What's the point of adding vertices to a straight line

[BrianBulla1]

I work with creating/editing features in a water network.  Someone has suggested that whenever we create a new valve (point feature) on a watermain (line feature), that we need to first create a new vertex where the valve will go, and then snap the valve feature to the vertex.

Currently we just snap the valve to the edge of the watermain.  Is there any added benefit to manually creating vertices as we digitize valves onto straight lines?

Thanks,

[JoeBorgione]

I avoid any extra vertexes in a straight line which of course is defined as the shortest distance between two points. It used to drive me crazy when street centerlines I managed had multiple vertexes on a segment that was one single straight line block. 

[DanPatterson_Retired]

over long distances, densification of the line is useful in order to maintain shape geometry should the layer ever need to be reprojected to a different coordinate system.  A 2 point line will only maintain the position of the start and end points.  The intervening space is assumed to be on a straight line with intervening locations being associated with differences in X and Y times a step increment.  This is never the case for geographic coordinates except along meridians and it may even fail for projected coordinate systems when one switches between projected coordinate system types (eg equal area, conformal etc).  The classic 'bad' projection is the Web Mercator projection where nothing other than location is any good

[KoryKramer]

I think you should ask the person who suggested that to you to explain their reasoning.  All of us can speculate, but we don't know why that was the suggestion 

If you were working with network datasets, whether there is a vertex or not can affect connectivity in the network - can traffic go from this edge to that edge? Depends on connectivity rules which can be based on whether the connection is at a vertex or not.

Since you say that you're working with water network, I'm assuming you have this modeled as a geometric network?

So my guess is that they have some reasoning that has to do with network connectivity: Overview of editing geometric network features—Help | ArcGIS Desktop But there would probably be a lot more questions about whether the edges are modeled as simple or complex and how junctions are defined in the network:

Maybe this helps?

[BrianBulla1]

Hi Kory,

Thanks for posting.  Actually, our water network is not modeled as a geometric network.  It is just a collection of points and lines, primarily used for cartographic needs and for keeping an inventory on what assets we have.

We do have some Topology rules laid out for identifying digitizing errors, but otherwise it's a fairly simple model of our water system.