I have a couple of packets waiting for magitoss to pick up. They are in the /fido/in directory that is set up in the config. However, magitoss acts as if they are not there.
The node number is 21:1/175.8 The packets are called: 0000p008.FR0 0000p008.SA0
The node number is 21:1/175.8 The packets are called: 0000p008.FR0 0000p008.SA0
I've just pushed an update that should work with these point files.
You'll need v0.9-patch or v0.10-alpha
To update, in your Magicka BBS directory..
git pull
git checkout v0.9-patch
then make www
Quoting apam to Dumas Walker <=-
I've just pushed an update that should work with these point files.
You'll need v0.9-patch or v0.10-alpha
I've just pushed an update that should work with these point files. You'll need v0.9-patch or v0.10-alpha
Dumb question... I run Magicka as a point and have never had an issue
what's the difference? (Pure curiousity on my part)
Dumb question... I run Magicka as a point and have never had an issue
what's the difference? (Pure curiousity on my part)
I believe it was because Synchronet adds the 'p' in the filename for a point system. I think that's where Magitoss was having an issue.
Andrew will correct me if I'm wrong, but that was my understanding. :)
The node number is 21:1/175.8 The packets are called: 0000p008.FR0 0000p008.SA0
Magimail is ignoring the packets because of the 'p' character in the file name. I will have to research what is the correct way of doing things.
Dumb question... I run Magicka as a point and have never had an issue
what's the difference? (Pure curiousity on my part)
The "p" in the filename designates a point address, and case
sensitivity shouldn't matter.
So I'm not sure which is correct. I'm thinking Synchronet is probably doing something non-standard, given that I've not seen this before in other BBS/Mailers I've had set up. On the other hand, the only mention of what a bundle name should be that I can find via google is the Synchronet documentation.
Quoting Black Panther to Tiny <=-
I believe it was because Synchronet adds the 'p' in the filename for a point system. I think that's where Magitoss was having an issue.
Andrew will correct me if I'm wrong, but that was my understanding. :)
apam wrote to Accession on 02-13-18 12:00 <=-
The "p" in the filename designates a point address, and case
sensitivity shouldn't matter.
Yes, in Synchronet, and according to Synchronet, bundle names can be anything at all (just the extension is specific)
Crashmail / Magimail expects bundle names to be hexadecimal characters (even for points), and will disregard anything that has a
non-hexadecimal character in the bundle name (such as 'p').
So I'm not sure which is correct. I'm thinking Synchronet is probably doing something non-standard, given that I've not seen this before in other BBS/Mailers I've had set up. On the other hand, the only mention
of what a bundle name should be that I can find via google is the Synchronet documentation.
So for now I've just allowed the 'p' (or 'P') in bundle names and the
rest should be hexadecimal. I guess for now that will do and if some
other program starts sending bundles with other non hexadecimal
characters I'll deal with it then.
May I ask why someone, other than the host route echomail/netmail?
As I said, Synchronet is in complience with a BSO style mailer but can,
if desired, create a mail attach depending on the switches used.
Yes, in Synchronet, and according to Synchronet, bundle names can be anything at all (just the extension is specific)
So I'm not sure which is correct. I'm thinking Synchronet is probably doing something non-standard, given that I've not seen this before in other BBS/Mailers I've had set up. On the other hand, the only mention of what a bundle name should be that I can find via google is the Synchronet documentation.
apam wrote to Bill McGarrity on 02-14-18 17:17 <=-
May I ask why someone, other than the host route echomail/netmail?
I don't know why the people who wrote crashmail chose to check that bundles had hexidecimal characters as the filename, my guess is to differentiate between bundles and other files that may arrive in the inbox, but that's just a guess.
As I said, Synchronet is in complience with a BSO style mailer but can,
if desired, create a mail attach depending on the switches used.
So it seems. Are you suggesting I remove the check altogether? I'm not keen on it, the crashmail people must have put it in there for a
reason, but then I'm sure they are human too.
You can compare that to FTSC documentation if you wish, but just supporting it as you've done would probably be better, since I doubt Rob is going to change Synchronet to conform to lesser 30yr old
documentation. ;)
Yeah, there's nothing there, the only 'technical standard' describes the flow files and stuff, not the names of the bundles themselves.
I don't expect Rob to change synchronet, at present I've changed things
so that it still checks for hexidecimal characters, but also allows for a p where synchronet puts one. What I'm wondering is if I should remove the check altogether and only check the extensions.
So for now I've just allowed the 'p' (or 'P') in bundle names and the
rest should be hexadecimal. I guess for now that will do and if some
other program starts sending bundles with other non hexadecimal
characters I'll deal with it then.
There were proposals for standardizing bundle filenames long ago but nothing came of it officially AFAIK. The only FTSC requirement was that they be in 8 format. Beyond that mailers just did whatever they deemed appropriate in ord to avoid filename collisions.
If multiple mailers/tossers have been taking a common approach, it's de-fact
Sysop: | Keyop |
---|---|
Location: | Huddersfield, West Yorkshire, UK |
Users: | 285 |
Nodes: | 16 (2 / 14) |
Uptime: | 70:00:29 |
Calls: | 6,488 |
Calls today: | 1 |
Files: | 12,096 |
Messages: | 5,275,483 |