New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Don't display banner after keypress #173
Comments
You just have to press any key and the PASS banner will disappear. Decreasing the size of the banner is not as easy as in Photoshop. This is ASCII and a smaller banner will not be seen easily from a distance. |
Would it make sense to add " [Press any key to remove the banner] "? Just a question if this is really "any" key? Some of them (Esc, space, etc) have different meeting. So, maybe: " [Press Enter key to remove the banner] "? |
Any key will remove the banner, including Esc, Space or F1. |
Yes, but I think esc will also terminate the test (which I think is intended)? As we may also use other keys in the future for other actions, I think it may make sense to "reserve" a dedicated for this, like Enter?
|
Really no needs for this now. Enter is within "any" and pressing any key to remove the banner is obvious. |
To be honest, I was not ware that the banner can be removed until I looked at the code. Given the question in this bug, I suspect I may not be the only one, which is why I suggested adding the message. The reason why I proposed a dedicated key is to avoid a situation where some setting get accidentally changed. For example, even today one may press space (which changes scrolling) or escape (that terminates Memtest86) and these seem like common choices. Happy to provide a patch, we can discuss it there, unless this is a "strong NO" to leave this as is? |
Ah, I'm blind, "Enter" is already used:
|
I can see having the PASS result visible when the screen image is small would be nice. And yes the picture after removing PASS would still include the word pass somewhere, but it would take more work to see it (opening the image to zoom in by the buyer). I'm against requiring one specific key. I don't see a reason to limit it. I don't follow the issue if the user happens to pick a key that normally does something else. I mean it shouldn't stop the run until the message is gone, but... it's not hard to hit it a second time. A note telling the user how to remove the banner might be handy for nervous people, but it's also obvious once you try to do anything. As another way to show the run finished at a distance, you could also change the background color or change part of it (make a ring on the outside reflect overall status). Though then you get into dealing with colorblindness settings to tweak them (red/green is common, but others happen too). You could even get fancy and as the runs progress the screen changes color. Pass 2 gets another change, etc. But I see this all as behind anything truly functional still needed. Even code cleanup tasks. |
The problem with "any" key is that if may have a function defined - either currently or in the future. Today, when a key is pressed, Memtest86+ removes the banner but also at the same time executes the action assigned to that key, such as terminates for Esc, changes scrolling for space / Enter, etc. Having two actions triggered by a single event is one of the basic UX anti-patterns. We can change the code to only remove the banner and stop there, but it means changing the actions depending on the context - for example, in one situation Esc will terminate the test, in other - just remove the banner. This is another UX anti-pattern. Especially that a user may not have an intention to remove the banner. IMO, the easiest solution that also guarantees no UX changes in the future would be to assigned a dedicated key, such as "x" or "tab" to remove the banner, and this is why I suggested that path. |
I had the same concerns about the banner covering up RAM information, and there's zero indication I could have pressed anything to remove it. Stumbling upon this issue is the only way I found out. |
Things are actually worse for the FAIL banner, in that it obscures the list of failures and which memory module they are in - and a keypress does not make it go away. ... what it says. This is from the v6 x64 release iso, btw. And similar, but not quite, #130, I guess. |
I will definitely remove the banner definitely after any keypress. |
Is a fix for this to be expected any time soon? |
Within 2 weeks. #34 is the priority right now but this will be added soon. |
In the latest version (6.01.036922a.x64 GRUB ISO) the FAIL banner still doesn't go away after any key is pressed. Or rather, it does, but some bit of code is going "Oh no, we've detected a failure but the banner isn't showing!" and putting it straight back again. This doesn't happen with the PASS banner, which goes away exactly as it's supposed to. |
Got it. Will check & correct this week. |
I still have the bug in version 6.10.ce2c29e.x64 . When is it planned that this is fixed? It overlaps the information about the error. Or which earlier version does not have the banner so I can use that to see the problem of my ram? |
You can use the |
Adding to the current issue, if you disconnect the keyboard and then connect it again, it seems the keyboard doesn't work anymore, so it's impossible to exit the pass banner, and it's not possible to take a picture with the memory module and name. The reason I connect/disconnect the keyboard is that I have a test PC, and usually run Memtest in this test PC instead of the main workstation. The monitor receives both inputs and I switch the keyboard as needed. |
Current PASS message, covers RAM module speed, model, and manufacturer.
Please consider making it smaller to not cover important information.
I made an example:
.
.
The reason is that a lot of people is using Memtest PASS screenshot to prove that RAM stick is good, during sell auction on internet. It always helps to convince buyer that RAM was tested.
Current large Pass message is giving a reason to buyer, to suspect that seller provides fake passed test using different RAM sticks (because true speed, model, manufacturer can be hidden behind that large message)
The text was updated successfully, but these errors were encountered: