I thought perhaps its time i made a few observations here to clear up some of the reasons why certain "improvements" are not really viable for this style of network, take for instance the torrenting idea, this is great if one person has a file and many others want that same identical file, but in a system like winmx of many sources and many variants of the same file of different size, bit-rate, encoding, etc its just not feasible to use, this is besides the fact that you will be hard pressed to locate torrents of any worth over a year or two old as there are no "seeders", this then is not a helpful model for us to use, I,m all ears to more feasible solutions but they need to be targeted to the fact that we are a p2p file sharing network and not merely a downloading community of the torrent or newsgroup sorts.
A popular misconception I hear time and again is that "torrents are faster", faster than what ?
WinMX can download a similar sized file in a faster time with the same amount of sources this is a fact, try it and see for yourself, this is why multiple src downloads where introduced and of course there is no similar upload overhead using WinMX, 30 users is 30 users whatever system you use.
Some folks may claim that this is not true and point to certain torrents, upon investigation its been seen that this is down to the use of a web server for uploading called in torrent circles a "seedbox", once again when mx is placed on a server the same speeds are seen, pound for pound you have an equivalent capability when compared like for like, so we can state simply that speed will not increase using that model here, it will in fact be worse.
New ideas and concepts are great to hear and we all like to learn and understand new information so if any of you folks do have a "brainstorm" jot it down and draw a theoretical overview or diagram, I for one am happy to look over any usable network scanarios and how best they can be utilised.
