Frame graph architecture - graphics

Since no new shader can be created during runtime, the full set is known ahead at compile-time. Each shader must reference a "pass" in which it will used to render.
To avoid frame-spikes during runtime, I'd like to pre-create all pipeline objects during startup. And to create a pipieline, the number of outputs and the format of each output attachment must be known - either to create a VkRenderPass or to specify the outputs for the dynamic rendering feature.
However, I'd also like to use the frame graph concept (speech by Yuriy O'Donnell) which in turn builds a graph of render passes with input-output specification and dependencies between them. Some passes are conditionally created (e.g. debug passes), some passes might be dropped from the graph (after "compiling" it).
Additionally, I need to support the "write on top" feature, so instead of specifying a new output during the building of the render pass, I can simply say that the output of this pass will use an output from a previous pass - this is useful for adding alpha-blended rendering, for example.
How can I match the two separate section of the code? In other words, how can I define all render passes during initialization but also use a dynamic approach of building the frame graph each frame without repeating myself?
This is what I'd like to avoid (pseudo-code):
struct Pass1Def
{
output1 = ImageFormat::RGBA8;
output2 = ImageFormat::RGBA8;
// ...
outputs = // outputs in order (corresponds to location in shader)
};
void init()
{
for_each_shaders shader {
passDef = findPassDef(shader);
createPipeline(shader, passDef);
}
}
void render()
{
auto previousResource = someCondition ? passA.outputResource1 : passB.outputResource2;
graph.addPass(..., [&](PassBuilder& builder, Pass1Data& data) {
// error-prone: order of function calls matter (corresponds to location in shader)
// error-prone: use the same format defined in Pass1Def
data.outputResource1 = builder.create(... ImageFormat::RGBA8);
// error-prone: the format depends on the outputResource of a previous pass
// however the format must be (and was) specified in Pass1Def
data.outputResource2 = builder.write(previousResource);
});
}

Related

Why would you use the spread operator to spread a variable onto itself?

In the Google Getting started with Node.js tutorial they perform the following operation
data = {...data};
in the code for sending data to Firestore.
You can see it on their Github, line 63.
As far as I can tell this doesn't do anything.
Is there a good reason for doing this?
Is it potentially future proofing, so that if you added your own data you'd be less likely to do something like data = {data, moreData}?
#Manu's answer details what the line of code is doing, but not why it's there.
I don't know exactly why the Google code example uses this approach, but I would guess at the following reason (and would do the same myself in this situation):
Because objects in JavaScript are passed by reference, it becomes necessary to rebuild the 'data' object from it's constituent parts to avoid the original data object being further modified by the ref.set(data) call on line 64 of the example code:
await ref.set(data);
For example, in MongoDB, when you pass an object into a write or update method, Mongo will actually modify the object to add extra properties such as the datetime it was insert into a collection or it's ID within the collection. I don't know for sure if Firestore does the same, but if it doesn't now, it's possible that it may in future. If it does, and if your original code that calls the update method from Google's example code goes on to further manipulate the data object that it originally passed, that object would now have extra properties on it that may cause unexpected problems. Therefore, it's prudent to rebuild the data object from the original object's properties to avoid contamination of the original object elsewhere in code.
I hope that makes sense - the more I think about it, the more I'm convinced that this must be the reason and it's actually a great learning point.
I include the full original function from Google's code here in case others come across this in future, since the code is subject to change (copied from https://github.com/GoogleCloudPlatform/nodejs-getting-started/blob/master/bookshelf/books/firestore.js at the time of writing this answer):
// Creates a new book or updates an existing book with new data.
async function update(id, data) {
let ref;
if (id === null) {
ref = db.collection(collection).doc();
} else {
ref = db.collection(collection).doc(id);
}
data.id = ref.id;
data = {...data};
await ref.set(data);
return data;
}
It's making a shallow copy of data; let's say you have a third-party function that mutates the input:
const foo = input => {
input['changed'] = true;
}
And you need to call it, but don't want to get your object modified, so instead of:
data = {life: 42}
foo(data)
// > data
// { life: 42, changed: true }
You may use the Spread Syntax:
data = {life: 42}
foo({...data})
// > data
// { life: 42 }
Not sure if this is the particular case with Firestone but the thing is: spreading an object you get a shallow copy of that obj.
===
Related: Object copy using Spread operator actually shallow or deep?

Where is my error with my join in acumatica?

I want to get all the attributes from my "Actual Item Inventry" (From Stock Items Form) so i have:
PXResultset<CSAnswers> res = PXSelectJoin<CSAnswers,
InnerJoin<InventoryItem,
On<CSAnswers.refNoteID, Equal<Current<InventoryItem.noteID>>>
>
>.Select(new PXGraph());
But, this returns me 0 rows.
Where is my error?
UPDATED:
My loop is like this:
foreach (PXResult<CSAnswers> record in res)
{
CSAnswers answers = (CSAnswers)record;
string refnoteid = answers.RefNoteID.ToString();
string value = answers.Value;
}
... but i can not go inside foreach.
Sorry for the English.
You should use an initialized graph rather than just "new PXGraph()" for the select. This can be as simple as "this" or "Base" depending on where this code is located. There are times that it is ok to initialize a new graph instance, but also times that it is not ok. Not knowing the context of your code sample, let's assume that "this" and "Base" were insufficient, and you need to initialize a new graph. If you need to work within another graph instance, this is how your code would look.
InventoryItemMaint graph = PXGraph<InventoryItemMaint>.CreateInstance<InventoryItemMaint>();
PXResultset<CSAnswers> res = PXSelectJoin<CSAnswers,
InnerJoin<InventoryItem, On<CSAnswers.refNoteID, Equal<Current<InventoryItem.noteID>>>>>
.Select(graph);
foreach (PXResult<CSAnswers> record in res)
{
CSAnswers answers = (CSAnswers)record;
string refnoteid = answers.RefNoteID.ToString();
string value = answers.Value;
}
However, since you should be initializing graph within a graph or graph extension, you should be able to use:
.Select(this) // To use the current graph containing this logic
or
.Select(Base) // To use the base graph that is being extended if in a graph extension
Since you are referring to:
Current<InventoryItem.noteID>
...but are using "new PXGraph()" then there is no "InventoryItem" to be in the current data cache of the generic base object PXGraph. Hence the need to reference a fully defined graph.
Another syntax for specifying exactly what value you want to pass in is to use a parameter like this:
var myNoteIdVariable = ...
InventoryItemMaint graph = PXGraph<InventoryItemMaint>.CreateInstance<InventoryItemMaint>();
PXResultset<CSAnswers> res = PXSelectJoin<CSAnswers,
InnerJoin<InventoryItem, On<CSAnswers.refNoteID, Equal<Required<InventoryItem.noteID>>>>>
.Select(graph, myNoteIdVariable);
foreach (PXResult<CSAnswers> record in res)
{
CSAnswers answers = (CSAnswers)record;
string refnoteid = answers.RefNoteID.ToString();
string value = answers.Value;
}
Notice the "Required" and the extra value in the Select() section. A quick and easy way to check if you have a value for your parameter is to use PXTrace to write to the Trace that you can check after refreshing the screen and performing whatever action would execute your code:
PXTrace.WriteInformation(myNoteIdVariable.ToString());
...to see if there is a value in myNoteIdVariable to retrieve a result set. Place that outside of the foreach block or you will only get a value in the trace when you actually get records... which is not happening in your case.
If you want to get deep into what SQL statements are being generated and executed, look for Request Profiler in the menus and enable SQL logging while you run a test. Then come back to check the results. (Remember to disable the SQL logging when done or you can generate a lot of unnecessary data.)

Creating Node.js enum in code to match list of values in database

I have a list of valid values that I am storing in a data store. This list is about 20 items long now and will likely grow to around 100, maybe more.
I feel there are a variety of reasons it makes sense to store this in a data store rather than just storing in code. I want to be able to maintain the list and its metadata and make it accessible to other services, so it seems like a micro-service data store.
But in code, we want to make sure only values from the list are passed, and they can typically be hardcoded. So we would like to create an enum that can be used in code to ensure that valid values are passed.
I have created a simple node.js that can generate a JS file with the enum right from the data store. This could be regenerated anytime the file changes or maybe on a schedule. But sharing the enum file with any node.js applications that use it would not be trivial.
Has anyone done anything like this? Any reason why this would be a bad approach? Any feedback is welcome.
Piggy-backing off of this answer, which describes a way of creating an "enum" in JavaScript: you can grab the list of constants from your server (via an HTTP call) and then generate the enum in code, without the need for creating and loading a JavaScript source file.
Given that you have loaded your enumConstants from the back-end (here I hard-coded them):
const enumConstants = [
'FIRST',
'SECOND',
'THIRD'
];
const temp = {};
for (const constant of enumConstants) {
temp[constant] = constant;
}
const PlaceEnum = Object.freeze(temp);
console.log(PlaceEnum.FIRST);
// Or, in one line
const PlaceEnum2 = Object.freeze(enumConstants.reduce((o, c) => { o[c] = c; return o; }, {}));
console.log(PlaceEnum2.FIRST);
It is not ideal for code analysis or when using a smart editor, because the object is not explicitly defined and the editor will complain, but it will work.
Another approach is just to use an array and look for its members.
const members = ['first', 'second', 'third'...]
// then test for the members
members.indexOf('first') // 0
members.indexOf('third') // 2
members.indexOf('zero') // -1
members.indexOf('your_variable_to_test') // does it exist in the "enum"?
Any value that is >=0 will be a member of the list. -1 will not be a member. This doesn't "lock" the object like freeze (above) but I find it suffices for most of my similar scenarios.

Blockly: How to obtain value of a Dropdown or Checkbox Block

I´m new to Blockly and can not find a way to obtain field value of a dropdown or checkbox.
Lets consider following scenario (generated with blockly-dev-tools):
Blockly.Blocks['feature'] = {
init: function () {
this.appendDummyInput()
.appendField("Feature") // just for label
.appendField(new Blockly.FieldDropdown([["manufacturer", "feature_manufacturer"], ["profile", "feature_profile"], ["glas", "feature_glas"]]), "category"); // for dropdown values
this.appendValueInput("feature_name")
.setCheck("String")
.setAlign(Blockly.ALIGN_RIGHT)
.appendField("Name");
this.appendValueInput("feature_prio")
.setCheck("Number")
.setAlign(Blockly.ALIGN_RIGHT)
.appendField("Priorität");
this.appendDummyInput()
.setAlign(Blockly.ALIGN_RIGHT)
.appendField("Versteckt")
.appendField(new Blockly.FieldCheckbox("TRUE"), "hidden");
now obtaining values from value inputs is not a problem, you can get thouse like this:
const featureName = element.getInputTargetBlock("feature_name");
if (featureName) {
console.log(featureName.getFieldValue("TEXT"));
}
const featurePrio = element.getInputTargetBlock("feature_prio");
if (featurePrio) {
console.log(featurePrio.getFieldValue("NUM"));
}
but dummy inputs hosting dropdowns or checkboxes have no methods to provide selected values.
It might be that this is my conceptual error to use dummy inputs to host the elements, but when using value input, you have always those nipples to the right, which are obsolate, since the values are provided by the checkbox or dropdown.
You should be able to skip the middleman and use element.getFieldValue. For example, to get the value from the checkbox field named "hidden", you could use element.getFieldValue("hidden").
P.S. You can also skip the element.getInputTargetBlock middleman and use Blockly.JavaScript.valueToCode (I.E., to get the value of the block in the "feature_name" input, you could use Blockly.JavaScript.valueToCode(element, "featureName", Blockly.JavaScript.ORDER_ATOMIC) or what have you). If you use a different generator than JavaScript, replace JavaScript with the generator you use (e.g. Python or whatever).

Why does this Context.Sync not work?

Why does this code snippet not write the values back to Excel unless I un-comment the range.values=range.values line?
$('#run').click(function() {
invokeRun()
.catch(OfficeHelpers.logError);
});
function invokeRun() {
return Excel.run(function(context) {
var range = context.workbook.worksheets.getItem("Sheet1").getRange("A1:B3");
range.load('values');
return context.sync()
.then(function() {
range.values[1][1]=99;
console.log(JSON.stringify(range.values));
//range.values=range.values
return context.sync();
});
});
}
Array properties are special. I have added a page on my website to describe the topic: Reading and writing array properties.
Summarizing from there, the way that the proxy-object model works, whenever you set a property on an object, the Office.js runtime has a hook into the setter and getter, which is used to intercept the call and add the command to the queue.
Let's take an example of a regular property first. Per the above, whenever you set something like range.format.fill.color = "red", the setter for the color property intercepts the request and internally adds a command into the queue to set the range fill color to red (to be dispatched with the next context.sync)
On the other hand, if all you had was var color = range.format.fill.color
(after a load and a sync, of course), the getter would fire instead of the setter, and the color variable would get the range's current fill color.
Now, that was regular properties. Whenever you set an element of the array, you are effectively accessing the array value as a getter. From a runtime perspective, this line is no different from a slightly more verbose version:
var array = range.values;
array[r][c] = '-';
Because the getter for range.values returns a perfectly plain JS array object, accessing it and then setting its value does nothing to propagate it back to the original Range object.
If you want the values to get reflected back, the best thing is to get a reference to the array right after the sync (i.e., var array = range.values, just as above), then set the values on the array as needed, and then finally set it back to the object: range.values = array.
It means you could also modify the values array in place, and then assign the values property back to itself at the completion of the loop (range.values = range.values). However, this looks awkward, as if it’s a no-op, whereas in reality it is not. So personally, I prefer to retrieve the array at the beginning and assign it to its own variable, then do any necessary modifications, and finally set the full array back.
UPDATE to clarify the above:
To be very clear, the arrays returned by accessing the .values, .formulas, etc., ARE pure vanilla JS arrays. That's actually the crux of the problem: that in order for Office.js to return pure objects, it means that those pure objects can't be "spiked" with the ability to reflect changes.
For what it's worth, we actually have an upcoming feature that should be rolling out in a month or two, where we will be introducing an object.set syntax, as in:
range.set({
values: [[1, 2], [3, 4]],
format: {
fill: {
color: "purple"
}
}
}
This will make it more convenient to set multiple properties on the same object, but it might also make the array properties easier to deal with.

Resources