{"thread":{"id":"51010","subject":"[PATCH 0/1] Support git:// with old MinGW","startedAt":"2019-04-29T22:04:39Z","lastAt":"2019-05-03T08:42:57Z","messageCount":10,"participants":["Johannes Schindelin via GitGitGadget","Thomas Braun via GitGitGadget","Eric Sunshine","Johannes Schindelin","brian m. carlson","Johannes Sixt"],"isPatch":true,"patchVersion":1,"patchTotal":1},"messages":[{"id":"374675","messageId":"pull.137.git.gitgitgadget@gmail.com","threadId":"51010","inReplyTo":null,"subject":"[PATCH 0/1] Support git:// with old MinGW","fromName":"Johannes Schindelin via GitGitGadget","fromEmail":"gitgitgadget@gmail.com","sentAt":"2019-04-29T22:04:35Z","receivedAt":"2019-04-29T22:04:39Z","isPatch":true,"sender":{"key":"johannes.schindelin@gmx.de","avatar":"https://avatars.githubusercontent.com/u/127790?v=4"},"body":"The MinGW project (which can now be considered \"old\", as mingw-w64 is the\nstandard thing to use nowadays, since it also supports 64-bit builds) had\nsome problems with our sideband feature, and in Git for Windows <2.x, we\nintroduced a config setting to still allow using the git:// protocol by\nforcing off the sideband channel.\n\nLet's contribute this patch to upstream Git, at long last.\n\nThomas Braun (1):\n  mingw: optionally disable side-band-64k for transport\n\n Documentation/config.txt          |  2 ++\n Documentation/config/sendpack.txt |  5 +++++\n send-pack.c                       | 14 +++++++++++++-\n 3 files changed, 20 insertions(+), 1 deletion(-)\n create mode 100644 Documentation/config/sendpack.txt\n\n\nbase-commit: 8104ec994ea3849a968b4667d072fedd1e688642\nPublished-As: https://github.com/gitgitgadget/git/releases/tag/pr-137%2Fdscho%2Fsideband-bug-v1\nFetch-It-Via: git fetch https://github.com/gitgitgadget/git pr-137/dscho/sideband-bug-v1\nPull-Request: https://github.com/gitgitgadget/git/pull/137\n-- \ngitgitgadget\n"},{"id":"374676","messageId":"31aa7bfcab834b753cc9f52fc9cc187f65e2d964.1556575475.git.gitgitgadget@gmail.com","threadId":"51010","inReplyTo":"pull.137.git.gitgitgadget@gmail.com","subject":"[PATCH 1/1] mingw: optionally disable side-band-64k for transport","fromName":"Thomas Braun via GitGitGadget","fromEmail":"gitgitgadget@gmail.com","sentAt":"2019-04-29T22:04:36Z","receivedAt":"2019-04-29T22:04:41Z","isPatch":true,"sender":{"key":"thomas.braun@virtuell-zuhause.de","avatar":"https://avatars.githubusercontent.com/u/1185677?v=4"},"body":"From: Thomas Braun <thomas.braun@byte-physics.de>\n\nSince commit 0c499ea60f (send-pack: demultiplex a sideband stream with\nstatus data, 2010-02-05) the built-in send-pack uses the side-band-64k\ncapability if advertised by the server.\n\nUnfortunately this breaks pushing over the dump git protocol if used\nover a network connection when using MinGW (but *not* when using\nmingw-w64).\n\nThe detailed reasons for this, are courtesy of Jeff Preshing, quoted\nfrom https://groups.google.com/d/msg/msysgit/at8D7J-h7mw/eaLujILGUWoJ:\n\n\tMinGW wraps Windows sockets in CRT file descriptors in order to\n\tmimic the functionality of POSIX sockets. This causes msvcrt.dll\n\tto treat sockets as Installable File System (IFS) handles,\n\tcalling ReadFile, WriteFile, DuplicateHandle and CloseHandle on\n\tthem. This approach works well in simple cases on recent\n\tversions of Windows, but does not support all usage patterns.\n\tIn particular, using this approach, any attempt to read & write\n\tconcurrently on the same socket (from one or more processes)\n\twill deadlock in a scenario where the read waits for a response\n\tfrom the server which is only invoked after the write. This is\n\twhat send_pack currently attempts to do in the use_sideband\n\tcodepath.\n\nThe new config option \"sendpack.sideband\" allows to override the\nside-band-64k capability of the server, and thus makes the dump git\nprotocol work.\n\nAs this only affects builds against MinGW, the default is still to use\nside band.\n\n[jes: split out the documentation into Documentation/config/, touched up\nthe commit message.]\n\nSigned-off-by: Thomas Braun <thomas.braun@byte-physics.de>\nSigned-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>\n---\n Documentation/config.txt          |  2 ++\n Documentation/config/sendpack.txt |  5 +++++\n send-pack.c                       | 14 +++++++++++++-\n 3 files changed, 20 insertions(+), 1 deletion(-)\n create mode 100644 Documentation/config/sendpack.txt\n\ndiff --git a/Documentation/config.txt b/Documentation/config.txt\nindex d87846faa6..e9b2d10e99 100644\n--- a/Documentation/config.txt\n+++ b/Documentation/config.txt\n@@ -406,6 +406,8 @@ include::config/reset.txt[]\n \n include::config/sendemail.txt[]\n \n+include::config/sendpack.txt[]\n+\n include::config/sequencer.txt[]\n \n include::config/showbranch.txt[]\ndiff --git a/Documentation/config/sendpack.txt b/Documentation/config/sendpack.txt\nnew file mode 100644\nindex 0000000000..e306f657fb\n--- /dev/null\n+++ b/Documentation/config/sendpack.txt\n@@ -0,0 +1,5 @@\n+sendpack.sideband::\n+\tAllows to disable the side-band-64k capability for send-pack even\n+\twhen it is advertised by the server. Makes it possible to work\n+\taround a limitation in the git for windows implementation together\n+\twith the dump git protocol. Defaults to true.\ndiff --git a/send-pack.c b/send-pack.c\nindex 6dc16c3211..fa9e8cf1fc 100644\n--- a/send-pack.c\n+++ b/send-pack.c\n@@ -38,6 +38,16 @@ int option_parse_push_signed(const struct option *opt,\n \tdie(\"bad %s argument: %s\", opt->long_name, arg);\n }\n \n+static int config_use_sideband = 1;\n+\n+static int send_pack_config(const char *var, const char *value, void *unused)\n+{\n+\tif (!strcmp(\"sendpack.sideband\", var))\n+\t\tconfig_use_sideband = git_config_bool(var, value);\n+\n+\treturn 0;\n+}\n+\n static void feed_object(const struct object_id *oid, FILE *fh, int negative)\n {\n \tif (negative && !has_object_file(oid))\n@@ -390,6 +400,8 @@ int send_pack(struct send_pack_args *args,\n \tconst char *push_cert_nonce = NULL;\n \tstruct packet_reader reader;\n \n+\tgit_config(send_pack_config, NULL);\n+\n \t/* Does the other end support the reporting? */\n \tif (server_supports(\"report-status\"))\n \t\tstatus_report = 1;\n@@ -397,7 +409,7 @@ int send_pack(struct send_pack_args *args,\n \t\tallow_deleting_refs = 1;\n \tif (server_supports(\"ofs-delta\"))\n \t\targs->use_ofs_delta = 1;\n-\tif (server_supports(\"side-band-64k\"))\n+\tif (config_use_sideband && server_supports(\"side-band-64k\"))\n \t\tuse_sideband = 1;\n \tif (server_supports(\"quiet\"))\n \t\tquiet_supported = 1;\n-- \ngitgitgadget\n"},{"id":"374680","messageId":"CAPig+cT1EP+HC5h4e0NGCu=1i6cZXPxERRgnnAdZJn+XeqpbFw@mail.gmail.com","threadId":"51010","inReplyTo":"31aa7bfcab834b753cc9f52fc9cc187f65e2d964.1556575475.git.gitgitgadget@gmail.com","subject":"Re: [PATCH 1/1] mingw: optionally disable side-band-64k for transport","fromName":"Eric Sunshine","fromEmail":"sunshine@sunshineco.com","sentAt":"2019-04-29T22:10:04Z","receivedAt":"2019-04-29T22:10:18Z","isPatch":true,"sender":{"key":"sunshine@sunshineco.com","avatar":"https://avatars.githubusercontent.com/u/163641?v=4"},"body":"On Mon, Apr 29, 2019 at 6:04 PM Thomas Braun via GitGitGadget\n<gitgitgadget@gmail.com> wrote:\n> Since commit 0c499ea60f (send-pack: demultiplex a sideband stream with\n> status data, 2010-02-05) the built-in send-pack uses the side-band-64k\n> capability if advertised by the server.\n>\n> Unfortunately this breaks pushing over the dump git protocol if used\n\ns/dump/dumb/\n\n> over a network connection when using MinGW (but *not* when using\n> mingw-w64).\n> [...]\n> The new config option \"sendpack.sideband\" allows to override the\n> side-band-64k capability of the server, and thus makes the dump git\n> protocol work.\n> [...]\n> Signed-off-by: Thomas Braun <thomas.braun@byte-physics.de>\n> Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>\n> ---\n> diff --git a/Documentation/config/sendpack.txt b/Documentation/config/sendpack.txt\n> @@ -0,0 +1,5 @@\n> +sendpack.sideband::\n> +       Allows to disable the side-band-64k capability for send-pack even\n> +       when it is advertised by the server. Makes it possible to work\n> +       around a limitation in the git for windows implementation together\n> +       with the dump git protocol. Defaults to true.\n\ns/dump/dumb/\n\nFor someone who hasn't read the commit message of this patch, \"work\naround a limitation in ... git for windows\" doesn't mean much. Perhaps\nthis documentation could explain in more precise terms under what\ncircumstances this option should be used?\n"},{"id":"374686","messageId":"nycvar.QRO.7.76.6.1904291916510.45@tvgsbejvaqbjf.bet","threadId":"51010","inReplyTo":"CAPig+cT1EP+HC5h4e0NGCu=1i6cZXPxERRgnnAdZJn+XeqpbFw@mail.gmail.com","subject":"Re: [PATCH 1/1] mingw: optionally disable side-band-64k for transport","fromName":"Johannes Schindelin","fromEmail":"johannes.schindelin@gmx.de","sentAt":"2019-04-29T23:17:48Z","receivedAt":"2019-04-29T23:17:56Z","isPatch":true,"sender":{"key":"johannes.schindelin@gmx.de","avatar":"https://avatars.githubusercontent.com/u/127790?v=4"},"body":"Hi Eric,\n\nOn Mon, 29 Apr 2019, Eric Sunshine wrote:\n\n> On Mon, Apr 29, 2019 at 6:04 PM Thomas Braun via GitGitGadget\n> <gitgitgadget@gmail.com> wrote:\n> > Since commit 0c499ea60f (send-pack: demultiplex a sideband stream with\n> > status data, 2010-02-05) the built-in send-pack uses the side-band-64k\n> > capability if advertised by the server.\n> >\n> > Unfortunately this breaks pushing over the dump git protocol if used\n>\n> s/dump/dumb/\n\nOf course!\n\n> > over a network connection when using MinGW (but *not* when using\n> > mingw-w64).\n> > [...]\n> > The new config option \"sendpack.sideband\" allows to override the\n> > side-band-64k capability of the server, and thus makes the dump git\n> > protocol work.\n> > [...]\n> > Signed-off-by: Thomas Braun <thomas.braun@byte-physics.de>\n> > Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>\n> > ---\n> > diff --git a/Documentation/config/sendpack.txt b/Documentation/config/sendpack.txt\n> > @@ -0,0 +1,5 @@\n> > +sendpack.sideband::\n> > +       Allows to disable the side-band-64k capability for send-pack even\n> > +       when it is advertised by the server. Makes it possible to work\n> > +       around a limitation in the git for windows implementation together\n> > +       with the dump git protocol. Defaults to true.\n>\n> s/dump/dumb/\n>\n> For someone who hasn't read the commit message of this patch, \"work\n> around a limitation in ... git for windows\" doesn't mean much. Perhaps\n> this documentation could explain in more precise terms under what\n> circumstances this option should be used?\n\nYou're right, this is confusing, especially since Git for Windows 2.x does\nnot have that bug.\n\nI simply dropped that sentence.\n\nThanks!\nDscho\n"},{"id":"374687","messageId":"20190429231928.GQ6316@genre.crustytoothpaste.net","threadId":"51010","inReplyTo":"31aa7bfcab834b753cc9f52fc9cc187f65e2d964.1556575475.git.gitgitgadget@gmail.com","subject":"Re: [PATCH 1/1] mingw: optionally disable side-band-64k for transport","fromName":"brian m. carlson","fromEmail":"sandals@crustytoothpaste.net","sentAt":"2019-04-29T23:19:28Z","receivedAt":"2019-04-29T23:19:38Z","isPatch":true,"sender":{"key":"sandals@crustytoothpaste.net","avatar":"https://avatars.githubusercontent.com/u/497054?v=4"},"body":"On Mon, Apr 29, 2019 at 03:04:36PM -0700, Thomas Braun via GitGitGadget wrote:\n> From: Thomas Braun <thomas.braun@byte-physics.de>\n> \n> Since commit 0c499ea60f (send-pack: demultiplex a sideband stream with\n> status data, 2010-02-05) the built-in send-pack uses the side-band-64k\n> capability if advertised by the server.\n> \n> Unfortunately this breaks pushing over the dump git protocol if used\n> over a network connection when using MinGW (but *not* when using\n> mingw-w64).\n> \n> The detailed reasons for this, are courtesy of Jeff Preshing, quoted\n> from https://groups.google.com/d/msg/msysgit/at8D7J-h7mw/eaLujILGUWoJ:\n> \n> \tMinGW wraps Windows sockets in CRT file descriptors in order to\n> \tmimic the functionality of POSIX sockets. This causes msvcrt.dll\n> \tto treat sockets as Installable File System (IFS) handles,\n> \tcalling ReadFile, WriteFile, DuplicateHandle and CloseHandle on\n> \tthem. This approach works well in simple cases on recent\n> \tversions of Windows, but does not support all usage patterns.\n> \tIn particular, using this approach, any attempt to read & write\n> \tconcurrently on the same socket (from one or more processes)\n> \twill deadlock in a scenario where the read waits for a response\n> \tfrom the server which is only invoked after the write. This is\n> \twhat send_pack currently attempts to do in the use_sideband\n> \tcodepath.\n\nSince this is a platform-specific issue, can we address this using a\ncompile-time constant instead of a config option? It would be better to\ndo the right thing automatically in this case and not have to have\npeople set a config option. It will also allow us to not to have to\nmaintain a config option indefinitely if MinGW becomes more capable in\nthe future.\n-- \nbrian m. carlson: Houston, Texas, US\nOpenPGP: https://keybase.io/bk2204\n"},{"id":"374694","messageId":"b8e799e2-356d-2a03-faa4-2792ec1b171d@kdbg.org","threadId":"51010","inReplyTo":"nycvar.QRO.7.76.6.1904291916510.45@tvgsbejvaqbjf.bet","subject":"Re: [PATCH 1/1] mingw: optionally disable side-band-64k for transport","fromName":"Johannes Sixt","fromEmail":"j6t@kdbg.org","sentAt":"2019-04-30T06:21:44Z","receivedAt":"2019-04-30T06:21:50Z","isPatch":true,"sender":{"key":"j6t@kdbg.org","avatar":"https://avatars.githubusercontent.com/u/14810926?v=4"},"body":"Am 30.04.19 um 01:17 schrieb Johannes Schindelin:\n> On Mon, 29 Apr 2019, Eric Sunshine wrote:\n>> On Mon, Apr 29, 2019 at 6:04 PM Thomas Braun via GitGitGadget\n>>> diff --git a/Documentation/config/sendpack.txt b/Documentation/config/sendpack.txt\n>>> @@ -0,0 +1,5 @@\n>>> +sendpack.sideband::\n>>> +       Allows to disable the side-band-64k capability for send-pack even\n>>> +       when it is advertised by the server. Makes it possible to work\n>>> +       around a limitation in the git for windows implementation together\n>>> +       with the dump git protocol. Defaults to true.\n>>\n>> s/dump/dumb/\n>>\n>> For someone who hasn't read the commit message of this patch, \"work\n>> around a limitation in ... git for windows\" doesn't mean much. Perhaps\n>> this documentation could explain in more precise terms under what\n>> circumstances this option should be used?\n> \n> You're right, this is confusing, especially since Git for Windows 2.x does\n> not have that bug.\n\nIf there is no bug, why do we need the patch?\n\n-- Hannes\n"},{"id":"374727","messageId":"nycvar.QRO.7.76.6.1904301832500.45@tvgsbejvaqbjf.bet","threadId":"51010","inReplyTo":"b8e799e2-356d-2a03-faa4-2792ec1b171d@kdbg.org","subject":"Re: [PATCH 1/1] mingw: optionally disable side-band-64k for transport","fromName":"Johannes Schindelin","fromEmail":"johannes.schindelin@gmx.de","sentAt":"2019-04-30T22:33:47Z","receivedAt":"2019-04-30T22:34:00Z","isPatch":true,"sender":{"key":"johannes.schindelin@gmx.de","avatar":"https://avatars.githubusercontent.com/u/127790?v=4"},"body":"Hi Hannes,\n\nOn Tue, 30 Apr 2019, Johannes Sixt wrote:\n\n> Am 30.04.19 um 01:17 schrieb Johannes Schindelin:\n> > On Mon, 29 Apr 2019, Eric Sunshine wrote:\n> >> On Mon, Apr 29, 2019 at 6:04 PM Thomas Braun via GitGitGadget\n> >>> diff --git a/Documentation/config/sendpack.txt b/Documentation/config/sendpack.txt\n> >>> @@ -0,0 +1,5 @@\n> >>> +sendpack.sideband::\n> >>> +       Allows to disable the side-band-64k capability for send-pack even\n> >>> +       when it is advertised by the server. Makes it possible to work\n> >>> +       around a limitation in the git for windows implementation together\n> >>> +       with the dump git protocol. Defaults to true.\n> >>\n> >> s/dump/dumb/\n> >>\n> >> For someone who hasn't read the commit message of this patch, \"work\n> >> around a limitation in ... git for windows\" doesn't mean much. Perhaps\n> >> this documentation could explain in more precise terms under what\n> >> circumstances this option should be used?\n> >\n> > You're right, this is confusing, especially since Git for Windows 2.x does\n> > not have that bug.\n>\n> If there is no bug, why do we need the patch?\n\nI thought you of all people (building with the ancient MSys/MINGW\ntoolchains) would benefit from it :-)\n\nBut if even you don't want it, I'll gladly drop it from Git for Windows'\npatches and be done with it.\n\nCiao,\nDscho\n"},{"id":"374728","messageId":"nycvar.QRO.7.76.6.1904301834230.45@tvgsbejvaqbjf.bet","threadId":"51010","inReplyTo":"20190429231928.GQ6316@genre.crustytoothpaste.net","subject":"Re: [PATCH 1/1] mingw: optionally disable side-band-64k for transport","fromName":"Johannes Schindelin","fromEmail":"johannes.schindelin@gmx.de","sentAt":"2019-04-30T22:37:06Z","receivedAt":"2019-04-30T22:37:24Z","isPatch":true,"sender":{"key":"johannes.schindelin@gmx.de","avatar":"https://avatars.githubusercontent.com/u/127790?v=4"},"body":"Hi brian,\n\nOn Mon, 29 Apr 2019, brian m. carlson wrote:\n\n> On Mon, Apr 29, 2019 at 03:04:36PM -0700, Thomas Braun via GitGitGadget wrote:\n> > From: Thomas Braun <thomas.braun@byte-physics.de>\n> >\n> > Since commit 0c499ea60f (send-pack: demultiplex a sideband stream with\n> > status data, 2010-02-05) the built-in send-pack uses the side-band-64k\n> > capability if advertised by the server.\n> >\n> > Unfortunately this breaks pushing over the dump git protocol if used\n> > over a network connection when using MinGW (but *not* when using\n> > mingw-w64).\n> >\n> > The detailed reasons for this, are courtesy of Jeff Preshing, quoted\n> > from https://groups.google.com/d/msg/msysgit/at8D7J-h7mw/eaLujILGUWoJ:\n> >\n> > \tMinGW wraps Windows sockets in CRT file descriptors in order to\n> > \tmimic the functionality of POSIX sockets. This causes msvcrt.dll\n> > \tto treat sockets as Installable File System (IFS) handles,\n> > \tcalling ReadFile, WriteFile, DuplicateHandle and CloseHandle on\n> > \tthem. This approach works well in simple cases on recent\n> > \tversions of Windows, but does not support all usage patterns.\n> > \tIn particular, using this approach, any attempt to read & write\n> > \tconcurrently on the same socket (from one or more processes)\n> > \twill deadlock in a scenario where the read waits for a response\n> > \tfrom the server which is only invoked after the write. This is\n> > \twhat send_pack currently attempts to do in the use_sideband\n> > \tcodepath.\n>\n> Since this is a platform-specific issue, can we address this using a\n> compile-time constant instead of a config option? It would be better to\n> do the right thing automatically in this case and not have to have\n> people set a config option. It will also allow us to not to have to\n> maintain a config option indefinitely if MinGW becomes more capable in\n> the future.\n\nI was really not sure at the time whether this would be fixed in MinGW at\nsome stage, and with the switch to mingw-w64 (by moving from MSys to MSYS2\nas of Git for Windows 2.x) we do not really have any concrete need for it\nanymore.\n\nI just thought that it might benefit somebody (it was my impression that\nHannes Sixt still built his own copy with MSys/MinGW, and we do have a\ntrack record of maintaining certain things even for single users, see e.g.\nour insistence on creating the .git/branches/ directory upon `git init`).\n\nBut if the consensus is that we do not need this at all anymore, I'll be\njust as happy to drop that patch.\n\nCiao,\nDscho\n"},{"id":"374732","messageId":"890578de-6286-1726-37fc-d05bf4b359df@kdbg.org","threadId":"51010","inReplyTo":"nycvar.QRO.7.76.6.1904301832500.45@tvgsbejvaqbjf.bet","subject":"Re: [PATCH 1/1] mingw: optionally disable side-band-64k for transport","fromName":"Johannes Sixt","fromEmail":"j6t@kdbg.org","sentAt":"2019-04-30T22:55:19Z","receivedAt":"2019-04-30T22:55:24Z","isPatch":true,"sender":{"key":"j6t@kdbg.org","avatar":"https://avatars.githubusercontent.com/u/14810926?v=4"},"body":"Am 01.05.19 um 00:33 schrieb Johannes Schindelin:\n> On Tue, 30 Apr 2019, Johannes Sixt wrote:\n>> Am 30.04.19 um 01:17 schrieb Johannes Schindelin:\n>>> You're right, this is confusing, especially since Git for Windows 2.x does\n>>> not have that bug.\n>>\n>> If there is no bug, why do we need the patch?\n> \n> I thought you of all people (building with the ancient MSys/MINGW\n> toolchains) would benefit from it :-)\n> \n> But if even you don't want it, I'll gladly drop it from Git for Windows'\n> patches and be done with it.\n\nAh, this is only for the ancient MinGW! I do indeed not mind if you drop\nit as I have updated my toolchain to the one you provide long ago.\n\nThank you for being considerate!\n\n-- Hannes\n"},{"id":"374862","messageId":"nycvar.QRO.7.76.6.1905031041590.45@tvgsbejvaqbjf.bet","threadId":"51010","inReplyTo":"890578de-6286-1726-37fc-d05bf4b359df@kdbg.org","subject":"Re: [PATCH 1/1] mingw: optionally disable side-band-64k for transport","fromName":"Johannes Schindelin","fromEmail":"johannes.schindelin@gmx.de","sentAt":"2019-05-03T08:42:47Z","receivedAt":"2019-05-03T08:42:57Z","isPatch":true,"sender":{"key":"johannes.schindelin@gmx.de","avatar":"https://avatars.githubusercontent.com/u/127790?v=4"},"body":"Hi Hannes,\n\nOn Wed, 1 May 2019, Johannes Sixt wrote:\n\n> Am 01.05.19 um 00:33 schrieb Johannes Schindelin:\n> > On Tue, 30 Apr 2019, Johannes Sixt wrote:\n> >> Am 30.04.19 um 01:17 schrieb Johannes Schindelin:\n> >>> You're right, this is confusing, especially since Git for Windows 2.x does\n> >>> not have that bug.\n> >>\n> >> If there is no bug, why do we need the patch?\n> >\n> > I thought you of all people (building with the ancient MSys/MINGW\n> > toolchains) would benefit from it :-)\n> >\n> > But if even you don't want it, I'll gladly drop it from Git for Windows'\n> > patches and be done with it.\n>\n> Ah, this is only for the ancient MinGW! I do indeed not mind if you drop\n> it as I have updated my toolchain to the one you provide long ago.\n\nOkay then, let's drop this!\n\nCiao,\nDscho\n"}]}