Addressing this to deon as I understand you're Oli's hub?

Anyway, I don't know about others but Oli's messages seem to come
through twice, the first one is mangled and makes it into my message
base (by mangled I mean half the paragraphs are missing) the second one
gets tossed into Bad as it's a dupe, but looks fine.

I'm not sure if it's my end, your end or somewhere inbetween.

My guess, and it's a stab in the dark really, is Oli doesn't wrap
messages at 80 chars, but rather each paragraph is one long line, and
the mangling is occurring when the paragraph line is too long (perhaps somewhere something is using a buffer with an arbitrary length)

As to why I also get a good message in the bad folder, I'm stumped.

This is the Path of the bad message
PATH: 1/151 3/100 1/100 126

This is the Path of the good message (in the bad folder)
PATH: 1/151 3/100 2/100 1/100 126

Hope that helps some, I'm not sure what else might be helpful.

Andrew

--- MagickaBBS v0.13alpha (Linux/x86_64)
* Origin: HappyLand - telnet://magickabbs.com:2023/ (21:1/126)

On Tue, 31 Mar 2020 20:05:00 +1000
"apam -> alterego" <0@126.1.21> wrote:

Hi Andrew!

Addressing this to deon as I understand you're Oli's hub?

Yes, I'm getting my echomail from 3/100.

Anyway, I don't know about others but Oli's messages seem to come
through twice, the first one is mangled and makes it into my message
base (by mangled I mean half the paragraphs are missing) the second
one gets tossed into Bad as it's a dupe, but looks fine.

That's weird. Is the dupe ("looks fine") really unmodified? Or did Mystic add CRs to the long line paragraphs?

I'm not sure if it's my end, your end or somewhere inbetween.

My guess, and it's a stab in the dark really, is Oli doesn't wrap
messages at 80 chars, but rather each paragraph is one long line,

That's correct.

and
the mangling is occurring when the paragraph line is too long (perhaps somewhere something is using a buffer with an arbitrary length)

which would be a new problem in the network.

As to why I also get a good message in the bad folder, I'm stumped.

This is the Path of the bad message
PATH: 1/151 3/100 1/100 126

This is the Path of the good message (in the bad folder)
PATH: 1/151 3/100 2/100 1/100 126

If Synchronet were mangling the messages, I would expect to occasionally get mangled messages too, but I don't.

My guess: 1/100 is running a newer Mystic pre-alpha then 2/100. g00r00 changed something in the word wrapping (line breaking) code for in-transit mails and for some cases it mangles mails (maybe a problem somewhere else in the code that didn't get trigger
This is just a first idea and the problem could be something completely different.

Does the mail in the dupe folder (that looks fine) has intact long lines?

I also wonder why 1/100 is getting a dupe from 2/100? Are 3/100 not adding SEEN-BY lines or is 2/100 ignoring SEEN-BYs?

Hope that helps some, I'm not sure what else might be helpful.

I think it's helpful information and a start.

---
* Origin: (21:1/151)

On Tue, 31 Mar 2020 20:05:00 +1000
"apam -> alterego" <0@126.1.21> wrote:

Addressing this to deon as I understand you're Oli's hub?

Anyway, I don't know about others but Oli's messages seem to come
through twice, the first one is mangled and makes it into my message
base (by mangled I mean half the paragraphs are missing) the second
one gets tossed into Bad as it's a dupe, but looks fine.

I sent a test message directly over Hub 1 in FSX_NET. Is the 960 character long
line still intact?

---
* Origin: (21:1/151)

On Tue, 31 Mar 2020 20:05:00 +1000
"apam -> alterego" <0@126.1.21> wrote:

Addressing this to deon as I understand you're Oli's hub?

Anyway, I don't know about others but Oli's messages seem to
come through twice, the first one is mangled and makes it into
my message base (by mangled I mean half the paragraphs are
missing) the second one gets tossed into Bad as it's a dupe,
but looks fine.

I sent a test message directly over Hub 1 in FSX_NET. Is the 960
character long line still intact?

Nope. There are 2 paragraphs exactly 256 chars in length. I just checked
one of your other messages, each paragraph is always truncated at 256
chars.

Andrew

--- MagickaBBS v0.13alpha (Linux/x86_64)
* Origin: HappyLand - telnet://magickabbs.com:2023/ (21:1/126)

Anyway, I don't know about others but Oli's messages seem to come
through twice, the first one is mangled and makes it into my message
base (by mangled I mean half the paragraphs are missing) the second
one gets tossed into Bad as it's a dupe, but looks fine.

That's weird. Is the dupe ("looks fine") really unmodified? Or did Mystic add CRs to the long line paragraphs?

I also got 2 copies of a message from Spectre to volker.

As to why I also get a good message in the bad folder, I'm stumped.

In my case here on BBBS I got one message from 3/101 3/100 4/100 that was intact.

I got another mangled one from 3/101 3/100 1/100 4/100

Both arrived in my message base.

If Synchronet were mangling the messages, I would expect to occasionally get mangled messages too, but I don't.

I got a good message via 3/100 and 4/100. Synchronet didn't mangle it.

My guess: 1/100 is running a newer Mystic pre-alpha then 2/100. g00r00 changed
something in the word wrapping (line breaking) code for in-transit mails and for some cases it mangles mails (maybe a problem somewhere else in the code that didn't get triggered before). Mails with long lines from 3/100 gets mangled at 1/100. Mails from 2/100 are always wrapped so they get forwarded unmodified by 1/100.

This is so. The last 2 prealpha versions are trying to not change mail in transit. I think they are successfully doing that. For some reason they are getting mangled when passing through 1/100. So, there is more to do.

I think it's helpful information and a start.

Yes, I hope g00r00 will be successful in tossing on mail unchanged.

--- BBBS/Li6 v4.10 Toy-4
* Origin: The Rusty MailBox - Penticton, BC Canada (21:4/106)

On Tue, 31 Mar 2020 21:40:50 +1000
"apam -> Oli" <0@126.1.21> wrote:

On Tue, 31 Mar 2020 20:05:00 +1000
"apam -> alterego" <0@126.1.21> wrote:

Addressing this to deon as I understand you're Oli's hub?

Anyway, I don't know about others but Oli's messages seem to
come through twice, the first one is mangled and makes it into
my message base (by mangled I mean half the paragraphs are
missing) the second one gets tossed into Bad as it's a dupe,
but looks fine.

I sent a test message directly over Hub 1 in FSX_NET. Is the 960
character long line still intact?

Nope. There are 2 paragraphs exactly 256 chars in length. I just
checked one of your other messages, each paragraph is always
truncated at 256 chars.

Pascal ;)

---
* Origin: (21:1/151)

Re: Oli's Messages
By: apam to alterego on Tue Mar 31 2020 08:05 pm

Hey Apam,

Anyway, I don't know about others but Oli's messages seem to come
through twice, the first one is mangled and makes it into my message

No, I havent noticed Oli's messages coming in twice. (I feed my BBS from 3/100 and 2/100 - I think technically I shouldnt get it from 2/100 (since I would be in the SEEN-BY as the message left 3/100) - but I think sometimes I do get stuff via 2/100.)

Is anybody else seeing it?

I have started seeing Spectre's messages twice again though - with one copy looking good (via 3/100) and another copy looking
"strange" (perhaps the strange paragraphs that you are seeing) - and I've only noticed that in the last few days (after I saw Avon do an update - since the strange messages comes via 1/100.) But that could also be Spectre's fat fingers
again :P

Anyway do you have an example? Date/Time and if possible a message ID - I'll then be able to track how 2/116 received it...

(If I need to, I could also generate a bad message I guess - one with soft-crs and one with a super long line, and see if we can figure out where the problem may lie...)
...deon


... Liberals are a Labour-saving device.
--- SBBSecho 3.10-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

Re: Oli's Messages
By: Oli to apam on Tue Mar 31 2020 01:07 pm

Hey Oli,

I also wonder why 1/100 is getting a dupe from 2/100? Are 3/100 not adding

SEEN-BY lines or is 2/100 ignoring SEEN-BYs?

3/100 is setting the seen-by lines - this is what the message looks like for me
(I get mail from 3/100 quicker than I get it from 2/100).

X-FTN-SEEN-BY 1/151 3/100 1/100 2/100 4/100 2/116 3/101 104

I'm not confident Mystic is honoring the seenby lines, because I often (havent looked in detail) get the same message from 2/100 that get discarded. I was wondering if it was because they are not in a "nice order".
...deon


... Me no wanna goto work. Me wanna bang on keyboard!
--- SBBSecho 3.10-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

Anyway, I don't know about others but Oli's messages seem to come
through twice, the first one is mangled and makes it into my message
base (by mangled I mean half the paragraphs are missing) the second one gets tossed into Bad as it's a dupe, but looks fine.

I have the same thing on my end with the exception of the second message getting tossed to Bad/dupe.

I'm using the latest Mystic prealpha.

--- Mystic BBS v1.12 A46 2020/03/18 (Linux/64)
* Origin: monterey bbs (21:1/168)

Hello alterego,

No, I havent noticed Oli's messages coming in twice. (I feed my BBS
from 3/100 and 2/100 - I think technically I shouldnt get it from
2/100 (since I would be in the SEEN-BY as the message left 3/100) -
but I think sometimes I do get stuff via 2/100.)

Is anybody else seeing it?

I see this regularly with a node running Mystic A43. I might get the same message from that node via 2 different paths.

That version has an issue with seen by lines, it doesn't start new lines with a
net/node pair. That being the case I don't report isues I see from that node.

You are right, you shouldn't get that message back from 2/100. I wonder what version 2/100 is running. The fact the message has changed in transit could also be a part of the reason you see it again.

Ttyl :-),
Al

--- GoldED+/LNX 1.1.5-b20180707
* Origin: The Rusty MailBox - Penticton, BC Canada (21:4/106.1)

You are right, you shouldn't get that message back from 2/100. I wonder what version 2/100 is running. The fact the message has changed in
transit could also be a part of the reason you see it again.

I think the hubs might be running in a fidoweb style config sending each
other all mail which IMO causes issues including dupes.

|15frank |08// |15netsurge
|07disksh0p|08!|07bbs |08% |07bbs.diskshop.ca |08% |07mystic goodness |11SciNet |03ftn hq |08% |07https://scinet-ftn.org

--- Mystic BBS v1.12 A46 2020/03/26 (Linux/64)
* Origin: % disksh0p!bbs % bbs.diskshop.ca % SciNet ftn hq % (21:4/154)

Hello Netsurge,

I think the hubs might be running in a fidoweb style config sending
each other all mail which IMO causes issues including dupes.

Well, yes they do. There is a similar mesh in fidonet, or more likely meshes.

A tosser shouldn't send a message to a node when that node is in the seen bys.

Ttyl :-),
Al

--- GoldED+/LNX 1.1.5-b20180707
* Origin: The Rusty MailBox - Penticton, BC Canada (21:4/106.1)

Re: Re: Oli's Messages
By: Netsurge to Al on Tue Mar 31 2020 04:09 pm

I think the hubs might be running in a fidoweb style config sending each other all mail which IMO causes issues including dupes.

Well, in theory it shouldnt right?

You are correct, Hubs 1-4 have the other hubs in its export list.

Thus a message leaving 3/100, will have 1, 2, 4 (/100) in the SEEN-BY list. In theory, the other hubs shouldnt (have a need) to forward on that message to other hubs. But I think it is.

To be honest, I havent tracked this issue deeply - since when I do get dupe messages they are discarded (so never a problem for me). It would be great though (in terms of effeciency) if the hubs didnt forward the same message to the other hubs...

Perhaps that's an opportunity to improve things?
...deon


... Immortality--a fate worse than death.
--- SBBSecho 3.10-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

I think the hubs might be running in a fidoweb style config sending each
other all mail which IMO causes issues including dupes.

Well, in theory it shouldnt right?

Yes it will. :)

In the case of the hubs they are meshing so each hub should send you a copy of each message that you are not listed in the seen bys. The hubs will mesh all the messages and in theory send all messages (just one) to linked nodes although the hubs will deal with a lot of dupes.

In North America we have what is called the backbone that meshes messages at the top and sends them (just one) to nodes linked to the backbone or regions or
nets connected to the backbone.

I'm sure there is a similar thing in Ozz. There is a similar thing in Z2 they call the fidoweb. Many nodes connect together so they have a huge mesh in Z2.

How that meshing happens isn't really important as long as it does happen.

--- BBBS/Li6 v4.10 Toy-4
* Origin: The Rusty MailBox - Penticton, BC Canada (21:4/106)

Re: Re: Oli's Messages
By: Al to alterego on Tue Mar 31 2020 08:19 pm

Well, in theory it shouldnt right?

Yes it will. :)

Well, I guess in theory there could be scenarios that I get dupes. EG: Somebody
in Hub 1 sends a message (which Hub1 will forward to 2,3,4), and Hub 2 sends me
a message, and as well as Hub 3 (since Hub 3 wont know that 2 sent it to me).

I was coming from the scenario that somebody on Hub 3 (Oli/Spectre, etc) sends a message, Hub 3 will send it to me, and Hub 2 shouldnt (since the copy that Hub 2 gets from Hub 3 will have my 2/116 in the SEEN-BY).

Now the fact that I was seeing dupes from Specture (3/101) with the soft-cr's, means that Hub 2 must have been ignoring me listed as a SEEN-BY (and thus I was
getting the modified message) - and the dupe checking had it as a new message (because the soft
Sadly, since the author doesnt think this is a valid usage scenario (based on email thread that occured about it), it may never be fixed :(
...deon


... File not found. Should I fake it? (Y/N)
--- SBBSecho 3.10-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

Well, I guess in theory there could be scenarios that I get dupes. EG: Somebod
in Hub 1 sends a message (which Hub1 will forward to 2,3,4), and Hub 2 sends me a message, and as well as Hub 3 (since Hub 3 wont know that 2 sent it to me).

Yep, being at the top there, you'll get upto three copies of some messages.. :)

--- BBBS/Li6 v4.10 Toy-4
* Origin: The Rusty MailBox - Penticton, BC Canada (21:4/106)

Some months back in Fido there were dupe messages from me, this was traced
back that one of the systems in the fidoweb was processing my message and due to some changes was not able to be seen as a dupe as normally expected.

Unfortunately I cannot remember the software that was doing this.

Terry

On Apr 01, 2020 02:19pm, alterego wrote to Al:

Re: Re: Oli's Messages
By: Al to alterego on Tue Mar 31 2020 08:19 pm

Well, in theory it shouldnt right?

Yes it will. :)

Well, I guess in theory there could be scenarios that I get dupes. EG: Somebody in Hub 1 sends a message (which Hub1 will forward to 2,3,4),
and Hub 2 sends me a message, and as well as Hub 3 (since Hub 3 wont
know that 2 sent it to me).

I was coming from the scenario that somebody on Hub 3 (Oli/Spectre, etc) sends a message, Hub 3 will send it to me, and Hub 2 shouldnt (since the copy that Hub 2 gets from Hub 3 will have my 2/116 in the SEEN-BY).

Now the fact that I was seeing dupes from Specture (3/101) with the soft-cr's, means that Hub 2 must have been ignoring me listed as a
SEEN-BY (and thus I was getting the modified message) - and the dupe checking had it as a new message (because the soft

Sadly, since the author doesnt think this is a valid usage scenario
(based on email thread that occured about it), it may never be fixed :( ...deon

... File not found. Should I fake it? (Y/N)
--- SBBSecho 3.10-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

... Platinum Xpress & Wildcat!..... Nice!!!!
--- Platinum Xpress/Win/WINServer v7.0
* Origin: The File Bank BBS! https://tfb-bbs.org (21:5/101) (21:5/101)

On Wed, 1 Apr 2020 14:30:55 +1100
"alterego -> Al" <0@116.2.21> wrote:

Well, in theory it shouldnt right?

Yes it will. :)

Well, I guess in theory there could be scenarios that I get dupes.
EG: Somebody in Hub 1 sends a message (which Hub1 will forward to
2,3,4), and Hub 2 sends me a message, and as well as Hub 3 (since Hub
3 wont know that 2 sent it to me).

I was coming from the scenario that somebody on Hub 3 (Oli/Spectre,
etc) sends a message, Hub 3 will send it to me, and Hub 2 shouldnt
(since the copy that Hub 2 gets from Hub 3 will have my 2/116 in the SEEN-BY).

Now the fact that I was seeing dupes from Specture (3/101) with the soft-cr's, means that Hub 2 must have been ignoring me listed as a
SEEN-BY (and thus I was getting the modified message) - and the dupe checking had it as a new message (because the soft-cr's were removed).

Sadly, since the author doesnt think this is a valid usage scenario
(based on email thread that occured about it), it may never be
fixed :( ..deon

My hope is that the new 256 char bug is a sign that he is working on removing the in-transit modifications of mails (and not a fuck you to long-lines or a completely unrelated bug).

---
* Origin: (21:1/151)

On 01 Apr 2020 at 09:34a, Oli pondered and said...

Sadly, since the author doesnt think this is a valid usage scenario (based on email thread that occured about it), it may never be
fixed :( ..deon

My hope is that the new 256 char bug is a sign that he is working on removing the in-transit modifications of mails (and not a fuck you to long-lines or a completely unrelated bug).

I am playing catch up. I agree things seem messy now. The current status is that NET 1 has a more recent build that 2 and 4. In the coming days I'll aim
to move all to the same build so we can get a level playing field to see
what's happening there. I don't wish to prejudge anything that is currently amiss.

--- Mystic BBS v1.12 A46 2020/03/26 (Windows/32)
* Origin: Agency BBS | Dunedin, New Zealand | agency.bbs.nz (21:1/101)

On Wed, 1 Apr 2020 20:46:16 +1300
"Avon -> Oli" <0@101.1.21> wrote:

On 01 Apr 2020 at 09:34a, Oli pondered and said...

Sadly, since the author doesnt think this is a valid usage
scenario (based on email thread that occured about it), it
may never be fixed :( ..deon

My hope is that the new 256 char bug is a sign that he is
working on removing the in-transit modifications of mails (and
not a fuck you to long-lines or a completely unrelated bug).

I am playing catch up. I agree things seem messy now. The current
status is that NET 1 has a more recent build that 2 and 4. In the
coming days I'll aim to move all to the same build so we can get a
level playing field to see what's happening there. I don't wish to
prejudge anything that is currently amiss.

Please don't. Mystic BBS v1.12 A46 2020/03/26 (Windows/32) is horribly broken. You should downgrade or test a newer build first.

---
* Origin: (21:1/151)

Well, I guess in theory there could be scenarios that I get dupes. EG: Somebody in Hub 1 sends a message (which Hub1 will forward to 2,3,4),
and Hub 2 sends me a message, and as well as Hub 3 (since Hub 3 wont
know that 2 sent it to me).

I got a dupe of this message. Insert rofl emoji here.

--- Mystic BBS v1.12 A46 2020/03/18 (Linux/64)
* Origin: monterey bbs (21:1/168)

The specific line wher eyour message was cut is below, and is the reason I
got a dupe:

SEEN-BY (and thus I was getting the modified message) - and the dupe checking had it as a new message (because the soft
Sadly, since the author doesnt think this is a valid usage scenario
(based on email thread that occured about it), it may never be fixed :(

--- Mystic BBS v1.12 A46 2020/03/18 (Linux/64)
* Origin: monterey bbs (21:1/168)

Re: Re: Oli's Messages
By: ryan to alterego on Wed Apr 01 2020 01:34 am

The specific line wher eyour message was cut is below, and is the reason I got a dupe:

...deon
--- SBBSecho 3.10-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

Re: Re: Oli's Messages
By: ryan to alterego on Wed Apr 01 2020 01:34 am

The specific line wher eyour message was cut is below, and is the reason I got a dupe:

Are you saying my message was chopped - so the reason you got it twice was because one copy was different to the second?

Can you quote the paths of the good and bad? Since you are on Hub 1, I'm thinking that its a Hub 1 thing (vs me) - since I havent changed anything (that
I can remember) for weeks now (and even then that changed only changed inbound
to me, not outbound I thought).
...deon


... Internal consistency is more highly valued than efficiency.
--- SBBSecho 3.10-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

Are you saying my message was chopped - so the reason you got it twice
was because one copy was different to the second?

Yep. Turns out it happened with this one as well :) I'll paste the relevant bits in the next two messages.

@PATH: 2/116 100 1/100

Can you quote the paths of the good and bad? Since you are on Hub 1, I'm thinking that its a Hub 1 thing (vs me) - since I havent changed
anything (that I can remember) for weeks now (and even then that changed only changed inbound to me, not outbound I thought).

Notice the path and the text above.

--- Mystic BBS v1.12 A46 2020/03/18 (Linux/64)
* Origin: monterey bbs (21:1/168)

@PATH: 2/116 3/100 1/100

Can you quote the paths of the good and bad? Since you are on Hub 1, I'm thinking that its a Hub 1 thing (vs me) - since I havent changed
anything (that I can remember) for weeks now (and even then that changed only changed inbound to me, not outbound I ...deon

--- Mystic BBS v1.12 A46 2020/03/18 (Linux/64)
* Origin: monterey bbs (21:1/168)

Perhaps that's an opportunity to improve things? ...deon

The moral of the story is, someone has to be at the top of the tree :)

Spec


*** THE READER V4.50 [freeware]
--- SuperBBS v1.17-3 (Eval)
* Origin: Scrawled in haste at The Lower Planes (21:3/101)