67216 2011-08-30 11:58:45 -0700 NRWT dumps text into diff.png files when images are 100% different 2011-10-24 15:28:03 -0700 1 1 1 Unclassified WebKit Tools / Tests 528+ (Nightly build) Unspecified Unspecified RESOLVED FIXED P2 Normal --- 34984 1 simon.fraser webkit-unassigned abarth dbates dpranke eric simon.fraser tony oldest_to_newest 459061 0 105661 simon.fraser 2011-08-30 11:58:45 -0700 Created attachment 105661 Zip file; diff image is bad I just ran NRWT with 12 parallel instances, and --pixel. Some of the resulting diff images were corrupted. See attachment. 459088 1 eric 2011-08-30 12:28:59 -0700 Thank you for the report! 459148 2 dpranke 2011-08-30 13:39:45 -0700 I assume this was using the Apple Mac port? Chromium and non-Chromium ports have very different approaches to diffing images, so it probably matters which one you were using. 459156 3 simon.fraser 2011-08-30 13:45:02 -0700 Apple port, yes. It's not clear that running in parallel is an issue. 465683 4 eric 2011-09-12 10:55:42 -0700 The backface-visibility-diff.png in his zip file has the content: diff: 100.00% failed it appears we don't handle the case correctly when the diff is 100% different. 489436 5 eric 2011-10-24 10:43:33 -0700 I feel like this may have been worked on in another bug. But maybe I'm thinking of the "displaying the diff % in results.html" bug. 489476 6 eric 2011-10-24 11:34:06 -0700 Are you still seeing this? It doesn't look like the current iteration of _read_image_diff could produce this result. 489484 7 eric 2011-10-24 11:40:16 -0700 I'm not able to reproduce 100% different. I tried with a test.html which just painted black, then changed it to just paint white, and saw only 86% different?? Similarly when changing from red to green, I only got 56% different. In neither case did the diff image actually show me anything useful... Maybe diff images are generally broken in NRWT on Mac? 489702 8 eric 2011-10-24 15:28:03 -0700 Without a way to reproduce this, there isn't much to do but close this. I believe this was fixed accidentally since filing, since we've been in that code several times. If you see it again, I'm happy to take a crack at it! 105661 2011-08-30 11:58:45 -0700 2011-08-30 11:58:45 -0700 Zip file; diff image is bad backface-visibility-results.zip application/octet-stream 3091 simon.fraser UEsDBBQACAAIAMyUHT8AAAAAAAAAAAAAAAAeABAAYmFja2ZhY2UtdmlzaWJpbGl0eS1hY3R1YWwu cG5nVVgMAL8yXU4QP1xO9QEUAO3a+VMTZxgH8E0MEJAjHEpNAQMCAiqQlASoIIEQ5D4iIGBVSALK FY4gchNQORQlFeSIIkVFx2IAUVARaqRY0IIHKAGpgIBaFDlLuQJtHGea/Qc60+PdX7L7TOb7vs/n 3cybndlcT/edCnJYOQiCFJwc7WkQtAYHQUg/tLS4wovPjBZ/GMZR/eIYB4MZ4exDkRCRQKLj6SFB JAs8gciwIJoRGARTM/MQkimRQQixJOk1GyqLEyZCKRRPJwoF5xkbFRIaEQxBCYhj3k/CVW3VVkms q0910AHhinUJXB0n/6qD1OekFWV6uHz/xgYql/mVUw+L9GFCv74yZ3T4mdEA9WOFzB8craQpponz sQCqwUCVAb1Cfdt0J8u1ofK94HHa7ID5ZGpU/CCkENtlNsdBjSq00S4oh0ak2IWJcqBTF2RtzfR8 +sfddaVtj7AXk9/1Jill7PH9MXn5cN1hPq+rgSdzdIU8tJPXZ1Q7qR4j56+CzLEarDBsi0MPjvTO jOdbuicm3K9X2mLd6q+TQiZue5zudntXp143+YijReFcDG3HOeGJQqGmTCgr5Kd2TLMH9/Ldjokq esU5998Vuc1lv6W53xjY0aLdsnYD38bh9snzc3kvbIaDXl9625plrfEF8beFwHW3sFcXiC1q+zFu my94x/QZK0yXVK/4zjhnRnMtHjGDOOwARw+hK2uhMVsqVV+oWjN82NfrdWK+6oGB568mqS+RNaaK mvJYeZ6CoyeGigoqqfUkZ1lh8WZ1sTabqqS/LEgmHNINuxtHt79um0z6xfqW16VXVqsGOjGFlHr0 aZHtqlHNQ483iYs/f1Pb6ztfyD/La5wZ2v7gTekdrJawu32E7bYv5czz+ILdO4uuJSY18caxNuO7 i/FvM4n5reza5elt+890FBvmqT0Y8Zl973L0UADefLA8In279bDxE+P6HjY/94S579niAythjssT 3zcVvPVZ9Ng3Ny060NnKGDbE68cff6edW/zAePp0r0vV2uHAyDZ02Zksw67Ib5d0XwQG8prmTPba 7ElodkiXn/Mq3NDR5Z07n95ixDQru6MYtZTf3yZSH50//q7cAGkSMdRe2hi5K2YSUxFcyWJXI9Xv PUudamCFX0yzEKH8rKrW82nEclwTQiM+O0nrV8OS7qKiTkta9tbygsaxR3Rah2c/hUYT6gmDXdbf KPWhj8X3lnZrsjQqLO3mW684Bxcs1Uwr167U0hYXr9H60tyxPRnFS1GCphWqIOFELgThICd7W+8E xMe+625MLzWp+8FSslxTnSAVQeAz6ej87iXfk8kO806VMQGzM7PLiyMUDoRAotAYHJkDZcigVndM ZW3+fPW5PLWfzFnQhhdKOFBi3ksKSvItNObeXxniEy1VDC7fKzlHEoxADJuTJQMJGEjkKD5MSRmW 2ogesBns2T2uJJeSmmKzwLsiib+JXEWEWu/dJEkYwmBwMqdyYTP3JXNunq+GBUaj0dw1bfaSQclb cOQADyGsNwsUqj2DoSOJ/W4dBse/Eg7rTEM8UQEeFvtUDs3t6SbCulVEIGKHSmCxB6VQjh/GsDCR bA5kGe2KgneEvDQXdxI2fwqZo2GhD4sNQCC6khodYGjaOLKihiosls+BAAtgASyABbD8k1maRKK0 U69gMFCGq8AGt6ESvoluRSDqjo7DCkwkUvdiHGwXzdDDkYu4ZfAUModZqAgbq1bsq1KnKilwZdHc vC5tWJdFHMjfLh5WICAQp72+hs1OoILBubS7w9ZRH0cmOGfDwP2QSKZsn6yk0I5COTqMYWBULmTO Vu3tsEIpB3pooAlb+0DZpmqTpvQf+DW9G9UWlhbSpogeyrA1Fa0Bdyn48QIWwAJYAMu/jiVhcXFF 4y4M5tMWmP73bKRorh5zs6SAEz+Bg38XAAWgABSAAlAACkABKAAFoAAUgAJQAMp/HOWIF4INQejl T+8CACBw1wAUgAJQAApAASj/C5SMdJk96y9HZ8ThIfHhRHW3r7YLzPwTUEsHCO7fLEnhBQAAbywA AFBLAwQUAAgACADMlB0/AAAAAAAAAAAAAAAAHAAQAGJhY2tmYWNlLXZpc2liaWxpdHktZGlmZi5w bmdVWAwAvzJdThA/XE71ARQAS8lMS7NSMDQw0DMwUFVIS8zMSU3hAgBQSwcIol9U3BcAAAAVAAAA UEsDBBQACAAIAMyUHT8AAAAAAAAAAAAAAAAeABAAYmFja2ZhY2UtdmlzaWJpbGl0eS1kaWZmcy5o dG1sVVgMAL8yXU4QP1xO9QEUAJ1US4/aMBC+51e4XBIEJGiPS4i0YpG6ElUrdS89OvEkWDUOtSc8 tOK/d5xACEukovqAUWa+h8fjib+8fl+8//qxZF/fv60SL17jRrkNuKANJSpIsnKzLa1EqYso5dnv nGcw2UkrU6kkHkOHiaMm14stHmkPFU9BfeSlxskeZLHG57RU4hRHTdyLo7NGWopj4r3KPAcDOgOW Au4BNJMbXoB9ZjFnawP5fNCjPRGEC7e6GCTuXxzxJE4NsQq5Y5ni1s5rmnc4YBJH9PU+tiCPXGow THDkky1pyUOv2iBZlVxQGcIwvJDZzMgtJl6QVzpDWepgyD48RmvHDSMyRRAQb41SpZHN2XRWJ1wQ l6wF1VkBQsvg1mjUwzFrwzJnQZ9IzJ6GbZJbBrAy+grEsigU1AgbDK/fLeCbRjA7roJuzpg9TafT hvPk9fqvE4PKqK7/usR0Zg171iR0xOpgyIVY7kDjSloEuofAd2z++HNd7oDWZMRMgtdAc8wmPus1 e3vwrlMqJVKfhO7HNQW4y5ozf3nYQoYgGv9+F1NX8g7D/JcMK67OgNlNenZpt9DAVlGDLdZSiYDX iBow7uTk0lisMzqnPzFQFh6w8cn5Q0bgjHnQSrfGruOdA5IWZVZtyEX4pwJz/AmKOEu62bB9jv6Z xYFakX8i28fahXdqRwQ33Xi1XwC+IBqZVvTC/M5TpwsdMX/SkLhZ0qW+qcb/k19oLvSnoXsGNAzP 0yOOmjFIU7EewH8BUEsHCCqpBGAzAgAAmAUAAFBLAwQUAAgACADMlB0/AAAAAAAAAAAAAAAAIAAQ AGJhY2tmYWNlLXZpc2liaWxpdHktZXhwZWN0ZWQucG5nVVgMAL8yXU4QP1xO9QEUAOsM8HPn5ZLi YmBg4PX0cAliYGBWYGBgiuBgAoqIioiqAynNEteIkuSM1OTs4tJcBsskgxRjS3NDE8OUNBMzs0Rz M4vkJENj42QzsxRzcwujk7LKNgwMXJc8XRxDKua8vWPI1+wgwOI4/+h+Qf349r4TUQmLV32ZxBkt fWfWTHEGhrnf7zIzMMxzS2BkOMIixuDgg8HIsa6OOL6r6Hj87LvxSQwHpnNjVTXKGBKMnLNn7koz MMgdZeVmUFDbwMCQxEQ2Y9tc2zXJsTsfrz9rV+7PwFDXMZoyhjCDmiljlDHKGGWMMkYZowxqMb4x bi8WXpm3I3AJsEnM4Onq57LOKaEJAFBLBwgYqryDAgEAAEALAABQSwECFQMUAAgACADMlB0/7t8s SeEFAABvLAAAHgAMAAAAAAAAAABApIEAAAAAYmFja2ZhY2UtdmlzaWJpbGl0eS1hY3R1YWwucG5n VVgIAL8yXU4QP1xOUEsBAhUDFAAIAAgAzJQdP6JfVNwXAAAAFQAAABwADAAAAAAAAAAAQKSBPQYA AGJhY2tmYWNlLXZpc2liaWxpdHktZGlmZi5wbmdVWAgAvzJdThA/XE5QSwECFQMUAAgACADMlB0/ KqkEYDMCAACYBQAAHgAMAAAAAAAAAABApIGuBgAAYmFja2ZhY2UtdmlzaWJpbGl0eS1kaWZmcy5o dG1sVVgIAL8yXU4QP1xOUEsBAhUDFAAIAAgAzJQdPxiqvIMCAQAAQAsAACAADAAAAAAAAAAAQKSB PQkAAGJhY2tmYWNlLXZpc2liaWxpdHktZXhwZWN0ZWQucG5nVVgIAL8yXU4QP1xOUEsFBgAAAAAE AAQAYAEAAJ0KAAAAAA==