[PATCH] hgweb: fix search hint border-radius on standards-based browsers and remove old gecko-specific property

[Augie Fackler]

On Thu, Oct 08, 2015 at 05:58:31PM +0100, Gijs Kruitbosch wrote:
> On 08/10/2015 17:42, Anton Shestakov wrote:
> >08.10.2015, 23:53, "Gijs Kruitbosch" <gijskruitbosch at gmail.com>:
> >>On 08/10/2015 07:04, Anton Shestakov wrote:
> >>>  08.10.2015, 04:31, "Gijs Kruitbosch" <gijskruitbosch at gmail.com>:
> >>>>    - -moz-border-radius: 5px; /* this works only in camino/firefox */
> >>>  Removing this isn't required for the fix (afaict), and if you believe http://caniuse.com/border-radius , -moz-border-radius-only browsers have almost twice the user base of -webkit-border-radius-only browsers. So I'd say deleting this is not worth it.
> >>Apologies for my ignorance, but how do you see which browser only
> >>supported the prefixed version, and then see the usage only for those,
> >>on caniuse.com ?
> >Browser versions which only support prefixed properties have that badge with yellow background. Usage percentage is in tooltips of the versions.
> >
> >>The unprefixed version was introduced in Firefox 4 / Gecko 2.0, which is
> >>now 4.5 years old and has been unsupported for *years*, nevermind the
> >>versions before that. I'm not familiar with the webkit/blink side of
> >>things, but I don't understand for which browsers on today's web we
> >>would need to keep moz-border-radius.
> >My point is why remove this one. On one hand - sure, nobody will miss -moz-border-radius, but on the other, -webkit-border-radius stays, even though it has even lower usage (for reference, it's currently 0.08% for -moz- and 0.05% for -webkit-). I'd say either remove both or remove none. Come to think of it, letting them both stay has no downsides (e.g. unlike some IE hacks that could look puzzling, these properties are self-describing and clear).
> The reasoning for not removing the webkit one is quite simple: I don't know
> the webkit situation (and it has a history of continuing to support prefixed
> versions for a long time even when unprefixed ones become available),
> whereas the gecko one is pretty clear to me.
>
> Anyway, it's not really worth having a massive discussion over this removal
> of one line of css. :-)
> Do I need to submit a new patch to the list in order to get the unprefixed
> version added without removing either of the prefixed variants?

No, but it kind of sounds like it'd be worth dropping the webkit
prefixed version of the border-radius rule in a followup.

>
> ~ Gijs
> _______________________________________________
> Mercurial-devel mailing list
> Mercurial-devel at selenic.com
> https://selenic.com/mailman/listinfo/mercurial-devel