258663 2023-06-29 02:21:43 -0700 Initial scale of a page is not respected when both initial scale and viewport width are set 2023-07-06 09:09:49 -0700 1 1 1 Unclassified WebKit Layout and Rendering Safari 16 iPhone / iPad iOS 16 RESOLVED FIXED https://bugs.webkit.org/show_bug.cgi?id=247636 InRadar P2 Blocker --- 1 juwagn a_protyasha a_protyasha bfulgham simon.fraser webkit-bug-importer wenson_hsieh zalan oldest_to_newest 1964208 0 466863 juwagn 2023-06-29 02:21:43 -0700 Created attachment 466863 viewport bug in meta tag on iOS 16.5.1 Safari Hello, <meta> tag name="viewport" and as example content="width=600***** stopped to work correctly after upgrading from iOS 16.1.1 to 16.5.1 Let's assume you have physic screen size on iPad 810x1080 and you want to set viewport width to 600px, then issue will occur in landscape and portrait mode, the browser reports then not expected 600px in width, but depending on orientation physic width. If you set viewport width to 900px then issue will occur only in landscape mode since physic screen width in portrait mode is 810 and that is lower value than 900px, and issue won't occur, but in landscape mode physic size 1080 is higher than 900px and issue will occur, reports then unexpected width. If you set as viewport width 1200 or any higher value then issue won't occur for both orientations, since 1200px is higher value than physic screen size in both - height and width (in the example 810x1080). Do you have any idea which degradation occured in iOS 16.5.1 that makes such issue? Attached file viewport600.html with width=600 if you open it on Safari iOS 16.5.1 then you will notice the smallest screen size is then the width of your physic screen depending on orientation. If you change that value from 600 as example to width=1400 then it will work fine in both orientations. Actually it is blocker since many users experienced wrong behavior by prefering viewport 600 in width on their devices. On Android devices with common browsers such error doesn't occur. Thank you in advance! Sincerely yours. 1964214 1 juwagn 2023-06-29 02:59:41 -0700 Just in addition, in example i have set initial-scale=1 but usually i set minimum-scale and initial-scale instead fixed dynamically and as their value divided physic screen size by viewport width lets assume physic screen width for wished viewport width 600 landscape 1080 / 600 = 1.8 portrait 810 / 600 = 1.35 <meta name="viewport" content="width=600, user-scalable=yes, initial-scale=1.8, minimum-scale=1.8" /> or <meta name="viewport" content="width=600, user-scalable=yes, initial-scale=1.35, minimum-scale=1.35" /> in such case it will cause issue on iOS 6.5.1 and work fine on iOS 6.1.1 Also it seems when minimum-scale/initial-scale is lesser than 1, like 0.79 then OK, if higher values than 1, like 1.35 then issue occurs. 1964270 2 webkit-bug-importer 2023-06-29 09:03:08 -0700 <rdar://problem/111515914> 1964657 3 466898 juwagn 2023-07-01 17:47:20 -0700 Created attachment 466898 New example for testing on iOS 16.5.1/16.1.1/15.7.7 and Android devices New example for testing on iOS 16.5.1/16.1.1/15.7.7 and Android devices 1964658 4 juwagn 2023-07-01 17:53:09 -0700 Hello, just uploaded new example to test. It shows error happening on iOS 16.5.1 when at same time iOS 16.1.1 and 15.7.7 are OK, same OK as Android devices, so regression is since iOS 16.5.1 in Safari. Example allows to set meta width on the fly and in orientation changing, you can switch from 600 to 1400 with buttons. PS: in my previous message i was wrong that error doesn't occur on higher width than device with but it occurs really with all values on iOS 16.5.1. 1964898 5 a_protyasha 2023-07-03 17:14:04 -0700 Retitled the bug to better represent the root cause. 1965043 6 a_protyasha 2023-07-04 17:34:24 -0700 Pull request: https://github.com/WebKit/WebKit/pull/15552 1965068 7 juwagn 2023-07-05 00:28:24 -0700 Hello Abrar Rahman Protyasha, on github you wrote that so called scale clamping gets performed on every page and that it isn't clear why the initial scale is not being respected, even when a developer may have had a valid use-case. But my personal question is, when this scale clamping bug was introduced exactly, was it possible to track it down? Just would like to be sure that my second test case attached here on the top is passing your new workaround!? https://bug-258663-attachments.webkit.org/attachment.cgi?id=466898 Thank you! Sincerely yours. 1965211 8 a_protyasha 2023-07-05 14:56:56 -0700 The behavior change occurred in https://webkit.org/b/247636, and the proposed fix here works for both test cases. 1965360 9 ews-feeder 2023-07-06 09:09:46 -0700 Committed 265804@main (524f1bcc6830): <https://commits.webkit.org/265804@main> Reviewed commits have been landed. Closing PR #15552 and removing active labels. 466863 2023-06-29 02:21:43 -0700 2023-07-01 17:47:20 -0700 viewport bug in meta tag on iOS 16.5.1 Safari viewport600.html text/html 1339 juwagn PCFET0NUWVBFIGh0bWw+DQo8aHRtbD4NCjxoZWFkPg0KPG1ldGEgbmFtZT0idmlld3BvcnQiIGNv bnRlbnQ9IndpZHRoPTYwMCwgaW5pdGlhbC1zY2FsZT0xIiAvPg0KPG1ldGEgaHR0cC1lcXVpdj0i eC11YS1jb21wYXRpYmxlIiBjb250ZW50PSJJRT1lZGdlLCBjaHJvbWU9MSIgLz4NCjxtZXRhIGNv bnRlbnQ9InRleHQvaHRtbDsgY2hhcnNldD11dGYtOCIgaHR0cC1lcXVpdj0iQ29udGVudC1UeXBl IiAvPg0KPG1ldGEgaHR0cC1lcXVpdj0iY2FjaGUtY29udHJvbCIgY29udGVudD0ibWF4LWFnZT0w LCBuby1jYWNoZSIgLz4NCjxtZXRhIGh0dHAtZXF1aXY9ImV4cGlyZXMiIGNvbnRlbnQ9IlR1ZSwg MDEgSmFuIDE5ODAgMTowMDowMCBHTVQiIC8+DQo8bWV0YSBodHRwLWVxdWl2PSJwcmFnbWEiIGNv bnRlbnQ9Im5vLWNhY2hlIiAvPg0KPG1ldGEgbmFtZT0ibXNhcHBsaWNhdGlvbi10YXAtaGlnaGxp Z2h0IiBjb250ZW50PSJubyIgLz4NCjxtZXRhIG5hbWU9Im1vYmlsZS13ZWItYXBwLWNhcGFibGUi IGNvbnRlbnQ9InllcyIgLz4NCjxtZXRhIG5hbWU9ImFwcGxlLW1vYmlsZS13ZWItYXBwLWNhcGFi bGUiIGNvbnRlbnQ9InllcyIgLz4NCjxtZXRhIG5hbWU9ImZvcm1hdC1kZXRlY3Rpb24iIGNvbnRl bnQ9InRlbGVwaG9uZT1ubywgZGF0ZT1ubywgYWRkcmVzcz1ubywgZW1haWw9bm8sIHVybD1ubyIg Lz4NCjxzY3JpcHQgdHlwZT0idGV4dC9qYXZhc2NyaXB0Ij4NCiAgZnVuY3Rpb24gdGVzdCgpIHsN CiAgICB2YXIgV0lORE9XID0gd2luZG93Ow0KICAgIHZhciBEID0gZG9jdW1lbnQ7DQogICAgdmFy IHQgPSAiIFVzZXJBZ2VudD4gIiArIFdJTkRPVy5uYXZpZ2F0b3IudXNlckFnZW50ICsgIjxiciAv PjxiciAvPlJFUE9SVEVEPiAiICsgRC5ib2R5LnBhcmVudE5vZGUuY2xpZW50V2lkdGggKyAiOiIg KyBELmJvZHkucGFyZW50Tm9kZS5jbGllbnRIZWlnaHQgKyAiPGJyIC8+IElOTkVSPiAiICsgV0lO RE9XLmlubmVyV2lkdGggKyAiOiIgKyBXSU5ET1cuaW5uZXJIZWlnaHQgKyAiPGJyIC8+PGJyIC8+ IFBIWVNJQz4gIiArIFdJTkRPVy5zY3JlZW4ud2lkdGggKyAiOiIgKyBXSU5ET1cuc2NyZWVuLmhl aWdodDsNCiAgICBELmJvZHkuaW5uZXJIVE1MID0gdDsNCiAgfTsNCjwvc2NyaXB0Pg0KPC9oZWFk Pg0KPGJvZHkgb25sb2FkPSJ0ZXN0KCk7IiB3aWR0aD0iMTAwJSIgaGVpZ2h0PSIxMDAlIiBtYXJn aW5oZWlnaHQ9IjAiIG1hcmdpbndpZHRoPSIwIiBzcGFjaW5nPSIwIiBzdHlsZT0id2lkdGg6MTAw JTtoZWlnaHQ6MTAwJTtwb3NpdGlvbjphYnNvbHV0ZTt0b3A6MHB4O2xlZnQ6MHB4O292ZXJmbG93 OmhpZGRlbjsiPjwvYm9keT4NCjwvaHRtbD4NCg== 466898 2023-07-01 17:47:20 -0700 2023-07-01 17:47:20 -0700 New example for testing on iOS 16.5.1/16.1.1/15.7.7 and Android devices viewport.html text/html 3692 juwagn PCFET0NUWVBFIGh0bWw+DQo8aHRtbD4NCjxoZWFkPg0KDQo8bWV0YSBodHRwLWVxdWl2PSJ4LXVh LWNvbXBhdGlibGUiIGNvbnRlbnQ9IklFPWVkZ2UsIGNocm9tZT0xIiAvPg0KPG1ldGEgY29udGVu dD0idGV4dC9odG1sOyBjaGFyc2V0PXV0Zi04IiBodHRwLWVxdWl2PSJDb250ZW50LVR5cGUiIC8+ DQo8bWV0YSBodHRwLWVxdWl2PSJjYWNoZS1jb250cm9sIiBjb250ZW50PSJtYXgtYWdlPTAsIG5v LWNhY2hlIiAvPg0KPG1ldGEgaHR0cC1lcXVpdj0iZXhwaXJlcyIgY29udGVudD0iVHVlLCAwMSBK YW4gMTk4MCAxOjAwOjAwIEdNVCIgLz4NCjxtZXRhIGh0dHAtZXF1aXY9InByYWdtYSIgY29udGVu dD0ibm8tY2FjaGUiIC8+DQo8bWV0YSBuYW1lPSJtc2FwcGxpY2F0aW9uLXRhcC1oaWdobGlnaHQi IGNvbnRlbnQ9Im5vIiAvPg0KPG1ldGEgbmFtZT0ibW9iaWxlLXdlYi1hcHAtY2FwYWJsZSIgY29u dGVudD0ieWVzIiAvPg0KPG1ldGEgbmFtZT0iYXBwbGUtbW9iaWxlLXdlYi1hcHAtY2FwYWJsZSIg Y29udGVudD0ieWVzIiAvPg0KPG1ldGEgbmFtZT0iZm9ybWF0LWRldGVjdGlvbiIgY29udGVudD0i dGVsZXBob25lPW5vLCBkYXRlPW5vLCBhZGRyZXNzPW5vLCBlbWFpbD1ubywgdXJsPW5vIiAvPg0K PHNjcmlwdCB0eXBlPSJ0ZXh0L2phdmFzY3JpcHQiPg0KICB2YXIgd2lzaGVkV0lEVEggPSA2MDA7 DQoNCiAgdmFyIFdJTkRPVyA9IHdpbmRvdzsNCiAgdmFyIEQgPSBkb2N1bWVudDsNCiAgdmFyIGRl dmljZURJTVMgPSBmYWxzZTsNCiAgdmFyIFdJTkVMRU0gPSBmYWxzZTsNCiAgdmFyIFNUID0gIiI7 IA0KDQogIHZhciBsYW5kc2NhcGUgPSBmdW5jdGlvbigpIHsNCiAgICByZXR1cm4gV0lORE9XLnNj cmVlbi53aWR0aCA+IFdJTkRPVy5zY3JlZW4uaGVpZ2h0OyAvL2ZvciBvbGQgQW5kcm9pZCBvbmx5 DQogIH07DQoNCiAgaWYoIm9yaWVudGF0aW9uIiBpbiBXSU5ET1cpIHsNCiAgICBsYW5kc2NhcGUg PSBmdW5jdGlvbigpIHsgcmV0dXJuICEoV0lORE9XLk1hdGguYWJzKFdJTkRPVy5vcmllbnRhdGlv bikgIT0gOTApOyB9Ow0KICB9DQogDQogIHZhciByZW1vdmVWaWV3UG9ydCA9IGZ1bmN0aW9uKCkg ew0KICAgICAgICAgIHZhciBtZXRhID0gZmFsc2U7DQogICAgICAgICAgdmFyIG1ldGFzID0gRC5n ZXRFbGVtZW50c0J5VGFnTmFtZSgibWV0YSIpOw0KICAgICAgICAgIHZhciBuaW4gPSAwOw0KICAg ICAgICAgIGZvcihuaW4gPSAwOyBuaW4gPCBtZXRhcy5sZW5ndGg7IG5pbiArPSAxKSB7IA0KICAg ICAgICAgICAgaWYobWV0YXNbbmluXS5uYW1lID09ICJ2aWV3cG9ydCIpIHsgDQogICAgICAgICAg ICAgIG1ldGEgPSBtZXRhc1tuaW5dOyANCiAgICAgICAgICAgICAgYnJlYWs7IA0KICAgICAgICAg ICAgfSANCiAgICAgICAgICB9IA0KICAgICAgICAgIGlmKG1ldGEpIHsgDQogICAgICAgICAgICB0 cnkgeyANCiAgICAgICAgICAgICAgbWV0YXMgPSBtZXRhLnBhcmVudE5vZGU7DQogICAgICAgICAg ICAgIGlmKG1ldGFzKSB7IG1ldGFzLnJlbW92ZUNoaWxkKG1ldGEpOyB9DQogICAgICAgICAgICB9 IGNhdGNoKGtvKSB7IH0NCiAgICAgICAgICB9ICANCiAgICAgICAgICBpZighbWV0YSkgeyBtZXRh cyA9IEQuZ2V0RWxlbWVudHNCeVRhZ05hbWUoImhlYWQiKVswXTsgfSANCiAgICAgICAgICByZXR1 cm4gbWV0YXM7DQogIH07DQogDQogIHZhciByZXBsYWNlVmlld1BvcnQgPSBmdW5jdGlvbihzdHIp IHsgDQogICAgICAgIHZhciBtZXRhID0gcmVtb3ZlVmlld1BvcnQoKTsNCiAgICAgICAgaWYobWV0 YSkgeyANCiAgICAgICAgICB2YXIgbWV0YXMgPSBELmNyZWF0ZUVsZW1lbnQoIm1ldGEiKTsNCiAg ICAgICAgICBtZXRhcy5jb250ZW50ID0gc3RyOw0KICAgICAgICAgIG1ldGFzLm5hbWUgPSAidmll d3BvcnQiOyAgDQogICAgICAgICAgbWV0YS5hcHBlbmRDaGlsZChtZXRhcyk7ICANCiAgICAgICAg fSANCiAgfTsNCg0KICB2YXIgc2hvdyA9IGZ1bmN0aW9uKCkgew0KICAgICAgdmFyIHQyID0gIjxp bnB1dCB0eXBlPSdidXR0b24nIHZhbHVlPSc2MDAnIG9uY2xpY2s9J3dpc2hlZFdJRFRIPTYwMDt0 ZXN0KCk7Jz4mbmJzcDsmbmJzcDsmbmJzcDs8aW5wdXQgdHlwZT0nYnV0dG9uJyB2YWx1ZT0nMTQw MCcgb25jbGljaz0nd2lzaGVkV0lEVEg9MTQwMDt0ZXN0KCk7Jz48YnIgLz4iOw0KICAgICAgdDIg Kz0gIldJU0hFRD4gIiArIHdpc2hlZFdJRFRIICsgIjxiciAvPlJFUE9SVEVEPiAiICsgRC5ib2R5 LnBhcmVudE5vZGUuY2xpZW50V2lkdGggKyAiOiIgKyBELmJvZHkucGFyZW50Tm9kZS5jbGllbnRI ZWlnaHQgKyAiPGJyIC8+PGJyIC8+IFBIWVNJQz4gIiArIFdJTkRPVy5zY3JlZW4ud2lkdGggKyAi OiIgKyBXSU5ET1cuc2NyZWVuLmhlaWdodCArICI8YnIgLz48YnIgLz4iICsgU1Q7DQogICAgICB0 MiArPSAiPGJyIC8+PGJyIC8+IiArIFdJTkRPVy5uYXZpZ2F0b3IudXNlckFnZW50Ow0KICAgICAg RC5ib2R5LmlubmVySFRNTCA9IHQyOw0KICB9Ow0KDQogIHZhciB0ZXN0ID0gZnVuY3Rpb24oKSB7 ICANCiAgICBpZighV0lORUxFTSkgeyANCiAgICAgIFdJTkVMRU0gPSBELmJvZHkucGFyZW50Tm9k ZTsNCiAgICAgIGRldmljZURJTVMgPSBbIFdJTkRPVy5zY3JlZW4ud2lkdGgsIFdJTkRPVy5zY3Jl ZW4uaGVpZ2h0IF07IA0KICAgICAgaWYoZGV2aWNlRElNU1swXSA+IGRldmljZURJTVNbMV0pIHsg DQogICAgICAgIHZhciB1ID0gZGV2aWNlRElNU1swXTsNCiAgICAgICAgZGV2aWNlRElNU1swXSA9 IGRldmljZURJTVNbMV07DQogICAgICAgIGRldmljZURJTVNbMV0gPSB1Ow0KICAgICAgfSANCiAg ICB9DQoNCiAgICB2YXIgVCA9IChsYW5kc2NhcGUoKSA/IGRldmljZURJTVNbMV0gOiBkZXZpY2VE SU1TWzBdKSAvIHdpc2hlZFdJRFRIOyAgDQogICAgU1QgPSAid2lkdGg9IiArIHdpc2hlZFdJRFRI ICsgIiwgdXNlci1zY2FsYWJsZT15ZXMsIGluaXRpYWwtc2NhbGU9IiArIFQgKyAiLCBtaW5pbXVt LXNjYWxlPSIgKyBUICsgIiwgbWF4aW11bS1zY2FsZT0iICsgKFQgKiAyKTsNCiAgICByZXBsYWNl Vmlld1BvcnQoU1QpOyANCiAgICBXSU5ET1cuc2V0VGltZW91dChzaG93LCA1MDApOw0KDQogICAg aWYoIVdJTkRPVy5vbm9yaWVudGF0aW9uY2hhbmdlKSB7DQogICAgICB2YXIgZiA9IGZ1bmN0aW9u KCkgeyB0ZXN0KCk7IH07DQogICAgICBXSU5ET1cub25vcmllbnRhdGlvbmNoYW5nZSA9IGZ1bmN0 aW9uKCkgeyANCiAgICAgICAgV0lORE9XLnNldFRpbWVvdXQoZiwgMTAwKTsgDQogICAgICB9Ow0K DQogICAgICBXSU5ET1cub25yZXNpemUgPSBmdW5jdGlvbigpIHsgDQogICAgICAgIFdJTkRPVy5z ZXRUaW1lb3V0KHNob3csIDUwMCk7DQogICAgICB9Ow0KDQogICAgICBpZihXSU5ET1cudmlzdWFs Vmlld3BvcnQpIHsNCiAgICAgICAgV0lORE9XLnZpc3VhbFZpZXdwb3J0LmFkZEV2ZW50TGlzdGVu ZXIoInJlc2l6ZSIsIFdJTkRPVy5vbnJlc2l6ZSk7IA0KICAgICAgfSANCiAgICB9DQogIH07DQo8 L3NjcmlwdD4NCjwvaGVhZD4NCjxib2R5IG9ubG9hZD0idGVzdCgpOyIgd2lkdGg9IjEwMCUiIGhl aWdodD0iMTAwJSIgbWFyZ2luaGVpZ2h0PSIwIiBtYXJnaW53aWR0aD0iMCIgc3BhY2luZz0iMCIg c3R5bGU9IndpZHRoOjEwMCU7aGVpZ2h0OjEwMCU7cG9zaXRpb246YWJzb2x1dGU7dG9wOjBweDts ZWZ0OjBweDtvdmVyZmxvdzpoaWRkZW47Ij48L2JvZHk+DQo8L2h0bWw+DQo=