OpenGeofiction

Luciano has commented on the following diary entries

Post When Comment
New waterway connection from the Vinn Sea to the ocean 18 days ago

Upvote.

A Lesson in Scale 18 days ago

@Greg_Rose - You could just bookmark the scale helper. Or use the link in the OGF:Dashboard that I made some years ago.

Thanks for the feedback on Ohunkagan. I'm approaching it "historically" - trying to map as the city grows in 20 year increments. Currently I'm doing the 1900-1920 phase, approximately. Need to expand streetcar suburbs substantially, get the elevated rail expanded, etc.

A Lesson in Scale 18 days ago

Sometimes new mappers aren't aware of the "scale helper" utility on OGF:

https://tile.opengeofiction.net/util/map_scale.html?map=C/14.34857283295594/-22.90641/148.28559&map2=OSM/x/38.88422/-77.02603&pxpt=0.2767&scale=25000

I actually don't personally use that tool that much. One thing I do is I will have a separate "sketch" layer open in JOSM where I have "traced" some real-world feature I'm interested in imitating. I did this a while back when developing the Ohunkagan streetcar network (historical, circa 1900). I had a tracing of the Minneapolis/St-Paul streetcar network from the same era (done from an old map), and I could compare things like curves, route density, overall extent... and get the right scale.

I forgot how bad iD used to be 19 days ago

The "large object" problem mentioned by Lithium-Ion is probably why I've decided JOSM is better for geofiction. I can open my entire territory in JOSM - or even more. Large scale mapping isn't necessary in OSM - the large-scale world effectively is a fixed object when talking about real-world mapping. It is quite rare for OSM mappers to need to edit anything larger than a few city blocks.

The requirements in geofiction are quite different. I spent the last few days tracing entire continents from old paper maps, for my Rahet project. JOSM is the only option, basically.

I forgot how bad iD used to be 19 days ago

I've always been one of those people who was irrationally hostile to iD. I'm not even sure why. When I first started with OGF, I tried the two browser-based editors and immediately chose Potlatch2. Potlatch is unsupported, I believe, at this point. I eventually transitioned to JOSM and although it's got its problems and drawbacks, I've become so comfortable with it at this point that I'd never imagine using a browser-based editor. Personally I dislike the "presets" (in iD or JOSM) because they feel too "real world" and constrain geofiction.

That said, I'm curious how hard it is, technically, to upgrade just iD within the rail port (osm-website) architecture. Clearly just upgrading the entire OGF rails port is out of the question, as it has been substantially customized for OGF. But iD has a standalone version, and I wonder how it might be upgraded without messing with the rails port.

On my Arhet server, I appear to have iD 2.7.1 enabled - this is just a consequence of that having been the active version of iD when I built the Arhet server off the rails port codebase 2 years ago. The current version of iD on github is 2.18.4.

I might play around with seeing if I could upgrade the iD editor on Arhet without breaking anything.

About tile updating 23 days ago

What Taka said is exactly right. Sometimes the render algorithm "misses" your edit, and the best way to get your edit to render is to go back and make another edit in the same area.

Move a single node 3 meters, add a new node to a curve, delete a node - something minor - and upload, and that will "kick the render in the pants" and it will work.

Backups broken, only contains nodes, no ways or relations about 1 month ago

To add some additional detail:

There is a substantial drop in backup filesize between 7/25 and 7/26. I expect this is where the failing backup first happened.

So I went and ran the same extract on the backup dated 7/25. Sure enough, that file is perfectly fine.

Backups broken, only contains nodes, no ways or relations about 1 month ago

I hadn't looked at an ogf backup in a while. I grabbed one this morning, did an extract for my contiguous territories in South Archanta using osmconvert, and sure enough, what you say seems to be true.

Unresponsive part of screen in edit mode. about 1 month ago

OK, I misunderstood your description in a comment with respect to the context of the original problem. Sorry. I didn't realize your focus was on Potlatch. You'd said you had it also with JOSM. and of course changing a JOSM parameter will only affect JOSM. I guess it's a laptop problem.

Unresponsive part of screen in edit mode. about 1 month ago

That makes me wonder if it's perhaps a latency problem - the OGF site is certainly slower in its response-times compared to OSM.

Have you messed with these settings in JOSM?

https://wiki.geofictician.net/wiki/images/0/0f/Luciano_Screenshot_JOSMtimeouts.png

Unresponsive part of screen in edit mode. about 1 month ago

JOSM gets weird sometimes. Personally, I think it's a Java memory-management problem, as opposed to being specific to JOSM. Your commment that P2 has it, but iD does not, is interesting. P2 uses Flash (which is also notorious for memory-management issues), and iD uses Javascript (which relies on the browser for memory management).

I have had issues similar to what you describe with JOSM, though nothing exactly like it. In the end, I have never had a problem with JOSM that wasn't solved by "shut down and restart."

If restarting JOSM (and/or your computer) doesn't solve the problem, I would definitely think that the problem was with the graphics card or video driver software on your computer.

Feature idea: custom highway/route signs on maps about 1 month ago

@EMKLI - your points are quite valid. I am sorry if my remarks seem "strange."

I suppose I was over-reacting. Partly, because it's frustrating when people ask for things thinking they are "simple" and so I want to explain that they are not, in fact, "simple." With respect to this specific feature being requested, I was already annoyed because I know other users have requested the exact same feature before, with exactly the same tone of entitlement, and with the exact same naivety with respect to the level of difficulty of what they're requesting.

I just feel there should be a more appreciative (thankful? polite? deferential?) tone, I guess, when asking for features or changes to a volunteer-run website. Regardless, your point about being a non-native speak of English is one that needs to be addressed.

Is it really true that "politeness" doesn't translate? It brings to mind my long experience as a US-born person living in South Korea. Many Koreans would observe, with surprise, that I was "more polite" than most Americans. This despite my far-from-perfect Korean language skills - Korean is a language famous for its complex forms of deference, humility, and levels of politeness. Trust me, I never mastered those, but I could still show politeness by speaking carefully and thinking about how my words might be perceived. I think that politeness is not about the structure of language, but rather it's about intention and culture.

That said, I am sometimes not very polite in my own comments on this website. There is something about online discourse that leads us to forget how to be polite. This is seeming like a universal tendency. For that, I must apologize. And perhaps also I should make allowances for others' shortcomings.

Beyond making requests, there are other ways to approach a desire for a new feature on a website where the users are also allowed to be creators. When I first joined OGF, there was no wiki. Then Thilo set up the wiki, but it was quite rudimentary - there was no "main page" showcasing users' work. Several users, including myself, felt it would be nice to have a "main page." Rather than announce "There should be a main page on the wiki" as if making a demand of the OGF admin team, I worked to learn how a wiki main page worked and created a trial implementation in my wiki sandbox. Thilo thought it was a good idea, and so later, that became OGF's wiki main page.

I'm not saying this to say that MisterBean's request for custom highway shields is equivalent to my desire for a wiki main page in 2014. What I am intending to say is that a little bit of online research would have helped that person to realize that making custom highway shields is a pretty complicated problem - sufficiently complicated that even OSM hasn't solved it, yet. There are some programmers who are trying to solve it, but it's nothing that has proven acceptable to the OSM admin team, yet. That information is already out there, online. Before making a request for a feature for OGF, wouldn't it make sense to do some research about how difficult it might be?

Is our carto style different from OSM's? about 2 months ago

@Alessa - that's a valid point.

Really that's up to Thilo. In the past I've got the impression from him (and he can correct us all if I'm off base) that he wants to have OGF's carto style to be "exactly" a replica of OSM's. This makes updates trivial, since he can just download the new version and it's done.

Because of the change to the database with the shapefiles, in fact this is already broken. So we can go two routes:

1) we can start doing piecemeal update, as you suggest - update those things which are easy, but with manipulations to the code (however minor they might be) that mean all future updates will also be fiddly and require effort

or

2) we can fix the coastline problem so we can go back to the "updates are trivial" process

I suppose a third option is to permanently fork the style and let it be frozen in time at the point when the fork occurred, letting it evolve in its own direction (or fail to, depending on the effort of the person maintaining it). Actually, this last is essentially what's happening, for now, on my Arhet server. I have two carto styles, one a "frozen fork" dated early 2018, and another "frozen fork" dated early 2020. Neither will likely ever be updated - rather, I'll eventually replace them if and when I figure out the new database system for storing shapefiles.

In some repects, that's what the histor style is on OGF, too - it was a fork from osm-carto in around 2014, and since then it's evolved independently.

In my utopian geofiction environment, there would exist an easy-to-edit "public" tool that the mappers themselves could update, fork, and maintain without having to have technical skill.

Is our carto style different from OSM's? about 2 months ago

@histor

I believe Toadwart was talking about this:

https://wiki.geofictician.net/wiki/images/d/de/Luciano_Screenshot_arcade_on_OGF.png

Is our carto style different from OSM's? about 2 months ago

OGF-carto is, indeed, a copy of OSM-carto. However, it is not the most up-to-date copy. It is a snapshot of at least several months ago, now. Normally Thilo was periodically updating the OGF-carto to match the OSM-carto, but this is a manual process that takes time, so it's possible this hasn't happened in a while.

If the change in icons was a recent change to the OSM-carto style, it won't be showing on OGF.

Another issue has arisen recently: OSM-carto recently made major changes to how they store coastline and border shapefile information. Where before, it was stored in file directories on the render server, that information has been moved over to the database. That means that much of the customization required to get OGF to handle its local coastlines and borders (rather than defaulting to Earth's coastlines and borders) will need to be re-coded. This is a somewhat formidable undertaking.

We might not be seeing an update to the OGF-carto in a long while, if I understand the issues correctly.

New Lucien - Any Feedback? about 2 months ago

I'm not sure how I feel about this.

Feature idea: custom highway/route signs on maps about 2 months ago

I dislike when people cavalierly ask for features on a FREE website with utter disregard for the technical issues involved and with utter disregard for the likely labor involved. Who, do they imagine, is going to put in the FREE hours to make this happen? If you want a feature, then how about you put forth and provide it?

It's the same as people who complain about lack of nominatum (name search), problems with the wiki performance, problems with the overpass api, etc.

Did anyone notice they aren't PAYING for OGF? So rather than complain about missing features, how about putting forth and helping create those features? And if you lack the technical expertise to contribute, how about you stop asking other people to put in FREE work to make it happen? Where does this entitlement come from?

Mapping Software 2 months ago

You need to give a lot more information for this question to have any kind of possible answer.

Import from what? from Handdrawn images? from OGF? from some other application?

Import to what? into OGF? into some image editor (e.g. Inkscape, MS paint, Gimp, Photoshop)? into some vector rendering software (e.g. Maperitive)?

Mapper's Challenge #30 — July 2020 — Contours and topography 2 months ago

When should I use the natural=coastline tag? Is it equivalent to ele=0?

I actually copy-paste my coastlines and lake-shorelines from an OGF layer to my contour file layer in JOSM. For coastlines, I explicitly add ele=0, for lakeshores, I add ele=. You can see this with my lakeshores in Makaska, for example, here:

https://opengeofiction.net/way/11426469#map=16/-42.4550/146.1073

Is ele in meters?

Yes. It is possible to use e.g. imperial system (US), but that would require changing settings on the server side. Best to keep everything consistent.

Do lines need to be a closed way?

Not at all. In fact break up lines is basically required when mapping smaller areas, unless you're doing islands.

Do lines need to be continuous?

Not at all. In fact, HORIZONTAL interval is more important for the conversion than VERTICAL interval. So I make a practice of adding intermediate lines where slopes are shallow, e.g. flat areas. I'll add short stretches of 5 meter (vertical) intervals where the distance between my 10 meter lines is large, or if an area is mountainous, I'll neglect every other line, thus 20 meter intervals or even 40 meter intervals.

What is the height within a closed way containing nothing else? Is it flat like a lake?

This is a bit problematic. I have taken to adding diagonal "filler" contours on lake surfaces, to prevent "glitches" in the render (holes and canyons). Here is a screenshot from Makaska (north end of Lake Ohunkagan, directly west of the city of Ohunkagan):

https://wiki.geofictician.net/wiki/images/0/08/Luciano_Screenshot_LakeOhunk_ContourSample.png

Ohunkagan's 1895 Exposition & World's Fair 3 months ago

@newflanders - "credit where credit is due", the Wychelle Cow is by mapper CartographerKing. It is, indeed, excellent, and I'm planning to make it into a long-lasting Ohunkagan landmark: The Wychelle Cow is a famous kitschy restaurant - "Come to the Wychelle Cow for that famous Ohunk style frybread! Take the tram to the beach and get off when you see the giant cow."

I would very much like to see more mappers participating in my invitation to map their state or country exhibits! It can lead to creativity such as this.