So if we want other environment variables to reflect the path to drop files, we'd keep adding more arguments? Why not just define a generic "path to dropfile" environment variable and use that instead?
--- SBBSecho 3.37-Linux
* Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
From Rob Swindell@1:103/705 to GitLab note in main/sbbs on Sat Aug 29 15:06:46 2026
We have C literals (e.g. '\t', '\r', ' ') and macros defined in src/encode/* and xpdev/* for these - why not use them?
--- SBBSecho 3.37-Linux
* Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
From Rob Swindell@1:103/705 to GitLab note in main/sbbs on Sat Aug 29 15:08:02 2026
And this looks like a function that belongs in maybe xpdev/unicode*
--- SBBSecho 3.37-Linux
* Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
From Rob Swindell@1:103/705 to GitLab note in main/sbbs on Sat Aug 29 15:08:36 2026
If I were going to define a new drop file format, it wouldn't be this. And there's other things I don't like about this commit (noted above).
--- SBBSecho 3.37-Linux
* Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
From Deucе@1:103/705 to GitLab note in main/sbbs on Sat Aug 29 21:25:40 2026
I would hope not, but the environment variable name for BBSDEV.DRP is defined by the standard, and it needs to be BBSDEV_DRP. I would expect that if we end up wanting arbitrary environment variables, we would add a char **environ argument, but that's more complex than is needed today.
--- SBBSecho 3.37-Linux
* Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
From Deucе@1:103/705 to GitLab note in main/sbbs on Sat Aug 29 21:29:52 2026
Well, '\t' is a char, not an `enum unicode_codepoint`, but assuming there's codepoint versions, it's mostly because I didn't know about them and if I did, I would be too lazy to look them all up.
No objection to it being renamed and moved into xpdev.
--- SBBSecho 3.37-Linux
* Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
From Deucе@1:103/705 to GitLab note in main/sbbs on Sat Aug 29 21:31:38 2026
I don't like to use names for endpoints of ranges, it hides the meaning behind manual lookups.
--- SBBSecho 3.37-Linux
* Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
From Deucе@1:103/705 to GitLab note in main/sbbs on Sat Aug 29 21:32:26 2026
If you're saying you want to revert it, go ahead, it's your software.
--- SBBSecho 3.37-Linux
* Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
From Rob Swindell@1:103/705 to GitLab note in main/sbbs on Sat Aug 29 22:41:46 2026
"The standard" isn't even 24 hours old, so would I think it could be amended. But being that no existing doors (e.g. that use other drop file types) will ever check/use any environment variable for this purpose, I guess it's a moot point.
Is there a reason you needed an overloaded sbbs_t::external() rather than just adding another argument that defaulted to nullptr?
--- SBBSecho 3.37-Linux
* Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
From Rob Swindell@1:103/705 to GitLab note in main/sbbs on Sat Aug 29 22:52:36 2026
I'm saying it looks to me like a missed opportunity to define something significantly better than the long standing drop file formats. I already added the nit-comments about the implementation.
--- SBBSecho 3.37-Linux
* Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
From Rob Swindell@1:103/705 to GitLab note in main/sbbs on Sat Aug 29 23:01:50 2026
com_base and com_irq are legacy/uninitialized values - not configurable or valid. We actually hard-code the com_base to 0x0f (for FOSSIL).
--- SBBSecho 3.37-Linux
* Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
From Deucе@1:103/705 to GitLab note in main/sbbs on Sun Aug 30 08:33:00 2026
"needed"? No, it's really just a matter of style. One hides the value for a parameter and the other doesn't. The underlying implementation is the same either way.
--- SBBSecho 3.37-Linux
* Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
From Deucе@1:103/705 to GitLab note in main/sbbs on Sun Aug 30 08:36:34 2026
Yes, I've brought this up before about the Synchronet CTerm version encoding. --- SBBSecho 3.37-Linux
* Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
From Deucе@1:103/705 to GitLab note in main/sbbs on Sun Aug 30 08:42:24 2026