're going to introduce a new proposal related to regions. I want to suggest one change as well. I don't want to directly add it in your draft Ryan as you might have opposition to it. We always divided regions (when they're big in size) into further sub-regions but it was such a pain for me to divide the Pakistani provinces because both practically and politically, subregions of Pakistani provinces doesn't make any sense at all. In-fact I realised people found Southern Sindh and Northern Sindh as offended and I've been attacked on-wiki for it few months back. Same goes for Southern Punjab which is also somewhat controversial. But I had no other good solution to divide both the provinces and so I came up with these. I propose to abolish the rule of "must have sub-region articles". We can off-course have sub-regions for those where they make sense both practically and politically but if they don't make any sense, lets not have them at all. They're totally useless and could confuse a traveller. --Saqib (talk) 15:22, 10 November 2014 (UTC)

In the case of Sindh, having a single region article would have been fine according to existing Wikivoyage policy, although it sounds like we need to make that clearer in our current documentation. Per Wikivoyage:Geographical hierarchy#Dividing geographical units:

"...when the number of subdivisions exceeds nine, it's time to start looking at an intermediate subdivision. That's not to say we never exceed nine! Sometimes there's just no logical alternative division, particularly at the lowest levels of the hierarchy. Remember: when dividing regions, use a breakdown that is most practical from the traveller's viewpoint, even if that creates a hierarchy that violates the 7±2 rule of thumb."

Contra Costa County is one such example - it was more confusing to break that county into smaller pieces than it was to have a region that contains 20+ cities, so per a VFD that I can't find we got rid of the sub-regions and just made one large parent region. -- Ryan • (talk) • 15:33, 10 November 2014 (UTC)

Well thats perfect. I was don't knew of it. I have spent year drafting suitable sub-regions for Sindh and Punjab. Julia spent effort working on Sindh sub-regions and if she have no opposition, I would like to abolish sub-regions of Sindh. --Saqib (talk) 15:38, 10 November 2014 (UTC)

Indeed, there is no requirement to subdivide regions, so yes, do what you think best, assuming you don't get any opposition after posting to Requests for Comment and trying to solicit comments. Ikan Kekek (talk) 20:35, 10 November 2014 (UTC)

Sure. Will do that soon. Anyways, eagerly waiting to see this bill to be introduced in the parliament Ryan. --Saqib (talk) 23:18, 11 November 2014 (UTC)

Agree with IK. I think we've sometimes treated 7±2 as a rule instead of the guideline that it's supposed to be (IMO). When I get the time, I'm also going to be raising a proposal to merge a number of subregions into one region of 15-20 cities because I think that's the best way to present it. -Shaundd (talk) 05:26, 12 November 2014 (UTC)

Cities

I like these suggestions as well. I've got one question... I don't see a section for Cities in the proposed region template, is this an oversight or did you mean to leave it out? I haven't given it a lot of thought, but my initial reaction is the important cities should still be highlighted. Or would the intent be to cover the important cities in the expanded subregion descriptions? -Shaundd (talk) 05:26, 12 November 2014 (UTC)

In the case of a bottom-level region like Westside (Los Angeles County) I would propose that each city should get the 1-4 paragraph treatment, which would provide a reader enough information to get a sense of which cities they wanted to read more about. For region articles that aren't bottom-level regions, either including a short city list (as we do today) or writing about the cities within the sub-region descriptions would make sense to me. I'd favor the latter since the cities within a region are often the main draw, but it's something that would be good to discuss if this proposal eventually gains some traction.

If there is enough interest I can solicit additional feedback in the pub now, my concern was that it will take significant effort to shepherd the resulting discussion into eventual action, so I wanted to let a few of the existing discussions die down, perhaps put together a sample region article for discussion purposes, etc. That said, if people want to start this discussion now, let me know and I'll solicit wider feedback. -- Ryan • (talk) • 05:51, 12 November 2014 (UTC)

I think it would be good to have some sample region articles before going to the Pub (at least a bottom-level and a mid-level). It should give some context to your ideas and hopefully focus comments more. -Shaundd (talk) 06:12, 12 November 2014 (UTC)

I agree with Shaundd. Also, one concern I'd have would be to avoid having regional articles that are treated as if they are second articles for the main city in the region or state. Any change we make should take that into account. But I think the main point is that regional articles are probably best treated, as a rule, as overviews, not lists. Except when they're not.

An additional point I'd make is that there are some state and even country articles that have some aspects of "city" articles. A good example is Niue, which absolutely needs to have listings, as the villages on the island are so small that Frank made the good decision to merge the articles about them back into the country-level article for the whole island. Some problematic examples are Mauritius and Goa. Ikan Kekek (talk) 06:17, 12 November 2014 (UTC)

I wonder if Mauritius has aspects of a city article or if it's more a sign that WV struggles to deal with rural areas well? Similar to Ryan, I had to rely on sources other than WV when trying to plan a trip a couple of years ago. Ryan's three points capture my experience well, but I'd add a fourth point that the fairly non-existent coverage of rural areas / regions with a lot of small villages but few (if any) large centers -- at least where I was travelling (Southwestern France) -- made it difficult to rely on WV. I'm not sure if this is due to fewer contributors travel to those destinations or if it's because our bottom level templates are focused around cities or something else. -Shaundd (talk) 06:10, 13 November 2014 (UTC)

We run into similar issues with places like Kauai or Big Island, where the towns are often not where the resorts or natural attractions are located - Talk:Nosy Be is a similar discussion that is currently ongoing. There have been some attempts to consolidate small towns into single "rural area" articles like Rural Montgomery County, but thus far I don't think that anyone has come up with a great solution to this problem. -- Ryan • (talk) • 16:25, 13 November 2014 (UTC)

I also think that there's a tendency for articles about islands or other relatively small areas (like the state of Goa) to be treated like city articles because they're clearly discrete places. I think that part of the psychology is that people much more commonly say "I'm taking a trip to Goa" than "I'm taking a trip to Panaji." Ditto for Mauritius and Port Louis. Ikan Kekek (talk) 22:07, 13 November 2014 (UTC)

It makes sense to me to consolidate rural areas into a single article of reasonable size if there really isn't much in any individual town/village. "reasonable size" might be different for different parts of the globe though. In the U.S. I can see covering rural areas of several hundred square miles in a single article if necessary since there are single city articles that cover that much land area (Jacksonville (Florida) for example). If a traveler is going to need a car to get to the place then I don't see much harm in covering large areas in a single article. A lot of rural counties in the U.S. could probably be covered in single articles.Godsendlemiwinks (talk) 01:57, 25 August 2015 (UTC)

Implementation

Hi Ryan! I also wanted to thank you for your interesting essay. We discussed it on Russian Wikivoyage and, while we don't quite share your ideas regarding TA, links to Wikipedia and so on, we all agreed that the descriptions of cities and regions could be more detailed and more catchy. Here you can see our ideas for the design. We were not sure that 4 paragraphs of text are necessary, and when we restrict ourselves to one paragraph, it nicely combines with a picture, which is another reason to stay with one-paragraph descriptions. Feel free to note it down as one of the possible implementations. --Alexander (talk) 09:42, 30 November 2014 (UTC)

Thanks Alexander - I've still been holding off on pushing these changes in English Wikivoyage given the traditional resistance to change and generally difficult process of getting any large change implemented here, but the fact that another language version has already implemented one of the suggested changes gives me a bit more hope that such changes might be possible. If you have any further suggestions based on your experience on Russian Wikivoyage (aside from limiting the text to a single paragraph) please share! -- Ryan • (talk) • 16:48, 30 November 2014 (UTC)

Well, we are still undecided about the old regionlist template. Should it stay in some form or give way to the new one, so that the map of regions becomes a standard image? Not clear yet. --Alexander (talk) 17:14, 30 November 2014 (UTC)

Update

A personal update, in case anyone is wondering why this proposal hasn't moved forward. I wanted to have a couple of example regions available for discussion, but the devil is in the details. For example, I looked at East Bay (Bay Area) and its two child regions: Contra Costa County and Alameda County. The goal should be to allow a reader to understand the available destination options and quickly get to the city articles, but in that example the "East Bay" article would then become little more than an index page with two items, and the county pages would still basically just be huge lists of cities that were difficult to navigate. A better solution from an efficiency standpoint would be to get rid of the county pages and just have an "East Bay" page that contains a lot of destinations, but that would be an awfully busy page. I've considered ways to implement Alexander's example from Russian Wikivoyage, but that also has drawbacks for some of our regions. If anyone else wants to put together an example region page please go ahead, otherwise I'll try to push forward once I can create an example that meets the stated goal of simplifying and improving a reader's ability to choose a destination based on our region articles. -- Ryan • (talk) • 16:05, 15 December 2014 (UTC)

Another solution is to remove regions like East Bay and North Bay that are probably excessive. The phrase "What the locals call "the East Bay" is really two fairly distinct regions" in the beginning of the Understand section renders the whole article pointless. --Alexander (talk) 20:22, 15 December 2014 (UTC)

It really doesn't. "East Bay" is a completely accepted and continually used designation locally. The fact that the climate varies drastically between the shore and the hills, etc., doesn't render the designation pointless at all. Ikan Kekek (talk) 20:36, 15 December 2014 (UTC)

I am fine with the designation, but it may not serve as a useful region. I will not argue about that because I don't know the place, but my general feeling is that a region with only two subregions may complicate navigation rather than simplify it. --Alexander (talk) 23:27, 15 December 2014 (UTC)

I do know the area. It would be a bad idea to try to create two separate regions. Where would the boundaries be? Within the city of Oakland in some cases. Ikan Kekek (talk) 23:44, 15 December 2014 (UTC)

I've been thinking of regions where I could test it out. Unfortunately, I don't have much time right now, but if I get a chance I'll take a crack at it. I like what Russian WV did, and I think it could work well in some if not many of our regions, but I wonder how it would work in a bottom-level region with many cities. -Shaundd (talk) 06:09, 16 December 2014 (UTC)

[unindent] I've taken a shot at drafting what a region page could look like using Ryan's suggested template, you can take a look at Thompson-Okanagan. Apologies in advance for my boring writing, I'm suffering from extended writer's block so it's a struggle to make good sounding prose right now. I'm hoping it's enough to get some discussion and feedback going. Some general thoughts on the process:

• I think the new format is already better for most of our region articles. There are probably tweaks that can make it even better, but I think it helps give a better overview of what's actually in the region. What they did at RU WV is worth a look, too.
• I also think (with a caveat that I'll note in the next bullet) that it's easier to complete than our current region template because the writer doesn't have to provide as much information. On one hand, you could argue that it makes it less useful, but I'm not sure how much detail we need on See and Do and Eat and Sleep if the region is not the bottom-level region. At first glace, including some of that info in the region description itself makes it easier to understand what a particular subregion is about and why to go there (or skip it).
• On the downside, I still find it hard to write that Regions section. The writer still needs a good knowledge of the region and its subregions (and cities) to create a useful and comprehensive description. As you'll probably see in the Thompson-Okanagan guide, my knowledge of each subregion varies quite a bit and so does the quality of what I wrote.

What are people's thoughts on the test article? Ultimately, I think Ryan is on to something and this format is more useful and needs further exploring. I'm not sure if it will solve the problem of us editors not writing region articles but I think it's quite possibly a better template to start from. -Shaundd (talk) 23:34, 5 January 2015 (UTC)

I like it. It's quite readable. Ikan Kekek (talk) 00:08, 6 January 2015 (UTC)

Why not add a "SEE" section and move the region detail there (except region list). --Saqib (talk) 00:34, 6 January 2015 (UTC)

That could work. And I certainly don't want _all_ region-level articles to lack See, Do, Buy and other hitherto standard sections. Is it really necessary to have only one type of template, or can we be more flexible in allowing some sections to be optional at the regional level? Ikan Kekek (talk) 02:02, 6 January 2015 (UTC)

Thanks Shaundd for pushing this forward - if anyone has the energy to open this discussion up to a wider audience please go ahead. When I was trying (and failing) to create some samples for this proposal, I basically found that we have three kinds of regions:

1. Article-worthy regions (examples include US state articles, where there is a lot to say about what to see or do). Some rural regions also fit this model. Example: California or Lake Tahoe.
2. Navigational sub-regions (mid-level regions that exist primarily to break up areas into smaller pieces). Example: East Bay (Bay Area).
3. Navigational bottom-level regions (bottom-level regions that exist to collect together groups of cities and other destinations). Example: Contra Costa County.

The first is something that is fairly easy to write articles about, but for the latter two the goal really seems like it should be to just help the user navigate as quickly as possible. Shaund's example does a decent job of that, although it would be good if we could find some way to make it even clearer that such articles are just intermediate steps to help readers get to the next level of the hierarchy.

When trying to create examples, a problem I kept encountering was in dealing with large mid-level and bottom-level regions - perhaps someone could try to put together a good example where the region contains nine or more children, since when I tried to do so it always come out as a mess. Ideally, whatever we propose might even allow consolidation of some regions so that we could (for example) combine East Bay (Bay Area), Contra Costa County and Alameda County into a single article and save the user an extra click.

Also, to Ikan's point about making most headings optional, that is exactly what I would propose - Napa Valley definitely warrants a "Drink" sub-heading, but that heading is basically worthless on Contra Costa County. -- Ryan • (talk) • 02:46, 6 January 2015 (UTC)

All of that sounds very sensible to me. Ikan Kekek (talk) 03:23, 6 January 2015 (UTC)

Count me in as agreeing that other headings should be included when they're useful.

To Ryan's issue with large regions, the next region I was going to tackle was Okanagan with its thirteen "cities" and handful of other destinations. I'm also wondering how a structure like this will work in large regions so I'm hoping this will be a better test. -Shaundd (talk) 03:56, 6 January 2015 (UTC)

Update #2

Every time I've tried to put together examples of existing Wikivoyage regions that use a template similar to the one proposed in this essay I've failed to come up with something I was happy with. Since it's been several months of failure for me, if someone else is interested in pushing this proposal forward please do so. I'll continue giving this some thought, but so far I'm not making any progress and don't want to stand in the way of someone else who has the insights necessary to move ahead with changes to our region templates. -- Ryan • (talk) • 04:43, 2 February 2015 (UTC)

Its definately a setback to see that the proposal cannot be move forward due to some obstacles. But I wonder whether it would be a good idea to allow usage of region list proposed here for some regions? For instance: Sindh region article can become a good guide if we use the proposed region template. --Saqib (talk) 12:50, 2 February 2015 (UTC)

I think we should continue to experiment. There's been lots of talk about weak regions over the years, but without creating some examples of how a different template could look, I think it's going to be difficult to move it forward (I recall the lack of examples killed at least one prior discussion). The more I look at Ryan's proposal and experiment with it, the more I like it over our current region template.

It's a work in progress, but I'm playing around with the Okanagan region article at User:Shaundd/Sandbox2. It's a larger region (13 "cities" and a few more other destinations) so I was concerned that if the city descriptions were longer like Ryan proposed, it would become an intimidating wall of text. I'm experimenting with a couple of ideas:

• one is merging "Cities" and "Other destinations" into one "Destinations" heading and then letting authors break it up as fits the region, and the second is
• highlighting the main destinations at the top with longer descriptions and a small picture (like they've done on Russian WV)

Eventually I want to have a couple of layouts plus the existing template so people can compare it. Like I said, what's there now is a work in progress, but it would be good to hear some thoughts.

Ryan -- do you have any examples of your different templates? Even if you don't think they work, it might help others come up with ideas if we could see them. I unfortunately don't have lots of time to work on it (and I want to keep drawing maps) but I'll keep chipping away at it. -Shaundd (talk) 05:38, 3 February 2015 (UTC)

I was looking at a few of the sub-regions of the Bay Area as described above in #Update, but I never saved any of them. The main problem was that whether I tried lists, sub-headings, or other layous, I kept running into the "wall of text" issue you've described, which had a huge negative impact on usability. -- Ryan • (talk) • 06:12, 3 February 2015 (UTC)

I wonder if we're making too much of the amount of text. Many of our well-written large cities have a lot of text and we still think they're quite useable and readable -- Chicago comes to mind, Buffalo#Eat has many well-developed bullet point listings that would be similar to what we have in mind (I think). I also wonder if we're focusing too much on difficult regions that are going to be hard to work with regardless of the region template used. Maybe developing examples that fit within our normal 5-9 cities / subregions would help better refine the idea further. -Shaundd (talk) 15:13, 3 February 2015 (UTC)

I agree with Shaundd that perhaps Ryan focus is difficult regions. Given that majority of regions actually falls within the parameters, I think the proposed region list is easily achievable. --Saqib (talk) 09:41, 25 February 2015 (UTC)

Yelp

Hi, Ryan. I use Yelp for certain purposes (I've had good luck with restaurant recommendations in California, but they suck in New York City, and maybe especially my neighborhood), but I would oppose allowing links to them, because of their history of corruption. They have tried to force businesses to pay for better reviews, beyond the listings that are described by Yelp as "sponsored." I don't know whether there have been any similar problems at Tripadvisor, and I hope not. Ikan Kekek (talk) 23:13, 15 November 2014 (UTC)

Fair enough - I put Yelp in there solely as an example of other user-generated review sites that people might be interesting in seeing included, but recognize that they've done some evil things so I'll pull that out to avoid distracting from the larger point about listings. I'm sure that there are other options for addressing the problem of listing data being insufficient for making informed hotel/restaurant/etc choices, but my solution during the recent trip was to use Tripadvisor, so this essay focused on that particular solution. Hopefully it's a problem others will be interested in, and that others might have better solutions for. -- Ryan • (talk) • 23:28, 15 November 2014 (UTC)

Go next

A silly comment. Does it really make sense to have "Go next" section in region articles. For instance. Pakistan have six region articles on Wikivoyage and we'll be mentioning Punjab in Sindh and Sindh in Punjab article. --Saqib (talk) 06:32, 15 December 2014 (UTC)

Maybe the section could be optional, but in a lot of cases, it definitely does make sense. Consider the US, with 50 states. If you're visiting Massachusetts, you'll want to know that the neighboring states are New York, Vermont, New Hampshire, Connecticut and Rhode Island, and one-liner listings for each of those states can give you some ideas of why you might want to go there rather than another state next, especially if you're driving. Canada, with fewer provinces than the US has states, is also a fine example, as it makes sense to give descriptions of neighboring provinces in "Go next." Ikan Kekek (talk) 06:36, 15 December 2014 (UTC)

In my opinion, country region list and map work great and more helpful. But thats just a opinion, I am not raising any opposition against the section. --Saqib (talk) 06:42, 15 December 2014 (UTC)

Considering, for example, that New York State is in the Mid-Atlantic region along with only two bordering states, New Jersey and Pennsylvania, I'd like to know what you think is more useful to the traveller than to have 1-liner listings for the other bordering states of Connecticut, Massachusetts and Vermont and the neighboring Canadian provinces of Ontario and Quebec. Ikan Kekek (talk) 15:51, 15 December 2014 (UTC)

Quality of listings

One issue that's awkward with "quality of listings"... how to spot the good venue gone bad in a timely fashion. For instance, the Radiator Springs research trip passes through Clinton (Oklahoma) in 2001 and the Pixar folks meet a "Doc" Mason, an old country veterinarian who ran a local Best Western at the time and donated the land for a Route 66 museum. Swell guy, looks like a nice place... but fast-forward ahead a few years. "Doc" sells the hotel in 2003 and dies in 2007 after a long and forgettable battle with Alzheimer's. Pixar acknowledges "Doc" and his advice on the "mother road" in the credits in their 2006 "Cars" film. The hotel, meanwhile, is going downhill quickly under new ownership and loses its Best Western membership, its restaurant sits closed, its pool empty or full of weeds, its plumbing backing up and the rooms not properly maintained. How prepared is Wikivoyage to spot that sort of slow, subtle decline if it doesn't actually merit a venue being flagged as CLOSED but makes what was once a good property no longer worth listing? Usually the locals will be the last to know, as they don't stay in the hotels. K7L (talk) 04:36, 7 January 2015 (UTC)

My experience in South Africa was that I was seeing a lot of entries on Wikivoyage that, like your Radiator Springs example, may have been added about ten years ago, but I had no way to distinguish stale listings from fresh ones, hence the suggestion to add a way to specify "last updated" to our listings. Also, as noted, I simply don't think our brief listing format is enough to choose among similar offerings - I suggest we expand our listings to include more links to a pre-determined set of other sources (like Tripadvisor and Wikipedia), thus making Wikivoyage the source I would have used to narrow things down to a subset that I would then research further, but I also recognize that's a controversial solution for many people, so any alternative suggestions would be welcome.

Are there other suggestions for addressing your concern about outdated info? If Wikivoyage had the number of editors that visit Wikipedia then our listings would likely be kept up-to-date by an armada of active editors, but absent a hundred thousand new contributors I don't have any good suggestions other than what I mentioned previously. -- Ryan • (talk) • 04:48, 7 January 2015 (UTC)

Great work so far and you're spot on that it would be nice to have lots more readers (and, consequently, many more casual contributors). How do you plan to get them? --180.191.104.64 05:14, 7 January 2015 (UTC)

One way is that some people post at least rates as of a particular year or month/year (e.g., "as of December 2014"). But the fact is, as you said, we just can't be sure. Ikan Kekek (talk) 07:25, 7 January 2015 (UTC)

Would it be helpful to have something like a "Have you visited this place recently?" button, or some other way of more directly soliciting feedback? It could lead to a "please update this page" message, or maybe add buttons for marking necessary updates (e.g., to flag something as being closed or bad), or to flag an entry as being current/correct (perhaps by updating a hidden date parameter in the listing template). WhatamIdoing (talk) 07:00, 3 February 2015 (UTC)

I suggested something similar at User:Wrh2/Why Wikivoyage was not my primary travel planning tool#Quality of listings. It would be very easy to add something like a "last updated" field to the listing template, and to have the listing editor automatically update that field. Similar capabilities, like allowing readers to flag a listing as closed with a single button click, would probably also not be too difficult, although we run into problems if we start trying to gather too much information and introduce complexity into the underlying data. I'm also fearful of the process of getting everyone to agree on what (if anything) they wanted to see implemented... -- Ryan • (talk) • 07:49, 3 February 2015 (UTC)

That's the note that made me think about it. How much of this can be done in the template (like the date), and how much requires more serious engineering effort? WhatamIdoing (talk) 23:31, 3 February 2015 (UTC)

I think anything that couldn't be implemented via a combination of template code and Javascript is probably out of scope. The above items are relatively simple to implement, although as mentioned the more data we try to capture, the more complex things become. As an example, if we wanted to store data about which users clicked a "closed" link, when they clicked that link, etc, we start to run into issues of having to store template data in a structured format (something like "user1:date1|user2:date2"), we would have to deal with bad data, etc. If however we're just storing a "last updated" date, that's a single template field and something that can be easily integrated into the Javascript of the listing editor. -- Ryan • (talk) • 23:57, 3 February 2015 (UTC)

I strongly support something like what you're talking about, but it wouldn't necessarily require updating. It should be an "accurate as of (date)" field. Ikan Kekek (talk) 07:59, 4 February 2015 (UTC)

Who do we know that can make these changes? Adding a date field to the template should be pretty easy, so perhaps we should start there. WhatamIdoing (talk) 16:53, 4 February 2015 (UTC)

Anyone with Javascript abilities can make these changes - I'm happy to do it if there is agreement, although I expect agreement will be hard to come by. -- Ryan • (talk) • 17:09, 4 February 2015 (UTC)

The drawback to dating listings is that it could serve to erroneously imply that we are more out of date than we actually are. Powers ^(talk) 02:25, 6 February 2015 (UTC)

I'm concerned that only listing "accurate as of (date)" would not encourage people to update it.

Powers, what do you think about recording this information (so that we know how old something is, without needing to crawl through the article history), but keeping it hidden? WhatamIdoing (talk) 00:57, 9 February 2015 (UTC)

I would say that it would be great to have a 'last edited' value that only appears in the listing editor (and the wiki markup). Also to have the listing editor automatically update the value every save. This would at least start to allow us to curate the listings and potentially report on them. Andrewssi2 (talk) 01:12, 9 February 2015 (UTC)

That solution would mitigate the issue of appearing out of date. Powers ^(talk) 02:04, 9 February 2015 (UTC)

Then let's start there.

Separately, if we have ideas that might be worth keeping or refining, especially if they might require a more significant engineering effort, then let's list them in Phabricator:, where both staff and volunteer devs have a chance to find them. I'm happy to add them, if someone else makes a list of which ideas are probably worth tracking (or any of you can, if you want to). WhatamIdoing (talk) 16:47, 10 February 2015 (UTC)

The problem with hiding "last edited" or "last verified" or whatever the field might be, is that it leaves the original problem unresolved - when I tried to use Wikivoyage guides there was nothing to indicate if a listing was showing current information or if it was ten years old and thus likely outdated. It might be useful to editors to be able to check a "last updated" date on listings, and we might be able to develop a culture in which editors ensured that we never had a listing older than X years, but in the interim it still leaves readers without any indication whether they are reading current information or something that was written in 2005 and hasn't been updated since. If the concern is with having "last edited/verified" lines next to listings that make our listings look more out-of-date than they are, can anyone suggest other ways to solve the problem of conveying to readers whether listings are up-to-date or not? -- Ryan • (talk) • 00:14, 11 February 2015 (UTC)

It doesn't solve the problem, but it might put us in a position to be able to solve the problem down the road. Seeing "Last updated last week" on 0.001% of listings isn't going to help readers much, either, and that's what we would get if we turned it on today. WhatamIdoing (talk) 00:51, 13 February 2015 (UTC)

The only thing I could think of would be to set a threshold age at which some sort of icon appears indicating a stale listing. The problem is that that age could vary widely depending on the destination and/or listing. I mean, the Louvre or the Statue of Liberty doesn't need regular updating, nor do the attractions at Walt Disney World. But small eateries and shops, especially in small towns, come and go regularly. Powers ^(talk) 02:13, 13 February 2015 (UTC)

In which case, we need to start collecting current dates, and can make it start displaying dates (maybe) next year. (It's my experience that small-town eateries last for decades. It's small eateries in cities that seem to come and go. I've lost track of the number of times the coffee shop near me has closed, and then be re-opened as yet another coffee shop.) WhatamIdoing (talk) 23:33, 13 February 2015 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘

What about the following:

1. Update the listing template to support a "lastEdited" date.
2. Update the listing editor to automatically update "lastEdited" when a listing is updated.
3. Where we have a "lastEdited" value we display it with the listing, and where we don't we display nothing - see a proposed UI at User:Wrh2/Why Wikivoyage was not my primary travel planning tool#Quality of listings.

To address Powers' concern about listings that don't change much, since we would only show this info for listings updated after the functionality is implemented it will only show up for new listings or listings updated after this change is implemented (added per Ikan). If, ten years from now, we have listings that have a "lastEdited" date ten years in the past, hopefully travelers will understand that a place like the Statue of Liberty is still there, but they might expect that admission prices have changed in the intervening ten years. Meanwhile, they might call ahead before visiting a restaurant whose listing hasn't been updated in ten years. Would that be an acceptable compromise? If so I can raise this discussion in the pub so it gets a wider audience. -- Ryan • (talk) • 23:48, 13 February 2015 (UTC)

I think that's a great idea! And the thing that makes it great it "we would only show this info for listings updated after the functionality is implemented." However, I would hope it would show up for updated existing listings, as well as new listings. Would it? Ikan Kekek (talk) 00:32, 14 February 2015 (UTC)

I haven't actually tried to implement it yet, so there might be unforeseen issues once I make an attempt, but what I'm envisioning is that any listing with a "lastEdited" value will display the date value in the listing. Thus if someone adds a new listing or updates an existing listing using the listing editor, that value would be populated and thus would display. For those who edit by modifying the listing syntax directly (i.e. without the listing editor) I'll look into whether we can use a simple, standard date format that to allow for easy manual updates. -- Ryan • (talk) • 01:05, 14 February 2015 (UTC)

I have concerns about visual clutter if we have a long set of relatively bare listings, all with last-updated dates, but it's hard to know for sure without seeing a lot of examples. And I still have concerns about publicizing just how out of date we can get at times. Yes, it's useful information taken in isolation, but in aggregate, it seems like we risk turning readers of 2020 off once they see lots of 2015s scattered all across the site. Powers ^(talk) 03:14, 14 February 2015 (UTC)

Pub discussion started at Wikivoyage:Travellers' pub#"Last edited" date for listings. -- Ryan • (talk) • 18:14, 14 February 2015 (UTC)

Some of these issues have been addressed

And I agree with some of your other conclusions (namely many region articles being among our worst embarrassments...) But what would you further suggest to make WV better usable a primary travel planning tool? Hobbitschuster (talk) 17:50, 17 August 2015 (UTC)

Assuming this question was directed to me, there are obviously lots of ways to improve Wikivoyage, and efforts such as Wikivoyage:Search Expedition are addressing important issues. This essay wasn't meant to capture everything that could be improved, it was merely meant to highlight three fundamental areas that aren't addressed elsewhere that were pain points for me in using Wikivoyage for travel planning. -- Ryan • (talk) • 19:48, 17 August 2015 (UTC)

Yes it was addressed to you and seeing that this essay appears to have shaped the development of this site quite a bit (whether consciously so or not), I am guessing that you could offer insight as to which points were addressed (the "last edited" thing for listings for one) and which ones are still not even close to covered... Hobbitschuster (talk) 19:59, 17 August 2015 (UTC)

Your latest edits

I approve of them. I think that the several-sentence formula, as long as it's accompanied by pithy writing and not an excuse for messy verbosity, is far superior to one-liner listings. Ikan Kekek (talk) 05:20, 24 October 2015 (UTC)

Thanks for the feedback. I've been doing some work trying to get Central Coast (California) (and some of the child regions) into better shape, and hopefully once that gets a bit further along it may be a useful example to use for further discussion. -- Ryan • (talk) • 05:25, 24 October 2015 (UTC)

I like the edits as well. It's something I've been wanting to do but haven't had as much time as I want to implement. Between the longer descriptions and bolding of the destination names, I think it makes the region page more useful while mitigating the potential "wall of text" that can come with longer descritpions. -Shaundd (talk) 22:02, 24 October 2015 (UTC)

Sample for discussion

Here is an example region article for discussion. This slightly different approach to regions does the following:

• Treat the region article as a navigational article moreso than an informational article, since we primarily use regions to organize the geographic hierarchy. The goal of the region article would change such that we would be focusing on helping users to select child cities and regions, versus today where our region articles focus on using sections like "See" and "Do" to give the user an overview of the region as a whole.
• Make the region article easier to "complete" by emphasizing the "Regions" and "Cities" sections. Our many skeleton region articles are a testament to the problems of creating a good region article, and I think that is primarily due to the fact that the current template emphasizes headings like "See", "Do" and "Eat" that require detailed knowledge of an entire region. The new approach simply requires an editor to do enough research to summarize the highlights of each child city/region in 2-5 sentences, and to then add a bit of transportation information to fill out "Get in" and "Get around".
• As noted above, the new approach makes headings like "See" or "Eat" optional so that we are no longer requiring comprehensive overviews of the entire region. Optional headings can be added when desired, such as a "Drink" heading for a wine region, but in the vast majority of cases we would leave them out and focus on simply describing the child cities/regions.

Thoughts and suggestions appreciated. I think the "Understand" section would make more sense as the first section, but left it in its current position pending feedback from others. If initial feedback for this new approach is positive I'll solicit wider feedback when I've got the time & energy available to try to shepherd what will likely be a lengthy and contentious debate. -- Ryan • (talk) • 02:33, 27 October 2015 (UTC)

I'm not sure how I feel about making some of those sections optional. In the example region, wouldn't it make sense to mention wine under "Drink" and locally-grown crops like artichokes and strawberries under "Eat"? But in some region articles, it might make sense to omit those sections, so I see your point. Ikan Kekek (talk) 04:12, 27 October 2015 (UTC)

Not having any eat or drink advice is weird... If I take some of the regions I have more intimate knowledge of, a couple of sentences on eat, drink or sleep may often be helpful. In Rio San Juan Region for example it is of value to our readers to mention the fact that almost all accommodation is rather basic, with a warm shower being a distinguishing mark of the more expensive hotels. In Franconia we would be amiss not to mention the beer and wine culture that are regionally dominant. And of course Franconian cuisine is rather different from Bavarian cuisine and delicious in its own right. I do agree, that those sections don't have to be mentioned (or detailed) in all cases though... The thing is: Just because a given travel-relevant area is treated by WV logic as a country does not mean there is more to say about it than about another area that is a region or a city according to WV logic... But I do like your more than one liner listings... Hobbitschuster (talk) 19:37, 27 October 2015 (UTC)

My thought in making the "See", "Do", "Buy", "Eat" and "Drink" headings optional is that the current region article template suggests using them to provide a broad overview of a region's cuisine/activities/sights, and a scan of our region articles shows that in most cases it doesn't really work. Yes, you can probably write a great "Drink" section for Napa Valley, but what about for Contra Costa County? In the example of Central Coast (California), you could put a few lines in the "Eat" section on how you would eat seafood along the coast and fresh produce in agricultural areas, but that's fairly obvious and not really of much value to travelers - the valuable content on the region's cuisine is in the articles further down the hierarchy. For other regions it's even worse - what would you say about food in Desert (California) or Gold Country?

In the majority of cases we use region articles solely as a way to organize the geographical hierarchy, and I think focusing foremost on the fact that a region article is a way to guide the user through that hierarchy would be a big improvement to the site. If some of those regions merit a "See", "Do", "Eat", etc section then it should definitely be added, but I think we do a disservice to readers when we put those sections in the default template and then see them lie empty in the majority of region articles as no one is able to come up with any useful content to fill them with. -- Ryan • (talk) • 02:51, 30 October 2015 (UTC)

I see your points. Which sections do you think should be standard for region articles? "Get in", "Get around", "Cities", "Go next"? Anything else? Ikan Kekek (talk) 03:20, 30 October 2015 (UTC)

User:Wrh2/Why Wikivoyage was not my primary travel planning tool#Regions is the template that I would propose (with this article being a sample implementation for discussion purposes), but obviously a change of this magnitude would require broad support, so if others feel strongly that "Eat", "Drink" or some other heading should be included then that would definitely merit further discussion. -- Ryan • (talk) • 03:28, 30 October 2015 (UTC)

What should go in "Understand" in this structure? Ikan Kekek (talk) 03:36, 30 October 2015 (UTC)

The "Understand" section would contain the same content that it does today. See the Central Coast article for a good example. -- Ryan • (talk) • 14:59, 30 October 2015 (UTC)

I think that I like this. The focus on picking a city is helpful.

In the particular example, I think that ==Eat== could support a paragraph about produce and seafood (and the fishing industry, which seems to have been overlooked), but having read about that in ==Understand==, I'm inclined to remove it. ==Drink== could include a paragraph about the wine appellations (there are lots), but again, that's already covered in ==Understand==, so it might be a bit pointless.

This particular example may be unusually convenient for talking about food. I don't think there's anything to be said about food and drink in any of the regions of states like Kansas or Nebraska that wouldn't be a mere copy-and-paste of what's in the state-level article. For that matter, I'm not sure what you would say in the state-level articles that wouldn't appear in the super-region article for the entire Great Plains: "Dinner means meat and potatoes. Meat is usually beef or pork. Drink is American beer, national brands of soda, or coffee, usually mediocre. Vegans are warned that options may be limited." I therefore support making these sections "optional" (not "deprecated"; just not always required). WhatamIdoing (talk) 15:04, 1 November 2015 (UTC)

We should keep in mind though, that the US are unusually huge in terms of largely homogeneous regions if you compare that to places in e.g. Europe, you will have varying - sometimes dramatically so - things to be said about eating and drinking (and other things) in the course of a few hundred kilometers. Sure all the Great Plain states are the same (at least that's what I take from your comments) but we'd be silly to omit an eat section in say Bavaria - and even some of its subsections differ wildly in eat or drink... in Bamberg and the rest of Upper Franconia for example beer is the drink of choice, whereas Lower Franconia is more known for wine... So maybe the US is a bad place to base what region articles "ought to look" off... That being said, I do see the rationale for sometimes excluding eat and drink if there is just nothing specific to the region to say about it... Especially if said region is Eastern County X Hobbitschuster (talk) 15:23, 1 November 2015 (UTC)

I think we agree. If it's "required", then it's "required, even when it's silly to include it". But if it's technically "optional" (even "optional, but encouraged"), then it can be present in the majority of articles (where it makes sense) and omitted in the minority (where it doesn't make sense). WhatamIdoing (talk) 16:47, 2 November 2015 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ I liked the proposal of Ryan so I'll support for it. --Saqib (talk) 15:01, 7 November 2015 (UTC)

Are further changes/clarifications needed?

I think the current proposal for changing how region articles are structured is more-or-less ready to open to a wider audience, but before doing so are there any remaining concerns? The specific changes being proposed:

1. Encourage adding 2-5 sentence descriptions to child regions, cities, and other destinations in the region article, changing the focus of the region from "region as a whole" and instead focusing on the region as a way to guide users down the geographic hierarchy. This change would likely require modifying Wikivoyage:One-liner listings.
2. Make the "See", "Do", "Buy", "Eat", "Drink" and "Stay safe" sections of the region template optional. If there is content to add, by all means include the section and write as much as desired, but these sections would no longer be included in the default template and would not be required for all region articles. This change would require modifying Wikivoyage:Region article template and Wikivoyage:Article templates/Sections.

If there are outstanding concerns with these proposals it would be good to address them prior to trying to shepherd proposed changes through the Wikivoyage consensus-building process, so any further feedback is appreciated. -- Ryan • (talk) • 20:00, 8 November 2015 (UTC)

I strongly support moving this forward. I find the current brief descriptions, even the good ones, don't convey much of the flavour of the sub-region/city/other destination.

I don't know if it will help much, but I've converted the Thompson-Okanagan region page based on my interpretation of this style. I also expanded the city and other destination descriptions in the Okanagan region article to 2-4 sentences, which I thought may be a good check to see if a "wall of text" happens in regions with many destinations (the region has 12-13 cities/towns). Some of it is still unpolished, but I thought it might be good to have a second sample as everyone has their own style so it might highlight some of the strengths and/or weaknesses of the proposed changes. Please feel free to comment/critique. One observation I have (based on my experience) is I don't think it will make writing the region pages much easier -- those longer descriptions require more knowledge of the destination and I often found it difficult to keep it brief, snappy and informative. That said, I think it's worth the effort because I find it more useful.

My only other comment is how would the new region style be evaluated according to the region article status guidelines? The last line currently reads, "The most prominent attractions are identified with directions", which we wouldn't be doing in the new format. Should this be amended to something like "The most prominent attractions are identified with the city or subregion they are located in"? -Shaundd (talk) 05:36, 9 November 2015 (UTC)

The Thompson-Okanagan example looks good to me, but I'm obviously biased; I'll make sure it is cited as an example in addition to Central Coast (California). As to whether or not this change makes writing region articles easier, I actually found it much easier to write a blurb about a city or child region than to populate the existing templates since I could just reference the child article and summarize its highlights, but that experience may differ for different destinations.

Regarding status guidelines, that's a good catch, and your suggested text works for me. I'll be sure update the proposal to note that the region article status will also need to be tweaked. -- Ryan • (talk) • 06:09, 9 November 2015 (UTC)

Region vs. country articles

I've just been having a discussion about 1-liner listings in the United States of America article. Some of the region descriptions are 2 sentences long, but I think that in an article about a whole big country, as opposed to a smaller region, some brevity is called for. Would you agree? Ikan Kekek (talk) 02:30, 7 November 2015 (UTC)

References: History of USA article since User:Mathew105601's edits and my reversion, User talk:Ikan Kekek#USA. Ikan Kekek (talk) 02:33, 7 November 2015 (UTC)

I saw that discussion, and despite having edited here for more than ten years that was the first time I realized we had a policy that mandates less than one sentence per listing for city and sub-region listings. I agree that some brevity is called for, but I think the current "no periods allowed" policy leads to descriptions that are far less useful than 2-5 sentences would be. For an example, in Turkey#Regions I think a description like "High and mountainous eastern part with harsh winters" is barely better than nothing, and not very useful in helping to plan a trip. That said, I'd be happy to leave country articles alone for now if doing so made it easier to make some progress on improving how we write lower-level region articles. -- Ryan • (talk) • 03:00, 7 November 2015 (UTC)

I agree that limiting descriptions to a single sentence without a period is a little excessive. Ikan Kekek (talk) 20:25, 7 November 2015 (UTC)

Please see Wikivoyage talk:One-liner listings#Relaxing the rules a bit. As a result of that discussion, I made this change that makes clear that subregion/district descriptions are not subject to the usual one-liner restrictions. Powers ^(talk) 01:38, 8 November 2015 (UTC)

Thanks a lot for the links! That discussion actually happened before I would have been aware of policy discussions I hadn't taken part in or asked someone about (I may have still been an IP user at that time). So basically, what that means is that if you use a template for regions, you don't have to use 1-liner listings, but it doesn't make clear what the standard is for length. Is it that 2-3 sentences you advanced in the discussion you linked? I didn't see a clear consensus for that, but instead, a preference for short phrases but a willingness to give you latitude to try something else. For the record, my position would be that we don't want to use more verbiage than reasonably necessary in sections like "Regions" where a summary is called for, but if the region requires as many as 3 tight sentences, so be it. Ikan Kekek (talk) 02:04, 8 November 2015 (UTC)