Good point @Johnny Carlsen. If we try to set clearer, stricter rules and they violated them, the banning would be more justifiable.
I agree with all of Johnny’s statements. That single point of contact is perhaps the most important point.
Hi all, and sorry for my absence. Cmoffroad reached out to me in a message, and I think it’s fair that I share the reasons for my inactivity with you all as well as my thoughts on the issue.
It wouldn’t be fair to blame Grab entirely for my disappearance, as there were other factors, though their corporate-sponsored editing certainly played a significant part. I was mostly busy with other IRL stuff when the mass corporate-sponsored edits began. Though I have since periodically had more free time, whenever I’d pop back in it seemed I’d only meet frustration over the sheer amount of low-quality edits now present. (See some of the changeset comments I’ve left for example.) Not all of them were Grab’s fault, but the majority surely seemed to be. Though in fairness, the frustration goes way back, to the Maps.me debacle and all that—it was quite inevitable that regression towards mediocrity would occur as OSM expanded after all. It didn’t help that the community had also practically broken down over the issue in the intervening period, despite many members’ best efforts at holding things together.
The frustration was mostly why I couldn’t maintain personal interest in the project. But to be honest, I’m not sure that I would, even without the mass editing happening. I’ve always had an agnostic view on the long-term prospects of OSM, at least in Thailand, and contributed purely as a hobby without really expecting that it would ever become complete enough for practical day-to-day use. It was just something I could enjoy doing. And personal interests do change, with or without external factors.
In any case, this distance has allowed me to look at things from a wider perspective.
I seldom use Grab (not for any related reasons; I just find hailing taxis the old-fashioned way to be more convenient than navigating their app), but recently I took one and noticed that the app’s navigation interface, which driver was using, was using OSM, and that had me thinking.
I did a bit of experimenting with OSM-based satnavs back in the early days, but I never trusted OSM enough to rely on unless it’s a route I’d personally mapped or checked entirely (which means I already knew it and didn’t need help navigating anyway). Seeing how Grab was now actually using OSM was a bit of a revelation. As hobbyists, we prided ourselves on the detail and accuracy by which our local areas were mapped (I always felt a little smug whenever I’d find an error in Google Maps that wasn’t present in OSM). But as a community, we were always too small for our coverage to come nearly close to being able to compete with the commercial providers. Grab’s mass editing may have introduced a huge amount of low-quality edits that irked us long-time contributors, but they’re also making it possible for the map to be put to practical use in real life, at least for certain uses. Their drivers are relying on it, so they do have at least some level of skin in the game, even if their bar of quality is way lower than we’d prefer. While I wish that Grab had forked the project and maintained their own thing (which would also have made a comparison of the results possible), I am wary of judging that their involvement has as a whole been a net negative to the state of OSM in Thailand as a whole, even if it clearly has been to the community. In any case, I do kind of wonder if it isn’t too late now to try turning things back.
I no longer feel informed enough on the situation, though, so I’ll refrain from voting.
@Johnny, @Beddhist, @Paul_012 great feedback, thank you for adding your voice again to the forum. much appreciated!
Some of you may be surprised but I will vote for (very) strict rules (A).
The community is obviously too small to review and monitor the quality of large-scale modifications done by remote organizations.
However, I believe that any genuine organizations wanting to provide quality contributions that benefit them on a small to medium scale, should be able to do so if their proposal meets the satisfaction of the local community (case by case basis).
For example, Facebook should be more than welcome to correct the names/phone numbers of businesses if they want to.
Grab should be able to improve road alignments with satellite imagery, and street names/directions with Mapillary.
However, since the road classification network is the greatest source of frustration and conflicts among the majority of voters here (B), and because most of the road network in urban areas has already been mapped, I will strongly advocate against any extension of the road network and modifications of its road classifications by remote companies and individuals who cannot rely on local knowledge or ground surveys.
Exactly that’s the most important point.
That was always my objective when I contributed here: I want to be able to use the map while I travel thru the world, mostly on bike (I gave up for car navigation in Europe - Google Maps is far better here for that purpose; the reason for the lacking usability of OSM is ideology: what’s a useful “routing speed”? “Subjective! F### off!”).
Hence, I do extended armchair mapping sessions typically before starting the holidays, and then some further contributions with the collected “ground truth” (more than 10,000 km of “identifiable” GPS traces, mainly on minor roads!).
With my background knowledge, I can estimate the risk of following facebook roads (but not yet grab roads - they are not yet tagged with “import”).
Never could I have achieved the mapping of so many missing roads in South Thailand, but thanks to Facebook they are on the map. With bad tagging, but that is just normal with other contributors here, too. When I travelled from Phatthalung to Nakhon Si Thammarat some 10 years ago, even major highway 403 was missing (but part of it was mapped as a railway…). And then west to Khao Phanom: I used the as-the-crow-flies routing, because nothing was available at all for almost 100 km.
Thank you everyone for voting and giving extremely valuable feedback.
To be honest, I didn’t expect so much participation and this shows that corporate vandalism is the most pressing issue in this community.
While we haven’t heard yet from a few more correspondents, I think there is enough data here to come up with conclusions:
-
First, there is a clear majority among the top and most active contributors that mass-editing corporate vandalism will not be tolerated anymore,
-
Second, the work of organizations/companies can still be beneficial to the community and a complete ban is not a sustainable solution.
-
Third, strict rules for organized edited teams and consequences for not respecting them must be decided by the local community.
-
Fourth, organizations must have a single point of contact who must proactively contribute to OSM, participate in discussions and respond to incidents.
So here are my 3 proposals:
A) Grab should suspend immediately all their mapping campaigns in Thailand until new rules for organized editing teams are being set up.
While the lack of rules is the fault of the community, Grab also bear some responsibility because it has in the last 4 years:
- repeatedly ignored feedback from the community and gave empty promises
- failed to provide any tangible solutions and guarantees that concrete actions have been taken
So this should be regarded as a symbolic ban, and I expect Grab to respect the voices of the community.
Should they violate this, I would involve the Data Working Group (DWG) with a strong case I have built over the last weeks.
B) Stricter rules need to be setup for organized editing teams
I propose that whoever voted (A): Mishari, nitinatsangsit, myself and any new volunteers, and with the input of Grab, work together after Songkran to come up with a first draft of stricter rules for organized editing teams, so that eventually Grab could rejoin under new rules.
C) The community needs a new setup
The community will never be able to function properly without clear written rules on how to:
- vote on issues and proposals
- respond to incidents
- work with organized editing teams
- effectively moderate the overall process
After Songkran, I will submit a full proposal to elect on a yearly basis 3 community moderators, a new role that will define and improve the community charter, and proactively moderate it.
Please let me know what you think of these proposals,
Until then, Happy Songkran!
I look forward to reading your proposal, and where I commend your efforts, I’m concerned about the stakeholders included in the process, that many will be excluded from decisions that affect them.
I don’t think “most active contributors” is sufficient as there is plenty of room for people of other roles as well, including evangelists. I have a modest amount of experience contributing and where this may give me some credibility, I don’t think it gives me any rights, moral or otherwise, over this project.
In order to save time, I recommend you put up broad principles in the Wiki for the community to comment on before hammering out the details of the proposal.
I also suggest you should also mention you’re doing this in osm-talk so we can gather feedback from the wider community as to how they organize things.
The community moderator role as you mentioned sounds interesting and I look forward to seeing your proposal. Do any of the other communities have such a role?
Lastly, as of time of 11 April, there are 8 votes in favor of (A) in the OSM Facebook poll, 2 for (B). Where I don’t think this changes your conclusions in a material way, it should be noted that even the Thai community is somewhat culturally split, and any effort needs to pay extra attention to inclusiveness.
I have just found the email. And to answer your PM, @cmoffroad: yes, remote-editing corporate vandals overwriting my on-ground survey works (Grab specifically in this case) as exactly the reason that I simply stopped participating in OSM entirely. I still have a centimeter-thick worth of field papers, that I never got a chance to enter the data, rotting in my backpack to remind me of it.
It’s probably out of voting time now; but I’m going to let everyone know that I only favor option (B). My Internet time is running out today, see you all later.
I vote B, Before that, I had seen that someone had changed the Ratchaprapha Dam Reservoir to a Commercial Area. I think should be limited to editing with a lack of maturity.
For a stricter rules, as an example, I suggest:
- Re-alignment of existing roads is not permitted unless it is clear that the existing alignment is incorrect, and the issue must be discussed with the community before proceeding.
- Adding a new road is allowed only for a small dead-end road. Connecting the existing road may have an impact on navigation and must be discussed with the community before implementation.
- Changing the highway classification is not permitted unless approved by the community.
- Adding a new street name is permitted, but changing an existing street name must be approved by the community.
Hello, I’m Jinal, a Program Manager at Grab since 2019 and an OpenStreetMap contributor since 2015. After reading all the comments and responses from everyone who participated in this thread, here are a few things that I would like to share with you all on behalf of myself and the team:
-
OpenStreetMap was new to us when we started mapping in 2018, we acknowledge that there were a lot of things that we did not know back then.
-
But in the last few years, all we have tried to do is work closely with all of you and find better ways to collaborate.
-
We have appreciated and valued your feedback and together, found ways we can improve the overall communication. Here are a few outcomes of the conversations:
- After a brief pause on the edits from our end, we kickstarted the conversation with some of the active community members on ways we can make our mapping better and re-review the data that was previously contributed, that’s when we came up with the [github master ticket](https://github.com/GRABOSM/Grab-Data/issues/49#issuecomment-691025559), where we shared about our objective of mapping and if we had any doubts we would open them up for discussion- This was prior to making OSM Forum the go-to place for sharing new campaigns
- This helped us get resolution and contribute to good quality data
- This has been something that has helped us collaborate with other community members across South East Asia too
- These were all the issues that we sought help from the community and the community members gave us advice and the best approach to handle the issues
I’m sharing this to let folks know that we have always tried our best to communicate with you all and find ways to do better. We have the same objective of contributing quality data, and while not yet perfect, I do believe things have significantly improved over the past few years.
We do acknowledge that there was an oversight on our end when it came to transitioning from Github to OSM Forum and this was because the individual who was handling these communications moved out and this was lost in transition. But once aware of the issues, we immediately modified our processes to ensure we are following the proposed guidelines. Specifically:
- Following up on the thread, we had open conversations and communication with everyone involved to understand and act on the feedback
- The result of this discussion was that we came up with a tag - ‘import=grabremote’ to enable people to differentiate between the edits made by our team and others
- We are still finalizing some nuances around when to use the tag and so on
I also want to address concerns raised about inactive accounts. Please allow me to share some context on how we’ve addressed this:
- We do ensure once the user is inactive, their accounts are updated with a message stating that this user account is not active and one can reach out to the email address in case there’s a query (example)
- Through the post, we realized that it’s not the best solution, as folks might want to just comment on the changeset without looking into the user profile. So we were trying to find a technical solution to ensure that we receive changeset comments from accounts of users who are no more a part of the team
- It was only recently we found that some firewall setting prevented these comment emails from reaching out to us, we have rectified this at our end, and going forward we should be able to respond to all the comments and messages
- One solution was to delete the account (which is an option available on OSM), but that would make it harder for anyone to communicate with us, so that was not the best solution
- One thing to note is that an OSM user name can be tagged only to a unique email address
But after some back and forth, we found the solution and now anyone can comment on the changeset and we will ensure that we respond to it
That’s why we have continued to collaborate with Mishari - to help us better understand and contribute to OSM in this region.
I want to reiterate what I said above that we know we aren’t perfect and we don’t get things right all the time. Every region is different from the other, some things work well and some things need to be tweaked. But the value of a community like OSM is the collaboration that helps us all do better. That’s why we have continued to collaborate with Mishari - to help us better understand and contribute to OSM in this region.
We do believe we have made and intend to continue to make important contributions to OSM in Thailand, and you have our commitment that we will continue being receptive to your feedback and acting on it as quickly as we can. I’m also happy to work with you all on setting up policies that are stricter for corporations like us as I’m an OSM contributor myself.
Just wanted to share this side of the story where it’s not our intention to create bad data. We are using this data on our platform to help provide livelihoods to millions of driver and merchant partners, and this is impacted if there’s bad data. Therefore, it’s in all our best interests to have good quality data in this region.
May I also ask if you would all be open to meeting in person? We would like to hear more from you and would appreciate also the opportunity to share about what we are working on and discuss with you the issues that we all are facing. This would help us all align on our common goals and find ways to communicate better going forward. Please do share the best time to do so and also the preferred location where we can meet most of the folks.
Best,
Jinal Foflia
This so restrictive and cumbersome that you might as well impose a ban.
@Beddhist If we’re going to set rules, this is what I’m thinking about. If these rules are too stringent and impractical, I believe mapping rules like these are not a viable option. Other agreements, such as single point of contact, that have already been covered in this issue, may be the focus of our attention.
@jinalfoflia Thank you for your response, and I appreciate your willingness to talk with the community.
I believe things will improve, but I recommend waiting and seeing.
To avoid further distractions, I would suggest that:
- rules for remote organized teams can be discussed by those who voted (A) in this new thread: https://forum.openstreetmap.org/viewtopic.php?pid=860881
- general communication with Grab should be continued in this existing thread: https://forum.openstreetmap.org/viewtopic.php?id=74519
New topics can be started in new threads.
Thank you for your vote! We can close this topic.
Well, actually mentioning it to the DWG might be a start. I quick scan of tickets created in the last year finds none mentioning GRAB by name (apart from a few where users from GRAB flagged up other users suspected of vandalism).
If there’s a problem, please tell us about it, and we’ll be able to help.
In addition to the normal rules that apply to everyone else, “organised” editors must also follow https://wiki.osmfoundation.org/wiki/Organised_Editing_Guidelines . If you believe that any group is not following that, please let us know and explain why you think that.
Best Regards,
Andy Townsend, on behalf of OSM’s Data Working Group.
Hello, I’m a bit late to the party. Only noticed a message to me from Cmoffroad now. I have not read all comments in depth, but sort of agree with Johnny Carlsons comments. Initially when I started road coverage was very poor/non-existant in some areas, so I wanted to just get basic roads covered as much as possible. Now, within coverage is actually pretty good, though the quality and detail may vary a lot. I presume some of the coverage was due to help of some corporate input. Openstreetmap is an open project, so in theory anybody can contribute. The key issue is quality of the contributions, and if lots of bad edits are made, that’s not good of course. I have noticed in the country side some residential roads going through fields and dirt tracks, which I assumed some newbie got wrong. Maybe it was Grab. In the case of Grab, it’s actually in their own interest to make only high quality contributions, because what’s the point if their drivers go get it all wrong and waste their time with bad maps. So my hope is to just have better coordination, standard rule sets, etc. And coorporations should not let loose total newbies for money that have no interest in quality.
I vote for B.
Seen lot of dirt track being classified as Residential here in Nan.