Hello,

I'm looking for the API for the IP stack for Linux? Does anybody know
the link?

Furthermore I'm looking for documentation/tutorial how to use the API.

Thank you in advance.

Markus

On Oct 9, 7:47 am, Markus <markus.hofm...@gmail.com> wrote:
> Hello,
>
> I'm looking for the API for the IP stack for Linux? Does anybody know
> the link?
>
> Furthermore I'm looking for documentation/tutorial how to use the API.


Sockets. A starting point:

http://en.wikipedia.org/wiki/Berkeley_sockets

I'd also recommend the book "Internetworking with TCP/IP, Volume III,
Client-Server Programming and Applications, BSD Sockets Version", by
Comer and Stevens.

Hi Robert,

maybe I must tell you a little bit more about my motivation.

I have to do a presentation about APIs. I want to show APIs in stack
programming. I want to show my colleges that a SAP (Service Acces
Point) does not change the behaviour of a stack if it is correct
programmed. Unfortenately everthing what I say I have to demonstrate.
To reach my aim I was thinking that I show them that I add a SAP
between IP and TCP layer under Linux. At the moment I am not really
sure if this approach is a good one.

They do not have the background knowledge of IP/TCP.

I will take a look in socket progamming but I was thinking it is
easier.

Thank you,
Markus

On Oct 10, 3:54 am, Markus <markus.hofm...@gmail.com> wrote:
> Hi Robert,
>
> maybe I must tell you a little bit more about my motivation.
>
> I have to do a presentation about APIs. I want to show APIs in stack
> programming. I want to show my colleges that a SAP (Service Acces
> Point) does not change the behaviour of a stack if it is correct
> programmed. Unfortenately everthing what I say I have to demonstrate.
> To reach my aim I was thinking that I show them that I add a SAP
> between IP and TCP layer under Linux. At the moment I am not really
> sure if this approach is a good one.
>
> They do not have the background knowledge of IP/TCP.
>
> I will take a look in socket progamming but I was thinking it is
> easier.


SAPs are an OSI concept and don't really apply to TCP/IP, and although
something performs a conceptually similar function in TCP/IP, it's
rather more ad-hoc. Nor are the layers really as separable as they
were (hoped for) in OSI. In some systems some form of "raw socket"
can be used to implement a transport on top of IP by applications, in
other systems such extension have to be done within the protocol stack.

Hi Robert,

I know that the comparision is not ok.

My goal is to show the people here (which are all telecommunication
engineers) that the layer would work without a SAP.

Do you know if I am right?

Do you know where I can found the definition of a SAP from the OSI?

Regards,
Markus

On Oct 12, 5:02 am, Markus <markus.hofm...@gmail.com> wrote:
> Hi Robert,
>
> I know that the comparision is not ok.
>
> My goal is to show the people here (which are all telecommunication
> engineers) that the layer would work without a SAP.
>
> Do you know if I am right?
>
> Do you know where I can found the definition of a SAP from the OSI?


First, this is Usenet, so please quote properly.

I'm not quite sure what you're asking. There already isn't a formal
SAP between IP and TCP, so it already works without one. Largely TCP/
IP does not define APIs, although there are certainly common ones.
For example, Sockets, provides a fairly common interface to UDP and
TCP. Many systems do provide an API to IP, some even make it
available to applications, but other than often following the Sockets
model, it's not consistent across platforms.

A SAP is OSI is really just the access point between layers, although
it's a bit formalized. The original IEEE-TOC paper provides some
decent context:

http://www.comsoc.org/livepubs/50_jo...eid=136833.pdf

Hi Robert,

First I want to say thank you. Now I know what a SAP is.

Second I do not understand in detail what you are meaning with:

"First, this is Usenet, so please quote properly."

How do I quote proberly?


Regards,
Markus

On 12 Okt., 19:44, "robertwess...@yahoo.com" <robertwess...@yahoo.com>
wrote:
> On Oct 12, 5:02 am, Markus <markus.hofm...@gmail.com> wrote:
>
> > Hi Robert,
>
> > I know that the comparision is not ok.
>
> > My goal is to show the people here (which are all telecommunication
> > engineers) that the layer would work without a SAP.
>
> > Do you know if I am right?
>
> > Do you know where I can found the definition of a SAP from the OSI?
>
> First, this is Usenet, so please quote properly.
>
> I'm not quite sure what you're asking. There already isn't a formal
> SAP between IP and TCP, so it already works without one. Largely TCP/
> IP does not define APIs, although there are certainly common ones.
> For example, Sockets, provides a fairly common interface to UDP and
> TCP. Many systems do provide an API to IP, some even make it
> available to applications, but other than often following the Sockets
> model, it's not consistent across platforms.
>
> A SAP is OSI is really just the access point between layers, although
> it's a bit formalized. The original IEEE-TOC paper provides some
> decent context:
>
> http://www.comsoc.org/livepubs/50_jo...nagement_eid=1...

Markus <markus.hofmann@gmail.com> wrote:
> Second I do not understand in detail what you are meaning with:
> "First, this is Usenet, so please quote properly."
> How do I quote proberly?

In brief it means making sure that the little ">'s" or whatever is
used to designate quoted text from previous posts line-up correctly
with the purported author(s).

For example, the way the ">'s" are above suggests that you wrote all
that text.

IIRC there should be a news.answers newsgroup that probably has a FAQ
on all sorts of things usenet/netnews. My dimm memory seems to recall
that FAQ's for usenet groups are archived on rtfm.mit.edu which one
can access via FTP and perhaps these days via http as well.

sincerely,

rick jones
--
a wide gulf separates "what if" from "if only"
these opinions are mine, all mine; HP might not want them anyway...
feel free to post, OR email to rick.jones2 in hp.com but NOT BOTH...

On Nov 8, 3:58 am, Markus <markus.hofm...@gmail.com> wrote:
> Second I do not understand in detail what you are meaning with:
>
> "First, this is Usenet, so please quote properly."
>
> How do I quote proberly?


Specifically I was complaining that you did not quote at all, leaving
no context for your responses/questions. This is Usenet, not
someplace that explicitly maintains threads like you seen on most
message boards. Yes, most news readers do put messages back into
threads, but it's not always possible, especially if the distribution
list changes over the course of the thread, and depending on delays in
the propagation of messages from various sources.

Second, do not top-post in technical news groups, either bottom post,
or put responses in-line. As a general rule, top-posting is always
frowned upon, but is often common in non-technical groups.

Third, always trim the quoted text as much as reasonable.

Perfection is not required, but some effort to meet those norms
improves Usenet for everybody, not least for you. In many technical
groups, blatantly violating either of the first two rules will often
get your post ignored.