I decided to start a thread on suggestions IRT the code repository rather than PM every moderator. That way it's open and any user can air concerns or suggestions about them.
1) I suggest deleting diku-linux.tar.gz and dikumud.zip from the DikuMUD directory. These are identical to diku-gamma.tar.gz and their presence may be confusing.
2) I also suggest deleting diku-alfa.tar.gz. This source is broken. It was apparently created from an OS/2 or Warp system and has 8.3 style filenames and OS/2 EOF markers. I will upload the original pristine diku-alfa.tar.gz. I had tried earlier but I got a file name collision.
I also have plans to submit copies of these bases which have been cleaned up for gcc4. IMC2 will not be included in these. If anyone has an objection to my doing this, it would be a good idea to speak up before any time is spent packaging and uploading them.
What's the preference for upload file format? Some of the latest Tiny/Mush/Mux code is being released in all three formats tar/bzip2, winzip, and tar/gzip formats. I myself release my windows/linux compatable ports in both tar/gzip and winzip format. I believe I uploaded most as tar.gz as Winzip apparently can handle it.
I don't see any reason not to include IMC in it, but if you feel that it's safer to leave that to just those who choose SmaugFUSS, then I have no objections.
As for deleting the duplications, while that's probably a good thing overall, but I do have to wonder if there is any way of verifying that the tar'd versions are Windows compatible too? It'd be a shame to force new potential admins into thinking we're all so anti-windows that cygwin seems like even it's not a real option since they can't untar the distributions without linux or major hassle. (I tried the other day to find a windows version of tar and found that while there are actually several out there, they suck royally.)
As for preference, personally it makes little difference to me as I have a linux box and a windows box on my desk and use them both regularly, though I tend to use the windows machine mainly for M$ Office stuff and the linux machine for pretty much everything else. But I also can only speak for myself. While I imagine that most of us are using linux or some environment like cygwin, I'm not sure how tar works under cygwin and it's relatives since I don't use them.
Winzip doesn't have any issues that I've noticed with opening gziped and tarred files. It's been so long that I can't remember when I first did it, had to be mid-90's. Nor does the unzip command seem to have any issues opening a winzip file on linux.
Since I work primarily for MS based firms, it's not an option for me. I've had to work from both ends as I prefer to work with my code offline and upload the modified files.
Yes it do but I found a problem. If you have the CRLF conversion option set, and last I checked it was the default, then all files will have LF's converted to CRLF's. This isn't a problem for sources and makefiles. It can be a problem for data files whether the be binary or not. That means Diku zone/area files, Mush/Mux/MOO/Cold databases , etc. have CRLF's which in many cases the code itself is dependent on LF. Not a big problem as you just switch it off.
Once you have it off, you're set and have no problems. It may no longer be the default as I changed it years ago. :-P
The software in tiny/muck/tinymuck-1.2-r10.tar.gz I named the same as the filename, "tinymuck-1.2-r10.tar.gz". It was one of my first uploads. It should probably sport the name, "TinyMuck 1.2 (port)" so there is no confusion with the original. Sorry my mistake.