Different SVN repository URLs for updating and committing?
I have access to an SVN repository that allows public reading via one URL, and private committing via another. Is there any way to use the corresponding URLs locally? I don't want to save or enter credentials just for updating, but I'm happy to type them in when committing.
I'm using TortoiseSVN if it matters, in case there's a specific solution for that.
(Git can do this via having different "Push URLs" from the normal URL.)
See also questions close to this topic
Using tortoiseSVN in parallel with native SVN
I have a checkout of a directory on a linux file system. The directory is mapped to my windows machine, and the checkout was setup using tortoise.
I'd like to be able to access this directory from a linux machine, and also do SVN things with it, from the command line.
I guess the first step is to tell SVN how this directory (which it thinks is unversioned) is related to the repository. My initial attempt was to use svn switch to point it at the repository, but this failed because the directory is unversioned.
Is there a way to have this single directory version controlled by both tortoise (through a mapped drive) and native SVN?
Prevent svn from outputtting unnecessary diff lines
For one particular source file, I generated a reference patch file using
TortoiseMerge. The patch file shows only lines where an actual change occurs. For example:
@@ -87 +87 @@ - Particles(I)%Mass + MassArray(I) @@ -91 +91 @@ - Particles(I)%Mass + MassArray(I)
However when I use the command line
diffoption this is what I get
svn diff --old FileName.FOR@6382 --new FileName.FOR@6383
@@ -84,11 +84,11 @@ end if TotalParticleMass = & TotalParticleMass + & - Particles(I)%Mass + MassArray(I) if (IsVirtualParticle(I)) then TotalVirtualParticleMass = & TotalVirtualParticleMass + & - Particles(I)%Mass + MassArray(I) end if end do write(TVMunit, '(I12, G12.4, G12.4, I12, I12, I12)') &
As you can see the file generated by the command line also includes lines which are not different. I really don't need this information for my current purposes.
How can I just get the modified lines using the command line? I need something identical to the patch file.
Code first w/existing database vs Database first and concurrent development with version control
I have been using EF6.x for quite some time and am very comfortable with it. However, as with most projects involving a team you have to stop every so often to "fix" something in the main branch (master branch/trunk etc) due to a bug or production change of some sort. If your code involves an Entity data model and your make changes related to a DB schema change, things can get messy if you're making these changes in trunk and trying to do an catch-up merge back into a development branch that has already been started. This has been my experience with using the DB-first approach, which includes the EF designer diagram.
With code-first, if a DB-schema change is to be handled in your main branch (production fix etc), then it seems my options are:
(1) make the updates to the EF-related classes myself and possibly use EF migrations, or...
(2) run through the EF-model reverse engineer process again, overwriting whats there. If you keep the model name the same you will first have to delete the existing files, else the wizard will complain.
With database-first, it seems im limited on doing this type of thing, due to the merge conflicts that will result. So the only option is to work only in a single branch and only code-related changes can be done in the mainline branch if you want to avoid merge issues.
If your team is working with Entity Framework and version control of some sort (which hopefully you are), how are you managing changes in this pretty common scenario?
[ update ]
Thanks to Erik, I think I have a solution. I had come across this as well, but didnt see much regarding the "round trip" ability after the 1st reverse engineer step. So I just went ahead and create a small dummy database along with a dummy console app with a class library. Then I pulled in this Reverse poco template ... changed one of the columns in the db, updated the tt file and soon as i saved (as mentioned in the tt file), my schema changes were reflected in the POCO classes. Awesome - this is probably my ticket to dumping the edmx file which I never use anyway and will likely enable code merging between branches.
How to add a jhipster entity to svn which can generated on an other users pc
What is the file to be copied to the svn such that it can be generated on other user pc s too.I have created the entity by using following,
jhipster entity Employee
what is the file i need to commit in the SVN
TortoiseSVN not asking for login credentials
I've set multiple repositories hosted on an Apache2 Server on my Raspberry Pi 3, created by the standard svn program.
I've created a svn_access_control file as shown below:
[/] * = r [repo1:/] Niko = rw * =
I've also created multiple users using the following command:
htpasswd -m /etc/apache2/dav_svn.passwd USERNAME
My dav_svn.conf has the following lines written in it:
<Location /svn> DAV svn SVNParentPath /home/pi/repos AuthType Basic AuthName "Subversion Repo" AuthUserFile /etc/apache2/dav_svn.passwd AuthzSVNAccessFile /etc/apache2/svn_access_control <LimitExcept GET PROPFIND OPTIONS REPORT> Require valid-user </LimitExcept> </Location>
Yet, when I want to access repo1 via ToirtoiseSVN on a Windows 8.1 Machine, I am not prompted with a login dialog. It just skips to the "Access forbidden" message.
Access to '/svn/repo1' forbidden
The Authentification Data is already cleared in the Client.
What settings do I have to change in order for TortoiseSVN to prompt me to log in?