32694 2009-12-17 19:50:23 -0800 WebKitLinkedOnOrAfter() ignores embedded frameworks 2009-12-18 11:20:49 -0800 1 1 1 Unclassified WebKit New Bugs 528+ (Nightly build) Mac OS X 10.6 UNCONFIRMED InRadar P2 Normal --- 0 opendarwin webkit-unassigned ap webkit.j oldest_to_newest 173025 0 45122 opendarwin 2009-12-17 19:50:23 -0800 Created attachment 45122 Sample Xcode project Overview: WebKit uses the function WebKitLinkedOnOrAfter(), defined in WebKit/mac/Misc/WebKitVersionChecks.m, to check the version of WebKit that an app is linked to. WebKitLinkedOnOrAfter() in turn calls NSVersionOfLinkTimeLibrary("WebKit"). However, NSVersionOfLinkTimeLibrary() only checks the main executable, and it returns -1 if the main executable did not link against the specified library. If you have an app that does not link to WebKit but that embeds a framework that does link to WebKit, NSVersionOfLinkTimeLibrary("WebKit") will return -1, but WebKit fails to check for that return value, and thus the value of WebKitLinkedOnOrAfter() has an unexpected result. The unexpected result of WebKitLinkedOnOrAfter() can lead to undesired behavior. For example, I discovered this bug because the WebUIDelegate method webView:contextMenuItemsForElement:defaultMenuItems: wasn't getting called when right-clicking in a text field. This is because in the file WebKit/mac/WebCoreSupport/WebContextMenuClient.mm, there is a function isPreVersion3Client() with the following code: static BOOL preVersion3Client = !WebKitLinkedOnOrAfter(WEBKIT_FIRST_VERSION_WITH_3_0_CONTEXT_MENU_TAGS); If the app does not link to WebKit but an embedded framework does, preVersion3Client gets the wrong value, YES. The result is that the wrong code is run in WebContextMenuClient::getCustomMenuFromDefaultItems(), and the WebUIDelegate method never gets called. This bug is important because it is a common situation for embedded frameworks to contain WebKit code. For example, the popular software updater Sparkle run WebKit code. A lot of applications use Sparkle, but the applications themselves may not link to WebKit. Our suite of applications embed a framework that runs WebKit code for several purposes, such as showing a support contact form to the user. Again, the app itself may not link to WebKit. Steps to reproduce: 1) Download, unzip, build, and run the attached sample Xcode project WebKitLinkBug. 2) Right-click in the "Last name" field. Expected results: The WebUIDelgate method webView:contextMenuItemsForElement:defaultMenuItems: gets called, returns an empty NSArray, and no contextual menu pops up. Actual results: The WebUIDelgate method webView:contextMenuItemsForElement:defaultMenuItems: doesn't get called, and a contextual menu containing default items pops up. Regression: This bug occurs on Max OS X 10.6.2 and Mac OS X 10.5.8. It also occurs with WebKit svn revision 51668 built from source. Notes: You can see in the console log when webView:contextMenuItemsForElement:defaultMenuItems: gets called. You can also see that NSVersionOfLinkTimeLibrary("WebKit") returns -1. However, if you go back into the WebKitLinkBug project, and link the WebKitLinkBug target to WebKit, then the bug does not occur, and NSVersionOfLinkTimeLibrary("WebKit") returns 34801920. 173032 1 mrowe 2009-12-17 21:14:53 -0800 I don’t think it’s possible to make these sorts of linked-on-or-after checks work correctly in the general case. WebKit has no information about which particular piece of code (whether it be application, framework or plug-in) happens to be using it at the particular moment. The application may have been written against a newer version of WebKit than a framework that it uses. An application plug-in may have been written against an older version of WebKit than the application that loads it. It seems like what you’re really asking for is an explicit way to opt in to the new behavior of one or more of these checks. One approach that could avoid the limitations of checking linkage information would be to add a method to the relevant WebKit delegate that the client could implement to say “Yes, I really am ok with having the new behavior for the WebView’s that I’m the delegate for”. A delegate that didn’t implement this new method would be determined by the linked-on-or-after check as per the current behavior. 173033 2 mrowe 2009-12-17 21:15:11 -0800 <rdar://problem/7483619> 173035 3 mrowe 2009-12-17 21:34:24 -0800 It’s unclear how we’d handle the smaller issue of WebKit being used by a framework from within an application that itself doesn’t link against WebKit. Should we assume it support the new behavior or should we be safe and give it the old? We’re effectively doing the latter at present. 173065 4 opendarwin 2009-12-17 23:17:36 -0800 I believe that something needs to be done, because it's just going to become a bigger problem over time. You could ship WebKit version 999.9.9, and I could build a framework against the Mac OS X 10.9 SDK, and WebKit will still think that the linked version is -1. Moreover, it seems pretty clear that this was an unanticipated bug. I've discovered a comment in the file WebKitVersionChecks.h: "A version of -1 is returned if the main executable did not link against WebKit (should never happen)." It should never happen, but in fact it does happen all the time, in shipping applications. That comment is like an assertion without the assert(). ;-) I'm not sure to what extent I'd want explicit opt-in methods. These version checks are all WebKit-internal backward-compatibility hacks. As a client of the WebKit framework, I shouldn't need to know the implementation details. I simply want my code to work as expected: e.g., if I set the WebUIDelegate, then I expect the webView:contextMenuItemsForElement:defaultMenuItems: callback. There's no problem with the application-specific, bundle identifier based checks (as long as the app developer is aware of them). But if it's not possible for the linked version checks to work correctly in the general case, then that kind of check should be avoided as much as possible. Don't change the behavior unless you're going to change the API. API-based checks are far superior. If an object still implements a deprecated method, then the old behavior must be assumed, and if an object implements a method that was newly introduced in a certain WebKit version, then new behavior can be assumed. I have no objection to 'opting-in' by implementing new API methods, but philosophically I'd be opposed to opt-in methods that are specifically tailored for embedded frameworks to implement in order to work around specific internal backward-compatibility hacks. 173105 5 webkit.j 2009-12-18 03:47:07 -0800 I’m confused as to why you’d make a test like that before calling a delegate method in the first place. My expectation as a Cocoa programmer is that the newest form of delegate method will be called if available, otherwise an older form will be used. AppKit and Foundation have linkage checks, but never, as far as I’m aware, for delegate methods. 45122 2009-12-17 19:50:23 -0800 2009-12-17 19:50:23 -0800 Sample Xcode project WebKitLinkBug.zip application/octet-stream 25569 opendarwin UEsDBAoAAAAAADOnkTsAAAAAAAAAAAAAAAAOABAAV2ViS2l0TGlua0J1Zy9VWAwAvO8qS6HvKkv1 ARQAUEsDBAoAAAAAACOikTsAAAAAAAAAAAAAAAAcABAAV2ViS2l0TGlua0J1Zy9FbmdsaXNoLmxw cm9qL1VYDAC87ypLIeYqS/UBFABQSwMEFAAIAAgAG6GROwAAAAAAAAAAAAAAAC0AEABXZWJLaXRM aW5rQnVnL0VuZ2xpc2gubHByb2ovSW5mb1BsaXN0LnN0cmluZ3NVWAwAzu8qSzbkKkv1ARQA09dS 8MlPTszJrEpNUShLLSrOzM8rVshPU/DMS8vXK8jJLC5RyE6tLFbQ0ufiAgBQSwcIpEm/Ii8AAAAt AAAAUEsDBBQACAAIACOikTsAAAAAAAAAAAAAAAAoABAAV2ViS2l0TGlua0J1Zy9FbmdsaXNoLmxw cm9qL01haW5NZW51LnhpYlVYDADO7ypLIeYqS/UBFADtfdty20iW7XP1VzD0ch6mWs77ZY5LHbZU 7lKXb22pynP6ZQIi0xLGFMABQV/K0xHzNfNh8yUnk6REikiC2CCghCV0R7hsCUjsXPu+c2fm0798 uR4PPplsGqfJTwf4EB0MTDJMR3Fy+dPBb+cv/qwO/nL0p6dRNryKP5lB/nVifjoYpteH0WQyNoen SW6yD9HQPJ/F45HJ6OFxOkyjw387fX6wGlYeYnRw9Kcfno6iPLL//eFpnOSDj+brTwenz0/S4eza JPnh2ddpbq7Po+zS5AdHGAn09Il9bv78NM8sRVte+X3xHffOMWf2rcXTZS9u0n07hKz2/rPJ5Nc4 X/swVYeEV3nzl9PzNB1fpF9uX2ZcHaK7X00v/sMM88FwHE2nPx28Pns1y6OLsTmJh7l9Jcq+HhTG fTueXcbJctCpQ3uDhtdnh/a/lsFH29m34N5iqHWCCiMtCHSDbQL2ZPGrsmk8y7LIM4OfR3FuRm/m 75ye3EzhwqK1ePRnJ5dm9D7Or/7t1ctj+/fs4Oj//Xz29Il7ZvG0lRdzabLBp2g8s4JK9MGTnVRt IWeBwomZmGRkVSI29ShaQgPHvAqSZQLxyuSR07daVJdhNLIfPZymmWXWr+br9GAQj346QIvPQL+z NsdK4jL/9py5070/uIeovkvTfCGoy+ljxFQjQB/Ppnl6vRj7ZmiCb+Z6VwuP3Uuvo2tzcGQ5ZIUr HkZOHO5qbinEvs9htvNzL+Jsmr8z00lqdSPb83sctTa9VyaZLT4jmJZaIKX83zqP87H9zrNXUZy4 l+58o5J0LD52at3RjWQCRWEL+W7E5RQ0w8zN4Wb8H55m5oPJrHEy6zQcDOyP12f85PZ576Tfmwvr y17Gycfns8u7Ey+8YhX+5/+cxZ/Wxrx15avfvkpHr6LpR+cXmeJS3PrxzRdeJeY6TeLhy3R4cEQw k0xRweTd5/0yZKUvnWVDcwv/m+T0Oro0C6woZ4JrTW6hKhWu+YubM9984+aDNy85qI+vzPDjdZR9 LMC2LpaASbyKv5jR2jw4IpxjIVTLE5l/9yyP8sLbhZncHezZ3AEdHE1nF9Yyzhb//NfCIH7FXI5x tnh3aQ+sQZAcMb51xlXEtq7K1tDa3YpLqOLE8lGufaVcd1cgPFl/xW+zLtJZPihHo4Ie19BM7yxu 1XA+kVs9fFI+9TWxn7+3kvu1FzdFcSfwFDEiLNVoX+BXImF1bHoST50ojbyysPnwmZlEWZSnfsnx 87SE6TsZV8kGP0RmC6SJnehtsFSb2V4te3s70PR///t/AAp29KPv4Z5hPx0whRVlmrFeOx8+s+08 BVX2f6IV9Twz2afY6mYDnu+hsKhepLaNk/5gTXKCBKFYrTMVzqI9orUfagVsPgEu0L2Ij//dTnlJ eTEx3KYLYN/FBCFSMNWbwu9Oz8DMlnYMrqnYOyr1qtkv8cjUzgaOrvpgZUsqwYimBNO9g5XtTHuT X5kM4sFqcMvGx0qwB88tjZTSmnLaTrBxlX4ePBuP+2Bjn3iQCztRLPvo/xFwW1AiiaT7M9urj3+f xfULYEf/+VhdXvEHW+Jft/ZifMHvxgib/9zhUaVW2Ma9tOHljBfxuFjEDr+KAWZ1PT63XZ6XBCFO baqyqzzvY0OnqvLSzoNhi2Blm7Sa+26b9Np8hhih5LEaoZ1csnkxs56gemYM4dKbiUmA5dy059S2 iI64rFbi6hU+KKsG78zQJHkfd99rkQ8jwQWSCMsKVb5yNgUo9O2uxmiGhGLo7rd2RKcrSJ5svOZF 5XhsomywpX5YUXxrSnA9Ia4rx/uI8paiLKBQu5C7my6tFuu1jFjyGRZ7e6U+ea0tavcXgkiBBSG0 ejEJ4teOx+m0GCiXBCDeuLLn0zwAoZpwpqr32UAYdRZ9AvHJW8nt+eRKOxJRS3c7+uTYNHgGbdI4 AzMLSy2YevDM4lJriSUmrTDrnflksnyQpwPHtlHft3bbg42YsJ+vvkTYxxffLbeJZoJohdqxh28t uYMzk88mh4eHEJP49pGYxAJuVs7P48l+rW3axouKtMRR+7Mc6N8mjzUYKbSv77NkoV0btxak6R0Y budbv2QBKmFBliyUllzdKfRXZ0Onliww4opIiQggw7md/G7D8lsy8la3txmVPx6rUdlZMdBIakZI 9eVuCJveGRib/vFIHHmdnmRECREckNz4+dSH2d3nNrGOm0ukAeu9AKU8nkGWpY6+9KbTzyUlENfc 2p92uJROvkLYNOzZtE2ZlE1XaPW9VhAuvY2mxX2oZWz61LNpizYRbeMQyDIVlE2DKBkNXkX58Gpw ln/1tPyUsO33vn15S/mASiuhWLTjqk7M2ID0qw8lNoviimKLNm0nDTuz/LHfhjWYH0W9DdziqjBB WGC2t6/qg/zuM1u5wiunkIUTgGa+iJMm1qgeCnfuo/lMC0o53mAojD0dbDpjTNoURzEJaTpbQVGp 58yh4V8lKBHZow9bnn/gnWd3KD2PLu2kioT5WtF27nIgQmL7QdoypwevzRdvx2UJr33bZB4lr0kz vCYSaUQ0bVurB28z8ylOZ96t9CX8/msdfvtrwd83v2kz/MZcI4UEVS3y+7epW713CYH1pYMPaTbY 5uVKGO9L9nYy/iEquoewOoxXiiiu7qbnDfP9b7Prybwv6ob3QJb/x+NkecUO8gZawLUkjCBJ28k0 ziZmPHY/duW1v2bR9XWU9ZnHvWYeUmjhcskqmUdFdnUwExFz966whtiyFTSVbNn8vISqEJXYNB9D awp4PRmvK+b7SHo9D6UF1lQTUNYB5Or87NPBzWabwevUtzGhjJ3/t2dnVXYqrplClItm+AmpZBaf 31HM3CYw5ULldV0Pna2WqVQyjkF7D2up6a31fX8Vj83g/OvE/qPfi9i41ZWCCSUaUtMSfi4d6MBF ALfM7fnZMD+l5pQJQUib/EyzzJFwq6HPZnl6HeXxMBqPfc0q9Vj63fGmnayNI6kIFqilldvZxTSP 85nLPvqjSO83XRNaKSUVrXQU6Q4+dTFP45hqJDDEEq0gAaRpu6B5HNZnZ+lXIyoUBjmG7ezo4/Fu cFUgjrHisLQZqmU2bMsHrhH1ybY+xxJu9OuzB82uz2IthXB/tM7yv8/S3H/6dQm7+yXaZpdohSQS KUVAGXctdp9E0ysoux+TrXVs4IxxUOG5FifcmaD9UnnYpXJOJdJW9RoKmLzcPjdf8sE7MxlHQ3O9 5fyxx6F77eTORCmrs1i1s5vgPIuS6Yc0u46ayp4faybMpVv60bjKwuVu1DuYC7urgYVmVrgg1mQF SyVr8ir6aAa/TSYmGxxH3hO6KopkPamsK5j7yGY9066k5srGVSBHXocZL9PPPTN2+VnBJUICFN4C eXEcTeI8Gsd/PGI2tONghTtokAnazravs4kxQ++NIX1VujVfLJX9tGUrrdZE5GdQB12wpFgqIhio BXIFRrXsLXfZm0Ul+tgvbrZQ7LR2gyDLxlZZmE56Dt6vl2n0xCvqDnVXAjV+Scc85djrzKvHdISV u/mdSY2232deimunDrGSmiglhay+BL+afYUtmmkj9wM83nhFWM4IzirEK1ug7mC0grm2s5Kw5doV FNVXzx0k0FLztjrlI3R4u70Rp1ozDgxZQHx8no6h26suHicHW1ukU5hRQTZuZmqYzaeuhOE73qmM 0XHP6GY3R1OkidCw0iGQ0b8l1m+M4wTaaLHtppYHzut6i3vU9VVCM8ltbOy7obrBVnfEkaCatmmG n8eXlwa6D+xfHqdqtrY4rxTnTNJWGX12HY3HYE7/uef0gtOFM+/qcVojjAQBNt30lrrrlloK+38s YU2MQAX+1WTAMwi+Vw7UrrNsY9SWaxUlZwwzsRk1gfHfr+DyQ92ay26xJJxooSlihS+WS+YKmSfF F/1B/tQMTsyHaDbeUk0pvrlNTH+oK6neWVUTVj8gleV1i9HYzSApbPJFdOEAm3YY9DrdloT13PG3 A0iOtSzcTNo8d87jy6vcbDMvPXN8to0qyjHmmxuymmfOyzSd9rzZ/WN/4FRnYVxjLpHCsMMzYNHU y/gyymfZI2mhuq+Iyl2Pg6XWm4lsLSZ0NqySNoXTkmCw6VnB08dVrToHLhRV1sZvWpCWONQHVsDA SlCOLH8KZqId9vgPcn+03GnZe0vrvglTBd1rtGodTQ18San33uXeWxFB7cyr+e5yFnTWdxN3tREp nijj59Ca6bkBp/fc7boGxiThjIM9N5Q/Z7OJyabDLJ70/AHwByMquTXUChz8whl00bMHyh7NsHIL TJsLEY1z510Ub9sG1nNmy12XXBKmEbiaCGXNfJtez5pdP24u4KVacKIKLqtf0+1MBF2r+wZhIpD9 o9WuDNe0fJyO0wzatXz8ODszal0LghnnmsN2q/cK2nW2uj0F1Kpom00X7piwrfdqlvDBe3HtD7vU 03+75kPnI9WUESok6GhFIB8XV6XWYaTvattHykjvz+7i9jq6tmj/++sztxHHscu3G6qBk3AwdxOW 1U/CgWxhO/ff5NVvYatWR2RMaI05rbKFbQvUHdzCpiR3B2hxUBCxgqKSlXo2ji+TwUvzAXqR3Lc+ GKzKR0ER5pyIzTJ3k3w8NkkObrX+r56HVXmoGRdSKQWK/IA8/NtsmscfHvH57zvDb2xjHGUhbd0g vnPdcUBt+mevTVUZiQXFWNo/mmFknx53g60U25EkbKcJUD3fZ3HufnoSuxs0wHckfq/cuKf1eM0F 0kKrTQ7ux4wur8wjrRGFrvyuYPIsJwCN0VZw31qjdJlFE+/pYL73HsO6ys4IgWPt7uKBtnmVcnQJ c/51Yn46uIimRrA//3b+Qh1scOz4t3fjf1y/wBe//L3nWXWeEaY5JUw2qoXVefZl/I/k75ejv6rL 367HSdTzDsI7gaSdDNPQ3UTN8O73v03+Qa7Q6S/vPp3+POdjz7vqvONI2XiNgXeCNe39oNF4kZve gLxnureOobDkWBUjvC6EPDsuPu856k26FEIUa90oR/uQp1WeKXfCpjvwqg95vjveWRa45g4ZRt/6 kKeMdy233WkbMSEuVUMLcn3ZshtlS5vAIMsk2AVhwLrlvOvu3Qx8HNJj4gOhWGEsSJtrbPPuqhp8 qNddRTGSBD221R1l6ceKNmUmS7qr6jCyVnfVw2RkBw6b50pwaYFaFSFKpaTyYfO/x+bzXkfNF2XA s1T7WM+mZ0JYjcB0zWNWZ0OnTqbHiGBuzQ6p3Na3mvvutr651z+39FxEPjMFOh+8Tvtn13sD4V2Y VCpmR0GtsOt4Ns3T6/gPc8Oz//3v/+kvT7onTyAxZUra3LphT/A+TkZp7wta8wWKchu33zm5HcKI jnkDjG0WwnT1i0pW099tX17FSXztv0Rvqyu4BruCB2JcdkaO0o5Dyd2D5Btj1T/S1It8b/qrcYdo TbmQd3P5OtyBFMlgFbKd5bGe2RWZLQgnNo5m7aji8/nPno3HgzwdvMiauezpwfKq+IMt26UW7njq 2zG1VxzHmBLaOlHRcBz3ixlP+svjqjZyaiKVlWi6KyjzodqxkIwjoRCu7khWc99tW96bi1/j/GWc fHw+uxx4wSiJzf7yWGOzykbGIQq2MBv/2jL0qyhOCkPfebUgWYv8/s38p0s9kYISxulNvL/xsWP3 5uKLdyTl2WRyYsbmMsrNXp+XnAuK5e158Ns/v9zeGiXR5t0p69+88/e7nz99vvjwsR3GImeVcfGV 0+cn6XBmLUd+uHhgqc5VbMAwTZJFw847M0yz0Y0lAFqBAqXHG+PeqH7hwYWhXD1eoGvlg9aRHUcX ZnxwNDHZhzS7dlmZOxPWJmZFk7uhMdN0lg1vdAW7YsVWRzAy0zxOogVZN8/fZJdPvIK/ZghWUzg9 OTii6/q/oR6BwIusHCSX5jSZR2TtAreKL6G46c7hNrE/aRkupgmhilAoWkp0Dq1slry1Fu9l9DWd tYwa0UwQrRActu4p53Bsosx+wVr1G+s+bRc9yTWzcRpDUPQw6R58dmSTze3aWR4loygbPbuw8vc2 SswYCiPBICGkihMbgYE9BGakczAu/evxOJ227FmlFFhYowfWXaxp52DL08vLsXFxWpzM0tn0bGLG 4+MrM/xon2wXR22TDck4BmsxId0Tv1kyStsO6LgiUiICVlcrq53Da5hOvraLlxKIa3caMhwu1j24 nEbOdbN9tRTWTml4QGc9Sedgm7jGrbYjOWUzeorhcHUvAJ7m6cQKWdS+7RcKEYKwVHDYuhfBDdvP FsR89ziHo6U6h9b0Kv3815mZTmtFuEAxs+afKYXBuTyhqHPAZabtEENqJDWz/g8OF+4cXNP5/r5n 45ZFjCvq7rGm8KiMdi+KneZRlt+PB7AuUyoiGNym0e5FsyMra20HGoxKjBgWNQDrYGC2yNZd40vL 6nnbuANFjXXPAyxRexEno7nz3LZ026jYMWmdqGJw+8a65xOG84MLb3d+nya/x9N4vs5kopazUUWU 234AR7F7yruoGf01i66vo+yeakVCCiaUACfzlHUvX1jgd2bBy0+Tqcnyk3twHwJxjBWH5/WUd1UC n83y9NrOcfj3WZobd59OHuez9k0i1lII9wccyu7l/BtQup6DEyuM9+BZJFKKM8bhC628ezo9jT6Z m9WvtuvAhGrCmYIv7IsOZhxruD1redkQY4moogJuAkX3Uo7MfHKeYwndeXpmgRy1HUtrLbHEBI5f 9yrq2SxZ7ru62Yg1n+rbyLri1mvGtxvLwEB21RcvsXR7ENv2vrebKMHgdc/7XsWjtleruQVL0zqy 1j0v6+B6k1+Z7NlkMo6H83m27DQoI5oSTME1USq6V3ufJQ7BZ+Px/eGnbZTntufAna7sXgEmMZ/v J8iTiFOG17esVEate6FKOjHJPcFGiGRccvCyIpXdi1Ci0ci1YZ9nUQyEbdXpDTF0VuS0ZnDsGOle oe+esVPYna9XQ11ZB1uZrtNR/OHrC3CjdT3kFOdMUloDuQ4auttWTodejfXtWhBid1GgrJGWsQ42 Ot2r8LmFW0F1HeHrXh6WuRurn0dTM46TlnOK1S3cYNy6l4KN3XXS94Pb2h3ZYOC6l4y5Fk64ooKL Tre3oIIh617+NZ1dTIdZPGm9REwll4IrcFeiTXY7CNrEZPcBm2BMEs7gqSrrYLdY7i4VM8mvJkva 7+aUkmMt4ftxWAf7xmaJDeLa9wWUIk2EhjemsA72ja0C3/ld7/fQ2bl2ez0Yv+4lDrOpK8q9jC/d 1mDTckVOCMqRYvBeFEa7ly+M03R6T1aOUEU5xhzuVLvYgOd2RrQfva3uPgdj1r1MYZbcRxRC3J4l Igg8ne/gxn1r2G52td6PjnKihaYIvFrDaPeyhTXw7sk1SJvVa0lwDXXtXt5g8UrefPhwT9Bxoah1 qgqePHSw0XgJ3b2orBQ2CiaawtcdOthgnJvs2k0PnDxwBArhKJFEUgk3c6yDenq3A/Fml+txmi0v em170xOnTNTxtx3s5lw7TGKtHfZettpxTDUScM/Bu+d4NyTyJJpe3V93sbCajZQi8GKK6KA9vIvk ufmSvzOTcTQ07TcDcCqRtljCFbuD7Z6zycRkw2hq3ttvtF1aEUhopuAnh7EO9nkOo0mcR+P4j3tA jgsuEYK32rEOttrNl8XuR+CUdcHKmj14KbSDLXbzwsqz6dtxFCfO3LUMHdGaEQGPnTt4/FqI3Y2U CImxgG8JYB08iC0EgNa/amSzN7ivUB3U3QAAYq6R60KpocLdS+Xc2R7uFN7WT8JantIMhayVA9je zPKxyetCNvId7bsbMdhJf6uTh8GIdW/pwgZzl8nxfCd3y5kYRZhzIsBVFo47aNtcXDK/wrDtje43 9zGCQeuePVukr/eA2upWWChqpHvV47mCvnP9Jy2nW5goobQG15x4BzvWXTfdPQja6tpbMGbdazuZ C9rfZnaOH+K2N79qxoVUStUArnvVpDlwL82HtrNTyV0FDn7AMO9iZ3r0cd5e/T6L7TQvT+Ll6sRr t7oYtX3KmPWOwp0CCEeyewHcNiSdQJ6n9+A4iA1rbTwswVk/72D/+jY05ziep+1ruUCS2vhFgzNY 3sGmdoemK9sF0XKqEKJYa3AhlHew130bkvem5Ypwrdy+PDia3ctEtqF5b1rOhX2BcwmXze3N8Xf/ sQnY8g6gN25J24zOTL5EbPHcxiVAd19+fbZ+c1C6GOHN/JHbC8Sgt4cV6z5rdKx4d4vC4nGHwPr8 i1AvnluiXNIFMryKxyP779VeovW9bkm8nMskss/kW5lUexJ/JtVnUV6eWlJYnPC62C/GW1xH9SIe m/8zHbz5nJjNi82bmx6GTK9MdWpML5vmg3dmOknd1oPWZkghMyzrRwLPcO20ibZmd8dk75id7xbI Tfvhu4DsRgFvXqpxAeEaXYvmo9t7yJ9sfUZohpmjt+QZ7S5ysflB2TNUaoWZILTkGXdEpZRMlY6D nGMVqGyc1b2cK4jvsrqqVDUmIRggIR62hJIQz+W1ICSLwt4YolwAdK4oxKEQxda6SY5YBxEluPyG 2DuQenQ+mBlTWvJuCqkCuD2PhQyGKEGIUylYBxEFhEq+aYRS++K5mT7bsHlEpMe3bd546OPe5jlE nmcKRzx5aXZnt8l1P1p4qHiXoOdjhTvLPPMiHAuGRRnRxSMgfURjt5a1vpgFk92iGjYmu5KDspiC wJTSXRT25nQOkEN6ZDgU2VJVJ7uoVsGoBqS6Hk0PJiMQsovGJxTZ+M5az26dLBjEcA5FcIFsliBr 2rk2RRgQR3r8RzCyAcGax6UFIxuQWnq8bDDNwwCL4fH8weAGGQxPMBLM0oGiD491CZWMFO5bhhk6 j9VuEFRAFcI3kXLCC0xoLtdHAGnwJdiBhKF4D5UnTShciuZzn4X7bD25DSYIC8xw2TOF2xF9z2zc 0ll8pHhfrI9mhighgpflP8Vb9Dwfc8Eup+tWqVjTlYQRJGnZM8I5QSZoKTOQVARbMS7DRymkFC6t +xb2msCMQLFc1aAqARyZR37LHUJB9RosSQPyI49OhaLbSiVoLamg58EIx4DQwWN7wgkKIML02MNw ggII6Is2OhzcAPn2+I1wcANiIZ8vC6eXkFio6F/D0Q0w4B6fHyqG04JSjmntCm2rkAJk2BMiBcuR XHuuZW8XIUWA6rFvHqGklCkqGcelQbLAmurSxZjiJeC+YTbvl/TEvzZ5tPEXL3umeJgMMEguCHSD igWIITzQl0tvQWoalF7AemNRHMKRDQg1PSIajG4CMBY+PxLIWBQ3ons0ePMSXk+GX9iOXXymeGiA p0pSuDYZZgiKwUKDhgASlBdhLTdgBYlokG6AZBZZHYxsAmktLYpfMLoRILb1qEQ4uiGrMR41DUQ3 h9RUPC1soewuVa6wx8vsbvEOQk8EVrgnzhPJFY499DzDbPRlg7kyAy6QttZBkrJyMFNYUaZZ2bco YkQIJsvm5ZafqLqzcdT3EBeWIFw6s+JNhMCWsEILZnOyCzAVHoEpX4MpyHpzdR4KiNE8QhyMbg6J 0YqKFQ5vSMdwUdmD0c0gsXzRAAWjG7QromgUw+ENiFA9hjoU3QQi3x7nEU4vQbufig4tXGMCQcId aVG3MaFNGQbons//hyMcUIv3xSTBhBhS5fQITindRaFvzmoAnItvN1CowN8FFUKSsrhWEIszsapQ 1p5g5YFQQsqewQhjyy2ma7daFjZ0NcY+Deg98GBWXgUqsLu5kB0QGRS5GIpqQgFW1SNZ4egGtSwW hT0U3RogJ57No8Ga1ISgFsfaq3Ft7vHTgCDRN49AkLpN/TYDpmWtb4RKxbAQdWEvSlCDsIN6lgpz LaW7yKYG6QbVVor4h6IbA3IKTxtjuK2s1mlYcmjdSKPN1gjIDh/PPAJBKiwcBGFZuhuS2t8TUX9T dkGCGuwGg8RKxbnuEJdNNjVIN6Qz3oN/ILopg3QAFRs3Qq2c2wwDECsVm7KDqae2ea1UtG4RpUWL RyFFFN88QkGKOMaKl7YlYS2FcH+UWUUbgXLGeFluXLh4yfcpRIUq3bRdvHXIk6kXLtQBRncFmW9O UCArIh7u7Fg42xSsBumGVF6LEhOObsjKWVGKg9ENOQDAc9BRuFZIzJjUtVWvxXyWSoi79swjVKCv iZUyIcu3n3HBtZZ1XWNRghqEHZDPeuYaLr8Slg7B63YxFCWoOUgVIIDzzSNUZYbbKFiWH8pCEada s9ITaRR2p6WWn0hDqQ0BhC6t1nNKnasvzffcVjZBdenGRes1OJO09BmNMBKkPEKygi8klqWRltCY Wz+FSw/RsQ8RplTpM1QLTlRpGxNGmAhk/yjdsepOW9Vcl5knx3dqhyqbWPEGdmB0XzAdDeobZIm0 KOQ7spJN/WyObg0INj2KF45uQLDpMQbh6AbUvTwGKhzdgFUfj9EMRzfkgI2iIQ9HNyC49ziXcHRD IuiiwwtHNyQELTrhYPGStNkoZrVPMmgVUkAI6olZwi0OckWw1Lpu/b5VTCEn/xZjvGCYKiIok3UT pRYRZaANpcWIOBzdsFPeC1F6OMIhTUbFzCEc3ZAIqJjNhKMbEAH5MqxgdENOGCval0CWjrjjM8id XdPFBJ0xSTjbkVhTyS2yqmwxRDOsXAS145wnLgnTqPZu/oIbaZDHgGjRg2x5n+umTDRINiBY9DA7 GNmAgMwnf8HoBu32L6hEKLIhe3y9WhqKblBk4QmWA5leafmuJcFlFpMLRZVNl0q3ZgrKkX2odv9O IZdpkDegg5ELeJSLVJGXDRIOiZ6KTApIOGS/nEdywhEOiZ88ZYRgARTRQlNUtn1YCqEJ0aUnW9i4 lGMtWVkXCKGKcox57QWFQiGoQfaBQqMCZuVyV2B3g3SDCmkFPoajG1KtKspWOLohxyl65D0U3ZCz eSTngmJZ+aDxFjoiGaQry3MEbShzapMmqonGHTzLjUEaxnzzCFXMRAIJzVTperaSmishy5sLhI0T Ue0dJkUpa5A1oCsyi3iUN08WeNkg4YA403MK3q5KSFsN1wzUhOg5mS8Y3ZBSSKG9N1TvJOOQdo1C y3EwsgWkNbjYBh2ObsjtHJ7W7GB0Q1bhixY/mBUUgOTB44VCka0gcBeP8A8Vx0B6qHy3yQaKY7TV MoWYqrs9sMVGbAZp7/LNI1hoyJFQqPZdwUXpaBBSUEhXnEcZ3UUWNEg3IDLSUlBiY8CqkWgbRgzS AeVryg8kusxdLYV57SP5WmxmZ5DmLN88AkGqJHeRTOl5koIizDkRZeVOzbiQO45d5JgooezMy5aC BcXYakjZMxRzLGRpD7pGyh19VHr5rGDM3VVWugfPnVmEsSg9dUcpKrCN7eoellNUsAalEnJhQlEU djiCTSlukG7IRtSieIaim0M6tTwqE45uSJmrqMbh6IYEDEXTEo5uyDXpBXMXLBrnAmmby+raoWOL iEJCsKJ3CEc35LSLoscKRzckhCx60XB0Q+6a8nj2YHQDPLlPTwOZjPnVaBrRsmYWzrF25d/S/XlM c5s2ydKmGCSpVRFdtuRucbTmlJUuy2uFJcdKlQWYVCFEsdZlzyi3G9lt6ypdX0Guf1jWNakF59Cg xEFWbYtsLs/EChLaHN2QnjaP6IWjGxCDedQhHN2AGMyjouHohrR7Fc1GOLohkU7RlIWjG3IWSNG8 hqMbEOl4TH44ugGRjs8N7UP3nX/d/ceWQOQkHuZxmkTZTTTyYRzluUnM6G2WTkyWx2a6nC8wKil8 cT3mGdnPHk7TLDejX83Xmy+AA5+n0zyLk8ujP9PD0+dvx7PLODkxE5OMLH5fnz5Z/vbusxgT4MPX kzmhL7L0+vQ52fYkYYBh3cMVh+WQYTlgWDOK8zR7Hyej9PNxmlimO/nOz74mw6ssTeI/IicZ7kdb BxEQ2kRl2jRkWF11WIoAw7qHKw+7P5IUQ2jDlWmDyCStLJMUwndame8MosXu4YrDQkBglUFgEMVk lRWTQYSfVRZ+DhF+Xln4IcRWplVDkNWVkdUQqdWVpVZLyLCy8rAKMqyqPCyIY1VZRpATr5/XzODL aJqvmcLS9yqSM3+4Mjl722SCADZ5/nBF2gDhx/zhisMCjOf84YrDAozn/OGKw/Ka8gIwDfOHK5PT gLwA7Mv84Yq0AezL/OGKwwLsy/zhisMC7Mv84WrDYoidwJXtBIaoOK6s4pAMg1TOMNyT+0sphtgJ XNlOYIidwJXtBIboO66s7xiiqriyqmKIquLKqoohqoorqyokeCOVozdCaoYCBKLipLKKkyZCAQKx E6SynYCoYmVNhCRnpHJyRihEWmhlaWH1hAVibSobG9aApOh684GAWxlbfbig4U0W24G3P9XArCFm ep6H1cEIItj24QamBbHnkISNMsBc5g9X4jhlIBqqegrK6kk1hdQyaOVahntyb95SSEGEVi6IUA7w FPOHKw4LMHjzh6sNKwF6S2U9vV28V/kjAJWjStWjSEHUREGEWEOkSkNkRQNyCqoBkQXVENmCWHoK sdgUYmwpyNhCal4MUpFikBIRgxR+GKScwyBFGgYpoTBITYNBKhUMUn9gkKoCg9QKGKQCwCB5PYMk 2gySPjNIUswgqS6DJLAMkpYySLLJIL6ZzT0u3AkwiKdmHMJ5DuE8xLMzDuEPh0AuIJALCBoCIq0C Iq0KMjLEMzNdU6YgHp1BPDqDuFI2d6V1yIeIF8QFM4gLBnC1asGPQ7w6h3h1DvHqHOLVOcSrc4hX 5xCvziFenaN6ETmHRAMcEg1wSDTAIdEAh0QDHBINcEg0wCHRAIdEAxwSDXDIwFXrGbyeHYN4SV61 Cm8f3LvowCFSXrU6IwFiKKuuNUmAuMqqpWkJEGtZtYYBKWHIqv4CUoWQVSVIArgvK3MfYAtl1Qqb AthMVbU0purFVApgklXVepp9cG9lVgC9U1X1TgH0TlXVO0igqXdQutFvXGF707zV91M0npm923yH 6fVhNJmMzeGpZVj2IRqa57N4bN85PE6HabSY45aZ1Xs5ts9emmwwn8BPB6tNrN/nsN++cUJ+HChM /vnj4Btm4scBof/8Zw9ZGWSMWpgU0g4ywX4ciB6wfth+2H7YBzjst2+S0x8HWEln7Yjk9u/43j3E t28CKWtoF25qToXq3VSggIFJJw9ozglu3R+5f1Z8+2YF8scBo2xOBbMSav/oBaIfNpRWrNsnLOay 2YtjP2xQp03wwmlTK5oYhXbaDAehoh92wYq5p6bznBVr+3cZQB44cYEbWhhJLUuJ+PaNKuTCPOGe ZtLJM7p3kqVFSgmynUjx40BLtSBRtUDimi45Bgq90GjcgoNZ1TcW8fVCbSlr71OPqH5nTfK8uLbM o4gVbUxDm2Rs+Sx6i9zCsCuTp52qknm2pGhHVclZZmdf53Z5bsYaT+n6l/uX+5f7l/uX+5f7lzv7 8k25X8tFLIBaCBD7l/uXu/+y1QRXrSBqkXG30hkR/GWbD8p5WWGxmuNKDLgP/Rcvt5sZ9i9/Ny9/ +6ZcTrxc8sSu4wc3Xj3tX+5fDmzPvn0jyno5TPSq1SPE0r4rRUm9XNkPQ0Q/bD/s9zjst2+ME7fM ulznc0FriIVf7PSWSbagQgeh4rsYtoEDumfJOB1G4/iPBo7o3jhrfPNU7o2bEO+n278EoiRejhNZ TD6Zlwsc5hs1ljRWRnG89u5jQ2+azrKhOT25wSy+OcX+OvoyP8Kerq6WWB+icCj+sfvLiZkOs/jC 0rIY5PT5STqcXZskP5z//mYqVeZ/i+jobWTlOhqvfWGyD6s2Kd8y/MFd5V/ez+Meeh1dm4Oj9+bi 1zh/GScfn88un00mJ2ZsLqPc+C3dEu2Z1dG1MV6fLe4S2HynBNw5bWdzrh3cYeHIohx/iN2MF6Pc /fR19B9pZmXx4MjOOEvd+ItRNr698VacLN/aNt3Dq03i97yFACYDv9ND8i9hBeH1mYVjHA/n9gPG /XdmOkkTS1yzAjCIR+4eDMIQ54jtlocXmaXnc5p9BEmEnbWViMMPNy8/+cVEdirTJ3cAgcjH/bCk CWw5klRJwsNie7YgObl8kCALwhTFmKsAIB+n4zR7GyVm/NCgvX8sfzHjyasoiWwM3IO5L5hvLYwv o6/pLO+x3BfL36Ymu03kTnNzfWaibHgVzp4+z9LP02IwUB5DuA3rWTr+3pmxnHsg5LeAWI7877H5 3ILfoxhpzJgkQfzeHIdAXLhJFptPobYn/9FwPZcEJxGeL1W426zGZ26LUBP7Z36D1L9usvz2scx8 Mtntc+fpWfTJjEoenyUr17L9sakdZfe31596Nq323HlaeK5wC2u9qsk+YMejrcSH/dXmHYU7rizs uvW/LRN1y/B8rzCGTgxv6Fi6kzE0pnlkJn04toHnOzM0K8NeYjUT83m3CU4nJqlmqJ+Nx9s/+jhM 8AM1pivtC2QGXlgClpl/+Kr4PLbWTDAqpAhRFF1DIxg/smt3o28HuNES/C/SWTJa1J39LFgCEIgB ryL7my+PMr9fTD0U7iaZhZf5uQXCiAsqiNI6BBMsDgFZ4GpsHWEDk0RzxTgOxAUHRShOpJ9i46td 3Xe1KwDyN1MPBP135EnLViCHQzOdxhfxOM6/dgrJjS6courP23BWK+O+TqyuEHm7xNxhIldLtB0m clVPD0Lk96bdq3JK8PzxgQCaRZeX4Wpx+6jOKl3uVadihh+yaeRhoGh//7urHj6Pk9H3qTZrOV6v NxU4/jq+eJlGnWP29wbjm1k+jpM+v9kXx7fRNDcXaZSNehj3gNF1HDw0b7jo/kWaM85UiAadc7fa 1mv4viim6fgiygIW4VoTTiIQkkqKEK3pvVw2sEb1zLWgfupz7r2B3KSxB3RPQH/OsrQXy73XoOOx CdsG8HCwvCkVHKd98tggnG8upib71CPaCKJm1Hv0htBczKSHsREYQ/cIPxw836ZZPm9j7ZHcE8l3 s+Rlmk56HPfEcaHbfQLUCqx90NkCqEufNDHDOS3THtZGYH1/lU7NuZnmPaD7Anp+ldm/9zDuCeNv 7172GO6P4XGaJGbYe/Vm0DxJPyfjNJh2t3QMVEtwliwA3Z6mcZZ/HYOOBrsPNNff+K53OZVw4Hby gcC/XZ9uZJNTxVbp2/X4lrutzs0X4NkUD2HThJt1KGmyn4YLk49N3yPoAZfTH4LcHo/jSccgrGbP VnvTWrZnD4HL72Zjk3WMzeXmaWt8Uk06Vg0+LUvH+zgZpd+9fJxk0edg0ZAfwpri8b0hv5h8p5D/ PiGstVK3/o87f7894XvtcO71E9BfpSOLIFqd+n0XtO3H7qyNtzgI/8TYdzIzOjETYwU6GX49MR+i 2Ti/OQV8HY/XZ4f2v4fooOIp+4fX0TCd3j1LoXDUPhLo4Hb4xTwOl4ew38GkgSl+MuN04n5zO9vb o/j3m+fmL2nplClC1aa8Oo9nN9/sTM6zaPjx7Os0N9fnUXZp8t+thM418c7JPXemujbwy2ia/5qk n5N3Zhy58/qXR5C/jfKrg6PDwyd3jhc//OJOCJrYR9bn6pXc0UKilrcffF1sTV5mkwdH9EaKnz4Z RXl09KenT6JFM8bRn/4/UEsHCIx9WruULQAAqLECAFBLAwQUAAgACACnoZE7AAAAAAAAAAAAAAAA FAAQAFdlYktpdExpbmtCdWcvbWFpbi5tVVgMAM7vKks65SpL9QEUAFPOzC3ILypRsHHOT85P1AeT ehl2XFyZeSUKuYmZeRoKIFZiUXqyjkJyRmKRghaQXRYdq6DJVc2lAARFqSWlRXkKfsGOBQU5mcmJ JZn5eb5gjRBNGsn5ecUlUL1amiDdCprWXLVcAFBLBwjJzB7rZgAAAHsAAABQSwMECgAAAAAAS6eR OwAAAAAAAAAAAAAAAAkAEABfX01BQ09TWC9VWAwAzu8qS87vKkv1ARQAUEsDBAoAAAAAAEunkTsA AAAAAAAAAAAAAAAXABAAX19NQUNPU1gvV2ViS2l0TGlua0J1Zy9VWAwAzu8qS87vKkv1ARQAUEsD BBQACAAIAKehkTsAAAAAAAAAAAAAAAAfABAAX19NQUNPU1gvV2ViS2l0TGlua0J1Zy8uX21haW4u bVVYDADO7ypLOuUqS/UBFABjYBVjZ2BiYPBNTFbwD1aIUIACkBgDJxAbAXElEIP4qxmIAo4hIUEM kiYiUB1zgJgfTQkjQlw8OT9XL7GgICdVLyS1osQ1Lzk/JTMvHShZWpKma2FtaGxiZGhuaWECAFBL BwiUVkhuXwAAAKsAAABQSwMEFAAIAAgARKKROwAAAAAAAAAAAAAAACQAEABXZWJLaXRMaW5rQnVn L015RnJhbWV3b3JrLUluZm8ucGxpc3RVWAwAzu8qS2DmKkv1ARQAdZLdboIwAIWv51N0jbdSyZZl MYhRqIkZKhHYz9XCoGIjtk0pKjF791U0W4ahN03b853T09QaHXc52BNZUM6G0DT6EBCW8JSybAij cNp7hiO7Y927Syf88DEQOS0U8KOJN3MA7CE0FiInCLmhC3xvFoRAeyCEFxDAjVJigNDhcDDis8pI +O4sLJAvuSBSVZ4262nASFUKdczF/d919G5KE2V37qwtqWxnOilZmhOX7EnOxY4wtSKZVlvofKxV hZL68jZmmTbbWOi6bvD4SJJSxV85aYLdE37HThSOJx7+XIzn+LvNY5bqdLqmRDY9dFGj4qXUs4hZ ZXRP/mrpRk5YGw7kOjH7D4/0l2+PYGvu6vq6YCyr18vDNNOejH4b78fJNs5IWImbntP520sbFmy4 VNe0oFY0abM9M6AZi1UpbxJHerRBLdXMP72FLv/AQvUvsTs/UEsHCAaEznVTAQAAvAIAAFBLAwQU AAgACAAYpZE7AAAAAAAAAAAAAAAAGwAQAFdlYktpdExpbmtCdWcvTXlGcmFtZXdvcmsuaFVYDADO 7ypLr+sqS/UBFABTzswtyC8qUbBxzk/OT9QHk3oZdlxcDsk5icXFCr6V4Zl5Kfnlzvl5JUX5OTmp RdZAucy8ktSitMTkVKC8W1Fibmp5flG2gpWCX7B/UlZqcglXNRcnpk4FLYX4XCzm1XJx6WqU5Wem aCoUZ+SXh6cmBSSmp4LsSc1L4QIAUEsHCDAryed8AAAAogAAAFBLAwQUAAgACAAYpZE7AAAAAAAA AAAAAAAAJgAQAF9fTUFDT1NYL1dlYktpdExpbmtCdWcvLl9NeUZyYW1ld29yay5oVVgMAM7vKkuv 6ypL9QEUAGNgFWNnYGJg8E1MVvAPVohQgAKQGAMnEBsBcSUQg/irGYgCjiEhQQySZk+gOuYAMT+a EkaEuHhyfq5eYkFBTqpeSGpFiWtecn5KZl46ULK0JE3XwtrQ2MTI0NzSwgQAUEsHCI19tDJfAAAA qwAAAFBLAwQUAAgACAClpZE7AAAAAAAAAAAAAAAAGwAQAFdlYktpdExpbmtCdWcvTXlGcmFtZXdv cmsubVVYDADO7ypLtewqS/UBFAB1j0EKwjAQRdfmFEO70YUeQBEKgruCuy6KSGxGG5xkShItIt7d 2hYq1G7/Y97/E2tTsQsQpc+9kwZrdrdVGQkRDyDTVnG9YxscE6FredJwQoM2yKDZws+9EMv5g7Va gEJJxIV4iVl+MiMRFMQej5sJ6pBQ9tzfqybpfU3yHkp8yXWG54O84rfon2kLeT7+AzoZaKvD5IjW 3oZrj3TpqhO0SnwAUEsHCMqUhBOjAAAAOwEAAFBLAwQUAAgACAClpZE7AAAAAAAAAAAAAAAAJgAQ AF9fTUFDT1NYL1dlYktpdExpbmtCdWcvLl9NeUZyYW1ld29yay5tVVgMAM7vKku17CpL9QEUAGNg FWNnYGJg8E1MVvAPVohQgAKQGAMnEBsBcSUQg/irGYgCjiEhQQyS5suhOuYAMT+aEkaEuHhyfq5e YkFBTqpeSGpFiWtecn5KZl46ULK0JE3XwtrQ2MTI0NzSwgQAUEsHCIUKPx9fAAAAqwAAAFBLAwQU AAgACADWppE7AAAAAAAAAAAAAAAAHQAQAFdlYktpdExpbmtCdWcvTXlXZWJXaW5kb3cueGliVVgM AM7vKkvz7ipL9QEUAO1dW3PiOhJ+nvkVlF/24cyAuYVkK+HUJAw77JBLBTJz9mlLYIXoxFguWSZh UvPft2UZ3w02JiHJel7GxN1S96dWX2RZPv7zca5XFphZhBonSr2qKhVsTKlGjNmJcjPufz5U/ux+ PEZsekcWuMKXJj5RpnReRaap4+rA4Jjdoik+tYmuYdasntEpRdW/BqeK32ynWleV7scPxxriCP7/ cEwMXrnHyxNlcNqjU3uODV4dLS2O52PEZpgr3bp6oB7XgM6htzgDiVJYfsh+BM9ZuwVcknodY1Ru r4lONv4vpvmd8EDHzcNqo52F89tgTKk+oY8ec6t9WFXDvdLJ33jKK1MdWdaJcjE6tzma6LhHphxY EFsqsXavdHtGDLdRS6AN7UygK0n5VQwq1n4SfvfX+fAMrpnS/c/X0XFN0EjqaK9fGEOrnjToumpR xrH2HS/d9nN2sMKkm24/0nykLkFEknh/4gmMweA0gfq4JnVJ1stFM6beAuk2LqhaxH7SbwRFDF1n kDYw7F81AkNy6fAMetuNO0wyPMOs4uh/ojSU2kahUqSRQ9HDJjY0cCMEbydQrqHe2qayoL5uxp1j joRDe/a5ViHaiaJuZZcvPhF2ZNbXlHJp1K76dVVVdwL0mW1xOpdte03XldBkla0DrWC6QHOsdM+X P4mh0YczanBGdR2zHA4nsc/mxj77hFn8GlsmhdlUuL/Wxv7ACmECkSkSJp+jO4nMGM9NHXHsddhe degF+xXpiC91fI6se4icbS/IJ1Geouk9CKF0G0GysAaS8lro2n16qh8dfKo0O78/VZ7ajfanSqvT /P074pKD3Yz7OpqBrbdbQNluNDd2NCZcB7Dkj0RfH6Z3ABbgJnIYRF+R/yDYJa6taw4MEGbIOXoc kV8gx1Oz2lIbhw38R/MQFPd/RJWOjproTgm3Lf/kDt+BO3wfjhm+xQzcOV5RX+BH3y5X0oZgXbio NtoHATwzeYGL0cieLECSlQPKHYtj/UD48FXrtBpgGPDPa36jhhW472JS85mS1O20guqmKzyCFHfF 3GNoNoa82td3C403hBQLryLKyqv6jNt05oVd4TFw5dv4fFgxESTjE4qY5pQJYfNLYLrq9fPzDM7G uZlurvMLByZT+eIWPdl5wZ1SnbJcHH2iYwP8r3XlMIw30Y+ci2zE+K9x5Xrc7+URyOEh07vKGOZA pU/ZHPHKAirDvI2MB/1+ZdHKxQiww2hZwv4cP5sNFNOeQNiqmsZsM5HN9ExEn8WYxClDkTD5D2Gv 3WdObH16aqifKg1VxKXWoYhLzVhc+pDkiUa2idnC8V6JXijRdUHa6Di8WrJUjkxOzE8h+Bejthkl iPnUK69vy3Uzwb8EMAuVwlAYcHJLAoEj3nh6/v0jmJpK1m28V/5Sd/vOgsYNVUwP3yJb533yiLU+ BHMRxhMsMoUrM8O/0QJ9NQSGWjbi0ZQRk58h4xKqN5kPWF9syCMhHZwiXV9ubuacGGRuz7MLKYsy K13O2OzatoopMnyR+rjeCFpuhvuiQYyM1f2Ly5z3o+3nYo8jmMGF+TDdWFjk4BAGHsAPD4kFSfbF ZQSkAP0XXQe7uQHrSUrLwh1FfyY4TjfH9TL5iImsc5erJDrUSaQPsHmMjVXpAA7a8c/1g0O4rEMi Hascwux5k/CMxfHgVCZpItVHxBAJaLQ+DmVxWWbElBoGdnwpaEuZtu0yV1TSs0i7XtEQU8mGcM59 8phcfrYfBFlHE6wr3YeE2ik+/Ba12RT7gTIwU6KkGrY4MZCUZEXf9muZiGl6ub4v76CndJvBYjFs aHsC6r8PstZ5Vqj8EionXu1Xh9fNoId1PEM8Fq3WIhZHIJuF1fMidpCG2LolmcGp6wEuGUxirPnV pqSLuIB1yRCVLUSqxrxxNMXBhUc2gAF117OVrhqqp6MYU3eJy4FXTR8LqON0DX77AxEk9tZgTAQ0 PHWItlbicyO7FuvnoithXOGgacv2VmuIOv6HVbl8MCIriLtUr55LveZO1WMWrySvke5Sw2YuDVs7 1DBtSXaH2uUbvqDLy5J4rKZeUraYMQkPiJRQA0fT14yQ7wy+fJP7YP/wJQWvPBiuzZS2hrGVHcYM 0Tcq8EGKwJljavqSxK2OOMcG1q4YheKDk62fIW+xKpE7ErsupF4dnIpnx5TJUn+ILO4+WhDlUNTV BLgiz3ijywJB0vCDIfmoOksn0UdK1QWxiJgYfIhsY3qXyijZLhmJbR8IEOGA2gFpRktomVGD/EJu 1pkmYCM7Cq2NpGvT4KyLG9tawdNTo3UEBethc93zskL7NnL3lbra4TXkPOJrtNQ0McVaawNah/rd WW0VC68d9Xm0yrRHYRdexjZ0OkU6+bUDPxOtVKL7DcKR8mVscg1EXn6OAJMFHkocZGVVy4eiHuD9 f0NPFrAQal3MvBg8R4/hUnPt2pTziLqHrSkjk6SlKec+zrE05eGpXSGwaqQHejCLDFRU8pTmlfCU dvOuzFs/IoyWWIAMcAef2ydzZ7Vd6ix0WMnLGRejKvxfVdesAEXIZbeC42ciw6bELgrjyF0fCdqa /6wnUYo5+psymDRKFwaHUdG+bGWt7HNiuFzxkanGcoOC2V4+Q/3RrDb+2K+1pm/i2WioqWV0saGX uy7aB0edRrPV3GwJzsL/A2X3uWxB7sit3q6Ya98wAlWsWgiQPPbxMkOyA2wbaqupqp2Gul9s5QNE YH+XINcPj9TOUcffFvWCIDubSq6QgfX3Bu3LY/kN6+Y5MtAsX7AowUwC8wpgHKIlZCYllkWxvLEw 8wrQAcfzEXbeutmbPz3Hhp0vgZBrfc/gfJuH7c5Bp9ncR4QTMOxpBJ4Dzz2kCNMptiwyITrhy1eF ZOIT3qDtydVmL3VN3lLyKoT0c8BXLKSfQ+1FyLc2cdyy9lVNmbeGob+Kst0iQQloFFCGZrP9JSXv A0SxQ3a/Rcj7wXGfdfH7QBHuO9vqT4mhvbaZnS2x8OuDMrHIMOAXZDKk6NWN9VuDUew8JQYWz05K HAutIq1eDythLADjCC1wGQyLojgWT/3KKV0URXHGDWJiGfMd4SiftjUO262jutreA6ylXWYBsE9t Q5OPexPXI+Vr3WXlVRTIqIwloAUB/coYLc2yKIrilZdycWU3WK6WBs5oWS3uEM7LiSXeUi4R3QWi WCsj+o7QlJqUMO4Exn3vh3s/eF5Rxp1dsyWSBZG8to0hpWaJY0Ec3cNyygLoOWAtk85nANWNSSae OrJYJaw7gfXnHbXwGFu8BLQooOM7BtcljAVhvLkelhgWx9A/UKhEsziaPfpg6LSc3ZuxlK+RJ+AI N0oUd4GiOIcCWFbnfpVgFgHTaVFsJirh3AWc4iBWcYDEoIw6hXCUh3CUGBbH0DuEtASzEJhUJ9Nl 6SR3geU1llKUYWdXiAaXiEokiyDpnye7Jxyf6RCPZ4JyzV43711o52tBrw3NIMfLvAj98iPgKb8n 8JMOJXprGJ7pxNzjrs33AKF4AX+PG4nfA4TXto7ZKzPD9R401fFmfB3e26T9zC+EJX3U7e3ZR4+h h725+WQItzSPt4a8VP5VIf82Idz3bqtiZ0RmtOf0wyGR83vj4ZDWHX2Qkv5zLfCh8yFJ9MtML3w0 5DNZzJ4PhEg5dXO9nWxORnKYSMGvamb8atoW59W7+s2o+PRUzFAD993PUqWTzNE9Ft8xHIqvyrPN dKM5EgaRgZAjQ4Ouxfef0qkZFo/rNt3vMzqXZ5anU1qcmu472+lEHIQbOZc310PR6hpSOpvpWEwB YtjUtkYm1vWzO+x8bncTG6DE+MCwMONitYPHEYh4iEzHqyZ+tG1764l5rfLWuiF6Ez593cpb3vIm +CN07Z2LHTjUOnhu+DlYXvDDRJkdcKC91TcBgIdhzf82gPuZx9Xp2clBPNuB9dU5mlLrMQRI7HB9 9UBV4hG/FsNkByouwNeZ4o6nrXeAfTE9ozeba1Vuim8/ZVHZdzybxw00GTOIVaOlxfF8LIIN/wF2 6aQBIRcVUjXQ8BBZ/LtBH4xrrCNxyr17HvYV4ncrux8S4/7UnlUfhR804X5Q0USz1aQ5uR8MWMqT +Ny8J/D1uuOahjjqfjyuIfkmTffj/wBQSwcIG/swixAMAAAZhgAAUEsDBBQACAAIALClkTsAAAAA AAAAAAAAAAAiABAAV2ViS2l0TGlua0J1Zy9NeVdpbmRvd0NvbnRyb2xsZXIuaFVYDADO7ypLzOwq S/UBFABTzswtyC8qUbBxzk/OT9QHk3oZdlxcDsk5icXFCuGpSWGZqeXWQIHMvJLUorTE5FQF38rw zLyU/HLn/LySovycnNQiBSsFv2B0Qa5qLk5PJ//SkpzUEphBCloK8eUwM2uBpqbmpXABAFBLBwij 5vlkaAAAAIUAAABQSwMEFAAIAAgAsKWROwAAAAAAAAAAAAAAAC0AEABfX01BQ09TWC9XZWJLaXRM aW5rQnVnLy5fTXlXaW5kb3dDb250cm9sbGVyLmhVWAwAzu8qS8zsKkv1ARQAY2AVY2dgYmDwTUxW 8A9WiFCAApAYAycQGwFxJRCD+KsZiAKOISFBDJLmJ6A65gAxP5oSRoS4eHJ+rl5iQUFOql5IakWJ a15yfkpmXjpQsrQkTdfC2tDYxMjQ3NLCBABQSwcIIBDG/F8AAACrAAAAUEsDBBQACAAIAMGmkTsA AAAAAAAAAAAAAAAiABAAV2ViS2l0TGlua0J1Zy9NeVdpbmRvd0NvbnRyb2xsZXIubVVYDADO7ypL yu4qS/UBFABlkNFLwzAQxp+XvyJ0CNvQ9r1TqTgH4tYHi/ZhyEib23rQJCW9Wof4v5usk4m+JHDf 3e+778aoGmOJB+tDjlqa/t5osqauwYZVwNj4pF/nUDwhRcMXVreMJU6qQYEmQWg0/09g7GqCcspR I7FPNmqh3vEbvmm7BuyxmiNVw1SKRSoUxInfBIqhGLzN2cgCdVZzPzxnX575bjy1P7YsUK6MkB6/ 2fZQvCL0rpfWAvXSOuLL88pBK6ImjqK+78NOQh2C7CIyDZZthJrAanDZ8jxqhWp2xqqwIlV7+6Nh mt1ZKw585lwHi3iSn7xm0xa0dHlKlxs+aA26eyRQ7dLYh+E+sQMssPRXEtZTYKhzCTvR1eeR+JfT X80nTLOV2U94ElwkwSVPs4ws6v3SGpU5ZEnGTvi2VJJP+fR8uc0PVPh3yJS4ndk3UEsHCCR3nWJF AQAA/QEAAFBLAwQUAAgACADBppE7AAAAAAAAAAAAAAAALQAQAF9fTUFDT1NYL1dlYktpdExpbmtC dWcvLl9NeVdpbmRvd0NvbnRyb2xsZXIubVVYDADO7ypLyu4qS/UBFABjYBVjZ2BiYPBNTFbwD1aI UIACkBgDJxAbAXElEIP4qxmIAo4hIUEMkpZ1UB1zgJgfTQkjQlw8OT9XL7GgICdVLyS1osQ1Lzk/ JTMvHShZWpKma2FtaGxiZGhuaWECAFBLBwiUa1M4XwAAAKsAAABQSwMEFAAIAAgAG6GROwAAAAAA AAAAAAAAACYAEABXZWJLaXRMaW5rQnVnL1dlYktpdExpbmtCdWctSW5mby5wbGlzdFVYDADO7ypL NuQqS/UBFAB1k1FvmzAUhZ+XX8FQXoMTbZqmihIRcKZIQKzgTO1T5IJDrIKxjGmKov33maSbWqh5 QcbnO/fcK667fK1K64XKhtX83l44c9uiPKtzxot7e4/Xs5/20pu4X8NtgB8RtETJGmWh/SraBJY9 A8AXoqQAhDi0ULRJsaU9AICJbdknpcQdAOfz2SG9ysnqqhc2AMlaUKm6SJvNNODkKrd1mZv7hzj6 a84y5U2+uM+084L1quV5SUP6QstaVJSrHS202gX9tVY1SurwHuSFNju54O084OErzVpFnko6BKcX +ACDPfZXETwkfgz/mDw2Wc3XbOxg1Oc6LTsyKoeEHozT1a3Ub0F450wvaLcN9wG+BriTx2wx//ad /efNkfixDvW49ECI7H7fBjms9sOZm/iEVJ8M5H0YY2lEsmdSUNyJkYOPUGTCUlZwolo5gpb6MUKn Wqq37tKrYkgvxj1Gacw4q9oq7RpFK8NwppfYD7bpwyGEKNo+xjDBB+zvfkFs7NzgtBjqkzQmjCfs 6bOfpr+KKW/HENLHjAlSBiVpmiGXpP0Gsoyoa4R/sAtuW+OC6055k79QSwcIoop+jLABAADqAwAA UEsDBAoAAAAAADqnkTsAAAAAAAAAAAAAAAAmABAAV2ViS2l0TGlua0J1Zy9XZWJLaXRMaW5rQnVn Lnhjb2RlcHJvai9VWAwAvO8qS7DvKkv1ARQAUEsDBBQACAAIABenkTsAAAAAAAAAAAAAAAA1ABAA V2ViS2l0TGlua0J1Zy9XZWJLaXRMaW5rQnVnLnhjb2RlcHJvai9wcm9qZWN0LnBieHByb2pVWAwA zu8qS23vKkv1ARQA3Rxrb+JI8jP5FSyaD5MoAzaYl3J3Oj8TawggA5nJaiXLQCfxDrGRMZNEq/nv 1203tttucNuE7Ormwwy4u6rrXdXlYhqN6m+fLmZTrXfx6bezv84qlrd4sn+CO+BtbNep/rvKX51V FitrswEb+A3u+AUfuPM/wcKPNwnt6GG466xxUZXAo+1Ux9J3aWuvlpq9AtUN3IAgLhpnlQqvKEq7 x3c4ReQVTmxxnNRqcs1+FcLeWrZzC5xt/dWeVyEWA2zcrbeANFw00AH2xoL/JHFfVR/g3wZ4QDSH mIVczBeNqyrip9Jsd0S5pYicJkjoE8fJWqvVCSC+gflX2x/Yzg9p+yiu1wpYgUfLB/VnRNmkAF34 lH7BUyIquxrkSO6JPKdIvCpIHNfutVotLuDrTfOsZ/Diej/qD7tPiMDoMRuN+AwuOENRufwz0uTJ IXndThI0ZO89KJOY0JNEKa0ugoI20E1Cye7CtY6iiYfY5K7Ia6rYFjSO53lZbMqSRMOeJkkLSRK7 TGqU3fXbR+tSEYSQRk3YB/qEaLsB1hLGAnaaFKGZjxjRsgG+bzuPQVARp1NDl2ZTdQK/fR579k/o HpfVc0QxQXQ7D3dhx8WIW/mIUwJsB9babAkKCfLNdpbui+w6vueuVsArK8d2nxl/mjKZEbKksNoi M/4UZZ0QUug2U5BgHgKXSwoYeZ8FeURRT+F5rtuUOKHXkVWpx3Gq0OX6AZDuPLjjlb3x6xvfC4y0 ME1cry/znbasqT2BUzlZ5pqiKHNdOvo0VQqNqmeY30pkpr7U77b4jqbImtjqC2EcU8UExsPumPBC TJ1Go+5vibhwRXWWeysRolZBhgn5BZ7ug+ex576+kUULDqm9MAnxRBKig4dMQdBKxFh2F5JbZbF7 PHY931rFamnR1ALhUL1VDcsuxCvCsUbYpm9rgEu3SsUDz64Prlfu3FrpyiSwpSg3dLWc3LDDG2JB NglhE+tokRDxYQGmZL1+Q8rYBGoZP1kbQBG2ArkOwq7cJnM3LR+mxJzBH/AyR1/F4Jxba/MDyZkX ukKv1RG6wYblxh9b/hNcqNV2Dybb+YO7guF5sgaLwA6DFWSYyB0+oy/vldkvEbLzAL8DHyM6UptC srytM3JWb5rrKWC9ct+egeOP3Y0PjQA6/iZUNEdR0UG5EzrSQscDHnAWKe3g4KXQgpfqPMLQ9ZT2 aAJb6NWqs3CXIaXCVRVeNPyvjvvioJ3YjH3w6tfXyUh4VcViwcdcVdehuvD3+gpK4M9GJoLCABbE xKkHAqH+69Fzt+v/1HDgKhZaDjFGY+PFs9ZrmO8iHBEXKdwRN43J2wY6UWNgzz3Le2vE6m9kQFKM WfONu9r6IOatpQpNTWpxXamlcVoHBWVV1fiQNw8oln869tLo2TjMQuUyGd772rR7H5tF0jhDVoqq AkarS14wcwwO3wd7he6D7+NUIV1wF6gv6jEr+09lY6X41fadWYHZcJHPzTMTNxoXcqO193Jjjj3w YL/W14uPU0vizDw2WKq6UyogPISJyqZAoxIqLX2vf7/AlEbOEpayMHlBCbPXprGnuVtnaaGUeiIW aQewsEmHy2O1TL/hEK/gFab9he0fZtV2FqvtEtamurMEr0HFs+OQemiKD2mmD6bm2BgpM3k6MRXd ILnhc7j5gsqMsD4pobpEcfP6vIoIr9EPqOW4UsEW2fvZWRp5xMh+E8uCEKxNlK/GaDQlOztMXaOT RWHiIBZNMHaMThmBibOYaC7UUjqhrLPnsZFfpO90Wslnj2TigLVFVb6OjSlMoswhLmzrQIOmtHWI +oQ9IBa+/xEhct+heUESM9LMZaQOo135DAX/hktBDj2YozJHsuWn+Dq/94pOXuSjyHuw2yJ3VVq+ O9xooeFm67RQ2yel3tVcxhjKvoJKtF0KdVailqdKM6mPEd0xTVdCdEe8XCwrv4Qp5xgpYdHXyKvT LSlO7XcURVObHK90FMgT+hN2OfBLbIoSAkxhC/YJHgsdKZJr6Sv6ZRK8+LU4oRPGsoeAYKo7khCF sj4BWCTfZhubWCvBoz1Re+dkuD3H0dpzSIpgmeNrB9RcrPVHOAxrsU1p6maorhWQQ5NG68h/Al55 MRS8iodyKNxavEyeVeReTBFhmmE2CfblZk8TZU3tK+0mpyQlOPbc5RaNsRSTXNGigjChQhfmrBB2 JLOwjltCVA0TdBbkny3yJjXPUzUf6jN+dUlAdGkQyTeuhGVR35pREh+jNRAaK9YUyKqMEHUBvTHJ rHh+K9zpJNRyuL+412kxwUwey2oAJVy26IXmEtt7sTf32NTYB9OSOZb1RpgVdSQbdiEz+E2JzMqc tS+TUKw5Lst4vJzPeVx/ZkpKouDE4zk59yeJFhrIyR5Cdhms5W9Ohce1iEKu5KjS8VX/QbkSGhjC muAnmFreI/CpwmcYZ8goIIk0lj1k+MF+3HpBFTJArh/N0wn4ED55SEB9dZGEqgYR48H1MqQn27u1 3YjFPOI/pddcm0qmpq5M25tNkni3QtudTr/MnYGELQTcGNuEkYbPl2ANnCVwFnZ6BXtuargETbSg FDw8uBr3P0pOPSZw4T5ObeE+B00cUMcLX3y4EoPFgRNfvju5meQ485M5XpM1QeR6Yr8ttJGTCzAp SMXNjyDqoAFi1vo01vYYFfPAzmXyBJl2QtoQmfssJQ1xV97194x3hfJTdpBvjLUdth/d2fjWarUb Lfr0+WZ0q56jV4q7RiEuhkiT34svafRF7x8FDD7RyMxmzH0RmQjbu1E1ImIXmm3LeA3eweQwahmH 2ZGwx1egrNYQaG6vbP8t/iFE7Tt6FVBt1flQk9CZJgvLccASJsVdqxuZWzibh4rksOaIZv2YbmY7 7SEKFdsjh9Xwc8N1/fihH2go49dMIYtIAgwZNnYJwlQoVkBYSRROcgqs3ASXMRYK5vJFVqnx5KNb y7kB+FRMHzUBTdx6Cv/Q5/jKMsekCPObMBlfTr2U0UIG6xHVfdHfNRDVfclx/6MNNyevn0ZgRQfk iWb9UT8CO95mD9ohYbGZaoQ2qV6glMnoIr0nkUmqhSfIQ7hwIjwq0gvN0WdEdUgAhKTuLM+2HJ/6 xqhYHycjoyTqfa1RljnpbB2ZHV9m6OEU6DCVYIVtwDbLCjETW6Afs09vsXK/y1Km+qP6QZN2aVfA nHIpo+GMo9Ak8RuYQCji4Jt4PzEnqmjIN+ZsohrmWJzeoJ/FDEcBXEUeje/N8Y04Uc3J1NDHiSXl fqBLpjy6hTC6pA/06b15pxoTfTSMysTdpplhqMNpdlkzxFv128j4mlgSw6VrWTaV+6F4q8vmcGSO dTlxNlq8HSnqwJzOhvrwGi5dt+Ol0Xiq3+q/Q7pGQ3Og3qmDXcCqVPShNhoP9MnU1PSBWj0wsYa3 T6biYBAIBm3+L3gFi61vzVfARHMQjXq9kX6JUxkbqqQPlZCwHc14EMKELKmUfsCvpNkF6o1cAxtC i2YIBgznKNT+LaZwr06wmlVpdm0i0Rq3odTDT0hiyxfLe/jyYvtPX5abt+fah1mPOhSlgQrV/N0U hwo8azjVhzOVzYz+qWZS+V01RuZAH36NYQjTwQYRGQ++Skq0q+Q/IIrEOsQIdrBh3K7UPn22nSfg 2T5Yntcu8cM/4OOJIaM5y/NGQFYj4OWPGt5yftoggtah9pD5QutAH5GV3aiiohpJv8D7iMW9I/F7 7G7/tBjV8vY3Z9KmlW3PHIhB2Ixkmhl9XAwqHmdOb18HTOj/3UT2xBrtdLEGanASshB8hOEEhnbR UMxW0+wIpqRPz2uxXGVzIA6vZ+K1Gm2EwI/Ott9n9/NvojE0RWk0m5qGOp3BL9P7sZpWYbBrNoRW q5h3oqGj1JPcMxoO7k1Rnup3qolIT65R8wAeI6+i36Is3M0rz9U7bE7a5k7ppCfQwPtI+AgpknaM q/i8Aj1Zymd7uLRK/v1fv6VVmaEj1iSxlHlvcPCSQQyQ5FShiSvUEjxY25VPHKxv7uyNDauiyMdo u4YHw8spXiMdL0iWOusyuTMnlX6AIN/59cK7yfBg/kjKMC/SnU6GUYw47PlI2p7r+qPwdVDh/xfh 19n/AFBLBwicBT6J/gsAAOhKAABQSwMEFAAIAAgAG6GROwAAAAAAAAAAAAAAACYAEABXZWJLaXRM aW5rQnVnL1dlYktpdExpbmtCdWdfUHJlZml4LnBjaFVYDADO7ypLNuQqS/UBFABtjbEOwjAMBfd8 xZM6dCM/gBjaDZBg6xiFxG4Mpa7cVOLzqZi54ZYbznvnPe5GLB8UipkMrIY4TVh1s0RgmWiFMmoh tAM9LlKvMr+6bWxRo41UIfPfupg+KdV94VwjnIkRwq079yE47DTyXtQqjr0mjf7nQzm5huYs7L5Q SwcINbrH53sAAACdAAAAUEsDBBQACAAIAJqkkTsAAAAAAAAAAAAAAAAoABAAV2ViS2l0TGlua0J1 Zy9XZWJLaXRMaW5rQnVnQXBwRGVsZWdhdGUuaFVYDADO7ypLxOoqS/UBFABTzswtyC8qUbBxzk/O T9QHk3oZdlxcDsk5icXFCr6VbkWJuanl+UXZ1kDBzLyS1KK0xORUhfDUJO/MEp/MvGyn0nTHggKX 1JzU9MSSVAUrBb9g/6Ss1GSgoX7BQJmczOTEksz8PJgKO65qLk4kcxW0FOJzka2pBVqUmpfCBQBQ SwcIz0ut9X4AAACcAAAAUEsDBBQACAAIAJqkkTsAAAAAAAAAAAAAAAAzABAAX19NQUNPU1gvV2Vi S2l0TGlua0J1Zy8uX1dlYktpdExpbmtCdWdBcHBEZWxlZ2F0ZS5oVVgMAM7vKkvE6ipL9QEUAGNg FWNnYGJg8E1MVvAPVohQgAKQGAMnEBsBcSUQg/irGYgCjiEhQQySptOgOuYAMT+aEkaEuHhyfq5e YkFBTqpeSGpFiWtecn5KZl46ULK0JE3XwtrQ2MTI0NzSwgQAUEsHCEWdgvNfAAAAqwAAAFBLAwQU AAgACADMpJE7AAAAAAAAAAAAAAAAKAAQAFdlYktpdExpbmtCdWcvV2ViS2l0TGlua0J1Z0FwcERl bGVnYXRlLm1VWAwAzu8qSyDrKkv1ARQAvZJNa8JAEIbP7q9YIoVYqkJ7sx/YIl6apgWlHmKQNZkk g/sRNqsipf+9mzRWLW0vgd5ml2dm9n3fbaPIlTbUmcHyEY2HcvWwTu/zfAQcUmaglzmEtGvqRrAo 66p+vONxL7s73D/txpoJ2Cq96h/VJUOGFuIgQBpmUEn62yZCuu5GYdyhLM85RhU9wniMEovMY2sZ ZSjTgetPfGUwqQl63mHHZ/JGWijN1eXC0A3ookRuqT95/ayfk3LzFAV4uNRM79y9dod2rknLn3gq denQWUKiNPzZOK8757Z1QM/QufjaWI4irYU4eGEfEQRH3lDGuYpCatWZ0NLBCVxkamuHv7AUwmpU EyksMaAbKHn/MZoZcj4FLVDa8P4rFbNf2CyYU7O1/YGsqIz+lphEXskfgozJB1BLBwgvM1f7MwEA ACwDAABQSwMEFAAIAAgAzKSROwAAAAAAAAAAAAAAADMAEABfX01BQ09TWC9XZWJLaXRMaW5rQnVn Ly5fV2ViS2l0TGlua0J1Z0FwcERlbGVnYXRlLm1VWAwAzu8qSyDrKkv1ARQAY2AVY2dgYmDwTUxW 8A9WiFCAApAYAycQGwFxJRCD+KsZiAKOISFBDJJmElAdc4CYH00JI0JcPDk/Vy+xoCAnVS8ktaLE NS85PyUzLx0oWVqSpmthbWhsYmRobmlhAgBQSwcIjnBpJF8AAACrAAAAUEsBAhUDCgAAAAAAM6eR OwAAAAAAAAAAAAAAAA4ADAAAAAAAAAAAQO1BAAAAAFdlYktpdExpbmtCdWcvVVgIALzvKkuh7ypL UEsBAhUDCgAAAAAAI6KROwAAAAAAAAAAAAAAABwADAAAAAAAAAAAQO1BPAAAAFdlYktpdExpbmtC dWcvRW5nbGlzaC5scHJvai9VWAgAvO8qSyHmKktQSwECFQMUAAgACAAboZE7pEm/Ii8AAAAtAAAA LQAMAAAAAAAAAABApIGGAAAAV2ViS2l0TGlua0J1Zy9FbmdsaXNoLmxwcm9qL0luZm9QbGlzdC5z dHJpbmdzVVgIAM7vKks25CpLUEsBAhUDFAAIAAgAI6KRO4x9WruULQAAqLECACgADAAAAAAAAAAA QKSBIAEAAFdlYktpdExpbmtCdWcvRW5nbGlzaC5scHJvai9NYWluTWVudS54aWJVWAgAzu8qSyHm KktQSwECFQMUAAgACACnoZE7ycwe62YAAAB7AAAAFAAMAAAAAAAAAABApIEaLwAAV2ViS2l0TGlu a0J1Zy9tYWluLm1VWAgAzu8qSzrlKktQSwECFQMKAAAAAABLp5E7AAAAAAAAAAAAAAAACQAMAAAA AAAAAABA/UHSLwAAX19NQUNPU1gvVVgIAM7vKkvO7ypLUEsBAhUDCgAAAAAAS6eROwAAAAAAAAAA AAAAABcADAAAAAAAAAAAQP1BCTAAAF9fTUFDT1NYL1dlYktpdExpbmtCdWcvVVgIAM7vKkvO7ypL UEsBAhUDFAAIAAgAp6GRO5RWSG5fAAAAqwAAAB8ADAAAAAAAAAAAQKSBTjAAAF9fTUFDT1NYL1dl YktpdExpbmtCdWcvLl9tYWluLm1VWAgAzu8qSzrlKktQSwECFQMUAAgACABEopE7BoTOdVMBAAC8 AgAAJAAMAAAAAAAAAABApIEKMQAAV2ViS2l0TGlua0J1Zy9NeUZyYW1ld29yay1JbmZvLnBsaXN0 VVgIAM7vKktg5ipLUEsBAhUDFAAIAAgAGKWROzAryed8AAAAogAAABsADAAAAAAAAAAAQKSBvzIA AFdlYktpdExpbmtCdWcvTXlGcmFtZXdvcmsuaFVYCADO7ypLr+sqS1BLAQIVAxQACAAIABilkTuN fbQyXwAAAKsAAAAmAAwAAAAAAAAAAECkgZQzAABfX01BQ09TWC9XZWJLaXRMaW5rQnVnLy5fTXlG cmFtZXdvcmsuaFVYCADO7ypLr+sqS1BLAQIVAxQACAAIAKWlkTvKlIQTowAAADsBAAAbAAwAAAAA AAAAAECkgVc0AABXZWJLaXRMaW5rQnVnL015RnJhbWV3b3JrLm1VWAgAzu8qS7XsKktQSwECFQMU AAgACAClpZE7hQo/H18AAACrAAAAJgAMAAAAAAAAAABApIFTNQAAX19NQUNPU1gvV2ViS2l0TGlu a0J1Zy8uX015RnJhbWV3b3JrLm1VWAgAzu8qS7XsKktQSwECFQMUAAgACADWppE7G/swixAMAAAZ hgAAHQAMAAAAAAAAAABAtIEWNgAAV2ViS2l0TGlua0J1Zy9NeVdlYldpbmRvdy54aWJVWAgAzu8q S/PuKktQSwECFQMUAAgACACwpZE7o+b5ZGgAAACFAAAAIgAMAAAAAAAAAABApIGBQgAAV2ViS2l0 TGlua0J1Zy9NeVdpbmRvd0NvbnRyb2xsZXIuaFVYCADO7ypLzOwqS1BLAQIVAxQACAAIALClkTsg EMb8XwAAAKsAAAAtAAwAAAAAAAAAAECkgUlDAABfX01BQ09TWC9XZWJLaXRMaW5rQnVnLy5fTXlX aW5kb3dDb250cm9sbGVyLmhVWAgAzu8qS8zsKktQSwECFQMUAAgACADBppE7JHedYkUBAAD9AQAA IgAMAAAAAAAAAABApIETRAAAV2ViS2l0TGlua0J1Zy9NeVdpbmRvd0NvbnRyb2xsZXIubVVYCADO 7ypLyu4qS1BLAQIVAxQACAAIAMGmkTuUa1M4XwAAAKsAAAAtAAwAAAAAAAAAAECkgbhFAABfX01B Q09TWC9XZWJLaXRMaW5rQnVnLy5fTXlXaW5kb3dDb250cm9sbGVyLm1VWAgAzu8qS8ruKktQSwEC FQMUAAgACAAboZE7oop+jLABAADqAwAAJgAMAAAAAAAAAABApIGCRgAAV2ViS2l0TGlua0J1Zy9X ZWJLaXRMaW5rQnVnLUluZm8ucGxpc3RVWAgAzu8qSzbkKktQSwECFQMKAAAAAAA6p5E7AAAAAAAA AAAAAAAAJgAMAAAAAAAAAABA7UGWSAAAV2ViS2l0TGlua0J1Zy9XZWJLaXRMaW5rQnVnLnhjb2Rl cHJvai9VWAgAvO8qS7DvKktQSwECFQMUAAgACAAXp5E7nAU+if4LAADoSgAANQAMAAAAAAAAAABA pIHqSAAAV2ViS2l0TGlua0J1Zy9XZWJLaXRMaW5rQnVnLnhjb2RlcHJvai9wcm9qZWN0LnBieHBy b2pVWAgAzu8qS23vKktQSwECFQMUAAgACAAboZE7NbrH53sAAACdAAAAJgAMAAAAAAAAAABApIFb VQAAV2ViS2l0TGlua0J1Zy9XZWJLaXRMaW5rQnVnX1ByZWZpeC5wY2hVWAgAzu8qSzbkKktQSwEC FQMUAAgACACapJE7z0ut9X4AAACcAAAAKAAMAAAAAAAAAABApIE6VgAAV2ViS2l0TGlua0J1Zy9X ZWJLaXRMaW5rQnVnQXBwRGVsZWdhdGUuaFVYCADO7ypLxOoqS1BLAQIVAxQACAAIAJqkkTtFnYLz XwAAAKsAAAAzAAwAAAAAAAAAAECkgR5XAABfX01BQ09TWC9XZWJLaXRMaW5rQnVnLy5fV2ViS2l0 TGlua0J1Z0FwcERlbGVnYXRlLmhVWAgAzu8qS8TqKktQSwECFQMUAAgACADMpJE7LzNX+zMBAAAs AwAAKAAMAAAAAAAAAABApIHuVwAAV2ViS2l0TGlua0J1Zy9XZWJLaXRMaW5rQnVnQXBwRGVsZWdh dGUubVVYCADO7ypLIOsqS1BLAQIVAxQACAAIAMykkTuOcGkkXwAAAKsAAAAzAAwAAAAAAAAAAECk gYdZAABfX01BQ09TWC9XZWJLaXRMaW5rQnVnLy5fV2ViS2l0TGlua0J1Z0FwcERlbGVnYXRlLm1V WAgAzu8qSyDrKktQSwUGAAAAABoAGgB0CQAAV1oAAAAA