v 10.0 BETA 3C test

[MikeGale]

Hi Albert,

I've just run a few batch tests on this version over some staging web sites.

Observations:

1) I didn't see that @import issue. Thanks.

2) I didn't see those stray hard to close dialogs, not meaningful without more starts.

3) I got a few

The text content of the "a" element begins and ends with whitespace. This may not be wanted.

(These come from content which has been updated by an online CMS system.) It struck me that these would be a lot clearer if the whitespace was shown as visible characters. (Paragraph sign, dots for spaces, guillemets for tabs etc.)

4) For the colour difference message I get something like

The color difference between the foreground color and background color is 408. This is below the recommended value of 500. [P3, 2.2] (NOTE: Sufficient contrast is Priority 2 for images, Priority 3 for text; assuming text here)

My first thought was, what are these priority levels? I think incorporating the words, must, should and may would make the meaning clearer to all. It is worth considering a switch on these messages, show none, show musts, show musts and shoulds... They have booleanised the tests, but clearly a more thorough implementation would give better "recommendation resolution" though I wouldn't really expect to see that in the spec...
5) The colour difference messages continue to flag CSS sheets that are intended to be that way.

[Albert Wiersch]

1) I didn't see that @import issue. Thanks.

Great! Thanks for reporting this issue and confirming that it's fixed.

2) I didn't see those stray hard to close dialogs, not meaningful without more starts.

Yes, I'm still not sure about this but I will run an allocation profiler and see if there is anything that might be causing some type of memory corruption.

3) I got a few

The text content of the "a" element begins and ends with whitespace. This may not be wanted.

(These come from content which has been updated by an online CMS system.) It struck me that these would be a lot clearer if the whitespace was shown as visible characters. (Paragraph sign, dots for spaces, guillemets for tabs etc.)

I've made the highlight more like the editor, with a black background and green text. I don't think I've ever received any complaints about those colors. Below is an example. Please let me know what you think.

4) For the colour difference message I get something like

The color difference between the foreground color and background color is 408. This is below the recommended value of 500. [P3, 2.2] (NOTE: Sufficient contrast is Priority 2 for images, Priority 3 for text; assuming text here)

My first thought was, what are these priority levels? I think incorporating the words, must, should and may would make the meaning clearer to all. It is worth considering a switch on these messages, show none, show musts, show musts and shoulds... They have booleanised the tests, but clearly a more thorough implementation would give better "recommendation resolution" though I wouldn't really expect to see that in the spec...

Thanks. I'll see if I can make these messages more clear.

5) The colour difference messages continue to flag CSS sheets that are intended to be that way.

Yes, I don't have a workaround or solution for this yet. One possible solution I have in might is to be able to disable that check using special CSS comments.

[Albert Wiersch]

5) The colour difference messages continue to flag CSS sheets that are intended to be that way.

I thought of another possible solution for this. What about an option to disable a message with a specific message ID that ALSO exactly matches the text of a specific message.

So, I could change the color contrast message text to include the actual color values. If it is then disabled with this new feature, then it would only be disabled when the message text exactly matches the disabled message text, so it would be disabled for only two specific colors. This feature would also work for any other message that has a message ID and it would allow for even more specific message disabling, so it has a more general use.

[MikeGale]

That is a very powerful idea.

Fine grained control.

I really like it. For example, if a developer has a lot of web sites, has similar CSS on more than one, then it just works for him.

(If messages are revised it has a glitch, but even that could, potentially, be automated.)

[Albert Wiersch]

(If messages are revised it has a glitch, but even that could, potentially, be automated.)

Yes, that sounds like the biggest downside... if there is any change in the text of a message, then it is re-enabled, but that's probably not a big deal. I'm glad you like the idea. I will see if I can get that in before the release.