Currently my apps are 32 and 64 bit Windows. However they are to be expanded to Linux and Mac as well. My question is, is there a way to have the compiled app dynamically named? For example a 32 bit version App32.exe A 64 bit version of the same app named App64.exe
While the apps are placed in their own folders after compilation, once there, I have not been able to find a way to easily identify if they are the 32 or 64 bit version. This problem extends to the Windows installer, which is actually what prompted the question.
Is there any set of best practices anyone can point to? I realize this can be a convoluted problem that can “grow” over time - especially as we have seen in the last year or so, Xojo has grown in its compilation abilities - and therefore the aggravation of this as a potential problem.
I appreciate any ideas, advice or just examples of what everyone is doing in a similar circumstances.
@Norman Palardy is there a way to the tell the IDE to build both 32 & 64 bit versions of the app? Or on Linux, 32, 64, & ARM versions?
I have some clients that have 32bit processors in lightweight devices (tablets/chromebook-like-devices) that I still have to build 32bit for. Other clients want 64bit as 64bit is twice as big as 32bit so it must be twice as good.
Thanks for the replies everyone. I will have a look at your blog too Travis.
But to clarify, I do use build folders. But I was looking more to actually giving the app name as _32 or _64 for examples. This way they are easily identifiable in the field. This may be helpful if for example the app fails and it cannot be run. Having the app name specific is a quick way for a tech to know what was installed. It is also helpful when assembling the installer too. Makes it “in your face” easy and simple to identify.
We are not recommending one over the other for all circumstances. Our comments were about what ended up being best for us. If you’re pressing Build in the IDE every time you actually want a build, then Build Automation may be exactly what you need.
By contrast, we use a completely automatic system now which only requires a human when something goes wrong. every time a Xojo engineer checks something into the framework, or the IDE, a build is automatically triggered, just to make sure we haven’t broken building the IDE on any of the platforms that we support. That way, when we do ask the system to build an alpha or a beta, we have a reasonable expectation that it’ll compile without trouble.
