您好，注意到您将 motorway_link 和主线的分/合位置改到了渐变段中间。然而这个连接点只有靠近导流线顶点，逻辑上才是正确的，因为在渐变段起点和导流线顶点之间，实际只是主线多出来了一根不分离的车道，出口匝道要到导流线顶点才开始/入口匝道到了导流线顶点就结束了。而且，商业地图也多用类似方式，OSM 上一些数据质量较高的地区（如德国）也普遍如此处理。如果认为不好看，可以对 motorway_link 做一些不穿越物理分隔的平滑处理，但还是希望保持这个连接点位置在导流线顶点附近。谢谢！

请遵照名称规范，name 属性在中国大陆非少数民族地区只能放简体中文名称，实地展示的英语名称请放入 name:en。

Hi Tronald,
"Dezhou Road Station" you've added in Shanghai about 11 months ago does not exist. It is not advisable to add anything that you are not sure about it. Thanks!

被 iD 忽悠加了 ford=yes？如果路是以一座简单的桥的形式跨过河的话，正确做法是把属于桥的这部分路切出来添加 bridge=yes 和 layer=1。ford=yes 是给野外的过水路面用的。

Thank you, Mueschel! I'll correct them.

Hi mueschel,
Thank you for your reply.

>> first the language, then the direction

Got it. Thanks!

>> tag only the languages which are actually listed on the sign

You are right. The signs do list English destinations, but not Chinese Pinyin (i.e. ISO 7098 romanization) ones. I usually provide romanized names for the purpose of helping non-Chinese users getting its pronunciation; however, it does not make sense for navigation purpose, as users are expected to see it instead of read it.

>> It's not necessary to have both a destination tag on the way and a destination sign relation.

Will having both of them cause troubles? I did it for compatibility reason, as the support for destination tags and/or destination_sign relations may vary among data consumers. If their coexistence are to cause adverse effects, I will prefer destination_sign relations, which is more flexible and has the advantage of associating destination information with actual signs.

Hi user_5359,
Seems like I made some error when copying and pasting tags in JOSM. I'll check and correct it later.

你好，服务区匝道节点请不要加出口编号，同时请加上 noref=yes。谢谢！

Zeehan Lin 你好，按照 OSM 的「实地求是」原则，尚未实际动工的工程项目不应标示。这里无关立场原则的事情，只是 G3 高速公路究竟以怎样的线路和形式越过台海，目前都并未确定，大陆的官方地图也不会标示。

I don't think it to be reasonable. I did notice the discussions regarding this, and I see that you took "destination:<something>:to" as an example of "ambiguities", but ":to" is just a proposed subtag; however, it is a widely followed convention to directly using ISO 639-1 codes as subtags for multilingual properties, and introducing ":lang:<lg>" here is creating inconsistencies; therefore I think it is ":to" instead of ":<lg>" that should be changed.