name resolution

[Xen]

Liam Proven schreef op 25-11-2017 17:24:
> On 25 November 2017 at 09:52, Xen <list at xenhideout.nl> wrote:
>>> 
>>> Apple was the first company with a mass-market computer with only 
>>> USB,
>>> and propelled USB onto the world stage.
>> 
>> 
>> I never heard of that.
> 
> Yes you did. A little machine called the iMac. Changed the entire IT 
> industry.

The entire IT industry. But nothing changed for me.

Well, USB, yes.

There were no people in my vicinity that had Apples. I had never even 
seen one.

> Second conclusion: you are not as knowledgeable as you think you are.

You only use your superior position of factual knowledge to claim you 
are always right.

> Learn to recognise when you are on a subject where others know more
> than you do. Then, stop trying to tell those of us who do know more
> than you about anything about that thing.

This was about Bonjour / Avahi and the idiocy of a zero compromise 
choice for claiming .local exclusively for mDNS.

That wasn't mandated by the RFC, I might add, and is also not logical.

Or necessary.



>>  From my perspective USB just came about because it
>> came about.
> 
> Nonsense.
> 
> Introduced by Intel in 1995. I first saw shipping hardware from IBM in
> mid 1996. It was unsupported by Windows or any other OS at that time.
> 
> Windows 95 shipped with no web browser, no TCP installed by default,
> not FAT32, and no USB support.

Why is that so relevant. If it was introduced in 1995, and there was no 
hardware until 1996, then how could windows 1995 have shipped with usb?

You seem to hold it against Microsoft. What's the point of this?


>> I saw no anomaly in the pace with which it was introduced that
>> would have anything to do with anything I could not see.
> 
> Again. You don't know. I do. Others do. It was my job to know about 
> this stuff.
> 
> Others know more than you do. Learn this.
> 
> Introduced: 1995.
> Shipped: 1996.
> Win95 OSR2: first OS to support USB, August 1996

Still entirely irrelevant to the topic.

Your topic was that Mac sped up USB, and now you cite completely logical 
'delays' in having drivers for stuff that was only introduced the same 
year.

>> Apple was never big in Europe too. Much bigger in America. Not so much 
>> in
>> Europe.
> 
> Wrong.
> 
> I was working for an Apple dealer in 1988. All schools in my country
> then switched to Apple Macs in the late 1980s.

What country?

I know Macs were popular for hackers or those who wanted to get on BBS 
in the 80s.

That was before my time.

I am talking about 1995-2000.

> Apple was _huge_ here.

I am talking about the lifetime of USB.

>> Sure and it was never very popular except for Apple desktops.
> 
> Wrong.
> 
> Included as standard by most Japanese laptop vendors. I have owned
> both Sony and Toshiba laptops with IEEE1394 as a standard feature. It
> was on millions of PCs. Standard support in Windows -- basic from Win
> 98SE, full from Win ME.

And after that first card I carefully picked motherboards that had it 
because I wanted to have it, but during the late 2000s this was 
(already) hard.

Practically all harddisks were USB, very few firewire.

But then, I never owned video equipment?

You can say what you want; no one in my vicinity ever used the stuff, 
and I went to university during that time, so I also saw hundreds, you 
could say thousands, of computers, and not one of them had it.

(Of course they were all the same computers lol).

 From 1995 till 2003 I was around a lot of boys my age who were all into 
computers.

Zip drives were huge for a while.

When I bought it USB was also available.

>> You mean in a home without a router.
> 
> No.
> 
>> I really never heard anyone ever laud AppleTalk before.
> 
> Again, you don't know as much as you think. How many professional
> network architects have you asked?

Relevance Liam. You are living in the ancient past.

I didn't say that no one did laud AppleTalk. I am saying *I* never heard 
it.

This is not a statement purely about AppleTalk, but about the world I 
lived in.

 From 1995 till 2003 when I was *VERY* busy with computers as hobby, with 
friends and at university, NO one every even mentioned the world 
AppleTalk.

Maybe it's in a book on networking I used to have (Tanenbaum).

Probably, as a historical fact.

I am not saying this to prove to you that no one liked it.

I am saying it to give the perspective of a computer-active teenager and 
adolescent in the Netherlands.

Who also went to university for Computer Science.

I am just giving you a glimpse of what my experience was.

I read computer magazines every month, 2 or 3 of them.

The point is that if it was so huge, it should have trickled into my 
world.

>> It has also been discontinued and replaced by Bonjour, but apart from 
>> that,
>> it never seems to have been an issue or lack of Windows computers.
> 
> Wrong. Bonjour is a name resolution tool; AppleTalk is a network
> protocol. Not even comparable.

AppleTalk included the ability of computers to self-organize.

That is the aspect we are talking about.


>> And Apple existed before that, I know that.
> 
> For 20y and shaped the entire PC industry.
> 
> DOS didn't include networking. LAN Manager used NetBEUI. Novell
> Netware used IPX/SPX. DEC Pathworks used DECnet. Many many more.

I know that DOS didn't. Stuff like that was unavailable to me.

When I read about Apples during the 80s I am jealous that I didn't have 
the tools to get online anywhere, no modem etc., when others did.

My first modem was around 1995/1996.

Yes I didn't have that world around me of (Apple) computers and whatnot.

I learned programming on a borrowed MSX when I was 8 or 9, then on a 
second hand IBM PS/2, and so on.


>> I also dislike, as I have said "automatic configuration" and the 
>> article you
>> linked as to why kids can't use computers plays into this completely.
> 
> Well you are the one complaining because you broke your
> automatic-configuration system and seem to think it's unfair.

Largely untrue. I run the dnsmasq server myself. I pick IP addresses for 
everything I want.

The clients are automatically configurated, because I already manually 
configurated them elsewhere.

We are all dealing with automation. I could not type these words if it 
was not for automation.

The question is whether you can be in control or not.

Your system, your beloved system (?) takes control away from me and from 
anyone.

Then we have to evade and seek shelter elsewhere.

You consider that fair?


>> So you currently laud it as beneficial but the article lauds it as
>> detrimental.
> 
> WTF?

The entire article is about how kids don't know computers because they 
don't have to do anything to make them work, and how the guy's family 
doesn't know anything about computers because he did everything for 
them.

That is what the Mac represents.

What are you WTFing me for?

>> Not having to configure stuff is not a good thing in the sense of also
>> having no _control_ over it.
> 
> You are contradicting yourself. You _praised_ the lack of  config in
> NetBIOS networks in the message I was replying to.

We still configured our TCP/IP networks manually, or something close to 
it.

NetBIOS never sat in the way of anything.

It left us free to do what we wanted, did not intrude.

This is different.

I cannot decide, in that sense, to make this Avahi and its machinery use 
a different name.

Yes I can configure that, but now I have to fight and automatic system, 
and it's not working for me.

Automatic systems are detrimental if you are needing to fight them.

At that point your personal power is gone.

You need to fight them when they configure _too much_ or make choices 
that intrude upon other stuff you are doing.


One example could be automatic cars but you can't take over the steering 
wheel.

Automation is nice when it serves you.

Not when it tells you you are not allowed to interfere.

Then you become the tool.


>> The TCP/IP issues never stopped the Internet from becoming popular, or 
>> home
>> networking for that matter, for the Windows world.
> 
> The Internet _is_ TCI/IP. It compelled everyone to support it.

Only if it was popular and doable to begin with.

I am saying the technological necessities were not so large that they 
prevented this thing from growing easily.

I am saying that if manual TCP/IP configuration, (or DHCP) is so bad,

then we would have seen that.


>> SMB/CIFS solved the issue on the file share level completely.
> 
> Only for Windows.

Sure. But just a moment ago you were...

Saying that it didn't.

Anyway.

Yes only for Windows and with Samba also for Linux.

>> Games used their own broadcasting system (or IPX) to find other 
>> players "in
>> the same room".
> 
> Incoherent, technically laughable.

I am saying they had no issues.

You have a problem with reality Liam when it does not agree with your 
technical standards.

Reality doesn't cease being just because you think it doesn't have 
enough diplomas.

>> So while you laud success here, AppleTalk has not had a huge impact 
>> directly
>> during the 90s on my personal life as far as I knew.
> 
> Because, as you have said, you were not a Mac user.
> 
> It appears from your last few lines that you still do not understand
> what AppleTalk *is*. Go read the Wikipedia page.

What I meant was that even though AppleTalk was successful its ideas and 
concepts did not permeate the Windows world.

If it was so amazing it would have done something, but it was 
unnecessary.

Unless you say that SMB/CIFS was inspired by it.

>> Of course Apple is a pioneer. That doesn't make everything they do 
>> good.
> 
> Never said it was.

The argument was whether Apple decided for the rest of the world or not.

I argue that in this case they do not because it is not compulsory to go 
along with their exact implementation.

I argue that the effect that Apple had during the 90s could not have 
been that huge except for hardware design.

I argue that even if hardware by default supports mDNS with a fixed 
.local postfix (not sure it is fixed) you are not necessitated to go 
along with that because these devices also acquire DHCP.

You should be free to choose between multicast and unicast.

All it requires is a router that does the translation. Now I know that 
many non-technical people do not have this.

You say below that routers do have it. Please tell me about it.

>> Yes people did not have names for IP addresses if they wanted to do 
>> more.
> 
> What?

What do you mean what. People did not have names on their local network.

For Windows computers the only name that mattered was the filesharing 
name.

Sometimes you needed a direct connection to a TCP/IP address for a 
server.

(Inside the home).



>> Routers could have done that a long time ago if there was a need for 
>> it.
>> 
>> They haven't.
> 
> Yes they did, but you don't understand how it works.

Then tell me about it.

Mr. Magic.


> They are examples of Apple adopting existing external standards which
> they did not invent and making them widespread across the IT business.
> 
>> AppleTalk was also never felt as a necessity and lost way to TCP/IP.
> 
> So?

I am saying that we did fine without their mDNS-like features.


>> If it was so important, why didn't the rest of the world adopt it?
> 
> They had their own protocols, but you do not seem to understand what
> the phrase "network protocol" means.

Do I really have to go quote Wikipedia for you Liam?

"Connected AppleTalk-equipped systems automatically assign addresses, 
update the distributed namespace, and configure any required 
inter-networking routing."

It is all about automatic, decentralized, peer-to-peer configuration.

The entire topic of this debate.

I am saying that the Windows world semi-mostly did without that yet 
still did fine.

You are now today saying that it is a MUST.

To have server-less auto-configuration.

Yet 90% of the world DID FINE.


>> You mention sudo. I can find no history about it. It has existed since 
>> 1980.
> 
> And nobody used it as standard before that.

Well Jobs has good ideas. And inspiration is free.

Inspiration doesn't come with compulsion.

If inspiration is compelled, it is not inspiration.

But the Avahi development is -- like most of systemd -- pretty cramped, 
no options, forced, I have seen quotes by mr. Poettering where he 
described his system as

"The superior technical solution that needs to be implemented in spite 
of political resistance".

(paraphrasing, but almost literally).

You are not free to do what you like.


Yes, you can get out of the car and walk the highway.

Or fight more and more systems that impose themselves upon you when you 
don't like what they're doing, you don't agree with them, etc.

- NetworkManager doesn't respect manual configuration of 
addresses/routes until you tell it to completely back off.

- DBUS is unintelligable to most people and can't be queried by the 
ordinary person.

- systemd has no support for shutdown scripts that are not the reverse 
order of the boot scripts

- udev is a system that while ideal in nature, in practice often leaves 
broken links in the chain because of some bug or incompleteness, and as 
every link is now the weakest link, there is no redundancy and earlier 
scripts (vgchange -ay)

would have completed the task when udev doesn't.

- gnome also pretty much decides that it knows better than you what you 
want.

- firewalld is a new Red Hat system that imposes a model on iptables, 
again, while in principle worth the effort, in practice these models 
often fail to be really pleasant.

   and it won't respect your meddling with the iptables rules.

- the systemctl, journalctl, timedatectl, etc., model is really annoying 
to use.

- journals in binary format are not all that great but a primary driver 
for that was immutability.

- nftables btw has a really bad syntax from my perspective

- we have not even spoken about systemd-resolvd

- PredictableInterfaceNames is a really bad solution for a worthy 
problem, but I now have to fight it because of the limited number of 
multi-nic machines out there that are more important than the millions 
of single-nic systems.



>> Windows users never felt a need for Bonjour.
> 
> As I said, MS has its own system which MS is pushing.

Of course yes. Samba. CIFS. NetBIOS. Etc.

>>> So in 2000, two Bills, Manning and Woodcock, defined a name service
>>> that would work for serverless networks
>> 
>> 
>> which don't exist.
> 
> I gave the link. It's adopted across the entire Unix world. What do
> you mean it doesn't exist?

I said "don't", plural. I meant that virtually no one runs a serverless 
system.

Unless you just want to connect phones, everyone has a router in their 
homes.


>>> It is _needed_. Before this, different machines with self-assigned
>>> network addresses could not find each other by name.
>> 
>> 
>> Nonsense.
> 
> You do not know what you are talking about, and you are attempting to
> argue with a trained network professional working on large
> mission-critical systems since the 1980s.

I was talking again about Windows filesharing :).

And you talk about the 80s.

I am not attempting to argue. I am arguing :).

> Do you have any clue how foolish this makes you look?

Not as foolish as you think.

Liam when you quoted "Nonsense" above you left out the part that 
indicated what I meant.

"\\john\files

Is certainly by name."

Why are you doing this?