# [PATCH/GSoC 3/3] Nousage message in error 5 messages from 2016-03-24 to 2016-03-24. Participants: Diwas Joshi, Pranit Bauva, Junio C Hamano. Thread: https://gitlist.dev/t/41803 ## Diwas Joshi, 2016-03-24 02:03 Subject: [PATCH/GSoC 3/3] Nousage message in error Message-ID: <1458785018-29232-1-git-send-email-dj.dij123@gmail.com> URL: https://gitlist.dev/e/1458785018-29232-1-git-send-email-dj.dij123%40gmail.com ``` - To show only error text instead of full usage message - Adds exits to callback function in parse-options-cb.c instead of returning -1 which results in display of usage message. --- parse-options-cb.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/parse-options-cb.c b/parse-options-cb.c index 239898d..b7321d1 100644 --- a/parse-options-cb.c +++ b/parse-options-cb.c @@ -85,8 +85,10 @@ int parse_opt_commits(const struct option *opt, const char *arg, int unset) if (!arg) return -1; - if (get_sha1(arg, sha1)) - return error("malformed object name %s", arg); + if (get_sha1(arg, sha1)) { + error("malformed object name %s", arg); + exit(129); + } commit = lookup_commit_reference(sha1); if (!commit) return error("no such commit %s", arg); -- 2.7.4 ``` ## Pranit Bauva, 2016-03-24 05:14 Subject: Re: [PATCH/GSoC 3/3] Nousage message in error Message-ID: URL: https://gitlist.dev/e/CAFZEwPN2vp%2BzOMdGY51LwNpgNcYxsGD4GBXHNOA8_qJ8vtE5OQ%40mail.gmail.com In-Reply-To: <1458785018-29232-1-git-send-email-dj.dij123@gmail.com> ``` This is my first review. It can contain some mistakes. On Thu, Mar 24, 2016 at 7:33 AM, Diwas Joshi wrote: > Subject : [PATCH/GSoC 3/3] Nousage message in error Mention about GSoC in the notes section (the one followed by the 3 dashes ie. "---") rather than in the subject. > - To show only error text instead of full usage message > - Adds exits to callback function in parse-options-cb.c instead of returning -1 which results in display of usage message. A general convention followed by git users it to write the commit message as "What he did to the code?" rather than "What problem was there in the code?" And of course after writing what you did to the code, you can definitely mention what problem in the code made you do this change. > parse-options-cb.c | 6 ++++-- > 1 file changed, 4 insertions(+), 2 deletions(-) > > diff --git a/parse-options-cb.c b/parse-options-cb.c > index 239898d..b7321d1 100644 > --- a/parse-options-cb.c > +++ b/parse-options-cb.c > @@ -85,8 +85,10 @@ int parse_opt_commits(const struct option *opt, const char *arg, int unset) > > if (!arg) > return -1; > - if (get_sha1(arg, sha1)) > - return error("malformed object name %s", arg); > + if (get_sha1(arg, sha1)) { > + error("malformed object name %s", arg); > + exit(129); > + } > commit = lookup_commit_reference(sha1); > if (!commit) > return error("no such commit %s", arg); Maybe you could describe a little more on why this change is required? Why would the user want to know "How to use the command?" when the actual problem is that SHA-1 checksum has been compromised? And I don't see any consumers of this method which *directly* interact with the UI. It seems that PATCH 1/3 and PATCH 2/3 are missing. ``` ## Pranit Bauva, 2016-03-24 07:24 Subject: Re: [PATCH/GSoC 3/3] Nousage message in error Message-ID: URL: https://gitlist.dev/e/CAFZEwPM_YBD%2Bgd0huxfi9c36imU5uG03S%2Bx6HQQrHEVwxnMXrA%40mail.gmail.com In-Reply-To: <1458785018-29232-1-git-send-email-dj.dij123@gmail.com> ``` On Thu, Mar 24, 2016 at 7:33 AM, Diwas Joshi wrote: > - To show only error text instead of full usage message > - Adds exits to callback function in parse-options-cb.c instead of returning -1 which results in display of usage message. I forgot to mention that this microproject is already picked up by someone else. Try attempting another microproject from the list. The thread for the discussion of this is : http://thread.gmane.org/gmane.comp.version-control.git/289312/focus=289336 ``` ## Junio C Hamano, 2016-03-24 16:23 Subject: Re: [PATCH/GSoC 3/3] Nousage message in error Message-ID: URL: https://gitlist.dev/e/xmqqh9fvalpm.fsf%40gitster.mtv.corp.google.com In-Reply-To: ``` Pranit Bauva writes: > A general convention followed by git users it to write the commit > message as "What he did to the code?" rather than "What problem was > there in the code?" It is OK for other projects to adopt a different convention, but The project convention here is different. We tend to write our log message like this in this order: - Explain relevant behaviour and code structure in the current code to refresh memory of readers to help them understand the next two items better. This paragraph is optional and you see it only in difficult patches. - Desribe the problem. What the end user would do and what she sees in response to it, why that is not a good outcome, and what would be a better outcome. For a patch to only improve code, replace "the end user" with "other codepaths that interact with the code being changed". - Explain the approach to implement a better outcome. This paragraph is optional and you see it only in patches that implement tricky solutions. - Describe the solution, as if you are giving orders to the maintainers to change the code this way and that way. ``` ## Pranit Bauva, 2016-03-24 16:28 Subject: Re: [PATCH/GSoC 3/3] Nousage message in error Message-ID: URL: https://gitlist.dev/e/CAFZEwPOd2RyfF5eMqB5XZqNg5Oc_fbDmERcOAMpgR81pmPOPQg%40mail.gmail.com In-Reply-To: ``` On Thu, Mar 24, 2016 at 9:53 PM, Junio C Hamano wrote: > Pranit Bauva writes: > >> A general convention followed by git users it to write the commit >> message as "What he did to the code?" rather than "What problem was >> there in the code?" > > It is OK for other projects to adopt a different convention, but The > project convention here is different. > > We tend to write our log message like this in this order: > > - Explain relevant behaviour and code structure in the current code > to refresh memory of readers to help them understand the next two > items better. This paragraph is optional and you see it only in > difficult patches. > > - Desribe the problem. What the end user would do and what she > sees in response to it, why that is not a good outcome, and what > would be a better outcome. For a patch to only improve code, > replace "the end user" with "other codepaths that interact with > the code being changed". > > - Explain the approach to implement a better outcome. This > paragraph is optional and you see it only in patches that > implement tricky solutions. > > - Describe the solution, as if you are giving orders to the > maintainers to change the code this way and that way. Thanks for making it clearer to me! I tend to write bad commit messages. Improving on it though. ```