Workarounds for CoPilot's problems.

[BMWBiker]

I don't see much real disagreement here.

Many people find CoPilot excellent (I think superb) for quick, easy, accurate routing as long as you like the major highways and the quickest route. True, this may be better in USA than UK.

And many of us are deeply disappointed in the bizarre routings for 'shortest' and other bugs that have been introduced in V5.

Some are bitter about the promised Live Traffic that didn't arrive, but many of us were not very interested in that feature anyway.

The lack of response from ALK is astonishing.

I recommended V4.5 to my friends, but when asked for a GPS recommendation now I just say I don't know.

But I need CoPilot, almost every day. It fulfills most of my needs and I am still ecstatic about most of the features.
I don't have time or money to buy and learn a new system, and for me CP 4.5 was just about perfect. That gives me hope that the bugs will be corrected, though I would have much more hope if they were acknowledged.

At this point I hardly care to complain or fix blame, I just want the bugs fixed SOON so I can get back to depending on CP the way I used to. I no longer care if that is a free patch for V5 or in a V6 that costs a little more, I gotta have it.

It has added a new dimension to motorcycle riding, I can't be without it now.

As for a counter-review, that sounds like a lot of effort and frankly I don't think I would find time even to read it. I would be more interested in seeing an accurate bug list, which we mostly have done.
_________________
Doug
CoPilot 6.0.1.24 (USA)
Compaq iPaq 3970, with PC2003
Navman sleeve (3400)
Blue Tooth GPS

[PONDEROUS]

It is because CP 5 will be very good, if it ever is fixed, that I, like BMWBiker have stuck with it, so I have no argument with him over that.

Having said this, I would be most unhappy on principle to pay for the fix. Also, although it would hardly break the bank for me, I can remember times when I could not have afforded it and not everyone will find the money easily.

[Skippy]

I also find that lack of and communication from ALK astonishing. These bugs have been well known for quite some time now and still no official word of a fix from ALK.

They are either in denial or they are quietly fixing the problems and waiting for CP6 to release the fixes. I will be very angry if they don't give a free upgrade to CP5 users after all the grief we have been though helping them beta test their buggy software.

And there is STILL no traffic integration.
_________________
Gone fishing!

[PONDEROUS]

Thinking again about the question of a user review, I think BMWBiker is right: it is too much effort – at least at this late stage, when, as Skippy suggests, the next version of Copilot may well be just around the corner.

Therefore, I think that the time for a user review is after some experience using the new version. At that stage, ALK will have had every possible opportunity to put their house in order - concerning both their software and their support service. Hopefully, the user review will be favourable, in which case I am sure that everyone will be happy. If not, then prospective purchasers will be able to make informed decisions.

User review or no user review, let no one be in any doubt. I and, I am sure, everyone else will say how successful ALK have been if and when they produce products and services that justify it.

[PONDEROUS]

BMWBiker mentioned earlier that routing was better on CP 4 on his laptop, and xda wanted to know if anyone had tried it on the pocket PC. I have now succumbed to curiosity and installed CP 4.
On the Scraptoft-Rearsby route mentioned earlier, CP 4 produces a distance of 7.7 miles, as compared with more than 10 miles for CP 5. This is obviously an improvement, but still unnecessarily longer than 6.6 miles for Tomtom 3 or Garmin and still suggesting a main road bias.

CP4 produces the correct route with the addition of a single waypoint wherever it is placed on the road between Hamilton and Barkby.

Having completed this exercise, I’m not sure what it tells us, other than that ALK have done better in the past.

[PONDEROUS]

Further to my post above, I think the most likely reason for the improved performance of CP 4 over CP 5 on this one route is different Road type bias settings; i.e. the infamous "tweaks" as discussed on the relevant thread.

The "tweak scales" in CP 4 are identical with those in CP 5, so apart from the settings it appears that CP 4 and 5 routing logic is going to be the same.

As there is no satisfactory combination of the road type biases in CP 5, I cannot imagine that there will be in CP 4. Therefore, I doubt that the solution lies in ALK reverting back to their CP 4 setup.

[BMWBiker]

Hmm, can't confirm or contribute, my laptop with CP4 is at home.

For benefit of newcomers, I think 'tweaking' can be summarized thus:

Under Trip Planner, Options, Advanced the user can exert some control over the route selection and time calculations. You can "favor" or "avoid" the different road types, and associate travelling speeds with each also, which are used to calculate ETA. The concept is good but most of us (I think) have observed bugs:
1. There seems to be no combination of "favor" and "avoid" that results in CP calculating the true shortest route.
2. Whatever road speeds you declare are valid only on the Desktop, these preferences are not transferred to the PPC.
3. (complaint from me only so far) The default roadspeeds for the rural roads are very low in CP 5, causing high time estimates for 'back roads'.

The default road speeds on the PPC are stored in the file deftrp.dat in a sequence of hex digits that represent the speeds for all the road types. "Tweaking" consists of setting the user-available options on the desktop as best you can, and then perhaps hacking the hex digits on the PPC to get more accurate routing or times there.

We discussed some tools to do the hacking, but lost interest when it was realized that all the 'tweaking' was not really solving the basic routing problem anyway.

Hopefully ALK in the next fixed version will be sure to transfer the default user preferences to the PPC.
_________________
Doug
CoPilot 6.0.1.24 (USA)
Compaq iPaq 3970, with PC2003
Navman sleeve (3400)
Blue Tooth GPS

[PONDEROUS]

That's a useful summary, BMWBiker. However, I think that it is a case of completely confused logic rather than a bug. In fact, trying to resolve the Shortest Routing problem by tweaking road type biases was never going to work.

If we see similar Favour and Avoid scales appearing in the next version, then we can expect trouble - unless truly shortest routes are produced when both Favour and Avoid are at zero for all road types. Also, the illiogical concept of % Favour and Avoid needs to be replaced by the choice of including or excluding particular road types - completely - in route calculations.

This issue has been discussed at length on the following thread:

Sticky: Copilot 5 ETA - Tweaking Road Speeds

[PONDEROUS]

I said in the above post that the road bias issue was discussed at length on the following thread:

Sticky: Copilot 5 ETA - Tweaking Road Speeds

I should have added that after the above mentioned thread was locked, the discussion conrinued on the following thread:

COPILOT 5 UK ROUTING

[PONDEROUS]

I have just tried the Scraptoft- Rearsby route on the PC version of CP 4 on my laptop and found that it produces a 10.3 mile journey - exactly the distance produced by CP 5.

I believe that this is conclusive; i.e. it confirms that the routing logic in CP 4 and 5 is, at least practically, the same and that the differences reflect where the road type biases are set on the "tweaking scales".

Therefore, I don't think there is any "mileage" in pondering further whether CP 4 is the key to satisfactory routing in Copilot.

[BMWBiker]

I had seen a clear difference last week, where CP4 took the desired short route on a small road in Pennsylvania USA, and CP5 could not be induced to take that road unless I dropped a waypoint on it.

Will have to re-try in a day or two when I get my laptop next to my desktop again.

BTW, I think the laptop is running CP 4.5.

Meanwhile I accept your finding Ponderous, though it puzzles me. Thanks for running the test. I will have to re-install CP4.5 UK maps to try the same.
_________________
Doug
CoPilot 6.0.1.24 (USA)
Compaq iPaq 3970, with PC2003
Navman sleeve (3400)
Blue Tooth GPS

[PONDEROUS]

Thanks, BMWBiker. I accept yours too.

All either of us can do is try it out, and that's what I did. The version on both PC and PPC was 4.0.4.107

It will be interesting to see your own results, but I think you can see what the conclusion has to be.

[PONDEROUS]

Apologies, BMWBiker. After checking again, I found that somehow I did the test with the PC version set on Quickest rather than Shortest.

Having corrected that, I get the same results for both PPC and PPC versions of CP 4. That is 7.7 miles from Scraptoft to Rearsby.

[PONDEROUS]

Having regrettably allowed my mistake to cloud what BMWBiker was saying (please see above post) I thought I should state the situation more fully.

The version of CP 4 that I tried out, i.e. 4.0.4.107, does produce better results on the Scraptoft – Rearsby test route, but it is still longer than necessary and still suggests a main road bias. I have now examined the settings on CP 4’s road type bias scales between 100% Favour and 100% Avoid and they are all at zero. This may mean that there is a calibration error that is much worse in CP 5 than CP 4. If so, it may be possible after all to calibrate a version based on CP 4 or 5 that will consistently produce Shortest routes. Perhaps V4.5, which I believe BMWBiker still intends to try when he has time, is such a version.

All of this is a big IF, as there could equally well be flaws that cannot be fixed by calibration.

One change that may not be achieved through calibration is calculation of Shortest routes regardless of what road types are included or excluded. If, for example, all roads except motorways are included, the Shortest route needs to be the shortest given that exclusion. Users need these options so that, for example, learner drivers can avoid motorways and heavy truck drivers can exclude very narrow or unmade roads. The way Copilot seems to have Shortest and Avoid and Favour all jumbled up together, it may not be possible to separate them without a complete re-write of the programme.

Re-write or not, the existing “tweaking scales”, between 100% Avoid and 100% Favour for each road type make no sense. What use is it to know that your route will avoid 30% of narrow roads or favour 50% of motorways? These scales have confused matters even more when taken as a means of adjustment to achieve Shortest routes. They need to be replaced with a simple “Include” or “Exclude” for each road type. Then, Copilot can calculate the Shortest or Quickest route accordingly.

There’s still a lot for ALK to be thinking about – and acting upon. I wonder if they are?

[lynnk]

Hello fellas ... allow me to throw in my 2 cents. I'm the bloke that started the ETA Speed Tweaking thread a long time ago. 8)

I've been following the discussions closely but staying quiet. After trying a lot of of different navigation packages, I settled on CoPilot as the product that had the most things I liked. Clearly, the product has some deficiencies and it effects each of us differently, depending on how we intend to use it.

I live in Maryland USA. My use has been typically for trips where I am going about 100 miles and need guidance to an end point destination. I have not wrestled much with shortest or quickest. In my case, both have given me essentially the same answer. I look at the map, decide how I want to go, go to the desktop and put in the necessary waypoints to take me on my desired route. It has always worked well for me when I use it that way. I did have some initial grips about the ETA being way off (hence the thread I started) but have since found that a few tweaks to the files in the PDA and it provides me with fairly accurate ETA's now. (A recent 3 hour trip was only off by about 10 minutes right from the start.)

My biggest gripe these days is the POI notification (or lack of it.) I would be a happy camper if they would fix that problem. And I agree about ALK support .... they are really hurting themselves by failing to communicate with their users. Just a few encouraging words would go a long way.

Keep up the good discussions ... who knows what eyes are watching .. maybe even ALK ????????

We see you’re using an ad-blocker. We’re fine with that and won’t stop you visiting the site.
Have you considered making a donation towards website running costs?.
Or you could disable your ad-blocker for this site. We think you’ll find our adverts are not overbearing!