Hmm, i would really like to make a donation but i don't like paypal. Is there any alternative way to support this bounty?

@Krashan: Thanks a lot and have fun!

Quote:

Same here, since I feel so happy about the recent news and developments (not least MUI 4), I just kicked in another $100 to this bounty (my first too)!

Even if MorphOS 1.5 would be released tomorrow, containing a TCP/IP stack of its own, I definitely think there is room for one more! Competition is good, and I am very confident in that Krashan will do a great job!

I hope that the development will continue after its first initial release, to also include things like multiple NIC's, routing, NAT, bridging, whatever (the full version of MiamiDX could perhaps be an inspiration), and why not ZoneAlarm alike features as well (firewall with the ability to control which specific applications and processes that is allowed network access or allowance to act as a server, and on which protocols and ports, etc). That would be super cool!

But let's just focus on releasing a first version for now, of course with rock solid DHCP for once!

The ezTCP stack has a modular design, resembling Reggae a bit in that objects may be inserted in the packet path. This way for example firewall, bandwidth limiter and other features are possible. The 'bounty release' will contain a basic router (limited to routing between a network interface and localhost), which will be extendable to support any number of network interfaces and include features like NAT and port forwarding. About DHCP - I know that some DHCP services in a popular operating system do not strictly conform to the specs. But I will make experiments needed and it will work.


[ Edited by Krashan on 2005/7/15 15:09 ]

Have updated the amounts, Krashan looks like the pot is considerably sweeter thanks to everyone (particularly takemehomegrandma).

One of the donations last night carried an email address and 'real name' that I'm not familiar with. If the donor could PM me, I'll fix the 'thank you' page (I used the first part of your email address in place of the username).

Thanks for taking the Challenge Krashan... Just added my Vote$
of support.... Keep up the Great work!!!

Wow its getting close to $600! I guess we all know what it is were wanting eh? I would love to see it hit $1000 lets do it!

Yeah, $570 already...and Krashan very well deservers it, and more. I hope we will break the 1000 dollars barrier.

Hehe, guess what people are willing to donate/pay for MOS 1.5/2.x

OT: Nice going Krashan!

Quote:

I'm pretty certain of that!

[ Edited by takemehomegrandma on 2005/7/16 14:56 ]

Well, some more internal work done, implemented a few functions... ezTCP has socket() and connect() now . Also one trivial thing like Inet_NtoA(). Thanks to extensive syslog debug ezTCP will be very useful for people writing network software (I've already found some small bugs in a few popular network applications ). The next step on the way is the name resolver module.

[ Edited by Krashan on 2005/7/16 15:09 ]

wow

Was the timeframe 8 month or 8 days

Damn! There is no stopping this guy!

This is great how Krashan is doing the Job!
I was first against the Stack but why not a little
competition with TCP/IP stacks!?!
Maybe this one from Krashan is better then AmiTCP-PPC.

@ Krashan

I hope my web server will work with your new stack. So far it works only with Miami. Doesnt work with Genesis and what I heard so far also not with the stack in MOS 1.5

http://main.aminet.net/comm/www/w3s_b6.lha

Krashan
Thanks for taking on the project and for the nice verbose progress reports. Saying that here are some things on my mind if you guys dont mind:

1. Its rather sad that there has been a working MOS tcp/ip stack for 1+ years now and NO Release

2. The recent news of potentiality of a 2006 MOS "1.5" release from a developer who is friendly with mos team.

3. Krashan have you consulted Zapek on some of the tcp/ip programming problems.? Also laire was working on it for a while AFAIK

4. I wish there would some form of statement from Mos Team about status of "1.5" and tcp/ip. Its a shame to see so much money go to something that is not *really* needed. For that kind of money we could have SSH port or Video Out on Radeon or something... What about a USB 2.0 driver?

magnetic

@ magnetic:

Its rather sad that there has been a working MOS tcp/ip stack for 1+ years now and NO Release

Yes, it's sad, but we as users can do nothing about it. We can sit and wait, or code and wait. I've choosen the second.

The recent news of potentiality of a 2006 MOS "1.5" release from a developer who is friendly with mos team.

But it is unofficial I guess? I've seen no official MOS Team announcement. And that means IMHO, the release date of MOS 1.5 is still unknown. It may be in two weeks as well as two years.

Krashan have you consulted Zapek on some of the tcp/ip programming problems.? Also laire was working on it for a while AFAIK

Well, I haven't encountered any TCP/IP programming problems so far. My best consultants are RFC and other documents (SANA-II specs, AmiTCP documentation...), Google and experiments. Also AFAIK Zapek only ported a ready stack (AmiTCP to be precise) and Laire worked on its GUI.

I wish there would some form of statement from Mos Team about status of "1.5" and tcp/ip.

I wish it too. But typical frequency of MOS Team official statements is one every two years .

Its a shame to see so much money go to something that is not *really* needed.

Well it is up to those who raised up the bounty. I guess they share my opinion, that there is no shame at all. On the other hand it's a shame that communication between users and MOS Team is so poor.

Hmmm ... I wouldnt say that communication with the
dev-team is poor ... there are just certain questions
that get no answers. I think all general questions
and not related to release-dates are answered quite fast.
(Even if the requests get nonpositiv answers: Like the request
for RMB in ASL.)

@ DJBase:

Sorry, it won't work. The reason is your server relies on outdated concept of inetd, the super-daemon. Inetd is supported only on unix-like platforms for compatibility reasons and is considered obsolete. Every modern server should be able to work in standalone mode, it means you should open port yourself, then listen() on it. Another solution is to write inetd as a separate application, it will solve problem for all TCP/IP stacks not having inetd built-in. Miami has it, ezTCP will not, 1.5 stack AFAIK hasn't it too.

@ Krashan

Well, thats sad to hear and as this is an ARexx server I dont think standalone will work.

@djbase

By using Alfies rxsocket.library it should be possible to add standalone mode into gw3s or
you could start using his Hserv without the need
for an inetd service.

http://users.libero.it/alfier/soft/hserv.html



But an inetd in eztcp would be fun, theres so many possibilities to play with very little work.


[ Edited by catohagen on 2005/7/17 12:57 ]

@ catohagen:

But an inetd in eztcp would be fun, theres so many possibilities to play with very little work.

As I said inetd may be implemented independently of TCP/IP stack. I do not say definite "no" for inetd module in ezTCP, but it for sure is low priority task. I think it is understandable.

@ catohagen

I looked at hserv and converted my homepage to be used with this server. Works so far. Problem solved for my case.

I'm sure ezTCP will be great when AROSTCP is released we may get a MOS port too, so we may end up with two TCP/IP stacks for 1.4.5 and three for 1.5(!)

ofcourse its understandable, unimportant stuff like this can be added later

1st priortity should be getting morphos networked, be able to surf internet etc.

Will you (later) implement your own API concerning addons communication ? im thinking of
panels for showing network traffic etc. Would be nice to use existing stuff like
GelbesPanel (http://amigazeux.net/gelbespanel/)

@ catohagen:

Will you (later) implement your own API concerning addons communication?

We can talk about additional features after the bounty release. Everything is possible for now.