I got two copies of the message I am replying to. The first had the
~M and the second didn't. I dunno why that would be.
@PATH: 3/101 100 4/100
This is the first one I recieved. It has the ~M.
@PATH: 3/101 100 2/100 4/100
And I also got this one, without a ~M.
I got two copies of the message I am replying to. The first had
the ~M and the second didn't. I dunno why that would be.
~M looks like it could represent a <CR>.
I don't see anything wrong with the message (in hexcurses), maybe the other one was catched by dupe checking on the path? Or is it a problem with your system or 4/100? It looks like you are getting most of these weird messages (or you are the one who reports it more often).
I wonder why the path in Spectre's message (@MSGID: 21:3/101 01051a14) says
this:
@PATH: 3/101 100 4/100 1/100 151
Doesn't 3/100 has a direct link to 1/100? Or is some tosser ignoring SEEN-BY lines?
Which PATH do you see on both messages? The same or a different one?
It looks like you are getting most of these weird messages. I wonder
why?
This is the first one I recieved. It has the ~M.
@PATH: 3/101 100 2/100 4/100
And I also got this one, without a ~M.
What is going on there? is 2/100 removing the ~M? Very weird.
1/100, 2/100 run Mystic, 4/100 switched to hpt, right?
Hello Oli,
I got two copies of the message I am replying to. The first
had the ~M and the second didn't. I dunno why that would be.
I'm unsure. Looking at those messages on the BBS they look
identical. Looking at them with golded all lines in a paragraph
start with ~M, except the first line.
Hello Oli,
I got two copies of the message I am replying to. The first had
the ~M and the second didn't. I dunno why that would be.
I'm unsure. Looking at those messages on the BBS they look identical. Looking at them with golded all lines in a paragraph start with ~M,
except the first line.
I wonder why the path in Spectre's message (@MSGID: 21:3/101
01051a14) says
this:
@PATH: 3/101 100 4/100 1/100 151
Doesn't 3/100 has a direct link to 1/100? Or is some tosser
ignoring SEEN-BY lines?
Yes, but it might reach 1/100 via 4/100 first.. :)
It should be trapped as a dupe if it is a dupe.
It looks like you are getting most of these weird messages. I
wonder why?
I don't always get two of them, just at times and I can't see a
pattern.
The last two i got both passed dupe checking at 4/100 and here. They
are both in my message base. The good one (without ~M) is about 150
bytes larger than the one with ~M according to the size indicated
below the subject in golded.
I dunno.. I don't understand.
I'm unsure. Looking at those messages on the BBS they look
identical. Looking at them with golded all lines in a paragraph
start with ~M, except the first line.
My guess is the packet has windows/DOS linebreaks in it (CR then LF)
and is getting imported into the jam base. Jam bases should only have
CR for a line break, so my guess is it's the LF that's showing up.
Which is odd because ^M is carriage return, not line feed.
I think Magicka filters out LF in jam bases (because they shouldn't be there) but I'm not sure off the top of my head.
Do you have DISPSOFTCR is golded.cfg? It's just a hunch, I don't know
if it has anything to do with it.
Could you check in the Jam .jdt file which control character that is?
It's obviously not a literal "~M" or it would be displayed in the BBS
too.
Yes, but it might reach 1/100 via 4/100 first.. :)
But 4/100 shouldn't export to 1/100 if it is already in the SEEN-BYs
(but maybe it is configured this way intentionally).
It should be trapped as a dupe if it is a dupe.
But what is a dupe? All message with the same MSGID? Or only messages
that are exactly the same? E.g. Crashmail only checks the MSGID.
More SEEN-BY lines.
Could you check in the Jam .jdt file which control character that
is? It's obviously not a literal "~M" or it would be displayed in
the BBS too.
I'll have a look and see, I'm not sure what I'm looking for but I'll
try.
I just put fsx_gen.zip in my web root. If anyone wants to have a look
they are at..
https://trmb.ca/fsx_gen.zip
https://trmb.ca/fsx_gen.zip
404
Hello Oli,
https://trmb.ca/fsx_gen.zip
404
I thought I had it in the webroot.
I think I have it in the root of the ftp server.
ftp://trmb.ca/fsx_gen.zip
ftp://trmb.ca/fsx_gen.zip
Okay, it's a 0x8D character aka soft CR.
@PATH: 3/101 100 4/100
This is the first one I recieved. It has the ~M.
@PATH: 3/101 100 2/100 4/100
And I also got this one, without a ~M.
What is going on there? is 2/100 removing the ~M? Very weird.
1/100, 2/100 run Mystic, 4/100 switched to hpt, right?
Hello Oli,
ftp://trmb.ca/fsx_gen.zip
Okay, it's a 0x8D character aka soft CR.
I wonder where that would come from? I wonder if it is in Spectre's original message. I don't think so, but I'm not sure.
What do you think about this explanation:
3/101 generates messages with a soft-CR (0x8d / 141) instead of a
regular CR (0x0d / 013). Mystic at 2/100 converts soft-CR to CR. hpt doesn't detect a dupe, because the text body is different. Golded
cannot display soft-CRs properly.
3/101 SuperBBS: everythings fine, soft-CRs are allowed
2/100 Mystic: changing the mail body is not okay (BUG!)
4/10? hpt: dupe detection is not perfect
4/106 Golded: soft-CR should be ignored, see FTS-0001 (BUG!?)
Maybe it's just a Golded UTF-8 / charset problem.
I wonder why the path in Spectre's message (@MSGID: 21:3/101 01051a14) says this:
@PATH: 3/101 100 4/100 1/100 151
Doesn't 3/100 has a direct link to 1/100? Or is some tosser ignoring SEEN-BY lines?
3/101 SuperBBS: everythings fine, soft-CRs are allowed
2/100 Mystic: changing the mail body is not okay (BUG!)
4/10? hpt: dupe detection is not perfect
4/106 Golded: soft-CR should be ignored, see FTS-0001 (BUG!?)
That's possible I suppose. But I don't ever recall seeing those ~M's before. I've only seen them in the last day or three.
My guess is the packet has windows/DOS linebreaks in it (CR then LF)
and is getting imported into the jam base. Jam bases should only have
CR for a line break, so my guess is it's the LF that's showing up.
Which is odd because ^M is carriage return, not line feed.
I think Magicka filters out LF in jam bases (because they shouldn't be there) but I'm not sure off the top of my head.
4/106 Golded: soft-CR should be ignored, see FTS-0001 (BUG!?)
Here we have SuperBBS -> SBBSecho -> Crashmail. It's unlikely that SBBSecho or Crashmail add soft-CR characters.
I did get doubles of a number of messages from Spectere, one had ~M and one didn't. How does that happen?
I did get doubles of a number of messages from Spectere, one had
~M and one didn't. How does that happen?
Sounds like Oli might have identified this. Since we are a mesh, this message went one way where soft-CR was ignored, and another way where
it may have been converted to CR. So when both messages arrived at
4/100 there were "different", and thus past on as not dupes.
...лоег
And I also got this one, without a ~M.
I did get doubles of a number of messages from Spectere, one had ~M and one didn't. How does that happen?
And I also got this one, without a ~M.
If you also replied/quoted the ~M it didn't arrive here with a ~M on it....
PS: As a test I've run this back through the message editor inside the BBS, do we have ~M's again?
I did get doubles of a number of messages from Spectere, one had
~M and one didn't. How does that happen?
The central hubs all poll each other, some-one modified it and its different enough not to pass as a dupe?
This is another one with the soft CRs. They get moved around when quoted
This is another one with the soft CRs. They get moved around when
quoted
I've been wondering if "The Reader" does it, or the bog standard
SuperBBS does it.
That ~M that I am using here is just the way golded represents the
soft CR, it's an 0x8D or ascii 141 I think.
I've been wondering if "The Reader" does it, or the bog standard SuperBBS does it.
It might be. If you take the reader out for a bit we can have a look and see. Regardless of were it comes from it needs to be dealt with differently.
I think it is a UTF-8 display issue. Maybe it just needs the the right configuration in golded.cfg. I never figured out how to configure
Golded for UTF-8 properly.
Hello Oli,
I think it is a UTF-8 display issue. Maybe it just needs the
the right configuration in golded.cfg. I never figured out how
to configure Golded for UTF-8 properly.
That is an issue too. When folks post with UTF-8 I don't think what I
am seeing is what they posted. I don't know why that is. When Michiel
posts in the UTF-8 area or Maurice sometimes posts in UTF-8 it
doesn't look right to me.
When you posted some UTF-8 not long ago in FIDONEWS also, I think it
was a small graphic.. I didn't see it. I'm not sure what's up there
or if there is anything I can do about it.
These ~M characters I've been seeing lately though are soft CRs.
I think that is more or less it. If your SBBS can change it's behaviour WRT soft-CR don't change anything just yet. I'd like to get the facts out if there are more to be had.
How long ago did UTF become a thing?
The Reader is from 1997, and SBBS is From 1993. I'm not surprised SuperBBS doesn't know anything about it, by the time this release
arrived, I think it was a given Aki was already on his way to the
Finnish army. So there wasn't a lot "new" in there, mostly fixes.
The Reader surprises me a little if it is using "Soft <cr>" I still
find that whole idea somewhat mind boggling. I can't find any
reference to UTF8 in its docs, so perhaps its just using the FTC of
the time for soft <cr>s. It's also somewhat interesting that its just become visible.
| Sysop: | sneaky |
|---|---|
| Location: | Ashburton,NZ |
| Users: | 2 |
| Nodes: | 8 (0 / 8) |
| Uptime: | 33:14:58 |
| Calls: | 2,116 |
| Files: | 11,149 |
| D/L today: |
315 files (11,412K bytes) |
| Messages: | 952,679 |