search

jbrukh / ggit

ggit
17 stars 2 forks source link

ggcompare fails on libgit2 git repo with incorrect date. #40

Closed jbrukh closed 11 years ago

jbrukh commented 11 years ago

found a mismatch in object 725 (23f8588dde934e8f33c263c6d8359b2ae095f863)... GIT ----------------------- object c5b97d5ae6c19d5c5df71a34c7fbeeda2479ccbc type commit tag v0.1.0 tagger Scott Chacon schacon@gmail.com Tue Nov 23 15:53:50 2010 -0800

tagging initial release of libgit2

GGIT ----------------------- object c5b97d5ae6c19d5c5df71a34c7fbeeda2479ccbc type commit tag v0.1.0 tagger Scott Chacon schacon@gmail.com Tue Nov 23 18:53:50 2010 -0800

tagging initial release of libgit2

diff ----------------------- object c5b97d5ae6c19d5c5df71a34c7fbeeda2479ccbc type commit tag v0.1.0 tagger Scott Chacon schacon@gmail.com Tue Nov 23 1 theirs: 5:53:50 2010 -0800 ours: 8:53:50 2010 -0800

MikeBosw commented 11 years ago

Yeah, I saw this too - I'm using git://github.com/twitter/bootstrap.git to reproduce it.

We should consider switching from cat-file -p to cat-file for ggcompare. The former is a "porcelain" command that isn't intended for API use from what I understand. I've started some work to this end.

On Thu, Nov 1, 2012 at 10:59 AM, jbrukh notifications@github.com wrote:

found a mismatch in object 725 (23f8588https://github.com/jbrukh/ggit/commit/23f8588dde934e8f33c263c6d8359b2ae095f863 )... GIT ----------------------- object c5b97d5https://github.com/jbrukh/ggit/commit/c5b97d5ae6c19d5c5df71a34c7fbeeda2479ccbc type commit tag v0.1.0 tagger Scott Chacon schacon@gmail.com Tue Nov 23 15:53:50 2010 -0800

tagging initial release of libgit2

GGIT ----------------------- object c5b97d5https://github.com/jbrukh/ggit/commit/c5b97d5ae6c19d5c5df71a34c7fbeeda2479ccbc type commit tag v0.1.0 tagger Scott Chacon schacon@gmail.com Tue Nov 23 18:53:50 2010 -0800

tagging initial release of libgit2

diff ----------------------- object c5b97d5https://github.com/jbrukh/ggit/commit/c5b97d5ae6c19d5c5df71a34c7fbeeda2479ccbc type commit tag v0.1.0 tagger Scott Chacon schacon@gmail.com Tue Nov 23 1 theirs: 5:53:50 2010 -0800 ours: 8:53:50 2010 -0800

— Reply to this email directly or view it on GitHubhttps://github.com/jbrukh/ggit/issues/40.

jbrukh commented 11 years ago

Still, what could explain the 3 hour difference? Is it some sort of git config setting that renders the time in a different zone or something? Or, are we calculating incorrectly?

On Thu, Nov 1, 2012 at 11:05 AM, MikeBosw notifications@github.com wrote:

Yeah, I saw this too - I'm using git://github.com/twitter/bootstrap.gitto reproduce it.

We should consider switching from cat-file -p to cat-file for ggcompare. The former is a "porcelain" command that isn't intended for API use from what I understand. I've started some work to this end.

On Thu, Nov 1, 2012 at 10:59 AM, jbrukh notifications@github.com wrote:

found a mismatch in object 725 (23f8588< https://github.com/jbrukh/ggit/commit/23f8588dde934e8f33c263c6d8359b2ae095f863>

)... GIT ----------------------- object c5b97d5< https://github.com/jbrukh/ggit/commit/c5b97d5ae6c19d5c5df71a34c7fbeeda2479ccbc>

type commit tag v0.1.0 tagger Scott Chacon schacon@gmail.com Tue Nov 23 15:53:50 2010 -0800

tagging initial release of libgit2

GGIT ----------------------- object c5b97d5< https://github.com/jbrukh/ggit/commit/c5b97d5ae6c19d5c5df71a34c7fbeeda2479ccbc>

type commit tag v0.1.0 tagger Scott Chacon schacon@gmail.com Tue Nov 23 18:53:50 2010 -0800

tagging initial release of libgit2

diff ----------------------- object c5b97d5< https://github.com/jbrukh/ggit/commit/c5b97d5ae6c19d5c5df71a34c7fbeeda2479ccbc>

type commit tag v0.1.0 tagger Scott Chacon schacon@gmail.com Tue Nov 23 1 theirs: 5:53:50 2010 -0800 ours: 8:53:50 2010 -0800

— Reply to this email directly or view it on GitHub< https://github.com/jbrukh/ggit/issues/40>.

— Reply to this email directly or view it on GitHubhttps://github.com/jbrukh/ggit/issues/40#issuecomment-9983144.

MikeBosw commented 11 years ago

Yeah. Not sure. I'm guessing the actual seconds are the same, but haven't confirmed.

On Thu, Nov 1, 2012 at 11:10 AM, jbrukh notifications@github.com wrote:

Still, what could explain the 3 hour difference? Is it some sort of git config setting that renders the time in a different zone or something? Or, are we calculating incorrectly?

On Thu, Nov 1, 2012 at 11:05 AM, MikeBosw notifications@github.com wrote:

Yeah, I saw this too - I'm using git:// github.com/twitter/bootstrap.gitto reproduce it.

We should consider switching from cat-file -p to cat-file for ggcompare. The former is a "porcelain" command that isn't intended for API use from what I understand. I've started some work to this end.

On Thu, Nov 1, 2012 at 10:59 AM, jbrukh notifications@github.com wrote:

found a mismatch in object 725 (23f8588<

https://github.com/jbrukh/ggit/commit/23f8588dde934e8f33c263c6d8359b2ae095f863>

)... GIT ----------------------- object c5b97d5<

https://github.com/jbrukh/ggit/commit/c5b97d5ae6c19d5c5df71a34c7fbeeda2479ccbc>

type commit tag v0.1.0 tagger Scott Chacon schacon@gmail.com Tue Nov 23 15:53:50 2010 -0800

tagging initial release of libgit2

GGIT ----------------------- object c5b97d5<

https://github.com/jbrukh/ggit/commit/c5b97d5ae6c19d5c5df71a34c7fbeeda2479ccbc>

type commit tag v0.1.0 tagger Scott Chacon schacon@gmail.com Tue Nov 23 18:53:50 2010 -0800

tagging initial release of libgit2

diff ----------------------- object c5b97d5<

https://github.com/jbrukh/ggit/commit/c5b97d5ae6c19d5c5df71a34c7fbeeda2479ccbc>

type commit tag v0.1.0 tagger Scott Chacon schacon@gmail.com Tue Nov 23 1 theirs: 5:53:50 2010 -0800 ours: 8:53:50 2010 -0800

— Reply to this email directly or view it on GitHub< https://github.com/jbrukh/ggit/issues/40>.

— Reply to this email directly or view it on GitHub< https://github.com/jbrukh/ggit/issues/40#issuecomment-9983144>.

— Reply to this email directly or view it on GitHubhttps://github.com/jbrukh/ggit/issues/40#issuecomment-9983362.

jbrukh commented 11 years ago

Yeah, I believe the seconds should be the same also. Otherwise, that would imply some sort of parsing bug of numbers, which is highly unlikely.

On Thu, Nov 1, 2012 at 11:12 AM, MikeBosw notifications@github.com wrote:

Yeah. Not sure. I'm guessing the actual seconds are the same, but haven't confirmed.

On Thu, Nov 1, 2012 at 11:10 AM, jbrukh notifications@github.com wrote:

Still, what could explain the 3 hour difference? Is it some sort of git config setting that renders the time in a different zone or something? Or, are we calculating incorrectly?

On Thu, Nov 1, 2012 at 11:05 AM, MikeBosw notifications@github.com wrote:

Yeah, I saw this too - I'm using git:// github.com/twitter/bootstrap.gitto reproduce it.

We should consider switching from cat-file -p to cat-file for ggcompare. The former is a "porcelain" command that isn't intended for API use from what I understand. I've started some work to this end.

On Thu, Nov 1, 2012 at 10:59 AM, jbrukh notifications@github.com wrote:

found a mismatch in object 725 (23f8588<

https://github.com/jbrukh/ggit/commit/23f8588dde934e8f33c263c6d8359b2ae095f863>

)... GIT ----------------------- object c5b97d5<

https://github.com/jbrukh/ggit/commit/c5b97d5ae6c19d5c5df71a34c7fbeeda2479ccbc>

type commit tag v0.1.0 tagger Scott Chacon schacon@gmail.com Tue Nov 23 15:53:50 2010 -0800

tagging initial release of libgit2

GGIT ----------------------- object c5b97d5<

https://github.com/jbrukh/ggit/commit/c5b97d5ae6c19d5c5df71a34c7fbeeda2479ccbc>

type commit tag v0.1.0 tagger Scott Chacon schacon@gmail.com Tue Nov 23 18:53:50 2010 -0800

tagging initial release of libgit2

diff ----------------------- object c5b97d5<

https://github.com/jbrukh/ggit/commit/c5b97d5ae6c19d5c5df71a34c7fbeeda2479ccbc>

type commit tag v0.1.0 tagger Scott Chacon schacon@gmail.com Tue Nov 23 1 theirs: 5:53:50 2010 -0800 ours: 8:53:50 2010 -0800

— Reply to this email directly or view it on GitHub< https://github.com/jbrukh/ggit/issues/40>.

— Reply to this email directly or view it on GitHub< https://github.com/jbrukh/ggit/issues/40#issuecomment-9983144>.

— Reply to this email directly or view it on GitHub< https://github.com/jbrukh/ggit/issues/40#issuecomment-9983362>.

— Reply to this email directly or view it on GitHubhttps://github.com/jbrukh/ggit/issues/40#issuecomment-9983427.

jbrukh commented 11 years ago

Since your change with cat-file, ggcompare passes on that commit.