Re: [git-users] GIT with large projects

[Wes Freeman]

What about having your long lived branches be in separate checked out
folders?

On Sun, Oct 20, 2013 at 6:39 AM, Blake McBride  wrote:

> Yes, of course, creating a totally new branch from where you are doesn't
> cause any problems because nothing changes.
>
> The problem is that, I and likely most of the world, have a handful of
> long-running branches - that already exist.  A full rebuild is out of the
> question when it takes two hours.
>
> I just took a look at GO.  I previously thought it was Google's
> replacement of GWT.  No, it's a language.  These days all that matters are
> web apps.  Anything else can be written in C/C++.  In terms of parallel
> programming, there are some really good tools for C/C++ such as silk.
>
> Take care,
>
> Blake
>
>
> On Sunday, October 20, 2013 3:18:32 AM UTC-5, Wes Freeman wrote:
>
>> When you create a branch, if you have already compiled object files, you
>> can keep them in your folder. You don't have to do a clean build on
>> creation of the branch. Then you can make your changes, do incremental
>> builds, and solve the problem/merge your code into the master.
>>
>> The problem lies in switching between branches that already exist, which
>> might force a clean build...
>>
>> As a side note, I'm really enjoying Go's ridiculously fast build times.
>>
>> Wes
>>
>>
>> On Sun, Oct 20, 2013 at 3:59 AM, Blake McBride  wrote:
>>
>>> Not sure what you mean about designed well, but in order to switch
>>> branches without having to do a full rebuild would involve:
>>>
>>> 1.  switching branches would have to auto-delete compiled modules
>>> (object files) for source files that aren't contained in the new branch in
>>> order to avoid link time collisions.  Or, your build process would have to
>>> detect left over object files from a branch switch and delete them at build
>>> time.
>>>
>>> 2.  from one branch to another there may be an include file (when using
>>> C/C++) that has a difference possibly necessitating a full rebuild.
>>>
>>> 3.  git would have to restore files using the current date and time (as
>>> opposed to their original date/time) in order for the build system to force
>>> a recompile on those modules (I checked - git does do this!)
>>>
>>> I am sure there are many other possible system-specific issues as well,
>>> i.e. many situations where switching branches would subtly necessitate
>>> a full rebuild.   They would present themselves as very hard-to-find bugs
>>> that would disappear when a full rebuild occurred.
>>>
>>> I can't imagine how any SCMS could solve problems like these.
>>>  (Although, I ask the question in case there is a solution that eludes me.)
>>>  Without a solution to problems like these, given a very large system, many
>>> of the cool features of a SCMS are not of much use.
>>>
>>> I bring all this up not to be difficult.  I read about many cool SCMS
>>> features, but I can't see how they could be useful in a very large
>>> environment that I use all the time.  I am wondering if there is a solution
>>> I am unaware of.  Thinking about it, I suppose there are some design
>>> decisions that could be employed that are driven by nothing more than an
>>> attempt to resolve SCM branch issues, but there is no way I am aware of to
>>> totally fix the fundamental problems.
>>>
>>> Thanks.
>>> Blake McBride
>>>
>>>
>>>
>>>
>>> On Saturday, October 19, 2013 5:52:00 PM UTC-5, Gergely Polonkai wrote:
>>>
>>>> Hello,
>>>>
>>>> according to your description, your project seems to be something like
>>>> the Linux kernel, and Git handles that just fine. Depending on your build
>>>> environment, Git branches may help you a lot, as, if it is designed well,
>>>> can prevent full rebuilds.
>>>>
>>>> Cheers,
>>>> Gergely
>>>> On 19 Oct 2013 23:40, "Blake McBride"  wrote:
>>>>
>>>>>  Greetings,
>>>>>
>>>>> I have a large application that takes about two hours to build.
>>>>>  Sometime I have to do partial-project commits in order to
>>>>> communicate development from one area to another (I can explain further 
>>>>> but
>>>>> it is irrelevant to the question).  I'd prefer (if I was using git rather
>>>>> than svn) to create a branch t

Re: [git-users] GIT with large projects

[Wes Freeman]

When you create a branch, if you have already compiled object files, you
can keep them in your folder. You don't have to do a clean build on
creation of the branch. Then you can make your changes, do incremental
builds, and solve the problem/merge your code into the master.

The problem lies in switching between branches that already exist, which
might force a clean build...

As a side note, I'm really enjoying Go's ridiculously fast build times.

Wes

On Sun, Oct 20, 2013 at 3:59 AM, Blake McBride  wrote:

> Not sure what you mean about designed well, but in order to switch
> branches without having to do a full rebuild would involve:
>
> 1.  switching branches would have to auto-delete compiled modules (object
> files) for source files that aren't contained in the new branch in order to
> avoid link time collisions.  Or, your build process would have to detect
> left over object files from a branch switch and delete them at build time.
>
> 2.  from one branch to another there may be an include file (when using
> C/C++) that has a difference possibly necessitating a full rebuild.
>
> 3.  git would have to restore files using the current date and time (as
> opposed to their original date/time) in order for the build system to force
> a recompile on those modules (I checked - git does do this!)
>
> I am sure there are many other possible system-specific issues as well,
> i.e. many situations where switching branches would subtly necessitate a
> full rebuild.   They would present themselves as very hard-to-find bugs
> that would disappear when a full rebuild occurred.
>
> I can't imagine how any SCMS could solve problems like these.  (Although,
> I ask the question in case there is a solution that eludes me.)  Without a
> solution to problems like these, given a very large system, many of the
> cool features of a SCMS are not of much use.
>
> I bring all this up not to be difficult.  I read about many cool SCMS
> features, but I can't see how they could be useful in a very large
> environment that I use all the time.  I am wondering if there is a solution
> I am unaware of.  Thinking about it, I suppose there are some design
> decisions that could be employed that are driven by nothing more than an
> attempt to resolve SCM branch issues, but there is no way I am aware of to
> totally fix the fundamental problems.
>
> Thanks.
> Blake McBride
>
>
>
>
> On Saturday, October 19, 2013 5:52:00 PM UTC-5, Gergely Polonkai wrote:
>
>> Hello,
>>
>> according to your description, your project seems to be something like
>> the Linux kernel, and Git handles that just fine. Depending on your build
>> environment, Git branches may help you a lot, as, if it is designed well,
>> can prevent full rebuilds.
>>
>> Cheers,
>> Gergely
>> On 19 Oct 2013 23:40, "Blake McBride"  wrote:
>>
>>>  Greetings,
>>>
>>> I have a large application that takes about two hours to build.
>>>  Sometime I have to do partial-project commits in order to communicate
>>> development from one area to another (I can explain further but it is
>>> irrelevant to the question).  I'd prefer (if I was using git rather than
>>> svn) to create a branch to commit the partial work, debug, and finally
>>> merge to the master when it is all done.  This way each commit on 
>>> masterwould be stable (contain no partial commits).  My problem is this.  
>>> Given
>>> the size of the project, I can't checkout different versions without
>>> causing a two hour build.  I am sure this must be a common problem.
>>>
>>> Stated another way - when you have a very large project, branching
>>> becomes a significant problem because of build times.  Are there common
>>> solutions to this sort of problem?
>>>
>>> Thanks.
>>>
>>> Blake McBride
>>>
>>>  --
>>> You received this message because you are subscribed to the Google
>>> Groups "Git for human beings" group.
>>> To unsubscribe from this group and stop receiving emails from it, send
>>> an email to git-users+...@**googlegroups.com.
>>>
>>> For more options, visit 
>>> https://groups.google.com/**groups/opt_out
>>> .
>>>
>>  --
> You received this message because you are subscribed to the Google Groups
> "Git for human beings" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to git-users+unsubscr...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.
>

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

Re: [git-users] error changes not staged for commit (windows)

[Wes Freeman]

Is it in your .gitignore file?

Wes

On Mon, Nov 26, 2012 at 9:34 AM, git-guy  wrote:

> i had tried git add .
>
> git add  path file
>
> but still no add the files, some help what happening? should add with git add 
> .  but i dont know what happening
>
>
> #git status
> # On branch master
> #Your branch is ahead of 'origin/master' by 1 commit
>
> # Changes not staged for commit:
> #   (use "git add ..." to update what will be committed)
> #   (use "git checkout -- ..." to discard changes in working directory)
> #
> #   deleted:   templates/file.txt
> #   deleted:   templates/test.jsp
> #
> no changes added to commit (use "git add" and/or "git commit -a")
>
>  --
>
>
>

-- 

Re: [git-users] git clone fails with "unable to overwrite old ref-pack file"

[Wes Freeman]

I agree, windows permissions are annoying. Actually it might be a
permission within the git_src folder getting copied down. Have you looked
at that?

Wes

On Tue, Aug 7, 2012 at 9:43 PM, Lisa Zorn  wrote:

> Thanks, I tried making sure I set things to not RO but I'll try more
> (windows permissions are confusing, I am never sure if I'm succeeding with
> things using the UI).
> The other thing is that before this, we git clone many other things into dir2
> and they all work fine, but this one fails.
> And it's re-producable; if I delete dir2 and start over, the same thing
> happens on the same git_src (ok clearly that's Muni_GenevaBRT_**
> HuntersPtToCandlestickIntercha**ngeOnly but I was trying to simplify my
> email) after succeeding with other git clones before that).
> -Lisa
>
>
> On Tuesday, August 7, 2012 6:37:34 PM UTC-7, Wes Freeman wrote:
>
>> Check your permissions on dir1, and make sure it's not set as read-only
>> or something.
>>
>> Wes
>>
>> On Tue, Aug 7, 2012 at 9:24 PM, Lisa Zorn  wrote:
>>
>>> Hello,
>>> I have been using git for years and have not run into this problem
>>> before.  I tried to git clone a repo that's on a local drive, and I get
>>> this error:
>>> dir1\dir2>git clone -b master Y:\git_src
>>> Cloning into 'Muni_GenevaBRT_**HuntersPtToCandlestickIntercha**
>>> ngeOnly'...
>>> done.
>>> fatal: unable to overwrite old ref-pack file: Permission denied
>>> fatal: The remote end hung up unexpectedly
>>>
>>> However, when I run it in another directory (not a subdir of dir1), it
>>> works fine.  But it's repeatable; if I run it from dir1\dirX it happens
>>> again.
>>> I am using the stable version of git, and I'm on windows.  What's going
>>> on with the "old ref-pack file"?
>>>
>>> Thanks!
>>> -Lisa
>>>
>>> --
>>> You received this message because you are subscribed to the Google
>>> Groups "Git for human beings" group.
>>> To view this discussion on the web visit https://groups.google.com/d/**
>>> msg/git-users/-/plrGfYp4gNkJ<https://groups.google.com/d/msg/git-users/-/plrGfYp4gNkJ>
>>> .
>>> To post to this group, send email to git-...@googlegroups.com.
>>> To unsubscribe from this group, send email to git-users+...@**
>>> googlegroups.com.
>>>
>>> For more options, visit this group at http://groups.google.com/**
>>> group/git-users?hl=en <http://groups.google.com/group/git-users?hl=en>.
>>>
>>
>>  --
> You received this message because you are subscribed to the Google Groups
> "Git for human beings" group.
> To view this discussion on the web visit
> https://groups.google.com/d/msg/git-users/-/MCq1e6ocdQMJ.
>
> To post to this group, send email to git-users@googlegroups.com.
> To unsubscribe from this group, send email to
> git-users+unsubscr...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/git-users?hl=en.
>

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To post to this group, send email to git-users@googlegroups.com.
To unsubscribe from this group, send email to 
git-users+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/git-users?hl=en.

Re: [git-users] git clone fails with "unable to overwrite old ref-pack file"

[Wes Freeman]

Check your permissions on dir1, and make sure it's not set as read-only or
something.

Wes

On Tue, Aug 7, 2012 at 9:24 PM, Lisa Zorn  wrote:

> Hello,
> I have been using git for years and have not run into this problem before.
>  I tried to git clone a repo that's on a local drive, and I get this error:
> dir1\dir2>git clone -b master Y:\git_src
> Cloning into 'Muni_GenevaBRT_HuntersPtToCandlestickInterchangeOnly'...
> done.
> fatal: unable to overwrite old ref-pack file: Permission denied
> fatal: The remote end hung up unexpectedly
>
> However, when I run it in another directory (not a subdir of dir1), it
> works fine.  But it's repeatable; if I run it from dir1\dirX it happens
> again.
> I am using the stable version of git, and I'm on windows.  What's going on
> with the "old ref-pack file"?
>
> Thanks!
> -Lisa
>
> --
> You received this message because you are subscribed to the Google Groups
> "Git for human beings" group.
> To view this discussion on the web visit
> https://groups.google.com/d/msg/git-users/-/plrGfYp4gNkJ.
> To post to this group, send email to git-users@googlegroups.com.
> To unsubscribe from this group, send email to
> git-users+unsubscr...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/git-users?hl=en.
>

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To post to this group, send email to git-users@googlegroups.com.
To unsubscribe from this group, send email to 
git-users+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/git-users?hl=en.

Re: [git-users] Beginner stuck in a commit

[Wes Freeman]

The "strange editing mode" is called vi. Check out this cheatsheet:
http://www.lagmonster.org/docs/vi.html

To edit, type "i" to switch to insert mode. To leave insert mode, hit ESC.
To save and quit, type :wq, which means "write, quit".

Good luck!

Wes

On Mon, Aug 6, 2012 at 10:57 PM, Jeffery Brewer wrote:

> I've slowly been trying to get git to work and just running into loads of
> problems.
>
> Using the windows bash I just tried to do a commit this evening and forgot
> to add a message (e.g. -m "my work for today") and sent the bash into some
> sort of odd editing mode that I can't seem to get out of. I finally just
> closed the bash and opened a new bash and tried to commit and got all kinds
> of error messages with a prompt to type (R) to recover. Typed R to recover
> and it took me right back into the strange editing mode that I can't seem
> to get out of now. I've backed up all the files in the directory (sensing
> an impending catastrophe) but not sure what else to do at this point to get
> git running again.
>
> Any help would be appreciated.
>
> More detail...
>
> If I open a new bash in the directory and run "git commit" I'm getting
> this error message:
>
> E325: ATTENTION
> Found a swap file by the name ".git\.COMMIT_EDITMSG.swp"
>  dated: Mon Aug 06 19:45:14 2012
>  file name:
> C:/Users/me/Documents/NetBeansProjects/foldername/.git/COMMIT_EDITMSG
>   modified: YES
>  user name: me   host name: my computer
> process ID: 10368
> While opening file ".git\COMMIT_EDITMSG"
>  dated: Mon Aug 06 19:55:29 2012
>   NEWER than swap file!
> (1) Another program may be editing the same file.
> If this is the case, be careful not to end up with two
> different instances of the same file when making changes.
> Quit, or continue with caution.
> (2) An edit session for this file crashed.
> If this is the case, use ":recover" or "vim -r .git\COMMIT_EDITMSG"
> to recover the changes (see ":help recovery").
> If you did this already, delete the swap file
> ".git\.COMMIT_EDITMSG.swp"
> to avoid this message.
> Swap file ".git\.COMMIT_EDITMSG.swp" already exists!
> -- More -- SPACE/d/j: screen/page/line down, b/u/k: up, q: quit
>
> --
> You received this message because you are subscribed to the Google Groups
> "Git for human beings" group.
> To view this discussion on the web visit
> https://groups.google.com/d/msg/git-users/-/XE0-ivLyDtAJ.
> To post to this group, send email to git-users@googlegroups.com.
> To unsubscribe from this group, send email to
> git-users+unsubscr...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/git-users?hl=en.
>

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To post to this group, send email to git-users@googlegroups.com.
To unsubscribe from this group, send email to 
git-users+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/git-users?hl=en.

Re: [git-users] git server problem

[Wes Freeman]

This happens sometimes if you have different default permissions on the
files, I've found. It seemed to happen going between a mac and a windows
client, for example. This might not be the same problem you're seeing, but
I figured I would chime in, just in case it is of use.

You can ignore changes like this on the client side with:

 git config core.filemode false

Wes

On Mon, Jul 2, 2012 at 2:15 PM, PJ Weisberg  wrote:

> On Monday, July 2, 2012, Avner Tamir  wrote:
> > ,Hi
> >
> >
> >
> > I’ve set git server on RHEL 5.6 box, running on httpd
> >
> > we’ve encountered the following scenario several times
> >
> >  User A pushes to remote
> >
> >User B pull from remote
> >
> >  User B sees the sources pushed by user A as to be committed in his
> local repository
>
> That sounds like a reasonable description of what "pull" does. What did
> you expect/want to see?
>
> --
> -PJ
>
> Gehm's Corollary to Clark's Law: Any technology distinguishable from
> magic is insufficiently advanced.
>
>  --
> You received this message because you are subscribed to the Google Groups
> "Git for human beings" group.
> To post to this group, send email to git-users@googlegroups.com.
> To unsubscribe from this group, send email to
> git-users+unsubscr...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/git-users?hl=en.
>

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To post to this group, send email to git-users@googlegroups.com.
To unsubscribe from this group, send email to 
git-users+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/git-users?hl=en.

Re: [git-users] Git gone :(

[Wes Freeman]

For the record, I think your question was perfectly reasonable for this
forum. It is related to git installation/configuration on a particular OS.
This group doesn't get much traffic, as it is--if we tell people to go
elsewhere to solve their problems, it will just continue to be a ghost town.

Here are some more general pointers (hopefully for the benefit of all):
It's usually easier to configure software in CentOS via yum, the package
manager. CentOS 5 doesn't have a yum package for git (CentOS 5 is pretty
old, you should consider upgrading to 6 at some point!), but there are
people who have kindly set up git packages for CentOS 5, such as:
http://www.webtatic.com/news/2011/09/latest-updates-git-1-7-6-1-httpd-2-2-20-php-5-3-8-nginx-1-0-6/

There are a couple of great books online for git specifically:
http://book.git-scm.com/

http://book.git-scm.com/2_installing_git.html
(this contains the proper "from source" installation commands with the
prefix:
$ make prefix=/usr all ;# as yourself
$ make prefix=/usr install ;# as root)

And there's "pro git": http://progit.org/book/ , written by one of the guys
at github. I learned a lot reading through the branching sections.

You should definitely set up a user account to push with via SSH. It's
possible this server isn't in the wild, which is fine, but if it is
accessible by the internet, I highly recommend disabling root ssh access
all together. See the section on disabling root SSH access in CentOS 5
here:
http://www.centos.org/docs/5/html/5.1/Deployment_Guide/s2-wstation-privileges-noroot.html

Happy git-ing.
Wes

On Tue, Apr 10, 2012 at 11:17 AM, Paul Hollyer  wrote:

>
>
> On Tuesday, April 10, 2012 3:21:04 PM UTC+1, Konstantin Khomoutov wrote:
>>
>> On Tue, 10 Apr 2012 06:46:12 -0700 (PDT)
>> Paul Hollyer  wrote:
>>
>> > > env $PATH
>> > > to see what your path looks like.
>> [...]
>> > Thanks for the quick replies, here is the result of env $PATH:
>> >
>> > env:
>> > /root/local/bin:/usr/kerberos/**sbin:/usr/kerberos/bin:/usr/**
>> local/sbin:/usr/bin:/bin:/usr/**>
>> > sbin:/sbin:/usr/local/rvm/bin:**/root/bin: No such file or directory
>> >
>> > I assumed re-installing git would make it add the required paths
>> > where necessary again if they had been broken.
>> Installing something by `make install` never touches the environment
>> (and for good).
>>
>> > Can you help me fix this?
>> >
>> > Git is at /usr/local/src I think,
>> Very improbable.
>> Possibly you unrolled the Git source tarball there
>>
> I did.
>
>
>> , but when you did
>> ./configure
>> it probably picked /usr/local as its "installation prefix",
>> so supposedly you do have main Git binary under /usr/local/bin, and
>> that's what you're supposed to add to your PATH.
>>
>> > so how do I add this to the search path?
>> The answer is "it depends".
>>
>> The most correct approach (to me) is to find binary package(s) for Git
>> matching your OS (and its version), and install them.  This gives you
>> two benefits:
>> 1) Git will be installed in a well-known place, where it will be
>>accessible without messing with the user's environment.
>> 2) You will have less problems when uninstalling or upgrading Git later,
>>as this will be routinely carried out by the package manager.
>>
>> If you insist on building Git from the source, there are two ways:
>> 1) Run `./configure --help` and see what it needs to install Git into
>>the standard hierarchy, that is, under /usr.
>>Usually this is --prefix=/usr
>> 2) Install as-is, but notice where the files are installed.
>>Just run `make install >/tmp/log` and inspect the generated log file
>>to see where the files have been installed.
>>
> This is what I did, following a recipe provided by someone using Git and
> Centos. I didn't think, this morning, when running make install, to pipe or
> > (append?) the output to a log file. These are the sort of common
> techniques that are yet to come naturally for me.
>
>
>
>> In the latter case you might have to fix the environment *of the user
>> under which account Git will run.*  This is important: from the output
>> you shown us, it follows that you run this command as root.  It's
>> highly unlikely you're pushing as root as well (if you're a sensible
>> person at least),
>>
> I'm not a sensible person, I need to fix this.
>
>
>>  and you have to tweak either the global environment
>> or the environment of the mentioned user.
>> How to do that depends on the shell which is spawned by SSH on the
>> server when you push.  Usually it's the same as the user's login shell
>> (the one used for interactive logins).
>> The config file to tweak depends on the shell.
>> For bash this will be
>> ~/.bashrc
>> and you'll have to add to it something like
>> export PATH="$PATH:/usr/local/bin"
>>
>> > Sorry if it's a basic linux question, but am learning by doing, and
>> > fixing...
>>
>> As you can see, you'll have to get a book on Unix and read it as the
>> question being discussed has little to do

Re: [git-users] Git gone :(

[Wes Freeman]

Yeah, that will probably fix it. The longer answer is: you probably
shouldn't be running git as root (especially over ssh).

Wes

On Tue, Apr 10, 2012 at 10:06 AM, Michael P. Soulier <
msoul...@digitaltorque.ca> wrote:

> On 10/04/12 Paul Hollyer said:
>
> > Thanks for the quick replies, here is the result of env $PATH:
> >
> > env:
> >
> /root/local/bin:/usr/kerberos/sbin:/usr/kerberos/bin:/usr/local/sbin:/usr/bin:/bin:/usr/sbin:/sbin:/usr/local/rvm/bin:/root/bin:
> > No such file or directory
> >
> > I assumed re-installing git would make it add the required paths where
> > necessary again if they had been broken.
> >
> > Can you help me fix this?
> >
> > Git is at /usr/local/src I think, so how do I add this to the search
> path?
> > Sorry if it's a basic linux question, but am learning by doing, and
> > fixing...
>
> If you're using bash, look at some basic bash documentation online.
>
> Short answer
>
> export PATH=$PATH:/usr/local/bin (I doubt it's src)
>
> Mike
>
> --
> You received this message because you are subscribed to the Google Groups
> "Git for human beings" group.
> To post to this group, send email to git-users@googlegroups.com.
> To unsubscribe from this group, send email to
> git-users+unsubscr...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/git-users?hl=en.
>
>

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To post to this group, send email to git-users@googlegroups.com.
To unsubscribe from this group, send email to 
git-users+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/git-users?hl=en.

Re: [git-users] Git gone :(

[Wes Freeman]

Yes, do:
env $PATH
to see what your path looks like.

It's possible that your .bash_profile was broken or something.

Wes

On Tue, Apr 10, 2012 at 8:43 AM, Michael P. Soulier <
msoul...@digitaltorque.ca> wrote:

> On 10/04/12 Paul Hollyer said:
>
> > Hi,
> >
> > I have a Centos5 server, with a number of repos set up to upload work
> to. I
> > have come in to work this morning, having pushed updates to the server a
> > few days ago, Easter Friday, and not touched the server since.
> >
> > When I tried to push work this morning, I got an error saying:
> >
> > git-receive-pack: command not found
>
> Have you checked your search path?
>
> Mike
>
> --
> You received this message because you are subscribed to the Google Groups
> "Git for human beings" group.
> To post to this group, send email to git-users@googlegroups.com.
> To unsubscribe from this group, send email to
> git-users+unsubscr...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/git-users?hl=en.
>
>

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To post to this group, send email to git-users@googlegroups.com.
To unsubscribe from this group, send email to 
git-users+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/git-users?hl=en.

Re: [git-users] How to merge two independent projects one commit by one commit?

[Wes Freeman]

I might be wrong, so be careful (you should make a copy of both repos
before you start, just in case), but I think something like this might
work:

1. Add your repository as a remote to the new repository.
2. Pull your changes in to the new repository.
3. From the new repository, git cherry-pick each commit.

Wes

On Fri, Oct 14, 2011 at 3:44 PM, www  wrote:
> Dear All:
>
> I am new to git and I am using linux.
>
> I am working on my project, using git and I have made about 20
> commits. All my files are at the directory mine/ Inside mine, there
> is .git directory. Now, I received the tar files from my partner. He
> just finished his part of work and we are working on the same set of
> files. He does not use version control. I put all his files in the
> parallel directory his/. I did the following:
>
> cd his
> git init
> git add   //all the stuff
> git commit . -m "received from partner and he has finished his part"
>
> Ideally, I want to step by step merge my revision one, my revision
> two, my revision three, etc  from mine/ into his/
>
> It is very likely that there will be conflicts since we are touching
> same files. So I want to do merging step by step. I hope to view each
> revision modification before committing.
>
> How can I do it? I have googled and studied the whole day and my brain
> hurts.
>
> Thank you very, very much.
>
> --
> You received this message because you are subscribed to the Google Groups 
> "Git for human beings" group.
> To post to this group, send email to git-users@googlegroups.com.
> To unsubscribe from this group, send email to 
> git-users+unsubscr...@googlegroups.com.
> For more options, visit this group at 
> http://groups.google.com/group/git-users?hl=en.
>
>

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To post to this group, send email to git-users@googlegroups.com.
To unsubscribe from this group, send email to 
git-users+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/git-users?hl=en.