)]}'
{"/PATCHSET_LEVEL":[{"author":{"_account_id":1000008,"name":"Lev Stipakov","display_name":"Lev Stipakov","email":"lstipakov@gmail.com","username":"stipa"},"change_message_id":"c41d64f6d96286583cdc2a851c94c4725913cef6","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":14,"id":"f7c242a5_3818f81a","updated":"2025-03-21 14:09:47.000000000","message":"Is split-dns already supposed to work? In my server config I have\n\n\npush \"dns server 1 resolve-domains fi\"\npush \"dns server 1 address 1.1.1.1\"\n\npush \"dns server 2 resolve-domains io\"\npush \"dns server 2 address 8.8.8.8\"\n\nso I would expect .fi go to 1.1.1.1 and .io to 8.8.8.8, but instead only one registry entry .fi is created with 1.1.1.1.\n\nAlso, with that\n\npush \"dns server 1 resolve-domains fi\"\npush \"dns server 1 address 1.1.1.1\"\n\nin addition to NRPT we set 1.1.1.1 on VPN adapter and many DNS requests not directed to .fi domains also resolved via 1.1.1.1. This contradicts to what documentation says: \"where only the given domains are resolved by the server\".","commit_id":"bee3a0251a5867e90df3f1a57853fd1082dfb23e"},{"author":{"_account_id":1000008,"name":"Lev Stipakov","display_name":"Lev Stipakov","email":"lstipakov@gmail.com","username":"stipa"},"change_message_id":"a726a1ffd82210af0bcaa30f813249f0cc6ccd19","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":18,"id":"af3fa437_f3237009","updated":"2025-04-09 13:54:44.000000000","message":"I couldn\u0027t find any more issues in the code, I tested it and it works like expected - I checked that the NRPT rules are created/removed and adapter-specific DNS suffixes are also respected in NRPT.\n\nProbably not directly related to this patch, but the behavior is now not according to man page. For example if server pushes\n\n  push \"dns server 1 resolve-domains fi\"\n  push \"dns server 1 address 1.1.1.1\"\n  \nI would expect only .fi domains being resolved via 1.1.1.1, however ALL domains are resolved via 1.1.1.1 - this is likely because we set 1.1.1.1 DNS on VPN adapter. why we do it in addition to NRPT? We should either change the man page or fix the behavior.","commit_id":"bfeafdc985a358fbebaa7fd16ed83f542f251897"}],"src/openvpnserv/interactive.c":[{"author":{"_account_id":1000001,"name":"Frank Lichtenheld","display_name":"Frank Lichtenheld","email":"frank@lichtenheld.com","username":"flichtenheld","status":"OpenVPN Inc."},"change_message_id":"dc78b0ca0cc47ae28ab62482409bc7efbc3c476b","unresolved":true,"context_lines":[{"line_number":2262,"context_line":"    if (IIDFromString(iid_str, \u0026iid) !\u003d S_OK"},{"line_number":2263,"context_line":"        || ConvertInterfaceGuidToLuid(\u0026iid, \u0026itf_row.InterfaceLuid) !\u003d NO_ERROR)"},{"line_number":2264,"context_line":"    {"},{"line_number":2265,"context_line":"        MsgToEventLog(M_SYSERR, TEXT(\"IsInterfaceConnected: \""},{"line_number":2266,"context_line":"                                     \"could not convert interface %s GUID to LUID\"), iid_str);"},{"line_number":2267,"context_line":"        goto out;"},{"line_number":2268,"context_line":"    }"}],"source_content_type":"text/x-csrc","patch_set":5,"id":"74358983_c1d55f71","line":2265,"updated":"2024-12-28 23:06:58.000000000","message":"This actually does not work correctly. You can\u0027t use multiple string literals as argument to TEXT()","commit_id":"92956c4ef0ac7db46d6071dadbf281941d8d70a0"},{"author":{"_account_id":1000001,"name":"Frank Lichtenheld","display_name":"Frank Lichtenheld","email":"frank@lichtenheld.com","username":"flichtenheld","status":"OpenVPN Inc."},"change_message_id":"9c8cfb6d42ff083d85887acf05e367634aac4ac9","unresolved":false,"context_lines":[{"line_number":2262,"context_line":"    if (IIDFromString(iid_str, \u0026iid) !\u003d S_OK"},{"line_number":2263,"context_line":"        || ConvertInterfaceGuidToLuid(\u0026iid, \u0026itf_row.InterfaceLuid) !\u003d NO_ERROR)"},{"line_number":2264,"context_line":"    {"},{"line_number":2265,"context_line":"        MsgToEventLog(M_SYSERR, TEXT(\"IsInterfaceConnected: \""},{"line_number":2266,"context_line":"                                     \"could not convert interface %s GUID to LUID\"), iid_str);"},{"line_number":2267,"context_line":"        goto out;"},{"line_number":2268,"context_line":"    }"}],"source_content_type":"text/x-csrc","patch_set":5,"id":"29071d80_c92d5c0a","line":2265,"in_reply_to":"2035b4d7_a8134d49","updated":"2025-01-10 13:11:20.000000000","message":"Yeah, it seems to work even though I\u0027m not exactly sure what the logic is that makes it work. Maybe when you have multiple string literals the L just gets applied from left to right? Either way I decided to instead get TEXT() removed completely from our code base, see my separate patch series. I will leave it to you whether you want to use it here and have it converted in my series or use L\"\" here directly.","commit_id":"92956c4ef0ac7db46d6071dadbf281941d8d70a0"},{"author":{"_account_id":1000006,"name":"Heiko Hund","display_name":"Heiko Hund","email":"heiko@ist.eigentlich.net","username":"d12fk"},"change_message_id":"5ad98be2f21c105c42299bbd33260a32bc4917e0","unresolved":true,"context_lines":[{"line_number":2262,"context_line":"    if (IIDFromString(iid_str, \u0026iid) !\u003d S_OK"},{"line_number":2263,"context_line":"        || ConvertInterfaceGuidToLuid(\u0026iid, \u0026itf_row.InterfaceLuid) !\u003d NO_ERROR)"},{"line_number":2264,"context_line":"    {"},{"line_number":2265,"context_line":"        MsgToEventLog(M_SYSERR, TEXT(\"IsInterfaceConnected: \""},{"line_number":2266,"context_line":"                                     \"could not convert interface %s GUID to LUID\"), iid_str);"},{"line_number":2267,"context_line":"        goto out;"},{"line_number":2268,"context_line":"    }"}],"source_content_type":"text/x-csrc","patch_set":5,"id":"2035b4d7_a8134d49","line":2265,"in_reply_to":"74358983_c1d55f71","updated":"2024-12-31 00:02:02.000000000","message":"How did you get to this conclusion? Works for me, and Compiler Explorer also shows that the string is stored of wide characters: https://godbolt.org/z/Md6cPGsEb","commit_id":"92956c4ef0ac7db46d6071dadbf281941d8d70a0"},{"author":{"_account_id":1000008,"name":"Lev Stipakov","display_name":"Lev Stipakov","email":"lstipakov@gmail.com","username":"stipa"},"change_message_id":"03d884a6f3ec6d292a7cc008b15bf6b6aaa21097","unresolved":true,"context_lines":[{"line_number":2599,"context_line":"                                  \"failed to set rule %s (%lu)\"), subkey, err);"},{"line_number":2600,"context_line":"    }"},{"line_number":2601,"context_line":""},{"line_number":2602,"context_line":"    free(wide_domains);"},{"line_number":2603,"context_line":"    return err;"},{"line_number":2604,"context_line":"}"},{"line_number":2605,"context_line":""}],"source_content_type":"text/x-csrc","patch_set":13,"id":"527430f0_4471f7d6","line":2602,"updated":"2025-03-19 13:58:01.000000000","message":"If domains is not set, we\u0027re freeing stack-allocated wide_domains (PWSTR wide_domains \u003d L\".\\0\";).","commit_id":"1e21d81fd5671c27081ba8c16a92a6a6e7ad38a2"},{"author":{"_account_id":1000006,"name":"Heiko Hund","display_name":"Heiko Hund","email":"heiko@ist.eigentlich.net","username":"d12fk"},"change_message_id":"3dcf58cc8d2674e6899cb1ca77171b96593fe9e6","unresolved":false,"context_lines":[{"line_number":2599,"context_line":"                                  \"failed to set rule %s (%lu)\"), subkey, err);"},{"line_number":2600,"context_line":"    }"},{"line_number":2601,"context_line":""},{"line_number":2602,"context_line":"    free(wide_domains);"},{"line_number":2603,"context_line":"    return err;"},{"line_number":2604,"context_line":"}"},{"line_number":2605,"context_line":""}],"source_content_type":"text/x-csrc","patch_set":13,"id":"b96c8df4_7beb686c","line":2602,"in_reply_to":"527430f0_4471f7d6","updated":"2025-03-19 22:45:41.000000000","message":"Done","commit_id":"1e21d81fd5671c27081ba8c16a92a6a6e7ad38a2"},{"author":{"_account_id":1000008,"name":"Lev Stipakov","display_name":"Lev Stipakov","email":"lstipakov@gmail.com","username":"stipa"},"change_message_id":"03d884a6f3ec6d292a7cc008b15bf6b6aaa21097","unresolved":true,"context_lines":[{"line_number":2779,"context_line":"    BOOL gpol_list \u003d FALSE;"},{"line_number":2780,"context_line":""},{"line_number":2781,"context_line":"    WCHAR iid[64];"},{"line_number":2782,"context_line":"    DWORD iid_err \u003d InterfaceIdString(msg-\u003eiface.name, iid, sizeof(iid));"},{"line_number":2783,"context_line":"    if (iid_err)"},{"line_number":2784,"context_line":"    {"},{"line_number":2785,"context_line":"        return iid_err;"}],"source_content_type":"text/x-csrc","patch_set":13,"id":"828a5a05_4a81dca3","line":2782,"updated":"2025-03-19 13:58:01.000000000","message":"We pass sizeof(iid) as count to wcsncpy and it should be count of wide chars, not the buffer size. This causes crash.","commit_id":"1e21d81fd5671c27081ba8c16a92a6a6e7ad38a2"},{"author":{"_account_id":1000006,"name":"Heiko Hund","display_name":"Heiko Hund","email":"heiko@ist.eigentlich.net","username":"d12fk"},"change_message_id":"3dcf58cc8d2674e6899cb1ca77171b96593fe9e6","unresolved":false,"context_lines":[{"line_number":2779,"context_line":"    BOOL gpol_list \u003d FALSE;"},{"line_number":2780,"context_line":""},{"line_number":2781,"context_line":"    WCHAR iid[64];"},{"line_number":2782,"context_line":"    DWORD iid_err \u003d InterfaceIdString(msg-\u003eiface.name, iid, sizeof(iid));"},{"line_number":2783,"context_line":"    if (iid_err)"},{"line_number":2784,"context_line":"    {"},{"line_number":2785,"context_line":"        return iid_err;"}],"source_content_type":"text/x-csrc","patch_set":13,"id":"8eed4464_e3ea759f","line":2782,"in_reply_to":"828a5a05_4a81dca3","updated":"2025-03-19 22:45:41.000000000","message":"Done","commit_id":"1e21d81fd5671c27081ba8c16a92a6a6e7ad38a2"},{"author":{"_account_id":1000008,"name":"Lev Stipakov","display_name":"Lev Stipakov","email":"lstipakov@gmail.com","username":"stipa"},"change_message_id":"c41d64f6d96286583cdc2a851c94c4725913cef6","unresolved":true,"context_lines":[{"line_number":2256,"context_line":"    if (IIDFromString(iid_str, \u0026iid) !\u003d S_OK"},{"line_number":2257,"context_line":"        || ConvertInterfaceGuidToLuid(\u0026iid, \u0026itf_row.InterfaceLuid) !\u003d NO_ERROR)"},{"line_number":2258,"context_line":"    {"},{"line_number":2259,"context_line":"        MsgToEventLog(M_SYSERR, TEXT(\"IsInterfaceConnected: \""},{"line_number":2260,"context_line":"                                     \"could not convert interface %s GUID to LUID\"), iid_str);"},{"line_number":2261,"context_line":"        goto out;"},{"line_number":2262,"context_line":"    }"}],"source_content_type":"text/x-csrc","patch_set":14,"id":"8e6ebec0_a562e6ce","line":2259,"updated":"2025-03-21 14:09:47.000000000","message":"At least on my machine I have 6 \"real\" network adapters (according to Get-NetAdapter | Format-List Name, InterfaceGuid) and 12 registry keys, apparently those are disabled/hidden etc. Because of that I got lots of scary \n\"IsInterfaceConnected: could not convert interface {F2D72071-CB76-4D6F-9DC8-EDAA3AC6B387} GUID to LUID\" messages. Let\u0027s don\u0027t display this error message - at least if ConvertInterfaceGuidToLuid returns ERROR_INVALID_PARAMETER.","commit_id":"bee3a0251a5867e90df3f1a57853fd1082dfb23e"},{"author":{"_account_id":1000006,"name":"Heiko Hund","display_name":"Heiko Hund","email":"heiko@ist.eigentlich.net","username":"d12fk"},"change_message_id":"25f84623bb9be5365243ea81bb8f36c13e5d06b3","unresolved":true,"context_lines":[{"line_number":2256,"context_line":"    if (IIDFromString(iid_str, \u0026iid) !\u003d S_OK"},{"line_number":2257,"context_line":"        || ConvertInterfaceGuidToLuid(\u0026iid, \u0026itf_row.InterfaceLuid) !\u003d NO_ERROR)"},{"line_number":2258,"context_line":"    {"},{"line_number":2259,"context_line":"        MsgToEventLog(M_SYSERR, TEXT(\"IsInterfaceConnected: \""},{"line_number":2260,"context_line":"                                     \"could not convert interface %s GUID to LUID\"), iid_str);"},{"line_number":2261,"context_line":"        goto out;"},{"line_number":2262,"context_line":"    }"}],"source_content_type":"text/x-csrc","patch_set":14,"id":"b99905f4_e88c89db","line":2259,"in_reply_to":"8e6ebec0_a562e6ce","updated":"2025-03-26 03:54:10.000000000","message":"Hm, I do not have adapters I can reproduce this on my system. Shall we do a joined session figuring this out, or do you have an idea already, why these adapters are listed in the interfaces registry hive, but fail to produce a LUID?","commit_id":"bee3a0251a5867e90df3f1a57853fd1082dfb23e"},{"author":{"_account_id":1000006,"name":"Heiko Hund","display_name":"Heiko Hund","email":"heiko@ist.eigentlich.net","username":"d12fk"},"change_message_id":"3ba80217c64952f64d332024121c9ed61f8d0d01","unresolved":false,"context_lines":[{"line_number":2256,"context_line":"    if (IIDFromString(iid_str, \u0026iid) !\u003d S_OK"},{"line_number":2257,"context_line":"        || ConvertInterfaceGuidToLuid(\u0026iid, \u0026itf_row.InterfaceLuid) !\u003d NO_ERROR)"},{"line_number":2258,"context_line":"    {"},{"line_number":2259,"context_line":"        MsgToEventLog(M_SYSERR, TEXT(\"IsInterfaceConnected: \""},{"line_number":2260,"context_line":"                                     \"could not convert interface %s GUID to LUID\"), iid_str);"},{"line_number":2261,"context_line":"        goto out;"},{"line_number":2262,"context_line":"    }"}],"source_content_type":"text/x-csrc","patch_set":14,"id":"d4afefe5_56002315","line":2259,"in_reply_to":"b879805b_55fa9a7f","updated":"2025-03-27 01:28:37.000000000","message":"The code fails now without error message if ConvertInterfaceGuidToLuid() fails. However, I wonder why it fails in the first place with those interfaces. Basically that means that for any interface, where this conversion fails, there will be no exclude rules ever, and not even a diagnostic message to explain this. Leaves an uneasy feeling behind for me.","commit_id":"bee3a0251a5867e90df3f1a57853fd1082dfb23e"},{"author":{"_account_id":1000008,"name":"Lev Stipakov","display_name":"Lev Stipakov","email":"lstipakov@gmail.com","username":"stipa"},"change_message_id":"0ceb772deedf0110682c8e9221997abdab71eabf","unresolved":true,"context_lines":[{"line_number":2256,"context_line":"    if (IIDFromString(iid_str, \u0026iid) !\u003d S_OK"},{"line_number":2257,"context_line":"        || ConvertInterfaceGuidToLuid(\u0026iid, \u0026itf_row.InterfaceLuid) !\u003d NO_ERROR)"},{"line_number":2258,"context_line":"    {"},{"line_number":2259,"context_line":"        MsgToEventLog(M_SYSERR, TEXT(\"IsInterfaceConnected: \""},{"line_number":2260,"context_line":"                                     \"could not convert interface %s GUID to LUID\"), iid_str);"},{"line_number":2261,"context_line":"        goto out;"},{"line_number":2262,"context_line":"    }"}],"source_content_type":"text/x-csrc","patch_set":14,"id":"b879805b_55fa9a7f","line":2259,"in_reply_to":"b99905f4_e88c89db","updated":"2025-03-26 08:55:03.000000000","message":"Sure. In my case it is probably Hyper-V virtual adapters.\n\nhttps://gist.github.com/lstipakov/b30a662ab8001d4a07ddda7406808eb5\n\nAgain, I think it is fine just to omit error message in this error case.","commit_id":"bee3a0251a5867e90df3f1a57853fd1082dfb23e"},{"author":{"_account_id":1000008,"name":"Lev Stipakov","display_name":"Lev Stipakov","email":"lstipakov@gmail.com","username":"stipa"},"change_message_id":"c41d64f6d96286583cdc2a851c94c4725913cef6","unresolved":true,"context_lines":[{"line_number":2507,"context_line":"    GetNrptExcludeData(data, _countof(data));"},{"line_number":2508,"context_line":""},{"line_number":2509,"context_line":"    unsigned n \u003d 0;"},{"line_number":2510,"context_line":"    for (int i \u003d 0; i \u003c _countof(data); ++i)"},{"line_number":2511,"context_line":"    {"},{"line_number":2512,"context_line":"        DWORD err;"},{"line_number":2513,"context_line":"        WCHAR subkey[48];"}],"source_content_type":"text/x-csrc","patch_set":14,"id":"427ad111_8562738a","line":2510,"updated":"2025-03-21 14:09:47.000000000","message":"Does it really make sense to create keys for all adapters, even if there are no domains/servers defined? \n\nCould we do it only if there are adapter-specific domains defined?","commit_id":"bee3a0251a5867e90df3f1a57853fd1082dfb23e"},{"author":{"_account_id":1000006,"name":"Heiko Hund","display_name":"Heiko Hund","email":"heiko@ist.eigentlich.net","username":"d12fk"},"change_message_id":"25f84623bb9be5365243ea81bb8f36c13e5d06b3","unresolved":true,"context_lines":[{"line_number":2507,"context_line":"    GetNrptExcludeData(data, _countof(data));"},{"line_number":2508,"context_line":""},{"line_number":2509,"context_line":"    unsigned n \u003d 0;"},{"line_number":2510,"context_line":"    for (int i \u003d 0; i \u003c _countof(data); ++i)"},{"line_number":2511,"context_line":"    {"},{"line_number":2512,"context_line":"        DWORD err;"},{"line_number":2513,"context_line":"        WCHAR subkey[48];"}],"source_content_type":"text/x-csrc","patch_set":14,"id":"44df0b7c_cc86afff","line":2510,"in_reply_to":"427ad111_8562738a","updated":"2025-03-26 03:54:10.000000000","message":"Yeah, if no domains or nameservers are defined the adapters are not added as exclude data, so there will be no RoutingX NRPT rules for those. I think that was what you were asking, right?","commit_id":"bee3a0251a5867e90df3f1a57853fd1082dfb23e"},{"author":{"_account_id":1000008,"name":"Lev Stipakov","display_name":"Lev Stipakov","email":"lstipakov@gmail.com","username":"stipa"},"change_message_id":"0ceb772deedf0110682c8e9221997abdab71eabf","unresolved":true,"context_lines":[{"line_number":2507,"context_line":"    GetNrptExcludeData(data, _countof(data));"},{"line_number":2508,"context_line":""},{"line_number":2509,"context_line":"    unsigned n \u003d 0;"},{"line_number":2510,"context_line":"    for (int i \u003d 0; i \u003c _countof(data); ++i)"},{"line_number":2511,"context_line":"    {"},{"line_number":2512,"context_line":"        DWORD err;"},{"line_number":2513,"context_line":"        WCHAR subkey[48];"}],"source_content_type":"text/x-csrc","patch_set":14,"id":"a11626e1_0dcbd6ed","line":2510,"in_reply_to":"44df0b7c_cc86afff","updated":"2025-03-26 08:55:03.000000000","message":"But there are: https://usercontent.irccloud-cdn.com/file/eJy2zlR6/image.png","commit_id":"bee3a0251a5867e90df3f1a57853fd1082dfb23e"},{"author":{"_account_id":1000006,"name":"Heiko Hund","display_name":"Heiko Hund","email":"heiko@ist.eigentlich.net","username":"d12fk"},"change_message_id":"3ba80217c64952f64d332024121c9ed61f8d0d01","unresolved":false,"context_lines":[{"line_number":2507,"context_line":"    GetNrptExcludeData(data, _countof(data));"},{"line_number":2508,"context_line":""},{"line_number":2509,"context_line":"    unsigned n \u003d 0;"},{"line_number":2510,"context_line":"    for (int i \u003d 0; i \u003c _countof(data); ++i)"},{"line_number":2511,"context_line":"    {"},{"line_number":2512,"context_line":"        DWORD err;"},{"line_number":2513,"context_line":"        WCHAR subkey[48];"}],"source_content_type":"text/x-csrc","patch_set":14,"id":"085ad92a_b096baa2","line":2510,"in_reply_to":"a11626e1_0dcbd6ed","updated":"2025-03-27 01:28:37.000000000","message":"Done","commit_id":"bee3a0251a5867e90df3f1a57853fd1082dfb23e"},{"author":{"_account_id":1000008,"name":"Lev Stipakov","display_name":"Lev Stipakov","email":"lstipakov@gmail.com","username":"stipa"},"change_message_id":"e280cf2c2285658142d421ae01a2ad2ed7235b9a","unresolved":true,"context_lines":[{"line_number":2184,"context_line":""},{"line_number":2185,"context_line":"    for (int i \u003d 0; values[i]; i++)"},{"line_number":2186,"context_line":"    {"},{"line_number":2187,"context_line":"        err \u003d RegGetValueW(itf, NULL, values[i], RRF_RT_REG_SZ, NULL, (PBYTE)domains, size);"},{"line_number":2188,"context_line":"        if (!err \u0026\u0026 *size \u003e one_glyph \u0026\u0026 wcschr(domains, \u0027.\u0027))"},{"line_number":2189,"context_line":"        {"},{"line_number":2190,"context_line":"            /*"}],"source_content_type":"text/x-csrc","patch_set":15,"id":"280dd3ed_8e292101","line":2187,"updated":"2025-03-27 10:05:22.000000000","message":"For existing but empty values, the size will be set to 2. As a result, on the next iteration we got ERROR_MORE_DATA if there is a non-empty value.\n\nHow about:\n\n```\nDWORD cbData \u003d *size;\nerr \u003d RegGetValueW(itf, NULL, values[i], RRF_RT_REG_SZ, NULL, (PBYTE)domains, \u0026cbData);\nif (!err \u0026\u0026 cbData \u003e one_glyph \u0026\u0026 wcschr(domains, \u0027.\u0027))\n{\n    *size \u003d cbData;\n```","commit_id":"6f2b6284841c2d4fa0b58451d5470aa4ee17013e"},{"author":{"_account_id":1000006,"name":"Heiko Hund","display_name":"Heiko Hund","email":"heiko@ist.eigentlich.net","username":"d12fk"},"change_message_id":"937e916354568ee0bbb8d12c81347018e83e86af","unresolved":false,"context_lines":[{"line_number":2184,"context_line":""},{"line_number":2185,"context_line":"    for (int i \u003d 0; values[i]; i++)"},{"line_number":2186,"context_line":"    {"},{"line_number":2187,"context_line":"        err \u003d RegGetValueW(itf, NULL, values[i], RRF_RT_REG_SZ, NULL, (PBYTE)domains, size);"},{"line_number":2188,"context_line":"        if (!err \u0026\u0026 *size \u003e one_glyph \u0026\u0026 wcschr(domains, \u0027.\u0027))"},{"line_number":2189,"context_line":"        {"},{"line_number":2190,"context_line":"            /*"}],"source_content_type":"text/x-csrc","patch_set":15,"id":"a8476c39_19850a6f","line":2187,"in_reply_to":"280dd3ed_8e292101","updated":"2025-04-05 04:11:35.000000000","message":"Very nice find.","commit_id":"6f2b6284841c2d4fa0b58451d5470aa4ee17013e"}]}
