Creating branches, switching branches, stashing the changes or committing the changes temporarily all within one workspace have worked for me. I also switch from debug to release build within the same workspace. I use release build to run the full regressions because it runs faster. I only clone when I need to work on different repositories on the same system.
We need to bring down the Hadoop/hbase instance to switch between debug and release build or branches. One issue I have seen is the T2 driver java client and the native side .so becoming incompatible because of version differences between debug and release even after clean build is done. Yes. we need to do clean build but this would reduce the space requirement on the development VMs than having a separate workspace for each of your task. Selva -----Original Message----- From: Anuradha Hegde [mailto:[email protected]] Sent: Thursday, March 3, 2016 5:57 AM To: [email protected] Subject: RE: Noob question on working with Git This has worked for me. Connectivity will work properly as well if dcs /rest install dirs are set correctly. If you do swstopall and start a new terminal and source env.sh then the install dirs for dcs/rest should be set correctly. Anu -----Original Message----- From: Sandhya Sundaresan [mailto:[email protected]] Sent: Wednesday, March 2, 2016 10:53 PM To: [email protected] Subject: RE: Noob question on working with Git Having 2 instances on the same system has worked for me if I completely shutdown one instance and use the other. You have to do an swstopall and ensure through "jps" and sqps that there is really nothing running . Then you can start your other instance. . I have not tried it in the past few weeks so I don't know if anything new will disallow it. But it's worked for me atleast to use SQL interfaces like sqlci - not sure about the connectivity part. Sandhya -----Original Message----- From: Dave Birdsall [mailto:[email protected]] Sent: Wednesday, March 2, 2016 4:17 PM To: [email protected] Subject: RE: Noob question on working with Git Having two or more cloned source trees on the same workstation doesn't seem to work well for SQL testing as the local_hadoop setup doesn't seem to like to be shared. -----Original Message----- From: Steve Varnau [mailto:[email protected]] Sent: Wednesday, March 2, 2016 3:13 PM To: [email protected] Subject: RE: Noob question on working with Git Yes, that last option means multiple cloned source trees. When you clone, you can give a destination directory name. It only defaults to "incubator-trafodion". --Steve > -----Original Message----- > From: Gunnar Tapper [mailto:[email protected]] > Sent: Wednesday, March 2, 2016 3:02 PM > To: [email protected] > Subject: Re: Noob question on working with Git > > Does this end my up with separate source directories on the > workstation? I guess I create separate source directories and then do > a git clone per source directory? > > On Wed, Mar 2, 2016 at 3:55 PM, Steve Varnau <[email protected]> > wrote: > > > Your github fork may have many branches. > > > > On your local development machine, your workspace may also access > > many branches, so you can switch back and forth (git checkout > > <branch>). Of course, you can also have multiple workspaces, so that > > you don't have to switch back and forth. > > > > --Steve > > > > > > > -----Original Message----- > > > From: Gunnar Tapper [mailto:[email protected]] > > > Sent: Wednesday, March 2, 2016 2:44 PM > > > To: [email protected] > > > Subject: Noob question on working with Git > > > > > > Hi, > > > > > > How do you folks work on different Jiras or features in parallel? > > > I don't quite get how parallel checkins work when I have a single > > > fork to work from? > > > > > > -- > > > Thanks, > > > > > > Gunnar > > > *If you think you can you can, if you think you can't you're > > > right.* > > > > > > -- > Thanks, > > Gunnar > *If you think you can you can, if you think you can't you're right.*
