I found a small bug running Hobo on Windows. Hopefully someone can take a look at it.
I normally run Hobo on FreeBSD. My standard development environment is:
- FreeBSD 10.3 on VirtualBox
- Hobo 2.2.6
- Rails 4.2.7.1
- JRuby 9.1.5.0
- PostgreSQL 9.3
I have a client who would like an app installed in a Windows environment using a SQL Server backend, so I decided to develop the whole thing on my Windows 10 development box, using the SQL Server 2016 database as a backend. These were the only changes from my standard development environment, When I used the command "jruby -S hobo g migration" on Windows 10, I successfully connected to the SQL Server database, but I ran into problems when I attempted to respond to the Hobo migration script:
"What now: [g]enerate migration, generate and [m]igrate now or [c]ancel? ^C"
My input wasn't recognized. I had to use ctrl-C to break out of the script, and the migration script wasn't saved, so I couldn't run Rake to complete the migration.
I am currently working around this issue by accessing the SQL Server database remotely. I have VirtualBox installed on my Windows 10 machine, I have my Hobo app installed on FreeBSD inside VirtualBox, and I have the database configuration pointing to the SQL Server database running on my Windows 10 machine.
As I have a workaround, this is not an urgent issue for me, but it could put off someone in an all-Windows environment who wants to try out Hobo.