Re: Intent to unship: mozmm CSS unit.
I thought there was also a legitimate use-case for displaying content "life size", e.g. if you wanted to draw a ruler on a tablet. But if the CSSWG doesn't agree after all this time, just drop it I guess. (Though I think there's something slightly broken about how Web developer needs are bubbling up to WGs. For example GeometryUtils hasn't been implemented by other browsers, who apparently detect no pressure from Web developers to solve the use-cases it solves, e.g. computing px offsets between arbitrary elements even if they're in a DOM subtree with a scale transform. Yet I ran into that problem pretty quickly while coding a Web UI. Maybe I'm just strange...) Rob -- lbir ye,ea yer.tnietoehr rdn rdsme,anea lurpr edna e hnysnenh hhe uresyf toD selthor stor edna siewaoeodm or v sstvr esBa kbvted,t rdsme,aoreseoouoto o l euetiuruewFa kbn e hnystoivateweh uresyf tulsa rehr rdm or rnea lurpr .a war hsrer holsa rodvted,t nenh hneireseoouot.tniesiewaoeivatewt sstvr esn ___ dev-platform mailing list dev-platform@lists.mozilla.org https://lists.mozilla.org/listinfo/dev-platform
Re: Intent to unship: mozmm CSS unit.
On 11/15/2017 11:51 AM, Jonathan Kew wrote: > On 15/11/2017 03:32, Emilio Cobos Álvarez wrote: > >> So, I've looked through and I haven't found any related minutes (I've >> looked for minutes that mentioned "physical", of which they were many, >> but all related to physical properties and scroll snapping, and for >> minutes that mentioned mozmm, of which they were none). >> >> I've found two threads in the w3c mailing list that look related, one >> which is basically the discussion that lead to all units being changed >> to be relative to the density of the display: >> >> https://lists.w3.org/Archives/Public/www-style/2010Jan/0058.html >> >> And someone interjecting on a thread about adding the Q unit: >> >> https://lists.w3.org/Archives/Public/www-style/2013Nov/0302.html > > It sounds like you may have missed the lengthy "[css3-values] Physical > length units" thread (spun off from "[css3-mediaqueries] DPI in > resolution media queries") that started in mid-Feb 2012: Indeed, thanks Jonathan! I bet the minutes at https://lists.w3.org/Archives/Public/www-style/2012Feb/0530.html were what David was referring to. > https://lists.w3.org/Archives/Public/www-style/2012Feb/0627.html > > I haven't re-read all that thread to see whether anything resembling a > consensus seemed to be emerging my (vague) recollection is that > there's generally little enthusiasm in the WG for a "true" physical > unit, with perhaps one or two vocal dissenters ___ dev-platform mailing list dev-platform@lists.mozilla.org https://lists.mozilla.org/listinfo/dev-platform
Re: Intent to unship: mozmm CSS unit.
On 15/11/2017 03:32, Emilio Cobos Álvarez wrote: So, I've looked through and I haven't found any related minutes (I've looked for minutes that mentioned "physical", of which they were many, but all related to physical properties and scroll snapping, and for minutes that mentioned mozmm, of which they were none). I've found two threads in the w3c mailing list that look related, one which is basically the discussion that lead to all units being changed to be relative to the density of the display: https://lists.w3.org/Archives/Public/www-style/2010Jan/0058.html And someone interjecting on a thread about adding the Q unit: https://lists.w3.org/Archives/Public/www-style/2013Nov/0302.html It sounds like you may have missed the lengthy "[css3-values] Physical length units" thread (spun off from "[css3-mediaqueries] DPI in resolution media queries") that started in mid-Feb 2012: https://lists.w3.org/Archives/Public/www-style/2012Feb/0627.html I haven't re-read all that thread to see whether anything resembling a consensus seemed to be emerging my (vague) recollection is that there's generally little enthusiasm in the WG for a "true" physical unit, with perhaps one or two vocal dissenters. JK ___ dev-platform mailing list dev-platform@lists.mozilla.org https://lists.mozilla.org/listinfo/dev-platform
Re: Intent to unship: mozmm CSS unit.
Hi David, On 11/15/2017 02:27 AM, L. David Baron wrote: > On Sunday 2017-11-12 16:11 +0100, Emilio Cobos Álvarez wrote: >> Hi, >> >> In bug 1416564 I intend to remove the mozmm CSS unit. >> >> This unit is Mozilla-only, has no spec, and is unused in all our >> codebase (except for two tests, one of those which tests the unit itself). >> >> This unit was introduced experimentally in bug 537890, our browser >> chrome code used it in bug 588464, and all that is gone since then. >> >> Given there's no spec, and no usage in the wild as far as I can tell, I >> think we should try to remove it. >> >> Thoughts? > > So I think this unit had a pretty strong use case: designing of > touch user interfaces, where touch targets need to be at least a > certain physical size in order to work well with human fingertips. > > There's also a risk from having physical units in CSS that we > learned the last time CSS had them: designs that work with some > ratios of physical units to other units and break with other ratios. > > I've been meaning to dig up the minutes from the time we tried to > get the CSS working group to add these units, but I haven't had a > chance over the last few days. I think if those minutes suggest > that the working group was receptive to adding such units to the > spec (or perhaps even agreed to do so, but then they were never > actually added by the editor), I'd tend to think we should leave our > implementation, whereas if the minutes suggest that the balance of > working group opinion was against them, then we should remove it. > Would you be able to try searching for those minutes? I know > there's been an in-person discussion (and I think it was at a > meeting at a TPAC a few yeas ago). So, I've looked through and I haven't found any related minutes (I've looked for minutes that mentioned "physical", of which they were many, but all related to physical properties and scroll snapping, and for minutes that mentioned mozmm, of which they were none). I've found two threads in the w3c mailing list that look related, one which is basically the discussion that lead to all units being changed to be relative to the density of the display: https://lists.w3.org/Archives/Public/www-style/2010Jan/0058.html And someone interjecting on a thread about adding the Q unit: https://lists.w3.org/Archives/Public/www-style/2013Nov/0302.html Then there's a relevant, more recent, csswg-drafts issue: https://github.com/w3c/csswg-drafts/issues/614 But from the comments in there it doesn't seem to be a strong sense about it being useful, but mostly the opposite (and actually [1] may be a legit bug we should look into if we keep it in tree, or behavior we should document). Also, some other alternatives for that use case that work in all browsers are mentioned, like resolution-dependent media queries. What do you think David? I landed the patch this morning, and I'd really prefer it to stick. Given the recent cssswg-drafts issue where the CSSWG members were not really excited about it, I think it's ok to remove it, but happy to back the patch out if you think it's not. -- Emilio [1]: https://github.com/w3c/csswg-drafts/issues/614#issuecomment-254679777 > > -David > ___ dev-platform mailing list dev-platform@lists.mozilla.org https://lists.mozilla.org/listinfo/dev-platform
Re: Intent to unship: mozmm CSS unit.
On Sunday 2017-11-12 16:11 +0100, Emilio Cobos Álvarez wrote: > Hi, > > In bug 1416564 I intend to remove the mozmm CSS unit. > > This unit is Mozilla-only, has no spec, and is unused in all our > codebase (except for two tests, one of those which tests the unit itself). > > This unit was introduced experimentally in bug 537890, our browser > chrome code used it in bug 588464, and all that is gone since then. > > Given there's no spec, and no usage in the wild as far as I can tell, I > think we should try to remove it. > > Thoughts? So I think this unit had a pretty strong use case: designing of touch user interfaces, where touch targets need to be at least a certain physical size in order to work well with human fingertips. There's also a risk from having physical units in CSS that we learned the last time CSS had them: designs that work with some ratios of physical units to other units and break with other ratios. I've been meaning to dig up the minutes from the time we tried to get the CSS working group to add these units, but I haven't had a chance over the last few days. I think if those minutes suggest that the working group was receptive to adding such units to the spec (or perhaps even agreed to do so, but then they were never actually added by the editor), I'd tend to think we should leave our implementation, whereas if the minutes suggest that the balance of working group opinion was against them, then we should remove it. Would you be able to try searching for those minutes? I know there's been an in-person discussion (and I think it was at a meeting at a TPAC a few yeas ago). -David -- 𝄞 L. David Baron http://dbaron.org/ 𝄂 𝄢 Mozilla https://www.mozilla.org/ 𝄂 Before I built a wall I'd ask to know What I was walling in or walling out, And to whom I was like to give offense. - Robert Frost, Mending Wall (1914) signature.asc Description: PGP signature ___ dev-platform mailing list dev-platform@lists.mozilla.org https://lists.mozilla.org/listinfo/dev-platform
Re: Intent to unship: mozmm CSS unit.
Also, another thought I just had about this unit: Probably if we don't unship it, then privacy.resistFingerPrinting should do something about it, since it allows to calculate the DPI of the screen trivially using CSSOM accessors. Filed bug 1416574. ___ dev-platform mailing list dev-platform@lists.mozilla.org https://lists.mozilla.org/listinfo/dev-platform
Re: Intent to unship: mozmm CSS unit.
On 12/11/2017 16:36, Emilio Cobos Álvarez wrote: On 11/12/2017 05:14 PM, Xidorn Quan wrote: I doubt if there is anything changed since then (except that we still haven't pushed this on csswg), so maybe we still shouldn't remove it. I'm definitely ok with adding something like that if there's an use-case and demand for it... But given I haven't been able to find any actual usage of it, not even internal, I suspect there isn't much? Related, but not exactly the same (so please carry on with whatever you want to do with mozmm): browser chrome at least has a usecase for a CSS unit that matches 1 device pixel on various dpis, because layout already adjusts some (but not all!) sizes to align to device pixels and this causes issues when we basically want "make a line that is 1 device pixel wide". We can workaround the lack of such a unit with CSS variables and something like: @media (min-resolution: 2dppx) { --mydevicepixel: 0.5px; } but it's tedious to do for dpi on windows (which has a lot more variations than on macOS) and probably a perf nightmare. ~ Gijs ___ dev-platform mailing list dev-platform@lists.mozilla.org https://lists.mozilla.org/listinfo/dev-platform
Re: Intent to unship: mozmm CSS unit.
On 12/11/2017 16:56, Gijs Kruitbosch wrote: because layout already adjusts some (but not all!) sizes to align to device pixels For reference: https://bugzilla.mozilla.org/show_bug.cgi?id=477157 and friends. ~ Gijs ___ dev-platform mailing list dev-platform@lists.mozilla.org https://lists.mozilla.org/listinfo/dev-platform
Re: Intent to unship: mozmm CSS unit.
On 11/12/2017 05:14 PM, Xidorn Quan wrote: > IIRC, we have discussed unshipping this somewhere before we added its > support to stylo (maybe the Taipei meeting this year?) and dbaron said > that mozmm provides an ability to size something based on physical size > which is not directly possible in any other unit, so this is something > we may want to put into spec rather than removing. Huh, I don't recall being part of that discussion, but that sounds plausible. > I doubt if there is anything changed since then (except that we still > haven't pushed this on csswg), so maybe we still shouldn't remove it. I'm definitely ok with adding something like that if there's an use-case and demand for it... But given I haven't been able to find any actual usage of it, not even internal, I suspect there isn't much? In any case, having features exposed to the web without any specification or other browsers supporting it looks like the wrong thing to me, and the kind of thing that may bite in the future (and that makes Servo's life harder, too), so I still would like to unship it if possible. If afterwards we figure out that this is needed, I'd be happy to see it added as something every browser agrees with. My patch keeps the ability to convert physical millimeters to CSS pixels, so it'd be trivial to reintroduce if Mozilla decides to push for this feature and other browsers agree. -- Emilio ___ dev-platform mailing list dev-platform@lists.mozilla.org https://lists.mozilla.org/listinfo/dev-platform
Re: Intent to unship: mozmm CSS unit.
On Mon, Nov 13, 2017, at 02:11 AM, Emilio Cobos Álvarez wrote: > Hi, > > In bug 1416564 I intend to remove the mozmm CSS unit. > > This unit is Mozilla-only, has no spec, and is unused in all our > codebase (except for two tests, one of those which tests the unit > itself). > > This unit was introduced experimentally in bug 537890, our browser > chrome code used it in bug 588464, and all that is gone since then. > > Given there's no spec, and no usage in the wild as far as I can tell, I > think we should try to remove it. > > Thoughts? IIRC, we have discussed unshipping this somewhere before we added its support to stylo (maybe the Taipei meeting this year?) and dbaron said that mozmm provides an ability to size something based on physical size which is not directly possible in any other unit, so this is something we may want to put into spec rather than removing. I doubt if there is anything changed since then (except that we still haven't pushed this on csswg), so maybe we still shouldn't remove it. - Xidorn ___ dev-platform mailing list dev-platform@lists.mozilla.org https://lists.mozilla.org/listinfo/dev-platform
Intent to unship: mozmm CSS unit.
Hi, In bug 1416564 I intend to remove the mozmm CSS unit. This unit is Mozilla-only, has no spec, and is unused in all our codebase (except for two tests, one of those which tests the unit itself). This unit was introduced experimentally in bug 537890, our browser chrome code used it in bug 588464, and all that is gone since then. Given there's no spec, and no usage in the wild as far as I can tell, I think we should try to remove it. Thoughts? -- Emilio ___ dev-platform mailing list dev-platform@lists.mozilla.org https://lists.mozilla.org/listinfo/dev-platform