Javascript Disabled Detected

You currently have javascript disabled. Several functions may not work. Please re-enable javascript to access full functionality.

Big Muscle

Member Since 25 Jan 2006
Offline Last Active May 05 2012 12:20 PM

Find content

Community Stats

Group Member
Active Posts 696
Profile Views 19,278
Member Title Expert
Age 26 years old
Birthday September 21, 1985
Gender
Male

Contact Information

1 Neutral

Friends

Latest Visitors

SandatRooess
18 Mar 2012 - 22:01
Crise
13 Feb 2012 - 18:02
exPERten
22 Nov 2011 - 16:21
RHD
24 Sep 2011 - 17:12
Lee
07 Sep 2011 - 09:13

Posts I've Made

In Topic: SFV checking

04 September 2011 - 07:03 PM

There is TTH for the purpose you request.

In Topic: Fake Shares

10 August 2011 - 09:57 AM

User being in 112 hubs? This could be a guide for you...

In Topic: Getting TTH for directory

30 July 2011 - 04:20 PM

Or you can look at bundles in current DC++ branch which have TTH for directories.

In Topic: Leverage resource consumption

30 July 2011 - 12:58 PM

Your opinion seems interesting but what about overall performance? I don't think so that it eats so much handles and memory now (not counting extreme cases like 200 hubs + 150 PM frames). Wouldn't there be big performance drop when it needs to clear and refill whole chat history and userlist content everytime you click on another tab? (however, StrongDC++ based clients have already optimized userlist memory usage, because other DC++ clients just create another memory structure for each userlist's item and therefore they eat twice more memory than really needed).

In Topic: My ideas for the next version

15 July 2011 - 07:47 AM

socialbreakz, on 14 July 2011 - 09:36 PM, said:

To point 2: Did I understand correctly that too fast speed access is not good for the HDD?

It's not so bad but current segmented downloading implementation is not optimal for performance - it creates and destroys handles too often which is not good. StrongDC++ testing version tries to improve it, but some more tweaks are needed (especially on the uploader's side). However, there's really nothing bad about segmented downloading - it brings positives only (when implemented correctly). But I would appreciate any suggestions for improvement :-)

To encryption: there's really no other proper way to implemented without using another TCP port. Yeah, you can use standard TCP port and transfer some header saying "transfer will be compressed", but it is not optimal - you need whole communication channel to be secure and not only part of transferred bytes. By the way, is it really such problem to forward another TCP port?