BUG #1739: fitToViewport uses dataset bounds instead of recomputing them#1744
BUG #1739: fitToViewport uses dataset bounds instead of recomputing them#1744doutriaux1 merged 1 commit intomasterfrom
Conversation
|
This is still work in progress. I need to replace all fitToViewport with the bounds version. |
|
@danlipsa do you want to post update to all of us? |
danlipsa
force-pushed
the
projected-data-bounds
branch
2 times, most recently
from
e3b9847 to
d1b4997
Compare
fitToViewport recomputes the dataset bounds using an 2D array of points. This is slow and imprecise and it does not take into account the bounds stored in the file. Now we use a new function fitToViewportBounds that receives the dataset bounds as parameter. The old function is still used for the outline and the continents. It is going to be removed in the future.
danlipsa
force-pushed
the
projected-data-bounds
branch
from
d1b4997 to
1583a77
Compare
|
@doutriaux1 did you review this branch? |
|
Not yet, was working on bots and lambert yesterday. |
|
@doutriaux1 let's merge this one as well soon so that we can release it today. |
|
there's no release w/o projections. Will take a look at this one when proj are in better shape. |
@doutriaux1 lets talk about this in the afternoon. I don't understand what you are saying here. Based on our discussions, we should release as things are better and not more broken than before. |
|
all i'm saying is I want to give a fair shot to the projection part, if we keep adding bits in the mean time it makes it harder to fix the proj issue. I will take a look at tihs when we fix or quit the proj issues |
|
@aashish24 @danlipsa ok this branch look good, but there's going to be baseline conflicts with lambert branch (whichever you decide on). LEt's merge lambert first and then update again this PR. Ok? |
|
once lambert is in master I will trigger a rebuild of this one so we can see which files need updating. |
@doutriaux1 we should merge this branch first as this was pushed before the lambert one. It would be nice if we follow the queue approach. Also, the lambert is the small one. I would prefer if you merge this one first. |
BUG #1739: fitToViewport uses dataset bounds instead of recomputing them
|
sure... |
|
@doutriaux1 did I say that you are awesome!! |
3 participants
fitToViewport used to recompute the dataset bounds using an 2D array of points.
This was slow and imprecise. Now we pass the dataset bounds as parameters.