On 23 January 2012 04:02, Shannon Whitty <shannon.whitty_at_gmail.com> wrote:
>>>> The changes I am making to my working files are not updating the
> >>> repository.
> >
> >>You are going to have to describe things a little clearer than this.
> >
> >>> My original files are stored in /projects/pr01
> >>> I created a repository for these files /repo/pr01
> >
> >>And you imported the files from /projects/pr01?
> >>Is /projects/pr01 a working copy or just the original files?
> >
> > As the second line clearly states /projects/pr01 are the original files
>
> But you didn't clearly state how you got them into the repository. The
> manual describes 2 different methods:
> http://tortoisesvn.net/docs/release/TortoiseSVN_en/tsvn-dug- import.html
>
> Plain import copies the files into the repository but your original
> files are still just that - original files. That folder does not
> become a working copy and so there is no link with the repository. The
> way we recommend in the manual is what is referred to as "Import in
> place" which is a 2-step process, but it makes your original folder
> version-controlled. If you only read the quick start section then you
> may have missed this, so perhaps there should be a cross reference
> from there.
>
> >>> I checked out the repository to /work/pr01
> >>> I made changes to /work/pr01/file01
> >>> I then comitted the file (/work/pr01/file01)
> >
> >>Is /projects/pr01 a working copy? Did you update it?
> >
> > /work/pr01 is the working copy, where I checked the files to as per the
> > instructions in the setup guide?
> >
> >>> I then uploaded the changed file from /projects/pr01/file0
> > 1 to the file
> >>> server and the changes were not present.
> >
> >>What do you mean by 'uploaded'? There is no such subversion command.
> >>What do you mean by the file server? Are you just talking about a
> >>plain file copy?
> >
> > I'm talking about my web server - I use an FTP client to update files
> from
> > my local PC to the web server - what's so hard to understand about that?
>
> It's not hard to understand, but you have to understand that we get a
> lot of people on this list describing things very imprecisely,
> referring to a working copy as a repository, a subversion server as a
> file server, etc. Since you were speaking about uploading a changed
> file which was not changed I had to ask for clarification.
>
> >>> It appears that when I commit changes from my working directory to the
> >>> repository, the original files are not updated
> >
> >>Why would they be unless it is a working copy that you updated? I
> >>think you need to read the introductory parts of the manual to find
> >>out how subversion works.
>
> > How else could i update a central copy so that when others try to
> access it
> > they get the latest copy? I did read the introductory parts and followed
> > them exactly... That's how I got to this point... Thanks for being so
> > condescending and helpful though!!!
>
> Sorry if that is how it has come across, that was certainly not my
> intention. But think about it. You copied files from a directory into
> a repository. You create a separate working copy, change it and commit
> back to the repository. How would the subversion server push them back
> out to your original folder without you asking? In your case you are
> using file:// access, but that is only recommended for testing and
> maybe sole developer use. In the general case the server will be a
> different computer, possibly on another continent. How would a remote
> server gain access to your PC to push changes back into your original
> file set? And if another developer committed changes, how would that
> server push those files out to your PC which might not be turned on,
> or might not even exist any more?
>
> > The manual said to create a repository from original files - which I did
> > It then said to check out to a local working copy - which I did
> > I then modified my local working copy
> > I then committed changes back to the repository - exactly as the
> > documentation suggested
>
> Yes, and thus far you have done everything right. No problems there.
> The only issue is that changes to a repository do not get pushed out
> to other places. AFAIK no version control system does that. Either you
> have to have a working copy which you update to pull the latest
> version from the server, or you export the files from the server to a
> folder, creating an unversioned snapshot of the files as they were at
> a particular repository revision.
>
> > If everything was working as expected I wouldn't be in here asking a
> > question - If I knew what I was doing I wouldn't need help!!! If
> providing
> > assistance is such a chore, why bother coming here? Or is it just to make
> > yourself feel smarter than everyone else because you learnt how to
> use this
> > software before I did... Is that how this place works - newbies aren't
> > welcome because they're not as smart as you?
>
> Again, sorry if I came across like that, I didn't think I had asked
> for any more than clarification, although maybe my last question could
> have been less abrupt. But there is really no need to retaliate with
> accusations like that. Newbies are welcome here, but in order to
> answer the question we very often have to ask for more information
> because we simply don't understand exactly what you have done or what
> you expect to happen.
>
> Simon
> -----------------
> I apologise for my tone. Thanks for your help.
>
> What I don't understand is what purpose the original files serve now?
> All it appears I have done is copied them to another folder for editing.
>
> My perception based on the instructions on this page, which is what I
> followed:-
> http://tortoisesvn.net/docs/release/TortoiseSVN_en/tsvn-qs-guide.html
> was that the original files where the "base" which would potentially be
> on a file server (original files) and I could check those out to my
> local PC (working copy) and all changes to the working copy would be
> pushed back to the base on "commit" for other users to pick up the
> latest changes...
I see where you are coming from now, but no, that's not correct. The
repository is the one true master and changes committed from a working
copy get pushed up to the repository. You can make as many working
copies as you like, and they will receive the latest version when they
are updated. It's a pull from the repository into the working copy,
there is no push operation.
>
> Are the original files now of no purpose?
As far as Subversion is concerned, no. They are still just plain files
with no link to the repository.
Subversion only operates on a working copy. Subversion's import
command is perhaps not the most well thought-out command and of
limited use. I think even the subversion developers would admit that.
Which is why the manual described the import-in-place method
(admittedly not in the part you were looking at). The manual section I
referenced describes this in a bit more detail.
Maybe using import in the quick start was a mistake too. I didn't want
to put the import-in-place method in the QS because it is a bit more
involved, but OTOH if you only read the QS with the intention of
ploughing through the rest of the manual only when you need to then I
guess you're likely to get off on the wrong foot. I'll see what we can
do to clarify this without making it overly complicated.
There is an open issue for implementing the import-in-place as a
separate dialog so you don't have to think about the 2 stages, but it
hasn't happened yet.
Simon
--
: ___
: oo // \\ "De Chelonian Mobile"
: (_,\/ \_/ \ TortoiseSVN
: \ \_/_\_/> The coolest Interface to (Sub)Version Control
: /_/ \_\ http://tortoisesvn.net
------------------------------------------------------
http://tortoisesvn.tigris.org/ds/viewMessage.do?dsForumId=4061&dsMessageId=2910501
To unsubscribe from this discussion, e-mail: [users-unsubscribe_at_tortoisesvn.tigris.org].
Received on 2012-01-23 10:47:10 CET