Let me see, were they the viruses, firewalls or some unspecified adjustments I should make to my network if I wasn't so stupid and could guess what they might be?

stief actually did help and I thank him for that.

As mentioned before, I am sharing a fairly large number of 'small' files...
(I am also sharing some 'largish' files - 16MB-99MB - but no more than about 15 in that size range. Then about 90 files in the 2MB-15MB range.)
The balance of my approximately total 1,600 shared files are about 15% > 1MB... In other words, the great majority of my shared files are less than 1MB.

I am allowing 5 Upload Slots and a Max of 4 for Queueing. (The Max of 4 for Queueing is not constant - I frequently alter the setting in order to observe the impact on 'output' over a 3-4 day period... 'Jury' is still out !)
I use such 'small number settings' because they result in far higher overall performance. The one factor of p2p file sharing which is not to be ignored is that of 'exposure'... That means that the longer it takes to complete a file transfer the more likely it is that there will be some problem. In and Out as FAST as possible is the big key to success. Only way to achieve that with a given bandwidth is to do as little as possible as well as possible ! Hmmm ? Big 'pretty numbers' are a fallacy.

I mention all of the above in order to explain why it is that I am able to observe 'client performance' in a way that is not particularly 'normal' as far as the majority of people are concerned. I have an enormous 'turnover' per 24 hours with a very wide variety of clients Downloading from me.

With the proviso that there are always exceptions, in the comments I am about to make 'exceptions' mean 'rare occurences'. I am also going to refrain from mentioning ALL of the clients I observe.

In terms of 'speed' and successful 'completions' of file transfers there is little to choose between LimeWire, BearShare, Gtk-Gnutella, the LimeWire 'derivatives' and Gnucleus.
They are all fairly high-performers with rarely any problems.
The latest BearShare 4.4.0.5x series may have a slight 'edge' over the others... For now ! Mañana ??? Who knows ?

The variations in performance which I can observe are almost always due to 'operators' trying to do too much with too little. (I say that because, with enough years of experience with this loony hobby, it's not too difficult to 'read' what is going on at the 'other end'.)


The only client which is very frequently a problem, in regards to 'speed' - or, occupying an Upload Slot for hours for a tiny file; failure to 'close' on completion; 'retry/wait states' on the order of 10-30 times more than other clients - is Shareaza. (Ditto for Downloads from Shareaza Users.)


Specifically, and on topic here, LimeWire functions well for the very great majority of users.

Shifting 850MB to 1.05GB Uploads per day of 263kB, 684kB, 1382kB, etc., etc. files makes it fairly easy to clearly see what's doing what !

Although I am not using LimeWire now, it is easy to state that


LimeWire works.

If there are 'settings' problems associated with LimeWire, than that is another issue.
But it is an 'issue' which applies to nearly all clients.

__________________
Fusion for REAL!--CDex 1.51--Wackyuses--Bitzi - Check Files Before Download !--.mp3 File Name Change Problems? Try Rename-It!--Alternate PORTS--Avast- Anti-Virus--WindowWasher--IrfanView--PC Pitstop - Test your PC !--Mac OSX Troubleshooting--GO .ape ! - For 'Lossless' Audio Compression--Port :6346 - Test If It Is 'Blocked'--Cole2k Codecs--OldVersion - Newer isn't always better !
Nod32 - Free Virus Scan: Use ESET's Online Antivirus Scanner

Tomorrow's forecast:
Sunni in places, Shi'ite in others...

Thanks for all the details s:-) ... And for finally mentioning your clients!!!! Now that didn't hurt, did it!

I take it you are using a far greater bandwidth than I have. That in itself has a substantial effect. As you say fast in, fast out is always best. That is self evident!

It makes it even more pertinent that I can do as well as I do with my narrow bandwidth.

I beg to differ on your above comment. Perhaps because you do not have a Mac, so you can not try Poisoned which took no setting up and just ran as expected.

stief thinks Poisoned may use port hopping to get its better results. I don't care how it does it, just that it does.

There was an issue with the last upgrade stuffing up existing daemons but I am able to run it here in my old Jaguar system perfectly well till a new version clears up the one in Panther. The upgrades come quick and fast so I'll let it mature a few more iterations before upgrading.

Not even a wee bit !

Because I did NOT mention what client I am using.
I only mentioned some of the clients being used by people who are Downloading files from me !
(Anyway, you KNOW which client I am using !) Taken incorrectly...

Me is humpin' along here on ADSL 256/128 which, in the real world, works out at an average max of 27.3kBs Down and 13.8kBs UP for my rather common set-up. With the client I am using I am running with ALL Default Settings... Wisely, there has been much progress with this client and the user is allowed to alter very few of the important settings. I only bitch about 2 of the 'user determined' settings because there are limits imposed which relate more to those who are sharing large files at the 'expense' of those who are sharing mainly small files... It's not really that big of a deal though... I mean, what can you do with 'statistics freaks' ?


Besides, they are doing what they do for the overall general well-being of the network. Sure beats fools with 2,000 Upload Queue slots as default or even allowable !
"I don't care how the oranges apple, just that they do."

Still need to compare like with like.

Found it!--that comment on slashdot about no viruses for OS X (search for "nada" to find the comment)

LOL--the thread discusses volunteer helpers, and has some pretty funny posts. e.g
http://slashdot.org/article.pl?sid=0...01&mode=thread

Let see who started the "unpolite" statements, this was your answer after etVoila tried to help:

And don´t forget you was the one who is the wannabe expert and know which client is faster...

You shouldn´t wonder when the people blame you after such comments, an other lesson...

Morgwen

True I was intemperate, but I reacted to etVoila saying I was in a "parallel Universe" just because our experiences are different. Mine is not real, his is.

Geez thanks, now I am a wannabe expert! Where did I claim expertise? All I related was my experience.

My experience was, and still is, that LimeWire is slow and unreliable at getting files. After stief helped me all that has happened is it is less slow and unreliable.

You are free to come around to my place and observe its performance. I share this experience with other posters here. The reason why some people get better results has not been explained to me yet.

The way you react... doesn´t matter what we said you ignored it and told us your experience, instead of reading and understanding what we tried to explain you. And who said that he knows which client is faster? I used almost every client and tested much and I never would say which client is faster because I can´t, think about it...

Because sometimes it can´t be explained by "outsiders", there is no allround solution for fast downloads. We tried to explain you the possible reasons and its you who have to check if it helps or not.

Morgwen

Ace !!!

You're ON, mate !!!!!!

btw, How much is 1st-Class airfare to Australia ?
Are you sure you can afford it ?
Will you meet me at the plane ?
You got any of that non-Koala hospitality ?
Is my room en-suite ?
Let's see... It's 2004 now... When would I have to leave ?

Hey, it won't be all take and no give on my part !

You show me your parallel universe and I'll show you mine !!!



btw2, What's for dinner ?

__________________
Fusion for REAL!--CDex 1.51--Wackyuses--Bitzi - Check Files Before Download !--.mp3 File Name Change Problems? Try Rename-It!--Alternate PORTS--Avast- Anti-Virus--WindowWasher--IrfanView--PC Pitstop - Test your PC !--Mac OSX Troubleshooting--GO .ape ! - For 'Lossless' Audio Compression--Port :6346 - Test If It Is 'Blocked'--Cole2k Codecs--OldVersion - Newer isn't always better !
Nod32 - Free Virus Scan: Use ESET's Online Antivirus Scanner

Tomorrow's forecast:
Sunni in places, Shi'ite in others...

Great! Will you be wearing that s;-) bikini Ursula?

btw I like it dripping wet.

Latest News from the Parallel Universe

Have been running the version of LimeWire stief put me onto for over 2 days now with no other network clients running and only occassional browser connection.

In trials with stief it ran smoothly and DLed from 6kbs to occassionally 20kbs. With real world files (shall we note the original post in this thread?) I have succeeded in DLing 5 files of from 6Mb to 10Mb. They all only intermittently DL. One never goes past 99% but keeps reconnecting. I have a balance of about 30 files. Those that are forever on 0% I cancelled, the others never get past 61% (2) and the rest stick at 7-14%

There is a clear pattern of LW cycling through DL and waiting, then after a few cycles needing more sources. Whereas Poisoned tends to connect and finish the job and do so at a much faster speed. I have observed my ULs in both LW and P and they follow a similar pattern, although ULs run much faster and steadily in both clients.

To me it seems LW's idea of distributing the load is a bit now & a bit later. Much like a waiter with several tables to serve, giving everyone a spoonful of peas then half a potato etc. Instead of serving them in order and giving each table one course when it is their turn.

Others have reported much the same although not in such detail. They have also noted the dramatically different behaviour when switching to P. This morning I started P again to check if LW changed with a second client running. Nothing noticeable, since LW is hardly using the full bandwidth. P in this relatively short time has DLed 5 files over 22Mb and one of 128mb. Obviously different source material but it follows the same pattern I had when aiming at the smaller files I use LW for, only because LW has zero chance of getting anything bigger on my machine.

P does this without having the many multi-server files that I concentrate on in LW. As I have said repeatedly, it is not like there is a little difference, there is a major difference between the 2 clients.

Now Ursula, if you are deciding which way to get to my parallel universe, I can recommend one which would almost certainly get you here quickly, and one which never would if you are over 20mb.