It is an idea I'd considered, but experience has shown it doesn't much work.   Look at our history of, say, the autocad/dxb formats, for example.   People will override safety latches and still fuss when things don't work as expected, even if you tell them that they won't actually work.  It results in huge support overhead.

These lists are the place to "remind people of its state and raise awareness for the need for contributions".  People have asked for help from those with similar equipment, needs, and equipment.  That's just not happening and I just can't replicate/support everything.  It didn't happen in beta, and its not happening now. With 5-800 downloads a day, three OSes, several hundred formats, and a couple hundred GPS devices - costing between $50 and $5000 each, we just can't catch 'em all.  Even the OP in this very thread turned into a personal "can you convert these files for me" convo because nobody else was interested in seeing if it worked for them.

Similarly, these listsa re the place to upload problem reports.  Whether it's because of bad/incomplete reports or because nobody with that combination of gear is interested in confirming/denying failure is hard to tell, but I can say that of the millions of GPSBabel users, the thousands of people on these lists (and this may be tainted by my perception of conversation/proposals/fixes while I was off list recovering), I'm coming to the realization that it's just more honest to cross bizarre combos off the list and quit offering them if they don't really work.

I'm being grumpy.  As much as I'd like to be able to personally support every one of our 5-800 new users per day with all their format needs, own every piece of equipment/programs out there, and personally take up the support traffic from the other developers to extract an actual developer-compatible bug report/feature request, that's just not realistic in a free program.   If other users or developers can't help out, frankly I'd rather cut features/support.

I still believe that open source can work.  Eleven years and many millions of users into this project, though, I've seemingly not reached critical mass.   I'm at a mental junction where I'm trying to decide between a major internal restructuring - costing me surely thousands of more personal hours of investment - and just letting the project coast to a natural halt.


On Tue, Nov 27, 2012 at 1:51 AM, Nils <internationils@gmx.net> wrote:
Hi Robert,

why not simply add a --experimental option to gpsbabel that will enable formats that are not stable, and will post a disclaimer that the format is not finished and invite contributions? If it works for some its unfortunate to dump it, so it might be more interesting to remind people of its state and raise awareness for the need for contributions.

What could be interesting is to point directly to a webpage / webform * where people can upload tracks that gpsbabel can't handle, and send those to a new gpsbabel-data mailing list... maybe that way more debug cases show up for others who want to have a go at a format.

Cheers
Nils
* OS, gpsbabel version, device name, device OS version, etc. and the data file


On 2012-11-27 07:56, Robert Lipe wrote:
We've apparently had about two people that have tried the lowranceusrv4
module, and both have failed.  The other user actually submitted a file
for others to confirm/refute success and it totally worked for me.
I've cc'ed the author of the new Lowarnce v4 code a few times and have
never gotten a response and other users/owners of the lowrancev4
equipment haven't chimed in to either let us know if this format ever
works for anybody other than Kris Beevers, the author.

I'm seriously considering pulling that format from future versions.
Between the author's disinterest, the miniscule user base, and the
apparent failure rate, and my Windows disk failing while I was
recovering from my surgery, we're doing nobody any favors providing a
format that appears to not actually work. Despite having many millions
of users, we've not really had any traction with lowrancev4 users.

It's very rare that I consider pulling a format once we've shipped it,
but if the author can't/won't help and the other users (if there are
any) can't won't help, and I can't repro success or failure with the
equipment I have (which, at the current donation rate, does not justify
me buying a Lowrance GPS...) I think we're doing a disservice and that
doesn't make me proud.

Hate to be gloomy, but without the interest of others to chase that, I
can't/won't carry the support burden on this.  I'd rather not support it
at all than continue to write "yeah, bummer, sorry this actually doesn't
work" emails for this format.