[Talk-us] Standard (mapnik) toolchain/processes: can we teach these better?
Simon Poole wrote: It has been pointed out in numerous places before, but just in case you missed it: there is an ongoing effort (since months) to remove all "catch alls" from the standard style. This implies that stuff you thought was rendered might vanish, but in fact it was just accidental that it was shown in the first place. If there is a feature that you feel strongly should be displayed with the standard style, then you should submit a feature request. Thanks to the talk-us community for entertaining my grumpiness about this, but I truly believe there is a direct connection between asking OSMers to "map well" and the visual feedback (rewards? yes, I think so) we get by doing so. Sure, it's great that beaches (and many other mapped objects, usually named polygons that describe an area, like a beach, shopping center or many other "things") can be easily found from OSM's main map via a simply-type-it-in Nominatim search: that IS good. But when we see rendered labels disappearing, even when this is explained by the reason given, it can be disheartening. I DID "miss that" news/memo about this "since months" effort. Where might I have learned this? I am (slowly, even after being an OSM volunteer for over five years) discovering there are ways to effect how our map looks (carto-issues bug reporting, the potential to enter a mapnik feature request -- where?). But I do think it would be helpful if these "assumed to be known by everybody" facts (they aren't!) were better promulgated. Either in our wiki somewhere, or with a link from the main page, or some other relatively easily findable method. I conscientiously read (and contribute to) our wiki pages, I follow talk-us, I explore code in github, I play around with rendering tools...yet about the machinations that make our map look and behave the way it does, on a day-to-day basis -- AND the changes that happen to it -- I seem to learn absolutely nothing. Until after the fact. Let's say I were to carefully consider that I DO think beaches (a polygon with tags natural=beach and name=*) should render in mapnik. What else? Polygons tagged landuse=commercial that also have a name=Shopping Center tag? (Maybe). And a hundred other potential things that used to (accidentally) render, but are now not being rendered in the interests of not rendering "catch alls." Do I enter a feature request for each and every one of them? Maybe, as that means I considered each and every one of them. But how do WE consider each and every one of them? Do we even do that? I ask sincerely. It seems many mapnik render decisions on an ongoing basis are made in a vacuum. That doesn't feel very OSM to me. In short: how might intermediate mappers like me better learn how our map is built and the processes which influence and effect changes within it? We should all have a stake in participating in these processes, should we wish to do so. That starts with better learning about them in the first place. I don't mean for it to seem like I think OSM's inner machinations are some big secret, I'm just asking for a bit of light to be shined along a path I can find this stuff out largely by myself. I'm pretty smart and resourceful, and can easily be pointed to the right places and told "Go." But I don't know a whole heck of a lot of what and where are these resources. Thanks in advance for a wide swath of pointers to get me (us) started. With my best attempt to remove any residual grumpiness, SteveA California ___ Talk-us mailing list Talk-us@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-us
Re: [Talk-us] Standard (mapnik) toolchain/processes: can we teach these better?
On Sun, May 25, 2014 at 12:46 PM, stevea wrote: > Thanks to the talk-us community for entertaining my grumpiness about this, > but I truly believe there is a direct connection between asking OSMers to > "map well" and the visual feedback (rewards? yes, I think so) we get by > doing so. Sure, it's great that beaches (and many other mapped objects, > usually named polygons that describe an area, like a beach, shopping center > or many other "things") can be easily found from OSM's main map via a > simply-type-it-in Nominatim search: that IS good. But when we see rendered > labels disappearing, even when this is explained by the reason given, it > can be disheartening. I DID "miss that" news/memo about this "since months" > effort. Where might I have learned this? > > I am (slowly, even after being an OSM volunteer for over five years) > discovering there are ways to effect how our map looks (carto-issues bug > reporting, the potential to enter a mapnik feature request -- where?). But > I do think it would be helpful if these "assumed to be known by everybody" > facts (they aren't!) were better promulgated. Either in our wiki somewhere, > or with a link from the main page, or some other relatively easily findable > method. I conscientiously read (and contribute to) our wiki pages, I > follow talk-us, I explore code in github, I play around with rendering > tools...yet about the machinations that make our map look and behave the > way it does, on a day-to-day basis -- AND the changes that happen to it -- > I seem to learn absolutely nothing. Until after the fact. > > Let's say I were to carefully consider that I DO think beaches (a polygon > with tags natural=beach and name=*) should render in mapnik. What else? > Polygons tagged landuse=commercial that also have a name=Shopping Center > tag? (Maybe). And a hundred other potential things that used to > (accidentally) render, but are now not being rendered in the interests of > not rendering "catch alls." Do I enter a feature request for each and > every one of them? Maybe, as that means I considered each and every one of > them. But how do WE consider each and every one of them? Do we even do > that? I ask sincerely. It seems many mapnik render decisions on an > ongoing basis are made in a vacuum. That doesn't feel very OSM to me. > > In short: how might intermediate mappers like me better learn how our map > is built and the processes which influence and effect changes within it? > We should all have a stake in participating in these processes, should we > wish to do so. That starts with better learning about them in the first > place. I don't mean for it to seem like I think OSM's inner machinations > are some big secret, I'm just asking for a bit of light to be shined along > a path I can find this stuff out largely by myself. > > I'm pretty smart and resourceful, and can easily be pointed to the right > places and told "Go." But I don't know a whole heck of a lot of what and > where are these resources. Thanks in advance for a wide swath of pointers > to get me (us) started. > +1 -- @osm_seattle osm_seattle.snowandsnow.us OpenStreetMap: Maps with a human touch ___ Talk-us mailing list Talk-us@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-us
Re: [Talk-us] Standard (mapnik) toolchain/processes: can we teach these better?
Am 25.05.2014 21:46, schrieb stevea: > > Thanks to the talk-us community for entertaining my grumpiness about > this, but I truly believe there is a direct connection between asking > OSMers to "map well" and the visual feedback (rewards? yes, I think so) > we get by doing so. Sure, it's great that beaches (and many other > mapped objects, usually named polygons that describe an area, like a > beach, shopping center or many other "things") can be easily found from > OSM's main map via a simply-type-it-in Nominatim search: that IS good. > But when we see rendered labels disappearing, even when this is > explained by the reason given, it can be disheartening. I believe that giving positive feedback to our contributors is important and that the "standard" map style is one of the ways we can do that. Unluckily there is just no way that "everything" that is correctly mapped can be displayed on a single map layer, there is work in progress to at least partially address the issue via a "objects in the vicinity search", but for at least for the slippy map it its current form we will have to live with the limitations of the medium. The standard style has, in the past, been very lenient it what it has rendered, for example it used to render everything that had a name. The downside of that, was that it didn't actually support "correct" tagging (whatever that is in an OSM context) and a conscious decision of what should be displayed on the map. Part of the work Andy has been doing is trying to clean that up and make the whole thing a bit saner. I DID "miss > that" news/memo about this "since months" effort. Where might I have > learned this? > > I am (slowly, even after being an OSM volunteer for over five years) > discovering there are ways to effect how our map looks (carto-issues bug > reporting, the potential to enter a mapnik feature request -- where?). > But I do think it would be helpful if these "assumed to be known by > everybody" facts (they aren't!) were better promulgated. Either in our > wiki somewhere, or with a link from the main page, or some other > relatively easily findable method. I conscientiously read (and > contribute to) our wiki pages, I follow talk-us, I explore code in > github, I play around with rendering tools...yet about the machinations > that make our map look and behave the way it does, on a day-to-day basis > -- AND the changes that happen to it -- I seem to learn absolutely > nothing. Until after the fact. > As a long time OSM contributor, actually longer than myself, you are surely aware how OSM works :-). People just "do" things. Andy in this case, while I may believe he may have covered it in his SOTM presentation, as the current main maintainer of the style, did exactly that. So I suppose the answer to your question is that thing to follow are the commits and discussions here https://github.com/gravitystorm/openstreetmap-carto I know that really cool and web 2.0ish response which doesn't help anybody who is not in a bubble of a certain kind. With a different hat on: yes it is a pet peeve of mine that we don't have a light weight way to push such information out to our contributors and, perhaps as a consequence, haven't developed a culture of actively informing before the fact. Simon signature.asc Description: OpenPGP digital signature ___ Talk-us mailing list Talk-us@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-us
Re: [Talk-us] Standard (mapnik) toolchain/processes: can we teach these better?
2014-05-26 12:07 GMT+02:00 Simon Poole :
> With a different hat on: yes it is a pet peeve of mine that we don't
> have a light weight way to push such information out to our contributors
> and, perhaps as a consequence, haven't developed a culture of actively
> informing before the fact.
>
+1
I think the german blog ("Wochennotiz") does a great job in filtering a lot
of interesting news about cartography, OSM, the mapnik style and a lot of
other things, and there is AFAIK also a translation by Pascal and Dennis of
the "main parts" (those potentially interesting an international audience)
into English (and some other languages) as well, called "Weekly OSM
Summary": https://blog.openstreetmap.org/
Not sure if the switch from rendering names as "catch all" to "selective
tags" was featured in the English version though. Anyway, a good read if
you are interested in OSM.
cheers,
Martin
___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us
Re: [Talk-us] Standard (mapnik) toolchain/processes: can we teach these better?
I appreciate Simon's response that it seems that the "really coolish" (people, processes...) happen in what often seems like a bubble: that is exactly what I was referring to. It's like the Cool Kids have their "insider club," a world of their own, THEN there are The Rest of Us. The +1 responses I got indicate I'm not the only one who feels this way. Again, I don't think Andy and "the other clockmakers" are a secret society -- indeed I have written a few emails to Andy personally and he has very kindly responded to my pointed questions with aplomb and grace, so I see no ill will being harbored, nor does it seem he/they wish to remain in the shadows (if they do, please remedy that!). But how they document their processes might either be done more openly, or just "more," period. Especially the "why, beforehand and decision-making" part of it. Maybe they just need to point to comprehensive block diagrams or something loosely resembling "the guts of OSM for Dummies" that The Rest of Us can easily find and digest. I realize this is a bit of a wish, but I think it is a high-value effort that would pay dividends in the near future: such sunshine in a project like ours seems a bit overdue, actually. Simon's description of future wishes for what can realistically be achieved with Standard rendering is excellent, and again, very much appreciated. I crave conversations about OSM like this. I just wish there were a better method to "pull it out" of the greater/wider knowledge of OSM than by a grumpy talk-us post complaining about what amounts to "a poor map to how our map works." I also appreciate Martin's +1 about this "lightweight way to push such information out to our contributors...(yet we) haven't developed a culture of actively informing before the fact." YES! EXACTLY! Let us endeavor to do exactly this. And Thank You, Martin, for the Wochennotiz. I have recently discovered the Weekly OSM Summary, which feels like a good start in this direction: like a small newsletter about people in OSM and the technical, social and interesting things they are doing RIGHT NOW in the project. This can only help gear up the inevitable even more questions than it answers. Now, we just need a forum (wiki pages? not really the best venue) where we can discuss such things. In my opinion, this is a critically missing component of a rich and vibrant project like OSM. I like our Help forum, with its interactive feel, I just wish there were a place we might discover intentions of what the future will bring: THAT seems to be the missing component. Simon did just that, but it felt like I was tugging on wild horses to learn it. In the hobby of amateur/ham radio, the usually older fellows who just know everything that you could talk to at a meet or on the air are called "Elmers." I know Elmers exist in OSM, I just don't want to bother them with every little question, patient as they usually are in answering each one I might allow to rise high enough to be worthy of asking them. If you are an Elmer, and have the time to spare, please help our community develop a way to share your deeper knowledge, feeding the cravings of intermediates like me looking to grow into more advanced contributors. If you are such a growing contributor (and who isn't?!) please help us to channel our questions and thirst for specific knowledge into better sub-communities of sharing, especially the one regarding future directions of our project. Thank you to everybody who read and participated in this thread! SteveA California ___ Talk-us mailing list Talk-us@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-us
Re: [Talk-us] Standard (mapnik) toolchain/processes: can we teach these better?
On Mon, May 26, 2014 at 2:24 PM, stevea wrote: > I appreciate Simon's response that it seems that the "really coolish" > (people, processes...) happen in what often seems like a bubble: that is > exactly what I was referring to. It's like the Cool Kids have their > "insider club," a world of their own, THEN there are The Rest of Us. [ ... ] Pssst. Hey, You. You over there feeling left out. Want to know the secret to joining the cool kids? The secret is, "you're already a cool kid." Disappointed? Don't be. You're already one of a small percentage of the world population who knows how to improve their local geo data and share it through OpenStreetMap. Think that isn't a select group? Think again. Only 30 - 50% of those who think they might like to contribute by signing up, actually contribute their first changeset. Only a few thousand people per day contribute, out of a planet of 7 billions. Pretty cool. Want to be even cooler? Become a coder of some sort. Contribute code to one or more OpenStreetMap-related software projects. You think mappers are a select group? They are. Now let's count coders who contribute on a daily basis. It isn't a few thousand per day. More like a few dozen[1]. And those are divided among dozens of projects. So pick a project that interests you; any one you like. Rendering, storage, UI, translations, accessibility, web site, QA, anything at all in the huge and varied OpenStreetMap tool chain and contribute. - find a long outstanding bug and check to see if it is (still) reproducible. - write some documentation for a beginner. - improve performance. - test a patch on different hardware. - triage a new bug. - compare some similar applications and write a review. Or even pick a project that you think needs to do more outreach, and help it do that outreach. Follow their project communication channels, and translate their bug reports, feature requests and design discussions into something suitable for a wider audience, then publish it to the appropriate wider comms channels. Learn more about what interests you. Share what you learn with others. An OpenStreetMap tag line from some of the early mapping party banners read, OpenStreetMap.org It's fun. It's free. You can help. [1] /me waves hands to distract from wild guess number. ___ Talk-us mailing list Talk-us@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-us
Re: [Talk-us] Standard (mapnik) toolchain/processes: can we teach these better?
On 5/26/2014 2:24 PM, stevea wrote: It's like the Cool Kids have their "insider club," a world of their own, THEN there are The Rest of Us. A small part of this is that communication takes place over many channels, partially related to the loose organization of OSM. I subscribe to 11 separate OSM email lists and the web and help forum. I don't want to deal with realtime chat, so I don't bother monitoring the IRC sections. Reading the OSM Weekly Summary blog will go far in addressing your feeling of disconnection from other parts of the project. A possible second step would be to subscribe to the OSM-Talk list, although there is a small amount of extra discussion that strays in there before moving to a more specific list. ___ Talk-us mailing list Talk-us@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-us
