The Complete History Of WinMX - Part 7

Part-1  Part-2  Part-3  Part-4  Part-5  Part-6  Part-7  Part-8  

We are nearing the end of our historical journey though the end of the frontcode era and into the twists and turns of a large filesharing community now cast adrift and free of all control but this is not at all where the story ends. Like any community over time factions and groups combine or split and this too occurred across many of the different WinMX community sites and venues, its human nature to want freedom from the decisions of others and the WinMX community is no different thus we have reached the time of the community fightback and the heroic efforts to keep alive the network.

In the first few days of "the shutdown" many users thought they had lost all of their chat friends of many years in some cases and this caused much upset and anguish, with no way to reconnect to the network it was impossible to share information on other venues or platforms to communicate to each other with, this then was the day of Sept 19th 2005 a date now infamous amongst the community and remembered annually. Whilst many users thought the sky had fallen in others of a more technical inclination saw this event as a challenge to face up to and set about looking for information and skilled programmers to see if anything could be done to replace the peer cache connection method that had more than a few traps for the unwary to stumble over, multiple efforts where directed at this problem and in some cases both duplicated efforts and frustration boiled over into public displays of verbal sparring, however across the next few days that felt like weeks to those involved a lead was taken in this effort by some of the programmers and "tech heads" from "the sourcecode" chatroom whose members had been of assistance some time before in creating a patch for a buffer overflow issue, this then gave them the respect necessary for others working on the problem to begin to share helpful information in a chatroom they set up and some form of organised chaos was born although it wasn't as professional and friendly as it could have been it was the firm basis of the beginnings necessary to recover that which was thought lost.

Around this time efforts where being made to publish chat room lists across many forums and sites as this allowed users to manually type in the room name and rejoin their friends, although nearly all of the other favoured WinMX functions such as browse and whois as well as searching, filesharing etc had become non operational, also "user herding" strategies where put into place in that selected other P2P networks where suggested as temporary refuges for whole groups of chatters, the idea was to move each whole chat room to another network until a fix could be put in place and it was envisaged this could be the work of some time to achieve given the unknown scope of the problem, users where directed to set up opennap servers and head to those as well as use other chat supporting networks such as Tesla / Filetopia, this then sets the scene for the next few days as efforts became more and more desperate and tempers flared.
Primary users still connected to the network where directed on many of the community forums and by chat room word-of-mouth to share their IP addresses so that when a peer cache fix did become available there would be a list of nodes to share out, by the hour this pool of users slowly diminished and some of those holding their machines on 24/7 instead of closing down while they slept as many did became upset and emotional when their machines accidentally rebooted losing them their much needed connection, many in the community had read articles on other sites declaring that WinMX was dead and Google made mention of the worldwide massive increase in searches that referenced WinMX, it became the most searched for term for some days, it seemed the size of the network had been massively under estimated and only now was the scale of the shutdown fallout becoming clear.

The problem of restoring operation of the network was essentially one of defeating two separate hurdles, the first was that all of the WinMX clients used a hardcoded URL ( to connect them to the peer caches, this proved to be a fairly simple issue to work around by initially using simple host file redirects to a new third party server, this method was soon improved on using a simple additional DLL file that sat between WinMX and the Windows Operating system and made the changes dynamically, this method also had the benefit of making it trivial to swap peer cache URL locations using several DNS routing addresses as whilst the Host file was simple its simplicity was also the cause of another problem, if a peer cache in the host file had to close down a whole new host file release was necessary, this was both cumbersome and inefficient to undertake with hundreds of thousands of users although great efforts where undertaken to pull off this feat.
The second hurdle to bypass was a very much more serious problem to resolve and could have spelled the end of the road for the network, the story behind the replacement peer caches hasn't been given much real publicity since the time it took place so this is a good time to introduce "Mr Drac".

Whilst the community developers had been making third party projects of varying complexity for some time there where still many parts of the WPNP protocol that WinMX utilised that remained a dark shrouded mystery, not much research had been undertaken into some of these areas for the simple reason it was never envisaged by anyone that Frontcode/WinMX Technologies would pull the plug and simply disappear, there was however one person that had already forseen this scenario a researcher/programmer from Japan where WinMX was still very popular. During the years of 2002 to 2005 "Mr Drac" had been working to document much of the protocol used by WinMX and had provided his fellow Japanese programmer Nushi with documentation for his chat client and server projects, this was then someone who had undertaken the creation of many packet dumps and analysed the transactions that passed between the client and frontcode when there was plenty of such traffic, such analysis was impossible at this stage, whether by miracle of will or sheer boredom with chatting, Mr Drac had spent his time analysing and finally creating a table of values that where used in the key exchanges when a client communicates with the peer caches, this was no trivial feat, this was raw magic to the dying network, however it might not have occurred so soon or at all if one of those from the development room had not been searching for both Nushi and Drac in the Japanese language chat rooms, it was after all well known that these two developers along with some of their friends had been experimenting with some aspects of the network architecture famously creating thousands of fake room listings at one point to demonstrate a bug in the chat room code implementation, such knowledge now gave the network its biggest single break.

When Mr Drac was located a single line of text asking for technical help was sent into the chat room alongside an  apology for using English, this respectful style of request set the tone as whilst some replies made it clear that this was not an English room, a few moments later a user asked some questions and Mr Drac himself started to join in the conversation, the host file fix was explained to him alongside the news that a development group had been formed and an invitation was offered to join that room as help was seriously needed to beat the peer cache hurdle, in perfect English his reply was simply to "wait here"    

Twenty minutes later a file was shared by Mr Drac and downloaded, after an initial look at the code it was clear that this code perhaps written some time ago but updated heavily in the last 20 minutes contained a working implementation of the "frontcode key table", alongside this was code to turn the whole thing into a tool to allow users to self connect to other users via IP directly, this then was the key part necessary to rebuild the network, however to utilise this one-to-one tool on a massive scale invited unwanted problems, a server based functioning peer cache was what was required.
The Drac src code was quickly taken to the development room and the code shared openly to ensure many developers where given a chance to utilise it, these developers included King Macro from the UK, Ramna From Italy, and Sabre and Hollow from USA and some others present, at first the importance of the code was overlooked but it was quickly thrown back into many conversations by the courier until its importance was recognised and the key developers started to drift off and set to work.  

On the morning of the 22nd of September 2005 users where able to reconnect using rudimentary host files. Two separate peer cache programs had been created, the battle to save the network was now in effect over, joy, excitement and proud boasts of defeating a multi million dollar enemy sprang forth across many chat rooms and forums, whilst many of those involved in the development arena had been present for 48 hours plus with a few hours here and there taken for sleep, it had seemed as though a week had passed, the actual time from closure to restoration was 3 days only, this demonstrated to many outsiders that the WinMX community of users was both technically proficient and ready to fight to defend their friendships and the right of free speech and association. The network users had won their freedom.      

As often occurs amongst any group of people who have never met in real life and thus had no "base" level of trust, suspicion, discourtesy and old allegiances took hold and the community that had worked hard together to snatch recovery from the jaws of defeat began to splinter due to the rise of a hierarchical structure that was imposed by some of the louder voices in the "Development group", this effectively silenced many others from having a say in the new direction the community should take

At the winding down stage of the development room Vladd of and Gemini decided to create a team to go forward in planning and operating network support, many thoughts and ideas had been put expressed during the shutdown days and with no one running the network now except the users it was seen as necessary to create something formal to maintain the network into the future and to discuss legal and other matters that might cause problems for the cache operators, and others involved in maintaining the network.
At the same time as Vladd was looking ahead so too where the operators of who already had a small team able and capable of doing the same thing already in place and operating their own cache and patch effort, however a document came to light originating from the newly created "pie team" that consisted of the development room hierarchy of Sabre and his supporters and vladds forum staff that made it clear to WinMXworld's operators that there was no room for their activities in the Pie Team masterplan, this document led the start of a split in the network.

Some of the community members now stepped into the empty middle ground and tried their best to weld the WinMX community back together however during a direct meeting between Ghostship (one of the operators of WinMXworld) and Vladd44 on instant messenger it became clear that Vladd44 was indifferent to WinMXworlds concerns and had no interest in cooperating with them for the greater good of the WinMX community, his last words on the subject where "so be it" when informed pressing ahead alone would cause a network rift and increased user fighting. Following on from this ill fated conversation a virtual war of words took place on both forums and the networks users suffered accordingly as time was wasted on verbal sparring and the inevitable demands of both sides to be heard following on from some post removals on vladd's forum.
This fighting went on for some weeks and the initial excitement of the elated users turned to disappointment when it became clear the network would be maintained by two separate groups, each eyeing the other with suspicion.

We now press on to towards the subsequent long haul to repair this great rift in the community that spawned thousands of angry posts and forced nearly all of the users of the newly resurrected network to choose sides, this small snowball of conflict was to have a continiuing impact on the resurrection and rebuilding efforts.

Part-1  Part-2  Part-3  Part-4  Part-5  Part-6  Part-7  Part-8  

©2005-2019 All rights reserved. Page last updated Sat Oct 03 2015