Zugg Software :: View topic

Wizard Joined: 25 Mar 2003 Posts: 1113 Location: USA

I don't know if this has been reported/discussed elsewhere recently. Basically, I'm seeing parts of ATCP codes coming through from the MUD and being displayed in the output window, usually when I send a whole string of commands and get back a sizeable chunk of text.

I've got an alias to remove, give, drop, and/or wear some of my artifacts. Just for testing, I executed it and tried to hand them all to myself. I knew it would give the "no one here by that name" messages, but it makes the ATCP bug appear pretty reliably for me.

Posted: Wed Mar 19, 2008 8:00 pm

First, I assume you have MCCP turned off when using ATCP?

Do you have a debug output from the first case that contains the "i" record (showing the raw MUD data)? Or is the problem that you never get an "i" record when it fails?

Wizard Joined: 25 Mar 2003 Posts: 1113 Location: USA

Yes, the first thing I did was double check that I have MCCP turned off because I know about the issues involving it and ATCP.

When I see the misaligned output, I don't get the 'i' record in the debugging window.

Posted: Thu Mar 20, 2008 12:23 am

I've been seeing this as well but I was never able to reliably reproduce it. Good work re-producing it.

Posted: Fri Mar 21, 2008 4:51 pm

If you are able to reproduce this, can you email me the raw log from #DEBUGFILE:

#DEBUGFILE test.raw test.txt

then do whatever on the MUD to make it happen and then email the two files to sales@zuggsoft.com. That will allow me to look for network packet boundaries.

Wizard Joined: 25 Mar 2003 Posts: 1113 Location: USA

Done! E-mail subject is the same as the topic for this thread.

Posted: Sat Mar 22, 2008 4:00 am

Hmm, haven't gotten any files yet. Did you send them to sales@zuggsoft.com?

Wizard Joined: 25 Mar 2003 Posts: 1113 Location: USA

Yes, I did.

Posted: Mon Mar 24, 2008 4:24 pm

Try sending them again to support@zuggsoft.com to see if that works any better. I checked our email logs on sales@zuggsoft.com and didn't see anything, so it doesn't even look like it was our local spam filter. But I know Vijilante has run into trouble when sending us attachments too. Make sure the test.raw and text.txt files are both attached to the email message directly...don't try to put them into any sort of zip archive or anything. Then create a second message where you put both files into a ZIP archive and send that. Between the two methods, I should get one of them.

And someday I wish I could figure out why some people have trouble sending me email like this...it's very worrisome.

Wizard Joined: 25 Mar 2003 Posts: 1113 Location: USA

Okay. I've sent two more e-mails to the support address, as requested. If you still don't get the files, I suppose I could post the logs up here...

Posted: Mon Mar 24, 2008 11:44 pm

Trouble is that the forums eat some of the characters in a raw log. You'd need to use yousendit.com or something (please god, not rapidshare).

Posted: Wed Mar 26, 2008 5:01 pm

Not sure what is going on with email. I haven't received anything from Larkin to either the sales account or the support account. I've looked at our email log files and nothing is even shown as being received. So it doesn't look like it's even getting to us or getting to the spam filter. This is similar to the problem I had with Vijilante (although recently I've been getting files in email from him without any problem). But I never figured out what was causing the problem. Maybe it's a problem with the ISP you are sending from somehow?

Anyway, the TEST.TXT file is human-readable and you should be able to post it to the forum. And for looking for packet boundary issues, that should be just fine. The "raw" file is only used for trying to replay the file using the #READ command, but that handles it all as a single chunk and doesn't recreate packet boundary problems.

It would probably also be useful to see the Script Debugger output that is associated with the specific TEST.TXT file that you capture.

Wizard Joined: 25 Mar 2003 Posts: 1113 Location: USA

I'll do something similar to what I did for the ANSI issue I just reported then. I'll try to capture the debug files and a screenshot and then post them on my web site.

As for the e-mail, I'm just sending from my Gmail account, and I send files just fine to many other people, so it sounds like your ISP is the one to investigate, maybe.

Posted: Thu Mar 27, 2008 8:06 pm

Well, I probably don't need the test files in this case. I just verified that if there is a packet boundary anywhere in the ATCP text, then CMUD terminates the ATCP and treats the following packet as normal text.

You'll notice that in one of your lines above, you got:

You cannot see that being here.
6279h, 3608m, 4498e, 10p, 25340en, 13730wH:6279/5460 M:3608/3138 E:4498/3912 P:10/10 N:25340/25340 W:13730/13730 elrxb-

and you'll notice that the ATCP text is missing the normal "NL:15/100" part. This looks like there was a packet boundary after the NL part, but before the "H:" part. So CMUD got the partial Vitals data, then a packet boundary, and then treated the rest of the Char.Vitals as normal text.

This is a bit tricky to fix, but I hope to have it done for the next version.

Posted: Thu Mar 27, 2008 8:09 pm

Posted: Thu Mar 27, 2008 9:07 pm

OK, I think I have the ATCP packet boundary stuff fixed. Man, this is always the trickiest thing to implement. It's one of those little details that I think a lot of "simple" MUD clients forget about. The fact that the network stream can end at any time, and between the time you receive the first packet and the time you receive the next packet, something else might be displayed to the screen (like a trigger #SHOW or something). So you have to save your current parsing state and then resume it when you get the next network packet.

And yet, at the same time, you still have to handle MUD prompts which just end with no warning (unless the MUD uses EOR or GA).

All of the older code already handled packet boundaries for ANSI, MXP, MSP and the other protocols. But when I added the Telnet Trigger support (and ATCP), I forgot to save/restore the IAC Suboption parsing state to handle this.

Wizard Joined: 25 Mar 2003 Posts: 1113 Location: USA

Sweet. I'm really looking forward to 2.21 now. Nice work!

Posted: Fri Mar 28, 2008 1:42 am

On a slightly related note I've noticed that trying to capture the ATCP codes and then display them (so that I can see what's being sent) doesn't work out quite how you would think. When I was first playing with them I'd print them out and it'd do all sorts of wacky things. I just figured something with ATCP wasn't conducive to displaying them and did my experiments by saving the text to a string list and examining the string list.

Anyway, looking at your explanation I was wondering if these were related. Also, if anyone else had a better way of experimenting with the codes...

Wizard Joined: 25 Mar 2003 Posts: 1113 Location: USA

The ATCP messages are sent at odd intervals sometimes, so displaying them as you receive them may yield unexpected results. The best way is to capture the ATCP values in variables and have some static method of displaying them, such as buttons (gauges) or the status window. Apart from that, they're really most useful in making sure your stats values are up to date when you do your curing and whatnot.

Posted: Fri Mar 28, 2008 3:32 am

Yeah, I was just finding that I was receiving some items (such as exits information) multiple times during a move. I was just attempting to display them so I could tell what order they were occurring in and exactly how many times they were being sent. Storing them in a string list accomplished that to some degree in that it gave me the quantity and order, the only thing I lacked was where in relation to the actual room name and such they were being sent.

Now that I mention it, it seems all this should be available in the raw log. I think I might try that.

Edit: Perhaps with the 2.21 changes I'll be able to display them directly. I'll have to test it out on the next version.

Posted: Fri Mar 28, 2008 7:58 am

I was told back when telnet triggers were first added that since ATCP codes are invisible but in the middle of lines of other text, using commands that display things will cause problems. Imagine what would happen if you tried to #sub stuff. It works well enough just to find out what the text is (it normally does manage to display all of it) and after that you can turn it off. If you're really concerned you could try raising an event or something to do the saying for you.

A more reliable way to see what's being sent (including separator characters) is with #debugfile - just start it before you connect and you should be able to watch the whole thing unfold.

Posted: Fri Mar 28, 2008 4:42 pm

You can also see the raw ATCP stuff in the Script Debugger window if you turn on the Raw Input message. Same as #debugfile.

Posted: Sun Apr 06, 2008 6:53 am

Version: 2.22

Looks like there may be a corner case this isn't being handled in the new 2.22 version:

What I see:

4730h, 6655m, 20396e, 26400w cexdb-se

Central Market of Ashtan.
A sewer grate looms darkly beneath your feet. A cute little humgii is sitting here placidly, attached to a leash. A runic totem is planted solidly in the ground. A gleaming wooden sign is posted here. There are 7 Knights of the Illuminati here. There are 8 monolith sigils here. There are 6 War Witches here. Seething with primal energy, the Well of Chaos stands here firm, a circle of divine power. Sprightly balancing on his hands, an Ashtan Joker cavorts here in a playful manner. A Zarathustran eagle perches atop a monolith, a golden plaque clutched within its talons. Brandishing a razor sharp blade, a Freedom Fighter stands here, twirling his long mustachios. Leaning against the wall, a Soldier of the Golden Apple Corps watches the goings on of everything around him. Plucky Duck Miwoe is here. He wields a small blackjack in his right hand. You see a sign here instructing you that WARES is the command to see what is for sale.
You see exits leading north, northeast, east, southeast, south, southwest (open door), west, and northwest.
4730h, 6655m, 20398e, 26400w55 E:20398/20400 W:26400/26400 NL:0/100 cexdb-s

South of Central Market.

From debugfile:

Posted: Sun Apr 06, 2008 7:00 am

Logged in with a clean session, no triggers:

Central Market of Ashtan.
A sewer grate looms darkly beneath your feet. A cute little humgii is sitting here placidly, attached to a leash. A runic totem is planted solidly in the ground. A gleaming wooden sign is posted here. There are 7 Knights of the Illuminati here. There are 8 monolith sigils here. There are 6 War Witches here. Seething with primal energy, the Well of Chaos stands here firm, a circle of divine power. Sprightly balancing on his hands, an Ashtan Joker cavorts here in a playful manner. A Zarathustran eagle perches atop a monolith, a golden plaque clutched within its talons. Brandishing a razor sharp blade, a Freedom Fighter stands here, twirling his long mustachios. Leaning against the wall, a Soldier of the Golden Apple Corps watches the goings on of everything around him. Chaote Borla is here. Nex is here. You see a sign here instructing you that WARES is the command to see what is for sale.
You see exits leading north, northeast, east, southeast, south, southwest (open door), west, and northwest.
4730h, 6655m, 20398e, 26400w26400 NL:0/100 exdb-

4730h, 6655m, 20400e, 26400w exdb-
Nex leaves to the southeast.
4730h, 6655m, 20400e, 26400w exdb-nw

Northwest of Central Market.
A runic totem is planted solidly in the ground.
You see exits leading north (open door), east, southeast, south, west (open door), and northwest.
4730h, 6655m, 20398e, 26400w exdb-se

Central Market of Ashtan.
A sewer grate looms darkly beneath your feet. A cute little humgii is sitting here placidly, attached to a leash. A runic totem is planted solidly in the ground. A gleaming wooden sign is posted here. There are 7 Knights of the Illuminati here. There are 8 monolith sigils here. There are 6 War Witches here. Seething with primal energy, the Well of Chaos stands here firm, a circle of divine power. Sprightly balancing on his hands, an Ashtan Joker cavorts here in a playful manner. A Zarathustran eagle perches atop a monolith, a golden plaque clutched within its talons. Brandishing a razor sharp blade, a Freedom Fighter stands here, twirling his long mustachios. Leaning against the wall, a Soldier of the Golden Apple Corps watches the goings on of everything around him. Chaote Borla is here. You see a sign here instructing you that WARES is the command to see what is for sale.
You see exits leading north, northeast, east, southeast, south, southwest (open door), west, and northwest.
4730h, 6655m, 20396e, 26400w exdb-
ÈRoom.Brief Central Market of Ashtans
South of Central Market.
A runic totem is planted solidly in the ground.
You see exits leading north, northeast, east, south (open door), southwest (open door), and northwest.
4730h, 6655m, 20398e, 26400w exdb-

Posted: Mon Apr 07, 2008 12:27 am

Another example in case this has a variety of mechanisms causing it:

Entering the Main Gate of Ashtan.
The city gates of Ashtan stand proudly here, black and gold banners waving defiantly at those who would challenge the Bastion of the North. A runic totem is planted solidly in the ground. A cute little humgii is sitting here placidly, attached to a leash. There are 8 Shadow Amazons here. A gleaming wooden sign is posted here. There are 7 monolith sigils here. Overflowing with blood and other trophies of the dead, a sacrificial altar to the Lord of Warfare has been erected nearby. Polished marble gleaming, the Effigy of Victory resides here, a testament to the favour of the Great Mother. There are 10 War Witches here. Ever-vigilant, a Zarathustran eagle monitors the area for threats. An obsidian eye sigil is here. A Rajamalan woman has been nailed upon a ridged, bloodstained cross and left to perish. Standing still as a statue, a blood steed has fixed its fiery red eyes on some movement in the distance. A black marble bench is here.
You see exits leading north, northeast, south, and northwest.
5060h, 6655m, 21900e, 26400w NL:0/100 cexdb-