No problem! Thanks for the help. You can post anything from any of the test games.

I'm just glad I'm not the only one seeing this hiccup.

How repeatable is this? If I log in to one of your accounts, should I be able to see this first-hand? Or is it pretty rare?

I see it all the time on game M, but not on C or L. Try that one first. I think M is the only game cosby is playing where he was able to replicate it.

Now, like Big D said, it may just be our telnet clients and not the game. I'm not the one to say for certain.

The thing to do wouldl get someone to make you a script to capture the data and save it to a text file, then compare the screen output to what is saved to file. That will tell you if it's an actually a data output error or screen ansi error.

Ok, I did a test displaying Cruncher's fig report twice on game M and didn't see any glitches. Two tests isn't much depending on how often it shows up. But it's also possible that my network conditions aren't suitable for recreating the issue. So like BigD said, it would help to verify that the data cap is the same as the screen output.

I just logged back on and ran the fig report and saw it again.

How can I run a data cap? If anyone has the script for twx or zoc, please send it on over.

Thanks

Can you use a standard telnet app like Windows telnet or putty and cap output with one of those? That would give us the least variables between TWGS and your output. If an output log for Windows telnet shows this glitch, I'd be pretty sure it's coming in as bad output.

Turn on logging in twx, it will record everything it sees. If memory serves me correctly it dumps it into a straight text format so you can open it in notepad.

Thanks T0yman. I've sent JP logs from Hyper Terminal and the TWX log file.

Both look like gibberish to me, but maybe he can see something in them that may help.

I looked over those cap files, Cruncher. I did see an occasional output glitch, but it was always the same thing, a dropped LF (character 10) following an ANSI reset. This is part of a legacy output sequence that I wish I could clean up, but it would effect helpers. At the start of every line of output, TW sends a CR (character 13), ANSI reset sequence (ESC-[0m), then an LF (character 10). This is an unusual sequence, especially with the CR/LF split like that, so it could be a problem with how that sequence is being filtered from TWGS to the client. I put in code to detect for the loss of this LF in the data stream on TWGS to verify that it's coming out of TW this way. If you can do some display tests again, I'll be able to tell if this glitch shows up.

Here's more examples of the glitches I'm seeing:

20928 1 Corp Defensive N/A
20929 1 Corp Defensive N/A
[1m N/A
21003 1 Corp Defensive N/A
21004 1 Corp Defensive N/A


21159 1 Corp Defensive N/A
21160 1 Corp Defensive N/A
21161 1 21492 1 Corp Defensive N/A
21493 1 Corp Defensive N/A
21494 1 Corp Defensive N/A


22602 1 Corp Defensive N/A
22603 1 Corp Defensive [1;36m 22698 1 Corp Defensive N/A
22699 1 Corp Defensive N/A
22700 1 Corp Defensive N/A

I'm trying hard to come up with a way to determine if you're really receiving output this way or if it's getting garbled somewhere after it leaves twgs. I think the best, cleanest test I can think of is using Microsoft Telnet. Not hyperterminal. You should be able to type "telnet twgs.classictw.com" from a C prompt and open a connection. If that doesn't work, you'll need to activate Telnet client for windows. I can help you with that.

The first thing I'd do is just watch the fig display and see if you can spot any glitchy displays like you've been seeing. If you can't spot something visually, you can use the -f option to specify a log file.

If Windows telnet exhibits this glitchy output, I'll continue to pursue this. Otherwise, without anyone else reporting this, and without being able to find any issues in extreme stress testing, I'll have to ignore this issue.

Well, you can teach an old dog new tricks... never knew telnet was available from the C prompt. If I had known that, I wouldn't have played via the java client when I first came back.

All looks good JP! Thanks!

If I remember correctly Windows 7 and Windows Vista you have to enable telnet client from within the add (control panel) programs section. I think by default they were disabled. Microsoft is not going to make it easy.

Ah.. I'm still running XP.