Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Add diet:vegetarian to some fast food brands. #4855

Closed
wants to merge 5 commits into from
Closed

Add diet:vegetarian to some fast food brands. #4855

wants to merge 5 commits into from

Conversation

Eric-Sparks
Copy link
Contributor

I've been adding the diet:vegetarian tag to some local restaurants but, according to the brand's website, these places universally have vegetarian options.

@UKChris-osm
Copy link
Collaborator

UKChris-osm commented Jan 10, 2021

Thanks @Eric-Sparks

There seems to be a random "Tr" floating around within the "A&W (USA)" entry, which is causing a build fail, any reason for this?

@Identitaet
Copy link
Collaborator

Identitaet commented Jan 10, 2021

Overall a good idea.

I think there's one problem here which is that many of these brands have different menus world wide.

For example while lots of restaurants have lots of vegetarian options for the Indian market the German market isn't as likely to have vegetarian options.
So that's something we have to keep in mind but in general a really good idea 👍(also really usefull for me and if one shop is universally vegan we can just do that as well)

@Eric-Sparks
Copy link
Contributor Author

Thanks @Eric-Sparks

There seems to be a random "Tr" floating around within the "A&W (USA)" entry, which is causing a build fail, any reason for this?

Umm, no? Didn't think I changed anything with A&W.

@bhousel
Copy link
Member

bhousel commented Jan 11, 2021

I feel pretty weird about this -
Does diet:vegetarian=yes really just mean "vegetarian options are available"?

That would cover almost all restaurants and fast food, right?
I could see tagging exceptions like only or no as interesting, but yes should maybe be assumed?
It reminds me of #4105/#3048 (cashback) and I really dont want to upset OSM.

@Eric-Sparks
Copy link
Contributor Author

I feel pretty weird about this -
Does diet:vegetarian=yes really just mean "vegetarian options are available"?

That would cover almost all restaurants and fast food, right?

So not all restaurants and fast food have a proper vegetarian option available. Offering a salad doesn't automatically make it a diet:vegetarian=yes. Think about Chipotle where you can get a full meal like a burrito with beans, guacamole, lettuce, and tomatoes.

From the OSM wiki:

yes
full options available : provides at least a proper choice for this option, constantly available
(i.e full meal in a restaurant, specific section in shop). If in doubt, see note below.
no
doesn't provide this option at all, or not reliably
(i.e side-orders or light snack only in a pub, temporary products in a shop).
only
only provides this dietary option (N.B. it may often be helpful to tag a vegetarian place with both diet:vegetarian=only and diet:vegan=yes)

Copy link

@akadouri akadouri left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Made some suggestions to change diet:vegetarian to diet:vegan based on chains I know have vegan friendly options or quick substitutions.

As to whether this is worth it, I think so. It's definitely subjective as to whether the tag is warranted, I wouldn't add it for a place that just had french fries but I would for a burger restaurant that offers a meat-free patty. I usually stick to vegan tags since vegan-friendly menus are a bit less ubiquitous than vegetarian ones and that tag encompasses vegetarian. It helps power OpenVegeMap which is an open alternative to things like HappyCow and until recently Google Maps/Yelp/etc didn't have reliable searching options for "Vegan". There's an active telegram where people talk about this as well.

data/brands/amenity/fast_food.json Show resolved Hide resolved
data/brands/amenity/fast_food.json Outdated Show resolved Hide resolved
data/brands/amenity/fast_food.json Outdated Show resolved Hide resolved
I wouldn't substitute vegan for vegetarian due to the way some searches may work but I'd totally add it.

Co-authored-by: Ariel Kadouri <ariel@arielsartistry.com>
Copy link
Contributor Author

@Eric-Sparks Eric-Sparks left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Ugh, okay, I clearly do better when just using git in a command-line environment...

So, I wouldn't change vegetarian to vegan due to the way some searches may work but I'd totally add diet:vegan=yes to those chains that support vegan.

data/brands/amenity/fast_food.json Show resolved Hide resolved
@bhousel
Copy link
Member

bhousel commented Jan 16, 2021

Still feeling pretty unsure about this PR. For example this change:

Screen Shot 2021-01-16 at 12 28 22 PM

This means:

  • Every White Castle in the USA gets the diet:vegan=yes tag.
  • Users of iD would be prompted to "upgrade" all the White Castles to have this tag.

This might be technically correct - I don't know - but it would surely be confusing and suspicious for users. If iD prompted me to make this change, I'd think something was wrong.

A good way to think about the name-suggestion-index is that we shouldn't add any tags to this index that we wouldn't also be comfortable making to the OSM with a mass edit. I'm very ok with updating tags like :wikidata and cuisine but less ok about tags like diet: or payment:.

@akadouri
Copy link

That makes sense, rethinking about it as a mass edit I don't know if I'd still make the change unless it was a core part of the chain (the diet:x=only case). Might be something to track on the wikidata page instead and check for that way.

@Eric-Sparks
Copy link
Contributor Author

Eric-Sparks commented Jan 16, 2021 via email

@bhousel
Copy link
Member

bhousel commented Jan 20, 2021

@Eric-Sparks - I've been thinking about this more.. What if we added another property like "impliedTags": { … } where people could put tags like this? They'd still be attached to the brand, but wouldn't be added to OSM automatically.

I agree with you that it's useful information to track somewhere, my biggest concern is just that adding them all en masse to OSM would be seen badly by some mappers.

A project that wants to know which brands have vegan offering could still use this index to say "If I match White Castle, I can imply that they offer vegan food". Or you could make/use/release your own distribution of OSM that has the implied tags already merged in?

@Identitaet
Copy link
Collaborator

@Eric-Sparks - I've been thinking about this more.. What if we added another property like "impliedTags": { … } where people could put tags like this? They'd still be attached to the brand, but wouldn't be added to OSM automatically.

that might be quite usefull but in my opinion more for values which conflict with that implied tag
so if a place wasn't offering vegetarian food and now is

I agree with you that it's useful information to track somewhere, my biggest concern is just that adding them all en masse to OSM would be seen badly by some mappers.

Yup probably not a good idea

A project that wants to know which brands have vegan offering could still use this index to say "If I match White Castle, I can imply that they offer vegan food". Or you could make/use/release your own distribution of OSM that has the implied tags already merged in?

Do you mean that as a general value for each catergory or for each entry on it's own ?

@Eric-Sparks
Copy link
Contributor Author

@Eric-Sparks - I've been thinking about this more.. What if we added another property like "impliedTags": { … } where people could put tags like this? They'd still be attached to the brand, but wouldn't be added to OSM automatically.

I'm not sure why you wouldn't want to add this information automatically?

I agree with you that it's useful information to track somewhere, my biggest concern is just that adding them all en masse to OSM would be seen badly by some mappers.

Why would this be seen as bad by some mappers?

What am I missing here? Adding diet:vegetarian to a restaurant is not new to OSM and has been used more than 30,000 times around the world. If a brand is openly advertising vegetarian options then it would seem like an easy thing to do to add this to their brand which then helps the millions of people that are looking for this information on the streets. Clearly there is a problem here that I'm not understanding as this seems to be rather straight forward.

@UKChris-osm
Copy link
Collaborator

If a brand is openly advertising vegetarian options then it would seem like an easy thing to do to add this to their brand.

Generally I would agree with you as it relates to this, but there are times when a brand will not necessarily have an entire menu available at every restaurant, which is sometimes marked as "at participating restaurants" or in White Castle's case "at participating Castles".

That being said, I wouldn't have any objection to the NSI listing "diet:*" based on a brands menu, as I think that is a general indication that it's likely to be available, and I can't imagine any restaurant or franchisee would remove that type of menu selection in full from the brand menu.

@ghost
Copy link

ghost commented Jan 20, 2021

I'd imagine a source tag could go a long way towards dispelling myths about iD meddling, since then the discussion becomes "well, argue with [restaurant brand] about whether they're offering what they say they're offering". This could be stored in each index entry, and seems easier to interpret for downstreams than an impliedTags as it would just be a link to the menu or part of the menu that features a full veg[eteri]an meal.

This was referenced Mar 12, 2021
@UKChris-osm
Copy link
Collaborator

I've noticed that #5006 was merged with "diet:vegan": "only", included, so would that have any bearing on this PR being accepted, or was #5006 merged incorrectly?

@bhousel
Copy link
Member

bhousel commented Apr 4, 2021

I've noticed that #5006 was merged with "diet:vegan": "only", included, so would that have any bearing on this PR being accepted, or was #5006 merged incorrectly?

#5006 is fine - there's a big difference between adding diet:vegan=only to a vegan restaurant and adding diet:vegetarian=yes to Burger King like this PR does.

@LaoshuBaby
Copy link
Collaborator

LaoshuBaby commented Sep 12, 2021

{{vote|no}}

Is it necessary to keep this Pull Requests still open?

This pr has already been pending for 8 months, and still have something controversial such as adding diet:vegetarian=yes for even Burger King or Subway, Taco Bell

@Eric-Sparks
Copy link
Contributor Author

Eric-Sparks commented Sep 12, 2021 via email

@bhousel
Copy link
Member

bhousel commented Sep 12, 2021

Is it necessary to keep this Pull Requests still open?

This pr has already been pending for 8 months, and still have something controversial such as adding diet:vegetarian=yes for even Burger King or Subway, Taco Bell

I'm ok with this PR staying open because I think it's a reasonable request, but I think tags like this should be in an impliedTags structure (tags that aren't applied automatically to all the matched features in OSM), and I haven't build that yet. We get criticism about setting brand:wikipedia tags, so having a separate impliedTags object could also solve that problem.

@bhousel
Copy link
Member

bhousel commented Sep 12, 2021

Well, this really shouldn't have been controversial. These restaurants are advertising having vegetarian options (not necessarily vegan, by the way, as this PR morphed into).

I agree with you, but if you were to ask any of the regular OSM channels (tagging mailing list, slack, discord, etc) and propose a mass edit to diet:vegan=yes to all the White Castles in the world, I feel like people would tell you no. I've seen way less controversial suggestions get rejected or reverted.

@LaoshuBaby
Copy link
Collaborator

having a separate impliedTags object could also solve that problem.

That will be a better way.
We will waiting for that so let this keep open. Thank you!

@LaoshuBaby LaoshuBaby added the waitfor Waiting for something before we can do this label Sep 12, 2021
@bhousel bhousel marked this pull request as draft November 1, 2021 14:27
bhousel added a commit that referenced this pull request Nov 24, 2021
(until #4855 is resolved we shouldn't do this)
@UKChris-osm UKChris-osm mentioned this pull request Dec 12, 2021
@UKChris-osm UKChris-osm mentioned this pull request Jan 7, 2022
@peternewman
Copy link
Collaborator

Generally I would agree with you as it relates to this, but there are times when a brand will not necessarily have an entire menu available at every restaurant, which is sometimes marked as "at participating restaurants" or in White Castle's case "at participating Castles".

Sort of linked to this, and also takeaway=yes tagging, how do we feel about adding food=yes ( https://wiki.openstreetmap.org/wiki/Key:food ) on e.g. amenity=pub ( https://wiki.openstreetmap.org/wiki/Tag:amenity%3Dpub ) like say Wetherspoons? Which can then trigger things like StreetComplete to ask if they have various diets available at each location?

It seems a bit safer to me than assuming a diet type will be available at all pubs from a brand, and I can't imagine a pub brand wouldn't have ANY food available in some establishments (given even the airport Wetherspoons serve food). Plus a bit like the takeaway aspect tagged on other food establishments, it's an important and relevant aspect of them.

@bhousel
Copy link
Member

bhousel commented Feb 27, 2022

Sort of linked to this, and also takeaway=yes tagging, how do we feel about adding food=yes ( https://wiki.openstreetmap.org/wiki/Key:food ) on e.g. amenity=pub ( https://wiki.openstreetmap.org/wiki/Tag:amenity%3Dpub ) like say Wetherspoons? Which can then trigger things like StreetComplete to ask if they have various diets available at each location?

I kind of disagree with this.. food=yes seems like a really unhelpful tag, and could be applied to almost anything.

amenity=pub should be tag that triggers StreetComplete to ask the "is there food here" question.

(as an aside, I sort of think the takeaway=yes tag is silly too, and we probably shouldn't have started doing it)

@peternewman
Copy link
Collaborator

I kind of disagree with this.. food=yes seems like a really unhelpful tag, and could be applied to almost anything.

amenity=pub should be tag that triggers StreetComplete to ask the "is there food here" question.

Yeah, we sort of got caught up with what is food, see streetcomplete/StreetComplete#3099 and https://lists.openstreetmap.org/pipermail/tagging/2021-July/062106.html onwards...

So tagging food=yes where we know there definitely would be would work around it for some places, without having to answer the awkward question?

See also https://en.wikipedia.org/wiki/Substantial_meal !

(as an aside, I sort of think the takeaway=yes tag is silly too, and we probably shouldn't have started doing it)

How so? A posh afternoon tea place is going to be takeaway=no or maybe takeaway=not_unless_you_bring_your_own_reusable_cup if it serves everything in the finest china, so therefore takeaway=yes is useful info...

@Eric-Sparks
Copy link
Contributor Author

Can we please get back on track with the original intent of this pull request? There are people out there that have dietary restrictions that are looking for ways to find food easily. When a brand openly advertises that they offer a vegetarian, or whatever, option, I think we should add that to the brand to help those with dietary restrictions find food. These places are mostly cookie-cutter where each place is going to offer the exact same food so there is no need to have each restaurant surveyed to determine if they serve the dietary option or not.

There has been a lot of speculation as to what "people" might say if you asked on a particular communications medium. I say that if there are people against such a proposal then they aren't thinking about the users out there that would actually use the data to obtain, you know, FOOD. Of course, no one has actually asked these "people", not that it really matters. This really isn't a huge deal and I'm not sure why people are making more of this than it has to be.

@Georift
Copy link
Contributor

Georift commented May 5, 2022

Hey I was linked across to this PR as I was making a change to a local brand who when checking all regional menus would have vegetarian and vegan options.

I totally understand the point you're making here @bhousel, it's a fine line to make sure people don't discount this index. For the implied tags feature, do you imagine there will be any integration with editors to help recommend the tag be applied? Perhaps a two step approach, first apply the tags, then offer some additional tags? I understand you can't speak for every editor but just your best guess.

At least for me, the only reason it would be interesting adding these tags to this index, would be to increase the usage of these tags across the map. Where this would help OpenVegeMap really flesh out it's data without needing significant investment or understanding by the editor about the nuance of this tag and our discussion right now.

Excited to hear your thoughts 🙂

@Adamant36
Copy link
Collaborator

I'm closing this as stale and probably not mergeable with how things currently are anyway. To quote a comment from another issue, "f it's an "attribute tag" that is true for every instance of this thing in OSM, there isn't a lot of value in actually adding it to every instance of the thing in OSM, and it just makes extra work for mappers if that attribute ever changes." Feel free to re-open this if anything changes about this in the future though 👍

@Adamant36 Adamant36 closed this Dec 13, 2022
@Adamant36 Adamant36 added wontfix Not planning to work on this at this time and removed waitfor Waiting for something before we can do this labels Dec 13, 2022
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
wontfix Not planning to work on this at this time
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

9 participants