{"thread":{"id":"14268","subject":"Re: [PATCH 06/12] connect: Fix custom ports with plink (Putty's ssh)","startedAt":"2008-07-03T03:07:09Z","lastAt":"2008-07-04T20:05:48Z","messageCount":3,"participants":["Edward Z. Yang","Johannes Schindelin"],"isPatch":true,"patchVersion":1,"patchTotal":12},"messages":[{"id":"82057","messageId":"486C425D.8090904@thewritingpot.com","threadId":"14268","inReplyTo":null,"subject":"Re: [PATCH 06/12] connect: Fix custom ports with plink (Putty's ssh)","fromName":"Edward Z. Yang","fromEmail":"edwardzyang@thewritingpot.com","sentAt":"2008-07-03T03:07:09Z","receivedAt":"2008-07-03T03:07:09Z","isPatch":true,"sender":{"key":"edwardzyang@thewritingpot.com","avatar":"https://gravatar.com/avatar/a805a0a3c1d7d36e7fe22270596e4d812723652933c59cac267e67c79126fdd0?d=mp&s=160"},"body":"\nJohannes Sixt wrote:\n > What about installing a wrapper script, plinkssh, that does this:\n > [snip]\n\nWell, the patch is shorter :-)\n\nJoking aside, it's a good question. I guess I prefer the patch because:\n\n1. It's been tested, it works. I haven't tried the script yet, so I \ndon't know if it works.\n\n2. Git historically doesn't use bash, so the script would have to be \nrewritten in Perl or plain sh or tcl or something.\n\n3. It's less brittle than the wrapper script if we decide to have Git \npass more params to OpenSSH.\n\n4. It's \"more native\".\n\nI don't know if these are compelling enough reasons, though.\n\n(cc'ed everyone else, whoops)\n"},{"id":"82113","messageId":"alpine.DEB.1.00.0807031313140.9925@racer","threadId":"14268","inReplyTo":"486C425D.8090904@thewritingpot.com","subject":"Re: [PATCH 06/12] connect: Fix custom ports with plink (Putty's ssh)","fromName":"Johannes Schindelin","fromEmail":"johannes.schindelin@gmx.de","sentAt":"2008-07-03T12:29:27Z","receivedAt":"2008-07-03T12:29:27Z","isPatch":true,"sender":{"key":"johannes.schindelin@gmx.de","avatar":"https://avatars.githubusercontent.com/u/127790?v=4"},"body":"\nHi,\n\nOn Wed, 2 Jul 2008, Edward Z. Yang wrote:\n\n> Johannes Sixt wrote:\n> > What about installing a wrapper script, plinkssh, that does this:\n> > [snip]\n> \n> Well, the patch is shorter :-)\n\nBut you have to do it for every SSH backend that you might want to \nsupport.\n\nAnd you have to recompile.\n\n> 1. It's been tested, it works. I haven't tried the script yet, so I \n>    don't know if it works.\n\nSorry, that argument does not fly.  \"My patch is better, because I did not \ntest your patch.\"\n\n> 2. Git historically doesn't use bash, so the script would have to be \n>    rewritten in Perl or plain sh or tcl or something.\n\nThat is so totally untrue.  We have Perl scripts and Shell scripts (for \nwhich we need the bash), and then we have the two GUIs which use Tcl/Tk.\n\nActually, we only have so few Perl scripts left that it might be possible \nto ship a version of Git on Windows without Perl.  The only script that \nneeds to be converted to a builtin is add -i.\n\nThe rest of the scripts are shell.\n\nSo this argument is totally bogus.\n\n> 3. It's less brittle than the wrapper script if we decide to have Git \n>    pass more params to OpenSSH.\n\nGranted, should we decide one day to use more elaborate features of \nOpenSSH, then we would have to change the script, too.\n\nBut most likely, Plink support would be broken by that update _anyway_, \nsince it does not grok the OpenSSH options directly.\n\nAnd guess what is easier to fix, a script that rewrites the arguments from \nOpenSSH syntax to Plink syntax, or a C program with over 78,000 code \nlines that has to be recompiled?\n\n> 4. It's \"more native\".\n\nWould it not be even more native if we just linked in libssl?  Would you \nwrite the patch?\n\nFurther, would you like to convert and maintain all people's wrapper \nscripts to C code inside Git?\n\nBTW what is the reason why Hannes' mail does not appear to be the mail \nyou replied to in GMane, but the patch Steffen sent?\n\nCiao,\nDscho\n"},{"id":"82254","messageId":"486E829C.608@thewritingpot.com","threadId":"14268","inReplyTo":"alpine.DEB.1.00.0807031313140.9925@racer","subject":"Re: [PATCH 06/12] connect: Fix custom ports with plink (Putty's ssh)","fromName":"Edward Z. Yang","fromEmail":"edwardzyang@thewritingpot.com","sentAt":"2008-07-04T20:05:48Z","receivedAt":"2008-07-04T20:05:48Z","isPatch":true,"sender":{"key":"edwardzyang@thewritingpot.com","avatar":"https://gravatar.com/avatar/a805a0a3c1d7d36e7fe22270596e4d812723652933c59cac267e67c79126fdd0?d=mp&s=160"},"body":"\n> Sorry, that argument does not fly.  \"My patch is better, because I did not \n> test your patch.\"\n\nJust tested, the patch works.\n\n> That is so totally untrue.  We have Perl scripts and Shell scripts (for \n> which we need the bash), and then we have the two GUIs which use Tcl/Tk.\n\nI came up with that conclusion by grepping the Git source code for the \nword bash; no results. Granted, it's still a null point because the \nproposed script doesn't use any bash-specific features.\n\n> Further, would you like to convert and maintain all people's wrapper \n> scripts to C code inside Git?\n\nI was under the impression that wrapper scripts were for fleshing out \nnew APIs and implementing non-performance critical functionality, \nwithout all the overhead of writing in C. There is little to no overhead \nfrom this patch.\n\nAnyway, Johannes still makes some pretty compelling points for the \nwrapper script, so you can count me +1 for the wrapper.\n\n> BTW what is the reason why Hannes' mail does not appear to be the mail \n> you replied to in GMane, but the patch Steffen sent?\n\nI actually did a \"Reply\" and so he was the only one who got the email at \nfirst. Then I resent it to the list, as well as the other CC'ed people.\n\n(Thus my comment at the bottom)\n"}]}