Yamar

GPS: $200 MacBook Pro: $1500 Parallels: $80 Windows 7: $100 GSAK: $30 Geocaching Premium Account: $30 Finding a film canister under a lamp post after buying all that stuff: Priceless! (actually, the only thing I think is silly in that list is needed to buy another operating system just to run GSAK; though I know of a lot of people that *only* bought windows so they could run GSAK)

I've always wanted to take my geoqo software (which does work on a mac, but is hard to get up and running), refactor the important awesome concepts into C++/Qt and make it easy to use on pretty much everything (mac, linux and windows). If there are people that want to volunteer to contribute time and energy (ie, code and documentation) into making such an open source project, it wouldn't take too much to get something up and running.

Almost 1500 installs in five years, of which 270 use the package? Am I really reading that correctly? Well, you need to read it in terms of trend (it's going up) and percentage (.29% of debian users are active geocachers). It's an opt-in program (I believe; I actually don't use debian and am not an expert) so you can't read the raw numbers as "I only have 270 users". The best you can do is draw conclusions from the percentages within the people that opted-in. Note that 0 people installed the documentation, which is probably where the "help support me" file is. But to point #3, the the vendors seem to do everything they can to isolate the user from the actual developer. What irks me even more is how the *packagers* often don't even submit bug reports upstream. The number of times I've gone patch-trolling looking for patches being applied that I've never seen... And I *always* find one. But yes, developers are far removed from the end-users. If it makes you feel better, it's actually very similar in the commercial world. See dilbert for references :-)

Unfortunately, measuring linux usage of any package is a real challenge. I am the manager of a number of successful open source packages, and my paying customers frequently ask me "how many people use it?". Well, it's extremely hard to count. Almost every non MS system out there has open source packaging systems that hide the true count from the developers. It's a very good thing to make it pre-packaged and available on demand, but it completely removes the ability to accurately gauge how many users you have. EG, the most successful of the software that I deliver is now part of pretty much every OS stack out there except windows. OSX, Solaris, every Linux distribution, etc re-packages my source code and distributes it. This makes it absolutely impossible to count my users. And really, I don't care too much. (It's my customers that care.) So what does this mean in terms of gpsbabel? It has the exact same issue: it's pre-packaged with every linux OS out there. That means counting the linux usage is pretty much impossible. Although some systems, like debian, does have some useful tools to help you see rising and falling trends of a package: http://qa.debian.org/popcon.php?package=gpsbabel but is only measuring a percentage of their actual users themselves. Still, however, it's interesting to see the usage in gpsbabel steadily climb! Now... You might think that paying contributions are a useful way to judge, but I disagree with that. There are different classes of users out there and linux users fall into their own camp. Linux users are: 1) end-users that are used to getting things for free and don't have the desire to pay for stuff. 2) developers that contribute freely with their own code and don't feel it's fair to pay for stuff when they're helping 'the cause' as well. 3) end-users that might pay but because of the above "apt-get install gpsbabel" have *no idea* that they're even being asked to contribute "a tip". Now, I'm not condoning the above three viewpoints. I'm just stating that it's unfortunately an apples and oranges comparison to a large extent.

You mean versus GTK? I think both QT and GTK are pretty portable these days. I kinda prefer Gnome over KDE, so I went the GTK route. That being said, multiplatform support is a real pain (do that enough in my day job ) so I'm not even going to bother. As Robert's pointed out in this thread, just trying to support the various linux distros is enough work. I think they're both portable, yes. And in fact geoqo has GTK support now for it's interface (through another abstraction layer). But... I don't think GTK ports or installs as easily as Qt does and I do think that Qt has lately passed GTK in terms of devices it runs on. EG, the other day I wrote a few simple Qt applications and to get it to work on my phone was simply a re-qmake and recompile without a single modification to the source. What I haven't tried yet is installing and compiling under windows. I'm hoping that easy too. But I'm not that naive to believe myself either though. (Qt has ports to androids and Iphones, I don't think GTK does but I could be wrong there)

I'm thinking of converting my geoqo application functionality to Qt (from perl) because I really like the power of geoqo's internal architecture but want a new spiffy UI on front of it. Plus Qt runs on pretty much everything and is probably the most portable UI development kit out there...

You should definitely check out geoqo then; it's command line interface is significantly better than its GUI interface and lets you do amazingly complex searches and data munging. # geoqo -s cache:state=Maryland -d top10:groundspeak_difficulty Top 10 'groundspeak_difficulty's Num % Value ---------------------------------------------------------------------- 180 32 2 105 18 1.5 105 18 2.5 79 14 1 61 10 3 16 2 3.5 9 1 4 3 0 5 1 0 4.5

Actually, since I don't think the attributes were the source of the problem, could we get them added back immediately? Then the Unknown issue could be solved later with the new needed checkbox or whatever.

I am puzzled by this. The app broke because the attribute spelling changed from "Unknown Cache" to "Mystery/Puzzle Cache." Ah... I didn't see that change and assumed it was broken because of the attribute additions. Changing data *values* like that definitely is a bad thing to hard-coded applications. What's worse is that changes like that require smart code to be able to deal with old and new PQs, which isn't a good thing. I wrote GeoQO to be generic so it can accept any type and you can create your own (ie, it handles much more than just Groundspeak data). But I realize that many applications would rather choose to hard-code type lists and things like that.

Sigh. Unfortunately whatever app broke (I haven't been following any "this broke" thread) is at fault. Most of the point of XML is surrounded by the fact that applications should ignore new things it doesn't understand (unless some specific spec called for it not to). Since GeoQO already supports parsing and using attributes and I was happily making use of the data, is there any time estimate for when it'll be turned back on.

I've built multiple garmin maps using OSM data. With modern garmin GPSes that use cards you can actually find software to make it happen. I suggest looking at the OSM wiki for help (search for "garmin"). It is not, unfortunately, easy for those who are just looking for a simple method. An alternative is to download pre-made ones from cloudmade: http://downloads.cloudmade.com/ They offer downloads of garmin .img files and .osm data files in large chunks. The .img files, though, are "debuging" mode .img files and actually don't look fantastic in the device (IMHO) which is why i end up re-rendering them myself.

As long as we're ranting, you know what I hate more: to go look into a distro that is repackaging your stuff to find that they're applying like 50 patches for bugs they've never even submitted to the upstream as issues in the first place! This compounded with the problems, all very true, that you mention above is a nightmare. Fortunately, some distros have gotten better about that. But many haven't. And most distros are user driven packaging too which means support varies from packager to packager.

Yeah, I hadn't either until someone recently pointed me at that page. It's certainly a useful system and just darn fun to play with :-) Actually they are. If you want "more users" and don't care about your "personal download counter" you'll get far more users if it's easy to find in the "add new software" box. Linux has gotten to the point where download and install by running configure is actually above the heads of 50% of the users now. (whether that's a good thing or a bad thing is still subject to debate, but if you want "more users" running linux, it has to be a good thing even though it comes with a lot of negatives). FYI, F10 for me works just fine. The only thing I had to do to make things work out of the box was solve the permissions problem. And yes, they should definitely fix that. But other than that no driver issues need to be solved (which isn't true of previous fedora releases like 8ish I think). That's actually why I created geoqo a while back. I needed something to manage waypoints on linux and nothing existed that even came close to my needs. I spent a lot of time on it while my work hours got slashed a while back and made a lot of progress (I still use it heavily now). But now I'm back at full time and don't have the same amount of time to throw at it. But, by far, the number one thing that makes it difficult is installation (which is why I've been working toward a single download installation executable that includes a packed gpsbabel to avoid all those problems with people that can't handle prerequisites). In the end, anything that involves device polling is a pain portability wise. That "other software" I talked about is heavily kernel tied and is ported to a gazzilion OSes. We've had our fair share of "ifdef hell" because of it and rapidly changing kernels make it a real pain in the neck to write portable always-working code. I understand and feel your pain. But in the end, is it better to be mad at the kernel developers that cause you the pain and take it out on the users by dropping support or is it better to try and do a service to the users? If we want more linux users out there (having just spent an hour on my wife's windows machine yesterday trying to solve her problems, I definitely want more linux users out there) then we need to provide good geocaching software that works around distro problems for them. I know too many people that have bought a windows machine just to run GSAK because they have nothing on OSX or something. That just makes me sick.

The numbers certainly aren't as high as one would like, but when you start getting included into distributions it gets very hard to measure. As a package author of another heavily used (non-geocaching) piece of software that is included with most major operating systems (except windows of course), I find it very very difficult to get a real perspective on use. The only good true metric is mailing list activity. But if you have mailing list activity you probably don't have perfect software (or else they wouldn't need to write you). Being included in distributions means your download count is close to meaningless. I've continually seen "drops" in downloads right after a package gets included into base OS systems. And since gpsbabel is included in most it would be hard to get a good measure of it's usage. That being said, there is one distribution that offers at least a perspective on how many people are using a package and that's debian. Here's a page showing the usage graph of people using gpsbabel, and as you can see the trend is increasing: http://qa.debian.org/popcon-graph.php?packages=gpsbabel You do need to take the raw number on the left with a grain of salt though, as it's an opt-in counting system so it won't count every user. But it does give a feel for trending that I think is likely accurate.

It can vary, but when I checked the most recent versions of Ubuntu (a year ago) and Fedora they both suffered from the permissions problem. Hence I mentioned it here. (I've actually been meaning to write a check into the geoqo code to specifically check for that issue and post a warning to the user)

I use gpsbabel all the time to talk to usb: But, the default is that it'll only work as root unless you do something to change the permissions. You can allow anyone (ie, your normal linux user) to use gpsbabel with usb: by creating (as root) the file "/etc/udev/rules.d/51-garmin.rules' and putting in it this line: SYSFS{idVendor}=="091e", SYSFS{idProduct}=="0003", MODE="0666"

As many people have said in the past: no. There is no list. Groundspeak has never created a list. Maybe they will in the future, but they haven't yet. (On very rare occasions they've make announcements about XML changes, etc, here.)

As mentioned earlier, I'm the author of "geoqo" which is designed to work on linux. The previous issues with how it works and gets installed should be resolved I believe. There is now an easy download executable that should make it simple to get installed. http://www.geoqo.org/ I use it extensively myself, but to be honest I don't use geoqo's GUI interface that much (except to print maps). Hence it doesn't get as much attention as it deserves. I'm more of a pure command-line driven guy myself, which is where much of the geoqo power really comes in. > geoqo -s 'log:text=awesome&&mem:sacsquare' -d count: Search/Set Count: 236 Apparently I have 236 awesome caches left to find in my local area.

And a lot more pain. I think the goal of a whole-state PQ is to not have to carefully manage 35 overlapping PQs. Yes you can create more premium accounts, but that isn't nearly as useful as a single PQ with most of the data you need. It's probably a lot less strain on the servers too.

You should look into the open street map project (http://www.openstreetmap/) which lets you edit maps for the entire world. Their maps can be compiled into garmin image files. And cloudmade even builds them for you ( http://downloads.cloudmade.com/) once a week (because the process isn't trivial). I've actually used OSM to modify roads in a country I was going to travel to and needed maps for (but shelling out $100 for a single trip wasn't worth it). The only downside is that they're not routable maps (yet). Your GPS can't direct you through turns. On the plus side, however, cities with active OSM editors often have walking and biking paths in the maps, which is something you'll never see in the commercial map sets.

In the terminal window, try typing: cd Desktop (I still need to install ubuntu to see what they're default setup is like... It's possible the above won't work)

Ah, if you can get to the shell you should just run it: ./geoqo.bin However, you'd need to 'cd' to your 'Desktop' folder first. And if you renamed it (eg, removing the '.bin' ending) you should run it however it's named now. (and FYI to everyone watching: it's a compiled perl executable and is a straight binary... It isn't a sh script or even a perl script at this point but is a raw executable)

Sorry for the delay (I was out of town). I think the problem is the .bin extension. Can you rename the file (right click on it to do so) and remove the ".bin" from the end of the name. At the same time, make sure the properties say it's executable and then double click on it again and hopefully it'll work. If that works, I'll create a new version this week that fixes some bugs (with importing; sigh) and also adds the ability to display caches (with dots) on a map!

Well, it took me a bit longer than I planned (it always does) because of lack of time. But! A new release of geoqo is available and it should be very easy for linux users to grab it and install it since it is a single-binary program with all needed prerequisites embedded internally. Main site: http://www.geoqo.org/ Binary Instructions: http://www.geoqo.org/wiki/index.php/Instal..._binary_release

Wow, I have no idea where that error is coming from. How odd (cause normally I'd know!) That's at the command line in a shell (terminal) window?