Author Topic: New Client Weekly Update (Read 90498 times)

achilles · « **Reply #60 on:** December 23, 2012, 04:34:10 am »

Harrara, many that have a good working knowledge of the WinMx protocols have not produced a patch after all this time. Months of work was put into developing a patch to mitigate the attacks before deciding a patch would not keep the attackers at bay very long. What coding knowledge do you have? The new client is being written in C++. Maybe you could donate code to the client. Maybe if they needed someone to work on certain features or added functionality. A new client is needed though regardless if a successful patch is created. If you can code a patch that has a long term effect on the attackers then that would be pretty impressive.

White Stripes · « **Reply #61 on:** December 23, 2012, 05:01:40 am »

if the 3rd coder talked about is who i think it is theres not much left before the new client is here... if harrara is able to make even a temporary kludge patch it wouldnt hurt....

harrara · « **Reply #62 on:** December 23, 2012, 05:53:45 am »

Hi all
What i sayd was if i found time for it, and i might got that.
If this work well for aprox 30 days, a little change in it will allowe us to run 30 more days +/-
If the new client is here in time stalling my little coding, well then i have get my self a free
repeat in coding C++
I guess i am rusty in thinking code first days, well see.
I've done worse exams, like bird strike in a CFM56 whining in full throttle, the ride last 4 and a half min, try to row a boat and the break a paddle...

Dont promise any, but gonna try.

If it's so easy to get the source material, you can help out and fill in with add's we do not want in it.

keep on the very good work, merry christmas all

MinersLantern · « **Reply #63 on:** December 28, 2012, 03:24:33 am »

From my extremely limited understanding of any patch, that would involve the controll of outgoing packets within winmx itself. Detect malformed packets, eliminate the data overflow issue, disconnect them, globally.

Since that part of winmx isnt fully understood, aint gonna happen, thus why the new client is being written.

MinersLantern · « **Reply #64 on:** December 28, 2012, 03:29:23 am »

I assume even outgoing primary data could be controlled, externally, but that would require a patch to the tcp/ip stack on windows itself. A bit of overkill. Even then it would consume too much cpu time to play well. That would be more of a workaround than a patch.

MinersLantern · « **Reply #65 on:** December 28, 2012, 03:30:34 am »

BTW. Can winmxworld remove this snow? Chistmas is over, Im cold and ready for the mid eighties. lol

harrara · « **Reply #66 on:** December 28, 2012, 04:01:55 am »

Tryed to "delete" the snow ?
filter in windows cant handle this, this flooding is on a secondary connected device.
Collected some IP and tryed block with ip filter, if i was only supernode, it works.
before i write an code i am going to run several of test and theories.

Allso need 3 more hours in our DeHavillan dhc6 to keep using it, time will show.

I will update you.

RebelMX · « **Reply #67 on:** December 28, 2012, 09:42:37 am »

@MinersLantern: The patch would modify Winmx packets just before Winmx passes them to the Windows stack. The patch hijacks the Windows stack calls made from within Winmx and manipulates the data held in memory before calling the real functions.

This part of Winmx is understood (perhaps not fully as without the src how can you know for sure theres not some additional meaning to some/all of the bytes in every packet?) and that is how the new client clone is being created based on this understanding. As new ideas are tested and new data held within packets is discovered it can be added (as was the way recently with region markers).

A patch in general is overkill as it effectively has to decrypt each packet modify it then re-encrypt it before sending it over the network. This is the reason a client is being worked on as it is much easier to start the process from scratch and change/fix this at source instead of hijacking functions. This is the moot point as a patch could be made to do anything the new client does, it would just be a slightly longer process.

I look forward to the new client, and may even attempt to pull out the security from the client and shoehorn it into a patch so that the old client can work safely on any new network at the same level of protection.

Bluey_412 · « **Reply #68 on:** December 28, 2012, 10:11:01 am »

but why would u wanna hang onto old client?
I see differing features that may be desirable, altho some may see some features as less than desirable
for years we have sat on 3.54b4, see the new client as the next version we have all waited so long for, just from different authors
(Do the authors of Windows 3.0 still write windows 8 code?)

Yes Plz Mr/Mrs/Ms forum admins, its now Dec 28, enuf of snow and silly red hats plz

Ace · « **Reply #69 on:** December 28, 2012, 02:19:14 pm »

win8 code is tooooooo advance to winmx

MSmithy · « **Reply #70 on:** December 29, 2012, 01:13:06 am »

Do we have an honest update on the overal client? Is there a project beta release date? Can we please have more info with hard deadlines?

silicon_toad2000 · « **Reply #71 on:** December 29, 2012, 03:56:25 am »

No, I think we've covered the fact that there are no hard deadlines.
The last update i had to hand was the screenshot I posted in this section of the forum.

GhostShip · « **Reply #72 on:** December 29, 2012, 04:44:16 am »

All anyone needs to know is that we have not given up nor will we, we have hit some issues that are delaying us but like other issues before it we will continue to work on and towards a release.

There really is little more to say folks aside from we are doing our best and will continue to do so, and yes we are aware of the ramifications of any and all delays on the community.

Thank you all your for patience.

MSmithy · « **Reply #73 on:** December 30, 2012, 01:47:45 am »

I'm so glad you are working so hard to get us this new client. I am very grateful. I apologise for being impatinet. Is there any long term PROJECTED release date? For example: aproxomate completion by June 2013? Anything would be extremely helpful and something to look forward to.

White Stripes · « **Reply #74 on:** December 30, 2012, 07:14:38 am »

every release date so far, long term or otherwise, has come and gone so... i dont think they'll do em anymore...

you could use june 2013 as a date if you wanted... just dont expect it...

achilles · « **Reply #75 on:** December 30, 2012, 07:10:27 pm »

I second that! I don't want to even hear of a release date anymore. The last one was in September, but nothing happened. No reason to give release dates. They can only bring disappointment. It will be released when it is released.

harrara · « **Reply #76 on:** December 30, 2012, 09:31:38 pm »

Maybe my work would be finished in time to be used for someone, maybe for a long time ?
Normally when someone creates a program, we gets a update frequently.
It does not happen here, beta testers, zero!
A coder likes to be criticized, this ensures that the product will be better, i hear words, that's all...
So if there is something happened here, share it with us, every week!

We do not bite, we just barks...

silicon_toad2000 · « **Reply #77 on:** December 30, 2012, 11:08:53 pm »

that would be very exciting hurrara, would be great to see your work completed.

As far as updates, when I have new developments at hand, I post them up. As I have outlined, the last new development I had knowledge of was the screenshot I posted recently.

achilles · « **Reply #78 on:** December 31, 2012, 12:30:43 am »

The biggest impediment with this project is the fact that other talented coders are not allowed to join the project due to not knowing who can be trusted due to the attackers. There has to be some way to bring in more talented coders. I just don't know how it could be done without the risk of exposing the source code to the attackers. The attackers themselves could claim to have great intentions, and want to joint the project. Since we would not know the attackers from anyone else then how do you know who can be trusted. If the client can make it to the point of being functional, and secure enough to be released as opensource then any flaws discovered could be quickly solved since there would be a much larger number of coders available to work on the project.

J a M e S · « **Reply #79 on:** December 31, 2012, 01:29:23 am »

I can't believe for 5 years I been on this site.