Add support for the "blink" graphic rendition attribute #7388
Comments
j4james
added
the
Issue-Feature
ghost
added
Needs-Triage
zadjii-msft
added
Area-VT
|
This should perhaps check SystemParametersInfo SPI_GETCLIENTAREAANIMATION, whose description specifically mentions blinking. The blinking frequency in the sample GIF is only one change per second, though. |
|
The current implementation is using the existing timer from the I should also mention that my current implementation only blinks at half the cursor rate (which is more or less what a DEC terminal would do), so in the time it takes the cursor to blink on and off, the blinking attribute would only have rendered half a cycle (i.e. it would either be on the whole time, or off the whole time). I think that works nicely for the DEC style of blinking, but may be too slow for a PC style blink. |
zadjii-msft
removed
the
Needs-Triage
|
For accessibility purposes, we will need to inherit the OS cursor blink rate configured in "Keyboard Properties" as blinking too fast can cause visual impairment but most users will not want it going too slow either. We can define the blink rate ourselves but I think OS inheritance will be better. |
I just said the current implementation is using the system cursor blink rate ( |
|
I honestly have no clue. But I don't believe conhost is responsible for the keyboard settings and properties so I just wanted to put that out there. Since only the Terminal team have access to Windows source code, they can tell if it calls the same code. I do think we should document how to access the setting when this feature is implemented. |
|
Personally I like the DEC style, even though most of the home computers and personal computers that I've used have (in text mode) used the IBM MDA style of just AND-gating the foreground pixels, and even though there's the argument that Windows Terminal should stick with its PC heritage here. Note however that the EGA in 16-colour graphics mode blinked pixels by switching between two sets of 8 palette entries, and could implement DEC-style blinking, so the PC heritage is not quite as one-sided as one might think. Also note the existence of DECBBSM, which is DEC private mode 116. If off, only the foreground pixels change intensity. If on, both foreground and background pixels change intensity. The far more important configuration option that you'll have to set up is the ability to turn blinking off. Observe that the ability to disable the blink graphic rendition for text is a selectable option in terminal emulators such as (among others) PowerTerm WBT, Reflection ZFE, Foxit KoalaTerm, Terminal.app in MacOS (Apple doco), iTerm2, Konsole, tilix, and PuTTY. |
Thanks for the feedback. I'm leaning that way too at the moment.
Yeah thanks, I am aware of that, but I'm not looking to do anything that fancy yet. I'm just trying to get us to the level of a basic VT100 terminal for now.
In the current implementation you can at least turn it off via the cursor blink settings, and if it's easy I might also add in the animation option that KalleOlaviNiemitalo was suggesting. If anyone still has complaints after that, we can always add more options, but I suspect the reality is that most people will never encounter a blinking attribute unless they're actively looking for one. Do you have any examples of software using blinking that some people might find annoying? |
|
Would EDIT: I see https://github.com/microsoft/terminal/pull/7490/files#diff-f30650c558f3ea3f151ac2c2bea3d9ebR135 already detects it but it's not being handled. This would be a really neat feature to have, given the emergence of "ansi dashboards" nowadays. It'd also set the windows terminal ahead of many (most) other emulators, too. |
ghost
added
Resolution-Fix-Committed
This PR adds support for the _blink_ graphic rendition attribute. When a character is output with this attribute set, it "blinks" at a regular interval, by cycling its color between the normal rendition and a dimmer shade of that color. The majority of the blinking mechanism is encapsulated in a new `BlinkingState` class, which is shared between the Terminal and Conhost implementations. This class keeps track of the position in the blinking cycle, which determines whether characters are rendered as normal or faint. In Windows Terminal, the state is stored in the `Terminal` class, and in Conhost it's stored in the `CONSOLE_INFORMATION` class. In both cases, the `IsBlinkingFaint` method is used to determine the current blinking rendition, and that is passed on as a parameter to the `TextAttribute::CalculateRgbColors` method when these classes are looking up attribute colors. Prior to calculating the colors, the current attribute is also passed to the `RecordBlinkingUsage` method, which keeps track of whether there are actually any blink attributes in use. This is used to determine whether the screen needs to be refreshed when the blinking cycle toggles between the normal and faint renditions. The refresh itself is handled by the `ToggleBlinkingRendition` method, which is triggered by a timer. In Conhost this is just piggybacking on the existing cursor blink timer, but in Windows Terminal it needs to have its own separate timer, since the cursor timer is reset whenever a key is pressed, which is not something we want for attribute blinking. Although the `ToggleBlinkingRendition` is called at the same rate as the cursor blinking, we actually only want the cells to blink at half that frequency. We thus have a counter that cycles through four phases, and blinking is rendered as faint for two of those four. Then every two cycles - when the state changes - a redraw is triggered, but only if there are actually blinking attributes in use (as previously recorded). As mentioned earlier, the blinking frequency is based on the cursor blink rate, so that means it'll automatically be disabled if a user has set their cursor blink rate to none. It can also be disabled by turning off the _Show animations in Windows_ option. In Conhost these settings take effect immediately, but in Windows Terminal they only apply when a new tab is opened. This PR also adds partial support for the `SGR 6` _rapid blink_ attribute. This is not used by DEC terminals, but was defined in the ECMA/ANSI standards. It's not widely supported, but many terminals just it implement it as an alias for the regular `SGR 5` blink attribute, so that's what I've done here too. ## Validation Steps Performed I've checked the _Graphic rendition test pattern_ in Vttest, and compared our representation of the blink attribute to that of an actual DEC VT220 terminal as seen on [YouTube]. With the right color scheme it's a reasonably close match. [YouTube]: https://www.youtube.com/watch?v=03Pz5AmxbE4&t=1m55s Closes #7388
This PR adds support for the _blink_ graphic rendition attribute. When a character is output with this attribute set, it "blinks" at a regular interval, by cycling its color between the normal rendition and a dimmer shade of that color. The majority of the blinking mechanism is encapsulated in a new `BlinkingState` class, which is shared between the Terminal and Conhost implementations. This class keeps track of the position in the blinking cycle, which determines whether characters are rendered as normal or faint. In Windows Terminal, the state is stored in the `Terminal` class, and in Conhost it's stored in the `CONSOLE_INFORMATION` class. In both cases, the `IsBlinkingFaint` method is used to determine the current blinking rendition, and that is passed on as a parameter to the `TextAttribute::CalculateRgbColors` method when these classes are looking up attribute colors. Prior to calculating the colors, the current attribute is also passed to the `RecordBlinkingUsage` method, which keeps track of whether there are actually any blink attributes in use. This is used to determine whether the screen needs to be refreshed when the blinking cycle toggles between the normal and faint renditions. The refresh itself is handled by the `ToggleBlinkingRendition` method, which is triggered by a timer. In Conhost this is just piggybacking on the existing cursor blink timer, but in Windows Terminal it needs to have its own separate timer, since the cursor timer is reset whenever a key is pressed, which is not something we want for attribute blinking. Although the `ToggleBlinkingRendition` is called at the same rate as the cursor blinking, we actually only want the cells to blink at half that frequency. We thus have a counter that cycles through four phases, and blinking is rendered as faint for two of those four. Then every two cycles - when the state changes - a redraw is triggered, but only if there are actually blinking attributes in use (as previously recorded). As mentioned earlier, the blinking frequency is based on the cursor blink rate, so that means it'll automatically be disabled if a user has set their cursor blink rate to none. It can also be disabled by turning off the _Show animations in Windows_ option. In Conhost these settings take effect immediately, but in Windows Terminal they only apply when a new tab is opened. This PR also adds partial support for the `SGR 6` _rapid blink_ attribute. This is not used by DEC terminals, but was defined in the ECMA/ANSI standards. It's not widely supported, but many terminals just it implement it as an alias for the regular `SGR 5` blink attribute, so that's what I've done here too. ## Validation Steps Performed I've checked the _Graphic rendition test pattern_ in Vttest, and compared our representation of the blink attribute to that of an actual DEC VT220 terminal as seen on [YouTube]. With the right color scheme it's a reasonably close match. [YouTube]: https://www.youtube.com/watch?v=03Pz5AmxbE4&t=1m55s Closes #7388 (cherry picked from commit d1671a0)
|
🎉This issue was addressed in #7490, which has now been successfully released as Handy links: |
Description of the new feature/enhancement
We already parse and store the
SGR 5blink attribute, but we don't actually "render" it. This is the only remaining DEC attribute that we don't yet support, and it's widely implemented by other terminal emulators. It's probably not essential, but it can occasionally be useful, and is fun for creating simple "animations".Proposed technical implementation details (optional)
I've been experimenting with this on the conhost side of things, and I found it could quite easily be implemented by hooking into the existing timer routine in the
CursorBlinkerclass, triggering a redraw of the render target every cycle or two. But since we don't want that overhead if nothing is actually blinking, I have a flag that is enabled whenever a blink attribute is encountered in the renderer, and only trigger the redraw if that flag has been set. The flag is cleared before every redraw, so once the viewport is free of any blinking attributes, the redraws will stop.As for how the blinking is actually rendered, there are two main approaches, demonstrated in the images below. On the left is what I call the "PC style", where the text actually disappears. On the right is the "DEC style" (the way I believe most DEC terminals worked), where the text is just dimmed. Looking at the two screen shots you can see the pros and cons of each approach. In the "Christmas Card", the blinking lights on the tree clearly look better when they just dim, instead of disappearing. The BBS advert, on the other hand, is completely dependent on the PC style to produce the changing text effect.
Having grown up with BBSes, I'm kind of partial to the PC style, but looking at things practically, I do think the DEC style is better. If you actually wanted to use blink to draw attention to some text, it's way better if you can still actually read that text. That's not so easy if it's constantly turning invisible, unless you also have a high blink rate, but that can be really annoying. For general text, a slow, subtle blink effect does seem more appropriate.
That said, I don't know of any other terminal emulators that actually implement the DEC style, so if we think it's important being compatible with other terminals emulators (rather than the actual terminals), then that's a factor to consider as well.
Ideally we'd one day have different compatibility modes, so we could actually support both options, but for now I do think we need to pick one to start with. I can't make up my mind which I prefer, so I'm happy to go with whatever the general consensus is here.
The text was updated successfully, but these errors were encountered: