Re: [gpsd-users] GPSD on Debian 10 (buster): allowing other hosts to access gpsd

[Bernd Zeimetz]

On 9/30/19 12:51 AM, Gary E. Miller wrote:

>> * Casual end users, e.g. the typical Android user. This user could not
>>   care less about systemd or its interactions with gpsd. We should be
>>   able to ignore them.
> 
> Two very separate end users here.  Android, which does not use systemd, 
> and Linux/MacOS users.
> 
> This is the class of user that has the most problems with systemd
> and come to gpsd-users and gpsd-dev because the systemd defaults
> and documentation are not doing what they want.
> 
> So the most critical need is for these folks.  At least until
> distros change their defaults and distro specific doc for them.

Debian does not have distro specific defaults.
It takes the systemd files as shiped in the gpsd release (with a fixed
path as we do not use /usr/local).
We also disable two udev rules as they match and break too many generic
usb<>serial adapters.


> 
>> * GPS receiver (and other data source) developers. Most of these want
>> to integrate their data source with gpsd. Their interest in systemd,
>> if any, is to allow them to do what they need to do. These are admins
>>   only by necessity.
> 
> And similar to the first group.  They want gpsd, as installed by the
> distro, to just work.  Which for sysgemd(ung) users is also not the
> case.

If it does not, fix the systemd files in your repository.
And/or add some test cases.
For me it just works fine.

Or point me to a proper bug report which is not the result of people not
reading the systemd documentation. I'm happy to commit fixes to the
systemd files in the gpsd repository...



>> Whoever works on actual systemd unit files should
>> probably know more about systemd than anyone else.
> 
> Unrelated to category.  But this is our real problem, few can make
> systemd(umb) work, and few of them stick around long enough to document
> their wounds.

I can, I'm happy to fix things, but so far I have not yet seen a proper
bug report.
If the socket is not listening on the public interface, its a feature,
not a bug, and common knowledge on how to change a systemd socket
configuration. Nothing to fix in the gpsd repository.

Please tell me what to fix, and it can be done.



>>>> And some links to the gpsd documentation.  
>>>
>>> Lost me?  We are talking about the gpsd documentation?  

s/gpsd/systemd/

>> Specifically, we are talking about documenting the intersection of
>> gpsd and systemd.
> 
> Yup, which sadly will be distro dependent.
> 
>> So Bernd's request for links I take to mean links
>> into other parts of the gpsd documentation.
> 
> Links gotta go somewhere.  Links are easy, we need the thing that is
> linked to that is helpful.  Once the doc is done, it can go in READMEs,
> HOWTOS, man pages, etc.
> 
>> I think links into the systemd docs may also be appropriate. I will
>> look into that as well.
> 
> For devs, yes, but most visitors to gpsd-users just want a simple
> list of steps to solve the problems of all known systemd(umbest)
> configurations to work with gpsd.  They already hate systemd(rek) they
> want to touch it as little as possible.

The only one who hates sytemd is you, other people just have not yet
learned how to use it.




-- 
 Bernd Zeimetz                            Debian GNU/Linux Developer
 http://bzed.de                                http://www.debian.org
 GPG Fingerprint: ECA1 E3F2 8E11 2432 D485  DD95 EB36 171A 6FF9 435F