FrostyCat
Redemption Seeker
Are there any plans for making GMS 2's project format more compatible with source control software?
So far with the beta, the source control integration has never been publicly tested. Like with GMS 1.x, the lack of such testing resulted in a project format that behaves badly with both SVN and Git. This hindered the adoption of source control as a standard in GM development, and led to a number of incidents when people lost projects in ways that SVN, Git or a comparable source control system could have easily prevented.
If using Tortoise instead of the built-in mechanism was enough, I wouldn't mind. But it still isn't enough because of views, both in their random naming scheme and the way they are enumerated inside the main project file. Conflicts between branches or working copies would be next to impossible to reconciliate.
With a strictly linear flow and only 1 developer, the current format works. But start branching or working in teams, and the whole approach falls apart. This is no way to move forward with a more professional audience.
With GMS 1.x, I think you've heard enough of my year-long whining about its SCM. I hope I don't have to whine like that again with GMS 2.x to make it done right.
So far with the beta, the source control integration has never been publicly tested. Like with GMS 1.x, the lack of such testing resulted in a project format that behaves badly with both SVN and Git. This hindered the adoption of source control as a standard in GM development, and led to a number of incidents when people lost projects in ways that SVN, Git or a comparable source control system could have easily prevented.
If using Tortoise instead of the built-in mechanism was enough, I wouldn't mind. But it still isn't enough because of views, both in their random naming scheme and the way they are enumerated inside the main project file. Conflicts between branches or working copies would be next to impossible to reconciliate.
With a strictly linear flow and only 1 developer, the current format works. But start branching or working in teams, and the whole approach falls apart. This is no way to move forward with a more professional audience.
With GMS 1.x, I think you've heard enough of my year-long whining about its SCM. I hope I don't have to whine like that again with GMS 2.x to make it done right.