On my copy of Safari—Version 4.0.5 (5531.22.7), Mac OS X Version 10.5.8 Build 9L30 - it has the amusing feature of loading windows with names visible rather than invisible.
Edit: Oh, and the formatting got a little amusing when I posted a reply with names hidden.
My apologies for the version which ended up being deployed, which is well below the level of quality I normally aspire to.
I’ve just committed a major revision to the AK script which I’d design-sketched a while ago but had shelved pending a better handle on how to write unit tests for LW.
The newer version should be compatible with IE, more responsive than previous versions, and should allow more fine-tuning in future versions.
However I can’t offer any guarantees as to when Tricycle are likely to pull the changes and deploy them to the live server. I don’t know yet how best to coordinate with the Tricycle crew when working on the LW codebase, and the current situation was also a consequence of that.
Morendil your latest patch has been applied and deployed. We took the liberty of inverting the preference for two reasons: Firstly it made more sense for the checkbox to be checked when the kibitzer was enabled and secondly we added a note that full compatibility requires Firefox for now, the text of which was hard to phrase with the uncheck to enable behaviour.
I note that the revision you’ve pulled is 0.5, which should be a lot more cross-browser than previous ones. I’ve done some testing under IE, Chrome and Safari.
The original relied on XPath queries to locate page elements corresponding to scores and names. Now that the script is part of the LW codebase, I was able to simplify the script’s operation by including a special CSS stylesheet when the Anti-Kibitzer is active, and acting directly on the display rules. This is faster and more compatible.
On my copy of Safari—Version 4.0.5 (5531.22.7), Mac OS X Version 10.5.8 Build 9L30 - it has the amusing feature of loading windows with names visible rather than invisible.
Edit: Oh, and the formatting got a little amusing when I posted a reply with names hidden.
My apologies for the version which ended up being deployed, which is well below the level of quality I normally aspire to.
I’ve just committed a major revision to the AK script which I’d design-sketched a while ago but had shelved pending a better handle on how to write unit tests for LW.
The newer version should be compatible with IE, more responsive than previous versions, and should allow more fine-tuning in future versions.
However I can’t offer any guarantees as to when Tricycle are likely to pull the changes and deploy them to the live server. I don’t know yet how best to coordinate with the Tricycle crew when working on the LW codebase, and the current situation was also a consequence of that.
Definitely needs more testing and tweaking: as I said above I wasn’t expecting these changes to get pulled in.
I’ve already committed a change that corrects the second issue, I’ve just reproduced the first on my local Safari, and am looking into it.
Morendil your latest patch has been applied and deployed. We took the liberty of inverting the preference for two reasons: Firstly it made more sense for the checkbox to be checked when the kibitzer was enabled and secondly we added a note that full compatibility requires Firefox for now, the text of which was hard to phrase with the uncheck to enable behaviour.
Great, thanks!
I note that the revision you’ve pulled is 0.5, which should be a lot more cross-browser than previous ones. I’ve done some testing under IE, Chrome and Safari.
The original relied on XPath queries to locate page elements corresponding to scores and names. Now that the script is part of the LW codebase, I was able to simplify the script’s operation by including a special CSS stylesheet when the Anti-Kibitzer is active, and acting directly on the display rules. This is faster and more compatible.