OpenCPN Version 2.1.0 Beta

[Algarve]

Hello and greetings to all.
Can someone tell me why the new version is so unstable?
The softwear is always going down?
Is anything wrong?

[raisin56]

Algrave
I'm pretty sure the problem is the four letters right after the version number.

Dan

[svaletheia]

LeaseOnLife and Cagney, yes the bells are for station-keeping use. It seems as though they simply don't exist, but the program should be playing them, increasing every half-hour to 8. See Watch Schedules and Ship's Bells - BoatSafe.com for some info.

MarkJ, one thing to keep in mind is, if your computer is asleep, by definition it will not be keeping a track. Period. Your computer MUST be on 100% of the time in order to process the incoming datastream from the GPS. This has the side effect of not disconnecting the GPS. Sleep mode on your computer doesn't just shut the display off - it literally shuts down the CPU. So the program is not running when its hibernating or sleeping. What you MIGHT want to consider is a modified power setting where the display shuts off (one of the biggest power consumers by far), the cpu scales back its frequency, and the hard drive spins down. But the computer should NEVER sleep or hibernate if you wish to keep your track plotted, the GPS connected, etc...

Another issue is that IF the computer does sleep/hibernate, many of them reset the I/O ports on awaken. This is sometimes a hardware issue, and while there may be a workaround that OpenCPN can attempt, it will not be guaranteed to work, if it exists at all. Sometimes its just the way your particular hardware behaves on sleep/resume. My Apple Macbook, for instance, typically does NOT do this, although I have seen certain USB-serial devices do this running on ALL computers/OS, as the USB-Serial converter typically loses power on sleep/hibernate and therefore the microproc inside it shuts down. Again, its sometimes hardware dependent, sometimes the fault of the host OS or its configuration. Its almost never the fault of the running application, although there is a small chance that the app can put the port into a mode that won't behave that way - again assuming such a mode exists for the hardware you have.

That puts what Hamish said in a little more context as he's using some properties of the Host OS's configuration to try and get around what's going on.

[MarkJ]

I have found the latest Beta version quite stable. It does close down when I click Course Up, but apart from that I find it fine for stability

[nohal]

Originally Posted by MarkJ

1) Hibernation/Sleep makes the GPS disconnect.
I NEED the GPS connection to keep connected in a low power state.
tell me what power state and how to adjust it. Remember at sea electricity is a problem. Computers are best not on 24/7

Me too, but unfortunately when the computer has to "do" something, it has to be turned on, which means not turned off, hibernated (=turned off) or sleeping (=turned off, but still consuming some power ) So bad luck here, the only solution is to turn off the automatic hibernation in your OS and keep the computer running at all times. When you are sleeping in your bunk you also can't plot your position on the paper chart

2) Track doesn't work properly

Very much connected to 1). Once you will understand, that the computer has to be on to do something, this problem will be way less important for you. I can also imagine some configurability here, but for me priority is low.

3)Projected Course not long enough/ adjustable

I agree here 100% - at low sailing speeds it's maybe even shorter than the organ you mentioned

4) No Course/Speed filtering

I think it's a question of personal preference. I have to say I like the way it is, but use PolarCOM to give me steadier readings at the same time.

5) "Delete All" button position wrong

I would personally move it completely away from the context menu to the Waypoint/Route Manager. One problem though: We still don't have a stable Route manager... You will probably have to wait for the geeks to fix it

Now just to tell you that these problems are not unsolvable let me point out: No chart plotter I've used has these difficulties. Track remains, No delete all where it can be easily used, no disconnections, and adjustable filtering. Nor does Maxsea or the computer based plotters I've seen or played with.

All of them have to be running (and consuming power) to do something
Yes, most of the concerns you have are not unfixable, but fixing them takes time someone has to contribute. So please let the geek people the right to choose what they invest the time to. And don't worry, they are listening to you

Pavel

[manimaul]

Originally Posted by Algarve

Hello and greetings to all.
Can someone tell me why the new version is so unstable?
The softwear is always going down?
Is anything wrong?

The "new" version is a beta release which means it is meant for testing new features. If you are not interested in testing and reporting problems you should be using 1.3.6

Will

[ColdFusion]

Originally Posted by MarkJ

1) Hibernation/Sleep makes the GPS disconnect.
...I NEED the GPS connection to keep connected in a low power state.
tell me what power state and how to adjust it. Remember at sea electricity is a problem. Computers are best not on 24/7

Agree with previous comments - your laptop needs to be on all the time. I'd suggest changing your power options to ONLY turn the screen off after X minutes. Laptop/PC screens consume lots of power so this should help conserve amps.

3)Projected Course not long enough/ adjustable
The red nose out the boats front is as long as a bees dick. It needs to be longer and/or adjustable.
My Chart Plotter has 10 mins, 30 mins, 1 hour, and 2 hours.

Echo that.

5) "Delete All" button position wrong
The Delete All button on the right click context menu is so close to everything else that in a seaway we are always in fear of deleting all!

Completely agree. Personally I'd like to see the Delete All button deleted from the context menu.

Keep up the excellent work

[Psyches]

Originally Posted by SethDart

Dave pushed the diff I sent and added both missing .po files. To my knowledge, there's no core code change awaiting push yet.

Good, thanks. I hoped for this outcome.

2/ Currently the following files are available on CVS:
Makefile.in aclocal.m4 configure
IMHO, they should be removed.

I agree with your .cvsignore comments, and especially some of the removals. Generated files should not be checked into cvs. I think it's important enough to reply here. Sound ok Dave?

Those 3 files should also be added to .cvsignore

Anything generated normally and can be accidentally checked in, should really be in .cvsignore. Else, it eventually gets checked in, and some folks waste time trying to directly manage it rather than re-generate.

Mark

[Belerion II]

I'm following the evolution from OpenCPN for quite a time. My complements to all who are involved in this project. You're doing a realy great job!

I have a few questions:
1) As you can see in the attached pictures, I'm missing the colors in the red buoys in "Paperchart"-mode. In "Simple"-mode the colors are there. It is in ver. 1.3.6 as well in 2.1.0 Beta. I'm running Windows XP SP3 on a Toshiba Tecra laptop and at a dedicated low-power PC aboard with the same results. The symbols are not conform IMO-regulations, but that should be no problem for the average user. (I'm familiar with Furuno and Transas ECDIS, so I am a bit spoild...).

2) for people who are not familiar with indepth ECDIS terminology, maybe you can ad a more readable explanation for the tags at the "Mariners Standard" in Toolbox -> Vectormaps.

3) the big "X" to stop the progam has no use for me. You can stop the program with the standard "x" in windows (and Linux).

Both versions are stable on both laptop and PC. Even with hundrerds, sometimes more than thousend(!) of AIS targets in the Rotterdam area.

Keep going the good work. Again I'm impressed!

Rob
CO32 Belerion II
nnnn

[nohal]

Originally Posted by Belerion II

1) As you can see in the attached pictures, I'm missing the colors in the red buoys in "Paperchart"-mode. In "Simple"-mode the colors are there. It is in ver. 1.3.6 as well in 2.1.0 Beta. I'm running Windows XP SP3 on a Toshiba Tecra laptop and at a dedicated low-power PC aboard with the same results. The symbols are not conform IMO-regulations, but that should be no problem for the average user. (I'm familiar with Furuno and Transas ECDIS, so I am a bit spoild...).

There are different sets of rules for the PAPER_CHART and SIMPLIFIED modes of resterization of point symbols. Part of it is defined by the involved standards and part of it is a bit of "magic" caused by the lack of information. I'm currently working on cleaning up the rules and a tool to edit the S52RAZDS.RLE (contained in <OpenCPNDir>/s57data folder), which is the culprit responsible for the rasterization behavior. The problem you see here is most likely caused by lack of resterizing rule for the particular buoy in the selected mode, so the more generic rule is selected to display it.

[bdbcat]

Sethdart, et al....

I agree (mostly) with your recommendations for CVS cleanup. I have made the updates you suggested, and some others, in the current CVS.

This latest CVS content update changes the build procedure; one must now have the autotools installed on the target system, and run the following:

aclocal
automake --add-missing
autoconf
./configure
make
sudo make install

Of course, this introduces all kinds of new possibilities for build problems. But, it does ensure that all intermediate files are built only on the target, so that any individual quirks of a distribution may be handled locally.

It would be helpful for interested linux users/developers to try a virgin-clean checkout from CVS to a new directory, and perform the full build cycle to verify this new plan. I'll look forward to reports.

Yet another reason maybe to move to cmake? Any volunteers?

Thanks, and good luck
Dave

[cagney]

Dave

Not a problem at all on Kubuntu 9.10 following your instructions.

I compiled last night as well, so I'm familiar with some of the error messages that might be generated.

Thomas

[cagney]

If you don't have automake and autotools installed, this command should take care of the installation.

sudo apt-get install automake autotools-dev

I think both these packages are necessary.

Thomas

[SethDart]

Originally Posted by bdbcat

It would be helpful for interested linux users/developers to try a virgin-clean checkout from CVS to a new directory, and perform the full build cycle to verify this new plan. I'll look forward to reports.

I just did so: clean CVS checkout and autotools then make. It worked perfectly. Thanks a lot Dave, you rock!

Yet another reason maybe to move to cmake? Any volunteers?

I'll try to get my hands in it and report in case of success.

[cagney]

Dave
This crash happened after running for many hours.
The situation: Just (15 sec) came out of quilting mode to zoom in on CM93.

Thomas