SAGA Removal & TM Update: Module 18 (Lesson 18.1-18.9) - #11264
hefniraera wants to merge 1 commit into
Conversation
DelazJ
left a comment
DelazJ
left a comment
There was a problem hiding this comment.
Thanks @hefniraera for taking on this big challenge.
I think it is however too big to get done in a single PR. Here is a fast and unfinished review and you will notice that there are some comments on paragraphs or texts you didn't touch. This chapter is at best QGIS 2 compatible and some of QGIS Processing "concepts" (may) have changed in the meantime, so they should be adjusted/double-checked. Let's then update the lessons incrementally and correctly. I would thus suggest to:
- Split the PR into smaller and self-contained ones, let's say one lesson per PR (or two if they are really connected or too small)
- In each lesson, turn the instructions into steps, using a numbered listing
- In each lesson, test the instructions we provide. E.g., having played recently with GRASS, I can say that the SAGA --> GRASS replacement looks really wrong and untested.
- There shouldn't remain screenshots from the QGIS 2 era. And considering that the next release will now be QGIS 4.4 LTR, and that version introduced some changes to algorithms dialog help area, I would suggest we use master to update the images so that it gets visually aligned with what users will get.
Hope that provides some guidance on how to handle training lessons update. And yes, IMHO this is the most time-consuming operation in docs writing. So, start with lessons you are the most comfortable with.
| * If there are input layers, it will use the CRS of the first layer. This is | ||
| assumed to be the CRS of all input layers, since they should have the same one. | ||
| If you use layers with unmatching CRS's, QGIS will warn you about it. Notice that | ||
| the CRS of input layers is shown along with its name in the parameters dialog. |
There was a problem hiding this comment.
Is this still true? Have you tested it?
|
|
||
|
|
||
| .. note:: In this lesson we will discuss how Processing uses CRSs. | ||
| We will also see a very useful algorithm: reprojecting. |
There was a problem hiding this comment.
There is no alg named "Reprojecting", right? Or do I misunderstand the sentence?
| contains a specific CRS field (as it happened in the last lesson with the | ||
| graticule algorithm) | ||
|
|
||
| Open the project corresponding to this lesson and you will see two layers named |
There was a problem hiding this comment.
Where is the project corresponding to the lesson? Afaict, it is not mentioned in the lines above. Can you point to it, please? Moreover, can you turn instructions into numbered list, please?
| @@ -43,19 +46,21 @@ | |||
| The other parameter of the algorithm allows to set how the algorithm uses | |||
| coordinates to calculate the new value that it will add to the resulting | |||
| layers. Most algorithms do not have an option like that, and just use the | |||
There was a problem hiding this comment.
| layers. Most algorithms do not have an option like that, and just use the | |
| layer. Most algorithms do not have an option like that, and just use the |
One output, right?
| You should get a new layer with exactly the same points as the other two layers. | ||
| If you right click on the name of the layer and open its properties, you will | ||
| see that it shares the same CRS of the input layer, that is, EPSG:4326. | ||
| see that it shares the same CRS of the input layer, that is, ``EPSG:4326``. |
There was a problem hiding this comment.
Can we add somewhere in the previous lines instructions to select that specific layer (for the first parameter)?
And generally speaking, I very much prefer we write the instructions people should follow than expect they rely on the algorithm screenshot. This makes instructions clearer IMHO and is also useful for visually impaired people that may rely on tools that read texts.
| as with any other algorithm. Since we are using the simplified interface, | ||
| you do not know which algorithms are based on SAGA or in another external | ||
| you do not know which algorithms are based on GRASS or in another external | ||
| application, but if you happen to double--click on one of them and the | ||
| corresponding application is not installed, you will see something like this. |
There was a problem hiding this comment.
I do not remember what the simplified toolbox looked at, but this paragraph looks quite outdated/wrong to me. AFAICT The provider of an alg is always shown in the tree hierarchy.
| Once GRASS is installed, you can launch a GRASS algorithm double clicking on its name, | ||
| as with any other algorithm. Since we are using the simplified interface, | ||
| you do not know which algorithms are based on SAGA or in another external | ||
| you do not know which algorithms are based on GRASS or in another external | ||
| application, but if you happen to double--click on one of them and the | ||
| corresponding application is not installed, you will see something like this. | ||
|
|
||
| .. figure:: img/first_saga_alg/missing_saga.png | ||
| .. figure:: img/first_GRASS_alg/missing_GRASS.png | ||
| :align: center | ||
| :width: 50% | ||
|
|
||
| In our case, and assuming that SAGA is correctly installed and configured, | ||
| In our case, and assuming that GRASS is correctly installed and configured, | ||
| you should not see this window, and you will get to the parameters dialog instead. |
There was a problem hiding this comment.
This whole block looks weird to me. I would suggest to drop situations where things do not work and assume that people have done things the way we instructed earlier, and THAT WORKS!
| you should not see this window, and you will get to the parameters dialog instead. | ||
|
|
||
| Let's try with a SAGA--based algorithm, the one called *Split shapes layer randomly*. | ||
| Let's try with a GRASS--based algorithm, the one called *Split shapes layer randomly*. |
There was a problem hiding this comment.
Happy to be proven wrong but it looks more like a search and replace SAGA by GRASS in this page... These are two REALLY different tools and can't be replaced this way.
There was a problem hiding this comment.
Not a GRASS alg I think.
And can you lowercase folder name, please?
|
Hi, @DelazJ Thank you very much for the detailed review. I'm aware there are plenty of details that can easily get missed during updates. I had a hard time keeping each lesson in sync. I was also unsure whether I should retain the existing writing style or come up witrh a more polished one, which could have created significant differences. However, your review has clarified everything. I'm happy to fix it 😀 |
2 participants
IN PROGRESS....
Mainly, the Linux part needs to be double-checkd. Will do later...
Goal:
Ticket(s): #