Microsoft Streets and Trips crashes unpredictably

I have run Microsoft Streets and Trips 2001, 2003, 2004, 2005 and now 2006, at different times on four different model Dell Laptops running Windows 98se or 2000, ME, or XP pro now with SP3.

The original Streets & Trips worked perfectly, but didn't update GPS fast enough (yes I found the patch), and the maps became outdated. When I began updating to more "advanced" versions I also began trying to use newer USB connected GPS(s) with the programs. One reason is the USB connected GPS uses the port for power and therefore does not need to be turned off and on separately. After fooling with this for almost six years now I am convinced that there is a problem with S&T's handling of the GPS data.

S&T will crash when running with GPS support at TOTALLY unpredictable times. Sometimes once a day, sometimes every 5 min. If I stop in the middle of everything and disconnect the USB GPS and connect one of the older serial connected units, the program will not crash as often.

I have three Garmin serial units (puck, Etrex and Legend), and two USB GPS "pucks" an Altina and a Holux. When the system does not crash, the USB units perform as well as the serial ones. One suspicion is the software conversion of the USB input to a "virtual" com port is the problem (I've tried several). Is there approved software for this purpose? Or is there another way of doing this. Is something else going on here? Is this a known problem? Is there a fix for this? The latest crashes will occur almost reliably whenever the zoom is at 9 mile ..not 8...not 10

One other less important problem is the map data from earlier versions of Streets & Trips seems never to import properly (even from 05 to 06). Weird addresses appear attached to pushpin locations. In a Microsoft tech support discussion we concluded that the "corrupted" map files were causing some problems, so I now have to start new map files whenever I upgrade S&T versions. Which means entering ALL data again to be sure I am not corrupting the new map file. Even when I have done this the strange names and addresses will sometimes attach themselves to existing pushpins. On one occasion I found all addresses on my pushpins were now the same. S&T is now unusable as far as I am concerned.

I have been reduced to using DeLorme which runs without crashing, but the maps are not accurate with GPS. I think I've done all the things suggested by Microsoft and anywhere else on the web I could find. Hardware acceleration off, update display, even changed the memory stick in the last laptop. Nothing works for long. Just about the time I think it's OK and I can trust it, the crashes start again. Do others have these problems, or suggestions?

Hi juiceclone,

Welcome to Laptop GPS World.

This may or may not help, but try to place Franson GPS Gate in-between your GPS receiver and your Streets & Trips. See if it helps in any way.

When Microsoft Streets and Trips crashes, does it display some error message?

How many hours do you need to use Streets & Trips before the crashing starts?

Check the following link regarding the second issue: Problem opening an old file in a new Streets & Trips version - and a solution

Yes I'm looking at the usb Franson gate now, deciding whether to buy it ...not that cheap. I tried a free prog like that about 3 years ago but don't think it was working correctly. As far as when S&T crashes ..it could be within a minute of starting or I could run it for days over the same route and not see any problems and then next day it would crash every few minutes.. Over the different versions and op sys I have used, the crashes were totslly unpredictable both in reason/operating conditions and the way it crashed. Sometimes all the street names would disappear, or the map would disappear and usually the program would exit shortly after. On a few occasions the laptop would shut down and reboot without warning.....Keep in mind this has been on three laptops ,each was used daily until I upgraded to the next. Currently I can reliably crash it by zooming to 9 miles...that will usually crash within 20 seconds...but that seems to be related to movement. No useful error message, just "MS S&T has encountered a problem and must close..send report yes/no restart yes/no." I use process explorer to shut down prgs non essential, but makes no difference.
I'm in the process of hooking one of my gps's out a window to different desktop pc's in the house to see if the problem shows up on them as well. Can't move the house but all gps fixes wander a little.

I'll look at the file problem next.
Thanks
.

It is rather strange for Streets and Trips to cause Windows to restart. It's simply not that deeply embedded into the OS. S&T may crash itself, but it should not crash Windows except under fairly unusual stresses.

This suggests either a hardware problem or a driver problem. Both are doubtful as you say it happens with all combinations of laptops and GPS units. There must be something unusual about your setup -- if it really crashes that much for everyone, then nobody would buy S&T.

P.S. You should always click "Send" when the Send/Do-not-send dialog comes up. (Exception: when you're, say, calculating your tax return.)

P.P.S. Paragraphs are the greatest thing since sliced bread. They make your writing ten times easier to read.

I agree on all counts...this should not be happening...I am beginning to wonder if there is something in the Dell laptop design causing this.
As far as the "send" goes. I usually do click send but have to wonder where it's going..I'm on the road and have removed the wireless card (eliminating aything that could cause a problem) but message still says "completed" ....go figure...

I have also noticed, since now running process explorer at all times, that CPU usage goes to approx 100% when crashes occur and does not drop till S&T closes. Post on another forum relating to this says "Microsoft uses a polling loop rather than interrupts to get GPS data," .

I don't know exactly what that means but it could be the reason DeLorme never crashes and S&T does ?

Unlikely. If that were the problem, then S&T would be eating 100% CPU (or 50% on a dual-core CPU) all the time, rather than just as the crash occurs. I also can't find that exact quote through searching -- are you quoting or paraphrasing?

The likeliest scenario is some kind of malformed data which S&T wasn't expecting, which causes it to get stuck somewhere. This does not explain the Windows crashes, only the S&T crashes.

Yes, a quote..in fact a ctrl/c copy and paste...I'll look for the forum and if I find it, include the url.

If the data or the delivery from the gps is intermittently corrupt, would not this show up in DeLorme ' or net stumbler or gps viewer also?

I don't worry about the complete reboot crash, that has happened about 3 times in 6 years. I only include it as information towards a diagnosis....

Hi
The url of the thread I quoted is:
http://forums.techarena.in/showthread.php?t=501234
should get you there

There is no reason that a polling loop would chew up an inordinate amount of CPU when it's working properly. In the case of a GPS receiver which transmits data, say, once every second the polling loop would only need to run once per second. It's such a tiny amount of code that each iteration of the loop would take a miniscule amount of CPU time to complete.

I've programmed dozens of data acquisition routines; probably split roughly equally between polling loops and interrupt driven, depending on the operating system and the driver and hardware capabilities. I was doing this back in the minicomputer days when most hardware devices didn't come with drivers because there were no "standard" operating systems. They were all proprietary to the computer they ran on. And even in those days of puny CPUs and system memory measured in kBytes there was little overhead for polling loops, properly coded.

In the context of modern CPUs there would be no visible difference on the CPU performance monitor between the two.

If the increase in CPU utilization is related to the data acquision routine - and it might well be - that's a symptom of either a buggy data acquisition routine or, more likely, something getting screwed up in the heap, a memory leak, or or something similar causing the data acquisition loop to go into hyperdrive.

...ken...

We're getting into the semantics of various computing terms here. What you're describing is timer-driven polling. A polling loop would be the stupidest form of while (1) {...}.

I read over the other post that juiceclone linked to. The guy is just speculating. In fact, it reads as though he's bumping into the same edge case that is being discussed in this thread.

Unfortunately, I rather doubt we'll find the cause of this problem going step-by-step like this. If MSFT tech support has already taken a crack at it, then we really don't have much to add. The exception is if someone happens to have experienced the same problem and found out how to fix it. These sorts of sporadic, unpredictable crashes are the hardest to troubleshoot.

Amen to that!!

...ken...

As I intended , I have connected both usb gps's to a desktop pc, A test bed using currently a net vista mb. Interesting ...it crashes also...The only common thread among all the crashes is the gps pucks.. I have begun to doubt their reliability.... You would have to live here in Florida to appreciate how damaging sunlight can be. I wonder if a year or so of exposure to that is destroying them. They seem to work well enough but after 10 to 60 mins a crash occurs and it is necessary to remove the usb plug and re-insert it to get anything working again. In any event, I have ordered newer gps, sirf III unit and when it arrives we'll see what happens.
At this point I'm relieved the laptop is not the problem.

juiceclone,

just to summarise your description of the symptoms:

1) Streets and Trips has been crashing for you unpredictably for years, even though you've used the following:

2) Microsoft Streets and Trips 2001, 2003, 2004, 2005 and now 2006,

3) at different times on four different model Dell Laptops

4) running 98se or 2000, ME, or XP pro now with SP3.

5) The original streets worked perfectly but didn't update gps fast enough (yes I found the patch) and the maps became outdated.

6) with USB GPS receivers it seems to crash more often than with serial units (It may not help, but I would try to use Franson GPS Gate, as suggested earlier. There is a 30 day free trial version available.)

7) DeLorme Street Atlas doesn't crash on the same setup

8) Streets & Trips has very high CPU usage (~95%) when GPS is enabled (which GPS features are need to be enabled for CPU usage to go that high?)

9) The latest crashes will occur almost reliably whenever the zoom is at 9 mile ..not 8...not 10 (Latest? When did you observe the relationship between the crashes and the zoom level?)

10) As far as when S&T crashes ..it could be within a minute of starting or I could run it for days over the same route and not see any problems and then next day it would crash every few minutes

11) Sometimes all the street names would disappear, or the map would disappear (Do you use the Create GPS Trail feature introduced in S&T 05?. The feature will cause 100% CPU usage after prolonged use, while driving. There is also some dependency on zoom level used. Symptoms include street names and GPS icon disappearing from map, as S&T takes very long to refresh map under such high CPU usage. Do not use the Create GPS Trail feature. If you do need to use it, then save (Ctrl+S) periodically to prevent high CPU usage from occurring.) and usually the program would exit shortly after. On a few occasions the laptop would shut down and reboot without warning. No useful error message, just "MS S&T has encountered a problem and must close..send report yes/no restart yes/no."

12) As I intended , I have connected both usb gps's to a desktop pc, A test bed using currently a net vista mb. Interesting ...it crashes also (Is the desktop also a Dell product, just like all the laptops?)


My guess would be that over the years there must have been various different reasons for the above mentioned problems. I don't think there is one common reason for all the issues. Perhaps one of the first things I would recommend would be to upgrade to the latest version of Streets & Trips, which is now S&T 08 - two years newer product than your S&T 06.