{"thread":{"id":"1858","subject":"[PATCHES] CURLDIR support, more usable cmd-rename.sh","startedAt":"2005-09-19T11:14:36Z","lastAt":"2005-09-20T06:25:52Z","messageCount":7,"participants":["Patrick Mauritz","Petr Baudis","Matthias Urlichs","Junio C Hamano"],"isPatch":false,"patchVersion":null,"patchTotal":null},"messages":[{"id":"8850","messageId":"1127128475.781.6.camel@divert","threadId":"1858","inReplyTo":null,"subject":"[PATCHES] CURLDIR support, more usable cmd-rename.sh","fromName":"Patrick Mauritz","fromEmail":"oxygene@studentenbude.ath.cx","sentAt":"2005-09-19T11:14:36Z","receivedAt":"2005-09-19T11:14:36Z","isPatch":false,"sender":{"key":"oxygene@studentenbude.ath.cx","avatar":null},"body":"hello,\n\nattached are two patches.\n\nthe first is an updated implementation of CURLDIR support (for those who\nhave libcurl outside their library search path)\n\nthe second changes cmd-rename.sh so that it only adds symlinks to\nexecutables that actually exist (eg. when building with NO_CURL,\ngit-http-pull would still be created).\n\nanother issue I had was that mailinfo uses strcasestr, which isn't in\nsolaris' libc (and not in posix according to\nhttp://lists.gnu.org/archive/html/bug-gnulib/2005-08/msg00076.html). I\ndisabled it in my build for now, and might come up with a patch later,\nunless someone beats me to it.\n\n\npatrick mauritz\n"},{"id":"8854","messageId":"20050919133731.GB2903@pasky.or.cz","threadId":"1858","inReplyTo":"1127128475.781.6.camel@divert","subject":"Re: [PATCHES] CURLDIR support, more usable cmd-rename.sh","fromName":"Petr Baudis","fromEmail":"pasky@suse.cz","sentAt":"2005-09-19T13:37:31Z","receivedAt":"2005-09-19T13:37:31Z","isPatch":false,"sender":{"key":"pasky@ucw.cz","avatar":"https://avatars.githubusercontent.com/u/18439?v=4"},"body":"Dear diary, on Mon, Sep 19, 2005 at 01:14:36PM CEST, I got a letter\nwhere Patrick Mauritz <oxygene@studentenbude.ath.cx> told me that...\n> hello,\n\nHello,\n\n> attached are two patches.\n\nseems that they didn't quite get attached in the end. ;-)\n\n> another issue I had was that mailinfo uses strcasestr, which isn't in\n> solaris' libc (and not in posix according to\n> http://lists.gnu.org/archive/html/bug-gnulib/2005-08/msg00076.html). I\n> disabled it in my build for now, and might come up with a patch later,\n> unless someone beats me to it.\n\nLinus already did in <Pine.LNX.4.58.0509181829310.9106@g5.osdl.org>.\n\n-- \n\t\t\t\tPetr \"Pasky\" Baudis\nStuff: http://pasky.or.cz/\nIf you want the holes in your knowledge showing up try teaching\nsomeone.  -- Alan Cox\n"},{"id":"8856","messageId":"1127139079.781.9.camel@divert","threadId":"1858","inReplyTo":"20050919133731.GB2903@pasky.or.cz","subject":"Re: [PATCHES] CURLDIR support, more usable cmd-rename.sh","fromName":"Patrick Mauritz","fromEmail":"oxygene@studentenbude.ath.cx","sentAt":"2005-09-19T14:11:19Z","receivedAt":"2005-09-19T14:11:19Z","isPatch":false,"sender":{"key":"oxygene@studentenbude.ath.cx","avatar":null},"body":"On Mon, 2005-09-19 at 15:37, Petr Baudis wrote:\n> seems that they didn't quite get attached in the end. ;-)\noops :)\nhere they are\n\n> Linus already did in <Pine.LNX.4.58.0509181829310.9106@g5.osdl.org>.\nyay :)\n\n\npatrick mauritz\n\n\ndiff -ur git-core-0.99.7.orig//Makefile git-core-0.99.7/Makefile\n--- git-core-0.99.7.orig//Makefile\tMon Sep 19 12:43:04 2005\n+++ git-core-0.99.7/Makefile\tMon Sep 19 12:45:45 2005\n@@ -42,7 +42,10 @@\n \n GIT_VERSION = 0.99.7\n \n-CFLAGS = -g -O2 -Wall\n+CFLAGS = -g -O2\n+ifdef CURLDIR\n+CFLAGS+=-I$(CURLDIR)/include\n+endif\n ALL_CFLAGS = $(CFLAGS) $(PLATFORM_DEFINES) $(DEFINES)\n \n prefix = $(HOME)\n@@ -261,7 +264,11 @@\n git-ssh-pull: rsh.o fetch.o\n git-ssh-push: rsh.o\n \n+ifdef CURLDIR\n+git-http-fetch: LIBS += -L$(CURLDIR)/lib -R$(CURLDIR)/lib -lcurl\n+else\n git-http-fetch: LIBS += -lcurl\n+endif\n git-rev-list: LIBS += $(OPENSSL_LIBSSL)\n \n init-db.o: init-db.c\n\n\n--- cmd-rename.sh~\tMon Sep 19 12:53:40 2005\n+++ git-core-0.99.7/cmd-rename.sh\tMon Sep 19 12:53:58 2005\n@@ -4,7 +4,9 @@\n while read old new\n do\n \trm -f \"$d/$old\"\n+\tif [ -x \"$d/$new\" ]; then\n \tln -s \"$new\" \"$d/$old\"\n+\tfi\n done <<\\EOF\n git-add-script\tgit-add\n git-archimport-script\tgit-archimport\n"},{"id":"8865","messageId":"pan.2005.09.19.14.37.11.966217@smurf.noris.de","threadId":"1858","inReplyTo":"1127139079.781.9.camel@divert","subject":"Re: [PATCHES] CURLDIR support, more usable cmd-rename.sh","fromName":"Matthias Urlichs","fromEmail":"smurf@smurf.noris.de","sentAt":"2005-09-19T14:37:14Z","receivedAt":"2005-09-19T14:37:14Z","isPatch":false,"sender":{"key":"matthias@urlichs.de","avatar":"https://gravatar.com/avatar/2708905af227313eba6f2b2ae0f7d0259b5ac5d71baef58fe5a13c699ce0bbf0?d=mp&s=160"},"body":"Hi, Patrick Mauritz wrote:\n\n> -CFLAGS = -g -O2 -Wall\n> +CFLAGS = -g -O2\n\nUmm, no, you don't get to unconditionally throw way -Wall.\n\n*Some* people think this flag makes sense and that the sane default\n(on sane OSes, i.e. those who have GCC and glibc) should be to include it. \n ;-)\n\n-- \nMatthias Urlichs   |   {M:U} IT Design @ m-u-it.de   |  smurf@smurf.noris.de\nDisclaimer: The quote was selected randomly. Really. | http://smurf.noris.de\n - -\nI must have a prodigious quantity of mind; it takes me as much as a\nweek sometimes to make it up.\n\t\t-- Mark Twain, \"The Innocents Abroad\"\n"},{"id":"8881","messageId":"7vhdchqan8.fsf@assigned-by-dhcp.cox.net","threadId":"1858","inReplyTo":"1127139079.781.9.camel@divert","subject":"Re: [PATCHES] CURLDIR support, more usable cmd-rename.sh","fromName":"Junio C Hamano","fromEmail":"junkio@cox.net","sentAt":"2005-09-19T16:34:35Z","receivedAt":"2005-09-19T16:34:35Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"Patrick Mauritz <oxygene@studentenbude.ath.cx> writes:\n\n> diff -ur git-core-0.99.7.orig//Makefile git-core-0.99.7/Makefile\n> --- git-core-0.99.7.orig//Makefile\tMon Sep 19 12:43:04 2005\n> +++ git-core-0.99.7/Makefile\tMon Sep 19 12:45:45 2005\n\nWhy not a patch from either git or cogito or stgit ;-)?\nAnyway, thanks for your patch.\n\n> -CFLAGS = -g -O2 -Wall\n> +CFLAGS = -g -O2\n\nNot relevant to the current topic as far as I can tell (unless\nyour non-standard curl include files trip under -Wall).\n\n> +ifdef CURLDIR\n> +CFLAGS+=-I$(CURLDIR)/include\n> +endif\n>  ALL_CFLAGS = $(CFLAGS) $(PLATFORM_DEFINES) $(DEFINES)\n>  \n>  prefix = $(HOME)\n> @@ -261,7 +264,11 @@\n>  git-ssh-pull: rsh.o fetch.o\n>  git-ssh-push: rsh.o\n>  \n> +ifdef CURLDIR\n> +git-http-fetch: LIBS += -L$(CURLDIR)/lib -R$(CURLDIR)/lib -lcurl\n> +else\n>  git-http-fetch: LIBS += -lcurl\n> +endif\n\nI do not have any objections to helping installations where CURL\nis installed somewhere nonstandard, but if that is the case I am\nwondering if assuming that they are all in the same directory\nhierarchy covers enough.  CURLDIR=/usr/local/ would be OK and\nCURLDIR=/opt/curl/ would be as well with this patch, so this\nprobably is fine.  Installations without ld.so.conf need -R, but\nis it safe to always force -R like this?\n\nAlso I wonder if we need similar customization for openssl\nstuff, but that is a separate topic.\n\n> --- cmd-rename.sh~\tMon Sep 19 12:53:40 2005\n> +++ git-core-0.99.7/cmd-rename.sh\tMon Sep 19 12:53:58 2005\n> @@ -4,7 +4,9 @@\n>  while read old new\n>  do\n>  \trm -f \"$d/$old\"\n> +\tif [ -x \"$d/$new\" ]; then\n>  \tln -s \"$new\" \"$d/$old\"\n> +\tfi\n\nNo questions about this part.  Thanks for the fix.\n"},{"id":"8887","messageId":"1127151642.781.15.camel@divert","threadId":"1858","inReplyTo":"7vhdchqan8.fsf@assigned-by-dhcp.cox.net","subject":"Re: [PATCHES] CURLDIR support, more usable cmd-rename.sh","fromName":"Patrick Mauritz","fromEmail":"oxygene@studentenbude.ath.cx","sentAt":"2005-09-19T17:40:42Z","receivedAt":"2005-09-19T17:40:42Z","isPatch":false,"sender":{"key":"oxygene@studentenbude.ath.cx","avatar":null},"body":"On Mon, 2005-09-19 at 18:34, Junio C Hamano wrote:\n> Why not a patch from either git or cogito or stgit ;-)?\nbecause my only use for git is with monotone's git import.\n\n> Not relevant to the current topic as far as I can tell (unless\n> your non-standard curl include files trip under -Wall).\nthe only issue with this is that sunpro's cc doesn't like -Wall\n\n> Installations without ld.so.conf need -R, but\n> is it safe to always force -R like this?\nthe only case that comes to mind is where you want to link git against\na library which is at a different location now than it is later, eg. in\nconjunction with DESTDIR or similar means.\nas long as git doesn't build any shared libraries by itself and links to\nthem, this shouldn't be an issue.\n\n\npatrick mauritz\n"},{"id":"8975","messageId":"7vfys08dcf.fsf@assigned-by-dhcp.cox.net","threadId":"1858","inReplyTo":"1127151642.781.15.camel@divert","subject":"Re: [PATCHES] CURLDIR support, more usable cmd-rename.sh","fromName":"Junio C Hamano","fromEmail":"junkio@cox.net","sentAt":"2005-09-20T06:25:52Z","receivedAt":"2005-09-20T06:25:52Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"Patrick Mauritz <oxygene@studentenbude.ath.cx> writes:\n\n>> Not relevant to the current topic as far as I can tell (unless\n>> your non-standard curl include files trip under -Wall).\n> the only issue with this is that sunpro's cc doesn't like -Wall\n\nSince you will be saying \"make CURLDIR=blah\" anyway, I'd ask you\nto say \"make CURLDIR=blah CFLAGS='-O -g'\" and will forget about\nthis part.\n\nThe rest of your patch I am keeping a variant based on it, done\nslightly differently, in the proposed updates branch.  I'll have\nit graduate to the master branch when I hear from somebody (can\nbe you) with SunPro that the it is OK (or an improvement patch).\n\nThanks.\n"}]}