[OpenIndiana-discuss] Change resolv.conf without using the NWAM GUI

[Richard L. Hamilton]

You have source code. Nobody has time to read and understand it all. But on [Open]Solaris you also have tools like dtrace to examine what's happening in user code, libraries, and kernel.

Is a proof of correctness possible? Probably not. SeL4 has that for at least the core OS (more like a proof that implementation conforms to specification language). Extending that to a massive user stack could be problematic.

But you didn't really have that even in the good old simple days of V7 Unix. I've read a fair portion of its code, and more than a little of it was sloppy - and that was written by the inventors of the OS and language.

One thing the System V lineage did right was produce the System V Interface Definition (SVID), a reasonably detailed specification independent of the implementation. And use of lint and other tools, and lots of work, and changes to C to support argument checking all made things, not perfect, but enough better that increasing complexity wouldn't collapse under its own weight.
 
Back in the 1500's or so was the last that any one person could know all discovered knowledge. Now, probably no one person can entirely know a complex computer system, down to every last instruction and register and transistor. But that's not how you work anymore; you have to develop a sense of what the least you need to know to fix a bug, solve a problem, add a feature, etc will have to be.

> On Jul 5, 2025, at 09:44, Reginald Beardsley via openindiana-discuss <openindiana-discuss at openindiana.org> wrote:
> 
> My thoughts exactly. When dual color and gray scale monitors, 64 MB of DRAM, Exabyte and 2 GB of disk was a deluxe system. Now we have tables of pointers to tables nested indenumerably deep behind a multitude of UI libraries consuming 2 GB to run a browser. 
> 
> I used to understand how everything worked. Now no one knows how anything works. The complexity level has reached absurdity. It's not possible to enumerate the pathologies in the typical desk top stack.
> 
> Have Fun!
> Reg
> 
>     On Friday, July 4, 2025 at 07:12:04 PM CDT, Carl Brewer via openindiana-discuss <openindiana-discuss at openindiana.org> wrote:  
> 
> On 4/07/2025 5:43 am, Gary Mills wrote:
>> On Wed, Jul 02, 2025 at 05:30:26PM -0500, Tim Mooney via openindiana-discuss wrote:
>>> nwamcfg works for configuring most of the rest of NWAM, though I
>>> don't know if I have an example of using it for nameserver setup.
>>> 
>>> It has a 'help' command, but you may want to do some refresher reading
>>> on what 'ncp' 'ncu', and 'loc' are and how they are related.
>> 
>> Ah, that was the command I wanted.  It worked perfectly for changing
>> the dns-nameservice-servers property and updating the /etc/resolv.conf
>> file.  The nwamcfg command works like the configuration command for
>> zones, making it easy for me to use.
> 
> vi used to be damn good.  Edit /etc/hosts, edit /etc/ifconfig.blah, edit 
> /etc/resolv.conf (and nsswitch.conf if you had to) and it would be 
> easily manageable without hiding all the stuff away behind inscrutable 
> command line options and obtuse GUIs.
> 
> I pine for SunOS 4.1!
> 
> In all honesty, how are all this extra layers of complexity helping? 
> This edits this, edits this, depends on this .. the config is where 
> exactly now? All so some poorly coded GUI can manage to not always wreck 
> your network at the cost of no-one knowing how it actually works anymore.
> 
> It's not an OI issue, we inherited it from decisions made at Sun and/or 
> whatever various interests were poking away at posix etc.  It'd be 
> awesome if it could be turned off - say for example /etc/rc.nostupidbs 
> then means that the OS just reads the /etc files like it should.
> 
> 
> 
> 
> 
> _______________________________________________
> openindiana-discuss mailing list
> openindiana-discuss at openindiana.org
> https://openindiana.org/mailman/listinfo/openindiana-discuss
> 
> _______________________________________________
> openindiana-discuss mailing list
> openindiana-discuss at openindiana.org
> https://openindiana.org/mailman/listinfo/openindiana-discuss
>