gpsd-users
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [gpsd-users] GPS fix last to long


From: Tomalak Geret'kal
Subject: Re: [gpsd-users] GPS fix last to long
Date: Thu, 06 Sep 2012 21:54:22 +0100
User-agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:14.0) Gecko/20120713 Thunderbird/14.0

On 06/09/2012 21:40, Gary E. Miller wrote:
Yo Tomalak!

On Thu, 06 Sep 2012 21:10:18 +0100
"Tomalak Geret'kal" <address@hidden> wrote:

While your intent is noble, 2.95 is not *that* old and out
in the real world where gpsd may be used via interface with
proprietary codebases and where (right or wrong) there is
insufficient top-down support for devoting time to support
updates, your "commandment" rings a little flat. I was
fortunate to have enough clout to drive an upgrade to 3.x,
but given that 3.0 wasn't even released a year ago, I
imagine I'm quite in the minority outside of academic circles.
Yes, 3.0 is not that old, but the older gpsd interface had been
deprecated for much longer than that.  Any distro that can't work with a
several year warning has problems of their own making.

There are other things in the world besides mainstream Linux distros that use gpsd.

And I refer not just to the interface itself but to code that may be broken within 2.95, like the problem that the OP found.

A slightly less extreme example would be the serious v3.2 control socket bug that I identified on this list back on 4th April of this year. Had the reaction then been "you shouldn't be using v3.2; use v3.4" then that would not have been very helpful; fortunately, it was not, and Eric both verified my 3.2 patch and fixed the problem for 3.5 onwards.

(As I initially began to write the previous paragraph I hoped that I'd found the bug in v2.95, as then the "couldn't upgrade gpsd due to no appropriate place in internal development plan to support interface change" argument would be somewhat more relevant. Alas, as I researched for it I found this not to be so. Oh well...)

No point taking patches for things that are already patched and fixed in
later revs, we are not as big as the Linux kernel where we can support
multiple forks.
I don't disagree with that. The project is (marginally) too small to bother with a v2-maintenance branch. For the most part, where no alternative is available, users can fork their own patched 2.95 as the require until such time as they can upgrade to and support a version of gpsd with the fix officially applied.

But that is quite different from the assertion that "no one should be using 2.95."

Tom



reply via email to

[Prev in Thread] Current Thread [Next in Thread]