{"thread":{"id":"7069","subject":"Cygwin and git 1.5.0","startedAt":"2007-03-03T18:31:02Z","lastAt":"2007-03-03T18:40:05Z","messageCount":2,"participants":["Ramsay Jones","Johannes Schindelin"],"isPatch":false,"patchVersion":null,"patchTotal":null},"messages":[{"id":"36192","messageId":"45E9BEE6.8000506@ramsay1.demon.co.uk","threadId":"7069","inReplyTo":null,"subject":"Cygwin and git 1.5.0","fromName":"Ramsay Jones","fromEmail":"ramsay@ramsay1.demon.co.uk","sentAt":"2007-03-03T18:31:02Z","receivedAt":"2007-03-03T18:31:02Z","isPatch":false,"sender":{"key":"ramsay@ramsayjones.plus.com","avatar":"https://avatars.githubusercontent.com/u/33702710?v=4"},"body":"Hi Junio,\n\nAt the end of last year I finally bought a new laptop (yay!). Much to my\nsurprise, I haven't yet got around to installing Linux (The Ubuntu live DVD\nis the only one I have found which \"works\"; but it's so *slow* as to be\nunusable, so I still don't know if all my h/w works with it).\n\nHowever, almost the first thing I did was to install Cygwin, which I have been\nusing on windows since about beta-20 (back in the days of cygnus solutions).\nCygwin has its problems, but I could not use windows without it.\n\nSo naturally I tried building git on Cygwin (copying my git repo over from\nmy old Linux system). I have to say that, after some initial problems, it\nwent rather well. (hint: don't use a working directory backed up from Linux\non cygwin; \"make clean\" didn't - among other things!).\n\nI recently upgraded git using the 1.5.0 tarball. (Yes, since I now have git on\nwindows, where I can access the internet, I could clone the git repo directly.\nHowever, since I'm still on dial-up, I can't spare the bandwidth. Maybe I will\ntry shallow clone sometime.)\n\nAs far as git 1.5.0 is concerned, the build/test went with only minor problems.\nI had to disable test t4016-diff-quote.sh, because it used tabs and LFs in filenames.\nI note that Alex Riesen already sent you a patch to cover that, so I won't send\nanother one.\n\nHowever, since I like to have \"ALL_CFLAGS+= -Werror\" in my config.mak file, I had\nto get rid of the gcc warnings. This resulted in the following patch series:\n\n[PATCH 1/6] Fix some \"printf format\" warnings.\n[PATCH 2/6] Unset (comment out) NO_C99_FORMAT on Cygwin.\n[PATCH 3/6] Fix a \"label defined but unreferenced\" warning.\n[PATCH 4/6] Fix an \"implicit function definition\" warning.\n[PATCH 5/6] Fix some \"comparison is always true/false\" warnings.\n[PATCH 6/6] Fix a \"pointer type missmatch\" warning.\n\nAs I note in the [PATCH 2/6] email, I should have deleted that setting from the\nMakefile, rather than commenting it out, but I didn't have the energy to re-do\nall the commits/emails. ;-)\n\nThe patches are against 1.5.0 tarball which, hopefully, is not so long ago as to\nmake them impossible to apply. Also, I'm hoping that my mail client (thunderbird 1.5.0.2)\nhas not caused any whitespace damage. (I can resend as attachments if required).\n\nAlso, I noticed that some recent changes seem to have made NO_C99_FORMAT somewhat\nredundant. I _suspect_ that the unconditional #include-ing of the C99 headers\n<inttypes.h> and (therefore also) <stdint.h>, has made it necessary to use a\ncompiler/library combination which has pretty good, if not full, C99 support.\nIOW, if it has those headers, it almost definitely supports the C99 printf format\nspecifiers. For example, my old Linux system would not be able to compile the\ncurrent git. (Has anybody tried on SunOS 5.8 and 5.9?)\n\nThis is not a problem, but I do wonder if it was a conscious decision.\n\nAll the Best,\n\nRamsay Jones\n"},{"id":"36193","messageId":"Pine.LNX.4.63.0703031938310.22628@wbgn013.biozentrum.uni-wuerzburg.de","threadId":"7069","inReplyTo":"45E9BEE6.8000506@ramsay1.demon.co.uk","subject":"Re: Cygwin and git 1.5.0","fromName":"Johannes Schindelin","fromEmail":"johannes.schindelin@gmx.de","sentAt":"2007-03-03T18:40:05Z","receivedAt":"2007-03-03T18:40:05Z","isPatch":false,"sender":{"key":"johannes.schindelin@gmx.de","avatar":"https://avatars.githubusercontent.com/u/127790?v=4"},"body":"Hi,\n\nOn Sat, 3 Mar 2007, Ramsay Jones wrote:\n\n> As I note in the [PATCH 2/6] email, I should have deleted that setting \n> from the Makefile, rather than commenting it out, but I didn't have the \n> energy to re-do all the commits/emails. ;-)\n\nIn these cases, I (carefully) edit the diffs. In your case, I would have \ndeleted the line starting with \"+\", and would have decreased the last \nnumber on the line starting with \"@@\" by 1.\n\nHth,\nDscho\n"}]}