I'm working on an app, that displayes huge (10000+ elements) graphs in a wpf canvas.
I'd like a feature like in Visual Studio, when you can split the view of an editor (so I can view two distant part of the same graph at the same time).
I have some constraints:
data binding (creating the bindings) of graph elements makes the loading of big graphs very slow, so I'm not using MVVM, the "VM" knows about the view and updates it directly when needed
the children of the canvas are frameworkelements, since I use the Tag property
because of the number of graph elements, I don't want to keep two different view for each element for the two part of the split view
So it should be like displaying multiple parts of the same canvas in different places. You can't set two parents for the FrameworkElements in WPF, so the easiest way is out of question :(
What are my options? Should I reconsider my constraints or there is some workaround for this?
Let me know if you need any more details (it's a big application, so I can't give you every information).
Edit: duplicating with visual brush is not an option since I need proper input event notifications, so both view must be editable.
Options:
Bind the same data to two controls.
Use a visual brush and duplicate input on the real control.
Create a custom graphing control that can output two parts of the graph at once.
If binding to two controls is too slow, then I think you need to rethink your application. The very fact that you have so much data displayed at once that you need a dual view to see separate parts is disturbing. That should raise a red flag. The red flag would notify you that, "What I need, and what I have is different." And you should go back to the drawing board and find out what you really need.
Otherwise, it might be best to create a custom control. The graph is rendering in its entirety even though you only need small portions displayed. If you had your own custom control you could speed up the entire app by rendering only visible portions at a time, and splitting within the control.
Related
I'm trying to improve the real-time log display of a rather complicated system. Currently there is a significant tradeoff between situational awareness (improved by displaying only high-level events) and reproducibility (which requires lots and lots of details). The obvious approach is to structure the details as a tree and allow their display to be expanded and collapsed.
The existing implementation uses just a RichTextbox. I'd like to use MVVM for my rewrite, not only because it's a best practice but also to enable behavior such as "Show details of an operation while it is ongoing, then automatically collapse them once it has successfully completed.".
As a further complication, the log records have some structure, which needs to be displayed in a columnar format. The nesting should be accomplished using the description column only (can be through pure indentation, as depicted, or using tree lines).
Can you recommend a way to map this data structure onto WPF template elements? There doesn't seem to be a way to produce a sequence of DataGridRow objects from a parent object. And because the rows do not have uniform height (some of the detail data is multiline), trying to use a WPF tree for the description and then synchronize the vertical placement of data in other columns doesn't seem feasible.
Will I have to give up on the idea of a data-bound model, and use more of a view-controller design with manual creation of WPF elements, with references to these elements such as DataGridRow stored explicitly within the in-memory tree structure, and implement collapsing and expanding by iterating through the child items and updating each descendant's Height property of the associated WPF element object?
And if I do break the data/presentation wall and use imperative coding style, is there something similar to WinForms SuspendLayout/ResumeLayout that I can use to achieve just a single layout pass per collapse, no matter how many rows changed in height?
Here is a quick mockup of the tree structure made using WPF+XAML. It's just a Grid without any document structure -- all the text elements are direct children of the Grid with manually set Grid.Row and Grid.Column attached properties. Clearly that would not be suitable for an implementation with new data and structure being generated in real time.
I've got a massive UI that I'm designing. The way that my employer wants it, there are at least 100 labels. Now, I've always thought that in cases like this, breaking up the UI into smaller custom controls was the ideal way to go. However, someone recently told me that custom controls are really only for code re-use. What is the actual suggested practice for this?
EDIT
The finished form will look like this:
Now, I'm using WPF for the UI, and I'm thinking of breaking this down into smaller bits.
Based on your image i see some repetitions, each of this repetitions could be a custom UserControl
But it depends on the usability is it easier to write a custom UserControl so do it but if it would reduce the readability of your code and it also adds additional complexity don't do it
here are an example of what could be separate UserControl's
the green ones are possible useful encapsulations of logic
the orange ones maybe need some not market stuff (don't know enough about your software)
the red ones are the maybe's based on the intern use (from the visual part they are repetitions so the should custom UserControl)
Since your UI is read-only, I'd suggest using a grid.
Are you new to WPF? To break the View into bits WPF offers you CustomControls and UserControls. They are two very similar things yet completely different from each other. CustomControls are Buttons, Labels, TextBoxes, DataGrids...etc. They are basically simple stand-alone controls. UserControls are groups of stand-alone controls serving a purpose such as example a Button and a ComboBox next to each other so user can select something in ComboBox and confirm that by clicking the Button.
If you wish to display data from database I suggest you DataGrid which will give you a table-alike look with rows and columns and all that. If you wish to place few buttons next to DataGrid on which the user may click to insert a new row or to edit a certain cell then I suggest you to wrap all that with a UserControl which you can reuse in other places where you have to display and change data from database too.
You should be using a datagrid and can customize its template to render individual cells as Textblock (lighter version of Label) from a rendering perspective. The main difference between Textblock and Label is very minor things such as access keys and disabled state behavior. But from a WPF object hierarchy - Textblocks are much lighter. But besides that point - from your employer perspective - once you have customized the grid template and render them (so as they look as textblocks/labels) - your employer should have no problems.
Also as somebody suggested above - if you want to logically break sections of the UI since they maybe coming from a different table in db - then User controls is the way to go (for maintainability of code)
Let me know if you are looking for more technical details or need help further technically.
There is nothing wrong in making and using custom controls or user controls or defining some data templates which will be reused depending on how your data is organized.
For sure the UI looks pretty messy and some sort of grid should be used with templates for example where there is similar data. I also have the suggestion and first think about the data and the functionality before starting and let the UI be driven by that. For sure you will the reuse controls/templates. If you think in front on the model and behavior the UI can afterwards more easily changed.
Create your viewmodel correctly, implement the functionality in commands, use bindings, after that the UI will come naturally, reuse controls, use several grids, make the UI more user friendly using several regions, tabs, windows or anything that makes the user more comfortable.
In case a screen of a WPF application contains lots of primitive controls, its rendering becomes sluggish. What are the recommended ways to improve the responsiveness of a WPF application in such a case, apart from adding fewer controls and using more powerful videocard?
Is there a way to somehow use offscreen buffering or something like that?
Our team was faced with problems of rendering performance. In our case we have about 400 transport units and we should render chart of every unit with a lot of details (text labels, special marks, different geometries etc.).
In first our implementations we splitted each chart into primitives and composed whole unit's chart via Binding. It was very sad expirience. UI reaction was extremely slow.
So we decided to create one UI element per each unit, and render chart with DrawingContext. Although this was much better in performance aspect, we spent about one month improving rendering.
Some advices:
Cache everything. Brushes, Colors, Geometries, Formatted Texts, Glyphs. (For example we have two classes: RenderTools and TextCache. Rendering process of each unit addresses to shared instance of both classes. So if two charts have the same text, its preparation is executed just once.)
Freeze Freezable, if you are planning to use it for a long time. Especially geometries. Complex unfreezed geometries execute HitTest extremely slow.
Choose the fastest ways of rendering of each primitive. For example, there is about 6 ways of text rendering, but the fastest is DrawingContext.DrawGlyphs.
Use profiler to discover hot spots. For example, in our project we had geometries cache and rendered appropriate of them on demand. It seemed to be, that no improvements are possible. But one day we thought what if we will render geometries one time and cache ready visuals? In our case such approach happened acceptable. Our unit's chart has just several states. When data of chart is changed, we rebuild DrawingVisual for each state and put them into cache.
Of course, this way needs some investments, it's dull and boring work, but result is awesome.
By the way: when we turned on WPF caching option (you could find link in answers), our app hung up.
I've had the same perf issue with a heavily customized datagrid since one year, and My conclusion is:
there is basically nothing you can do
on your side (without affecting your
app, i.e.: having fewer controls or
using only default styles)
The link mentioned by Jens is great but useless in your case.
The "Optimizing WPF Application Performance" link provided by NVM is almost equally useless in my experience: it just appeals to common sense and I am confident you won't learn anything extraordinary either reading. Except one thing maybe: I must say this link taught me to put as much as I can in my app's resources. Because WPF does not reinstanciate anything you put in resource, it simply reuses the same resource over and over. So put as much as you can in there (styles, brushes, templates, fonts...)
all in all, there is simply no way to make things go faster in WPF just by checking an option or turning off an other. You can just pray MS rework their rendering layer in the near future to optimize it and in the meantime, try to reduce your need for effects, customized controls and so on...
Have a look at the new (.NET 4.0) caching option. (See here.)
I have met a similar problem and want to share my thoughts and founds. The original problem is caused by a virtualized list box that displays about 25 complex controls (a grid with a text block and a few buttons inside displaying some paths )
To research the issue I used the VisualStudio Application Timeline that allows to how much time it takes to render each control and PerfView to find out what actually WPF is doing to render each control.
By default it took about 12ms to render each item. It is rather long if you need to update the list dynamically.
It is difficult to use PerfView to analyse what heppens inside since WPF renders item in the parent-child hierarchy, but I got the common understanding about internall processes.
WPF does following to render each item in the list:
Parse template using XAML reader. As far as I can see the XAML parsing is the biggest issue.
Apply styles
Apply bindings
It does not take a lot of time to apply styles and bindings.
I did following to improve performance:
Each button has its own template and it takes a lot of time to render it. I replaced Buttons with Borders. It takes about 4-5ms to render each item after that.
Move all element settings to styles. About 3ms.
Create a custom item control with a single grid in the template. I create all child elements in code and apply styles using TryFindResources method. About 2ms in the result.
After all these changes, performance looks fine but still most time is spent on loding the ListControl.Item template and the custom control template.
4. The last step: replace a ListControl with Canvas and Scrollbar controls. Now all items are created at runtime and position is calculated manually using the MeasureOverride and ArrangeOverride methods. Now it takes <1ms to render each item from which 0.5ms is spent on TextBlock rendering.
I still use styles and bindings since they do not affect performance a lot when data is changed. You can imagine that this is not a WPF solution. But I fave a few similar lists in the application and it is possible not to use templates at all.
In order to make a convenient UI for an .Net 2.0 Winforms application I am working on, I have need for a control that I'm pretty sure goes beyond the "out of the box" behavior of any standard control. A mock-up of what I'm trying to achieve follows:
Mock up http://www.claware.com/images/temp/mockup.png
Essentially, this part of the application attempts to parse words into syllables from tribal languages (no dictionary to refer to; any and all unicode characters are possible.) By the time the user gets this far, he has already defined the vowels / consonants in his language and some other configuration. There is then an iterative process of (1) the application guesses which syllables exist in the language based on some rules, (2) the user refines the guesses, selecting the correct parsings or manually parsing a word, (3) the application "learns" from the user's feedback and makes smarter guesses, (4) repeat until the data is "good enough" to move on.
The control needs to present each word (the grey headers), then all the syllable break guesses (the white areas with dots separating the parts of words.) There is also a way to manually enter a parsing, which will display a text area and save button (at the bottom of the mockup.) When the user hovers over a guess, the background changes and "accept / reject" buttons appear. Clicking on the accept, or entering a manual parsing, removes the entire word from the list. Clicking the reject button removes just that item.
I'm by no means 100% sold on the formatting I have above, but I think you can get a general idea of the types of formatting and functional control I need. The control will also scroll vertically--there may be thousands of words initially.
My question for you experienced WinForms developers is: where to start? I would really, really like to stay within the .Net core framework and extend an existing control as opposed to a third-party control. (At the risk of starting a religious war: yes, I suffer from NIH-syndrome, but it's a conscious decision based on a lot of quick-fix solutions but long-term problems with 3rd party controls.) Where can I get the most "bang for my bucK" and the least reinventing the wheel? ListView? ListBox? ScrollableControl? Do I need to go all the way back to Control and paint everything manually? I appreciate any help that could be provided!
[Edit] Thanks everyone for the ideas. It seems like the most elegant solution for my purposes is to create a custom control consisting of a FlowLayoutPanel and a VScrollBar. The FlowLayoutPanel can contain instances of the custom controls used for each word. But the FlowLayoutPanel is virtual, i.e. it only contains those instances which are visible (and some "just out of scroll"). The VScrollBar events determine what needs to be loaded. A bit of code to write, but isn't too bad and seems to work well.
I would look at the TableLayoutPanel and FlowLayoutPanel controls. These will let you organize a series of controls with moderate ease in a vertical fashion. I would then create a UserControl that consists of a label and 2 buttons. The UserControl will expose properties like Text and events that are exposed for the button clicks.. For each entry in the list, you will create an instance of the UserControl, assign the text value, and handle the click events. The instance will be placed in the Table/Flow panel in the correct order. Both of those layout panels do allow for inserting items between other items so you can add/remove items from the list dynamically.
Edit:
Given the length of what you are trying to render, I would consider using the DataGridView and do some custom rendering to make it perform how you want it to work. Using the rendering events of the DGV you can merge columns, change background colors (like highlighting the dark gray lines), turn on/off the buttons, and handle changing the grid into edit mode for your rows to allow modification or inserting of new values. This method would easily handle large datasets and you could bind directly to them very easily.
Well, this certainly looks like a candidate for a custom component that you should be creating yourself. You can create this using standard .Net drawing commands along with a text-box, and a regular button control.
Now you want to find out where to start.
Create a Windows Forms Control Library project.
Drop in the textbox and the button control.
The panel drawing code should preferably be done by code. This can be done using the regular GDI+ commands.
Edit:
Here's another idea, and one that I've practically used in my own project with great success.
You could use a web-browser control in the app, and show your data as html. You could update the source of the web-browser control based on the input in the textbox, and clicking on the links in the web browser control will give you the event that you can trap to do some action. Your CSS will work.
I used this technique to build the 'desktop' in an app I made called 'Correct Accounting Software'. People loved the desktop so much that it is one of the best loved features of the app.
Here's how I would do it:
Create a custom control. In this custom control, have a ListBox atop a LinkButton, and when the LinkButton is clicked you can make it give way to a TextBox. The ListBoxes will have the top row unselectable... you can probably get the rest from there. When you get your list of words, fill a Scrollable of some kind with one control for each word:
(foreach String word in words){
myScrollable.add(new MyComponent(word));
}
From there, I'm not sure what you want to do with the boxes or the data, but that's my initial idea on the UI setup.
Use the WebBrowser control and generate the HTML markup into it using DocumentStream or DocumentText.
My company's product is, at its core, a framework for developing metadata-driven UIs. I don't know how to characterize it less succinctly than that, and hope I won't need to for purposes of this question, but we'll see.
I've been trying to come up to speed on WPF, and have been building UI prototypes here and there, and recently I decided to see if I could use Expression Blend to help with the design of these UIs. And I'm pretty mystified at this point.
It appears to me as though Expresssion Blend is designed with the expectation that you already know all of the objects that are going to be present in the UI at design time. But our program generates these object dynamically at runtime.
For instance, a data row might be presented in a horizontal StackPanel containing alternating TextBlocks (for captions) and TextBoxes (for data fields). The number of these objects depends on metadata about the number of columns in the data row. I can, pretty readily, write code that runs through a metadata record and populates a StackPanel dynamically, setting up the binding of all of the controls to properties in either the data or metadata. (A TextBox's Width might be bound to metadata, while its Text is bound to data.)
But I can't even begin to figure out how to do something like this in Expression Blend. I can manually create all these controls, so that I have a set of controls that I can apply styles to and work out the visual design of the app, but it's really a pain to do this.
I can write code that goes through my data model and emits XAML for all these controls, I suppose, and then copy and paste it. But I'm going to feel really stupid if it turns out there a way to do this sort of thing in Expression Blend and I've dropped back and punted because I'm too dim to figure out the right way to think of it.
Is this enough information for someone to try formulating an answer?
I think expression blend is a very good choice ESPECIALLY if you want to design dynamic, datadriven UIs. You can use Blend to desgin DataTemplates that define how a single dataobject is to be presented to the screen. For example, if you have an object of type Person you can define the controls like textboxes, border etc. that then are to be generated automatically for each Person in your list.
You can then fill your ItemsControls (DataGrid, ListView, TreeView ...) with those dynamic databojects and WPF knows how to render them. If a Persopn is removed form the list, the generated visual item will be removed too. This is just a simple example the whole notion of dynamic data is deeply baked into WPF and you can access these features using blend.
To be able to design the datatemplates in expression blend you need to provide sample designtime data.
Also to be effective with all this it is of high advantage if you stick with the MVVM design.