quote:Originally posted by Akor

I guess enough info has been gained for me to rap up this thread for the archives. I have found that currently:

3. When ztm is done, you export your TWX data using by right-clicking the TWX icon found on the lower right half of your screen in what is known as the Notification area or systray and mouse point to "Data" then "Export TWX File" then left clicking to export the TWX database to a file. The saved file will have a .twx extension. Then in SWATH, click "File" then "Import" to import that file into SWATH.


As long as the "Parse Incoming Text" button in SWATH is turned on (don't know why you wouldn't have it on) then SWATH "sees" the ztm in progress and you don't have to import the TWX data into it. You can watch the GUI map update as the twx ztm script works. That is as long as you are in computer mode. If using the interogation mode than the import method would be correct. You could use the twx import if you were swaping data between corpies. Corpies doing their own respective tasks collect different data and that is a way of updating everyones data. Make sense? [^]

quote:Originally posted by Akor
I really am starting to wonder now as to who gave it an award for the ztm. I wonder if the granters knew how to play the game. That is very misleading.


The trade wars awards are mainly for fun and frequently don't reflect reality.

On average, in a 5k sector game, rev's ztm will miss less than 10 warps doing 3 passes. I assume the same applies to supg's, since he did it the same way. In nearly every case, knowing those warps won't gain you any significant information. I guess if you had a script that would find them for not a lot of extra time mapping, it couldn't hurt to know them, but it seems to me that you could spend your time better watching paint dry than developing that script.

quote:Originally posted by Akor

quote:Originally posted by Kavanagh


The one closest to the Observer warps is the most accurate. Time taken is secondary, but clearly of interest.


Why the one closest to the Observer's ztm?

As for time, we have to factor ping time too. Players with poorer connects would of course have a slower ztm.


quote:
I can run one using twassist, but it cannot handle more than 5K sectors.


DOS Trade Wars Assistant? That brings back memories.


Ping time is irrelevant to ztm, unless you live on the Moon (2500 ms ping, line of sight assumed, and a decent laser to a commsat).
John Pritchet changed the CIM report speed, I assume to avoid the crashes we used have when several were ztming at the same time, pre twgs.

I use diallup. Even with a 300 ms ping, 5K takes 50 mins +/-5
It takes the same length of time if I bang a game on my local machine 127.0.0.1:xxxx, zero ping, unless one wants to consider micro or nanoseconds.

Speed is certainly a consideration, after accuracy, hence my suggestion that we set up a test game, NOT as a competition between players, just to check :

A) Accuracy
B) Speed

I wrote earlier that I use Twassist. I like it. I did NOT write it, no kudos to me if it the fastest/most accurate, no shame to me if it it is the opposite. Likewise the other ztms. I'm just curious as to which might be best.

A small codicil to what kav said. While connection speed shouldn't matter, if the ztm script it poorly written, it can make a significant difference. I would suspect swath's ztm is of that sort.

quote:Originally posted by Harley Nuss (teamEIS)

A small codicil to what kav said. While connection speed shouldn't matter, if the ztm script it poorly written, it can make a significant difference. I would suspect swath's ztm is of that sort.


Having reread my post, I was unclear. The intent was to say that latency (ping + microsludge garbage) is irrelevant to ztm speed in the case of comparing rev's at 50 ms ping with rev's at 300 ms, not when comparing different methods.

I do agree with what you wrote, a really poor algorithm would suffer with high latency.

Don't have SWATH, but have heard that the ztm problems were accuracy related rather than speed? IIRC SWATH can run REXX, dunno why users do not use Rev/Supg ztm with SWATH, mebbe it doesnt get into SWATH dBase?

This topic is confusing me. I haved used Swath ZTM on ver's 1.8, 1.7.1 amd have had no problems or inaccuracies. Running swath through TWX..last version. Have gone and checked after completion and have not found the results incorrect.. Used on a 20k and a 5k

Do a bubble search or dead-end search and then goto one of the given areas that you have yet to explore, you should find that it is not actually a deadend after all.

Kav, in TWAssist, you copy/paste the commands for several hundred plots at a time (if memory serves), and so they are queued against the server, avoiding any lag of any sort since the server isn't awaiting the next plot requests to come in. If a ZTM is written to send plot requests and await results before sending the next plot requests, it will suffer from slow ping times, and will take longer than necessary to complete. A ZTM written for time efficenty needs to que 2 to 3 plot requests on the server to avoid latency penalties. The reason to avoid significantly more (TWAssist) in the queue is because plot 1 may eliminate the need for plots 90-100.

One caveat - a good ZTM should have a one-way check at the end, once the passes are complete, and this will typically have no plots queued against the server. But since this check takes only a few minutes, only an insignificant penalty is incurred.

Regarding Swath's ZTM and its accuracy: missed warps are going to be difficult to pinpoint Sparky. Truly, unless you are running a TW server, you'll probably only find missed warps if you do a warp count comparison before and after probing the universe. Still, this is a quest for accuracy, and the method implemented by Swath just isn't as effective, efficient, or as accurate as Rev's approach (at 6 passes).

+EP+

Now my question is when to do the ztm. I usually do it right after I get all the sector info of FedSpace. But since it takes so long to do it I always get a late start. The main reason I do it is to prevent myself from warping into a bubble. When would be the best time to do a ztm?

I answered your question in your other thread..but to make it quick here, ZTM after you run your turns, that way you have a small grid already for later use, credits to get a ship/twarp and whatever else you want to pimp out your ship, and MCIC ports for later use.

I'm still not understanding why you want to prevent warping into a bubble. I'm sure i'm just missing something here and not understanding it, but what's the big deal? go in a bubble, go out of one.

I thought that one-ways were entrances to a bubble and a bubble is called that because it has only one way in to a bunch of sectors and no way out; like a bubble. Am I wrong?

quote:Originally posted by Akor

I thought that one-ways were entrances to a bubble and a bubble is called that because it has only one way in to a bunch of sectors and no way out; like a bubble. Am I wrong?


It is more like a balloon - you can usually go in and out of the same gate. It is also possible that there will be only a one-way in, but there will be at least one one-way warp out. If there is no way out, then you have ran into a black hole and will most likely be stuck.

               / Promethius / Enigma / Wolfen /

"A man who has no skills can be taught, a man who has no honor has nothing."

For the most part, you'll never even know that you've been in a bubble until you run the bubble finder in Swath and some of your figs show up in the bubble map. Bubbles present no danger on day 1, and generally present no danger anytime. They are certainly not a cause of alarm.

+EP+

heh now I understand why you were talking about bubbles. Ok. EP is right, it's no danger if you go into them. And what Promethius said about blackholes is VERY rare in a game. To my knowledge, I have never been in a game and found one unless the universe was edited by the sysop.

BTW, I like that idea of a bubble as a ballon. I'll have to remember to use that example next time...fits it perfect.