[Tuxman]

I was urged to release it here, so here it is...

February 24 2009
eMule 0.49c AnalyZZUL v3.0
modified by Tuxman, inspired by pesti

WHAT   | BY		  | DESCRIPTION
-------+-------------+---------------------------------
BASE   | zz		  | eMule 0.49c ZZUL 20090222-2320
IMPV   | Tuxman	  | grant friend slot from more lists
IMPV   | Tuxman	  | toggle friends from all lists
FIX	| Tuxman	  | small Analyzer display fixes
FEAT   | Tuxman	  | Spread Priority

Spread Priority is my own implementation of an automatic
upload priority. If enabled (switchable in Options -> Ex-
tended), it will set all files' priority according to their
complete sources count. This does NOT apply to partfiles
or files with the built-in "Auto" priority!

August 08 2008
eMule 0.49b AnalyZZUL v2.1
modified by Tuxman, inspired by pesti

WHAT   | BY		  | DESCRIPTION
-------+-------------+---------------------------------
IMPV   | Tuxman	  | added support for more file types
FIX	| Tuxman	  | the Analyzer was broken
IMPV   | CB		  | chunk dots are not drawn for large files

August 02 2008
eMule 0.49b AnalyZZUL v2.0	modified by Tuxman, inspired by pesti

WHAT   | BY		  | DESCRIPTION
-------+-------------+---------------------------------
BASE   | zz		  | eMule 0.49b ZZUL 20080801-2355
FIX	| WiZaRd/gidi | some Client Analyzer fixes
IMPV   | Tuxman	  | removed some bans (we have CA instead)
IMPV   | Tuxman	  | added support for more file types
IMPV   | Tuxman	  | hid AntiLeech label if not applicable

May 15 2008
eMule 0.49a AnalyZZUL v1.1	modified by Tuxman, inspired by pesti

WHAT   | BY		| DESCRIPTION
-------+-----------+---------------------------------
FEAT   | Moonlight | SUQWT

May 13 2008
eMule 0.49a AnalyZZUL v1.0	modified by Tuxman, inspired by pesti


WHAT   | BY		| DESCRIPTION
-------+-----------+---------------------------------
BASE   | zz		| eMule 0.49a ZZUL 20080512-0047
FEAT   | WiZaRd/CB | Client Analyzer 1.5
CHGE   | Tuxman	| version string, eMule logo

Please note that there are most probably horrible bugs I accidentally added...

It is a "testing mod", actually, not a "final state" mod, so I want to declare the following:

As I am AFAIK not really required to give any support here, ask any questions regarding this modification here; I won't surely visit this thread too often anymore.

Binaries
Sources

eD2K links:
ed2k://|file|AnalyZZUL%203.0%20Binary.7z|1722813|D991F37B872857C67BB58DFF41F01BA3|/
ed2k://|file|AnalyZZUL%203.0%20Source.7z|1637700|BF7C5388DA020026362ED555831BD928|/

(These files may, additionally, be removed from the HTTP server when I cancel the project or if there is any new version or anything... please share them also via eD2K/Kad!)

This post has been edited by Tuxman: 23 February 2009 - 11:38 PM

[LorenzoC]

I have been testing this mod for a day on my PIII 500 Win2K SP4 system and It works flawlessy, no visible issue.

On a side note, I think this "simple" mod could be useful for releasers who also want the CA, so I hope Tuxman will go on with it.

Somehow it walks a similar path as the Bastard.

Thanks Tuxman.

This post has been edited by LorenzoC: 14 May 2008 - 03:35 PM

[Tuxman]

I don't really have any precise future plans (more features, ...) for this project.
There is no need for a second BastarD mod.



I thought v1.1 will be .49a with all fixes, if needed, then v2.0 will be .49b, v3.0 .49c and so on.
Nothing more...

Provided that I'll continue it, of course.
Thanks anyway.

This post has been edited by Tuxman: 14 May 2008 - 03:46 PM

[LorenzoC]

See, this is only my opinion but:

1. CiccioBastardo could stop developing the Bastard. He is missing somewhere after he said he was leaving for Sweden and he sounded a little de-motivated lately and/or not having time for modding.
2. The Bastard does not use the CA but the FineCS that actually works pretty much the exact opposite, so the philosophy about the "anti-leeching" part is completely different. A different user base I would say.
3. My personal opinion is there is always place for a SMALL, EFFICIENT and NOT BLOATED mod. Your Beba is ok but the project philosophy is to not implement the releasing features that are instead present (the basic and really needed ones) in ZZUL. So either you decide to add releasing features to Beba or you should better add some basic features to ZZUL (SUQWT ?) like you did. It makes a lot of sense from here. Of course if you have time for it.

This post has been edited by LorenzoC: 14 May 2008 - 04:00 PM

[Tuxman]

Thanks for your posting.
Well, the beba mod is mainly filled with the features I need, while the AnalyZZUL was actually requested by someone else...

Continuing the AnalyZZUL would mean that I'm gonna develop sth I don't actually use myself.
Additionally, as I stated elsewhere, I'm not quite familiar with the ZZUL code itself, so I could never provide much support for ZZUL-related problems... and I would have to disregard my beba mod.



I wouldn't, however, have called it AnalyZZUL if I had planned to add more features at all... if I'd want to turn it into a "full" releaser mod, there were dozens of feature requests again, and if I'd follow them, this mod would end in another all-purpose high-quality but very, very bloated thingy...

So what to do?

This post has been edited by Tuxman: 14 May 2008 - 04:12 PM

[LorenzoC]

See, it is your time and your work.
I am an user and I use your time and your work.

I like the Beba because it is simple.
I really hate the bloated mods.

I like this ZZUL+CA mod because it adds the antileeching to the ZZUL and nothing more.
IMHO you don't have to produce another full featured mod, it could simply stay this way our you could add some more features you (or the others) may think could be useful but (ZZUL + a very minimum set of features) is pretty much everything is needed.

I see this "experiment" like an "easy" complement to Beba not as an alternative, given that you aren't experimenting anything new here.
Of course I don't have any idea of the difficulties in handling completely different code bases while modding, so at the end it is your strategical decision.
IMHO, if you can spend some time on it, I find this mod a good idea.

Oh, as user I could care less of names...

This post has been edited by LorenzoC: 14 May 2008 - 04:36 PM

[Tuxman]

LorenzoC, on May 14 2008, 06:34 PM, said:

(ZZUL + a very minimum set of features) is pretty much everything is needed.

Tell that to those ppl who can't live without high-color client lists.
(I'm afraid that there is no unique definition of "minimum"...)

LorenzoC, on May 14 2008, 06:34 PM, said:

I don't have any idea of the difficulties in handling completely different code bases while modding

That's not the point, as I manage them separately.
(Updating to a new code base always takes some effort, though.)

Alright then... (ZZUL + a very minimum set of features) is pretty much everything is needed ... I'll take you up on that.
For now.

This post has been edited by Tuxman: 14 May 2008 - 04:48 PM

[LorenzoC]

Tuxman, on May 14 2008, 06:45 PM, said:

Tell that to those ppl who can't live without high-color client lists.
(I'm afraid that there is no unique definition of "minimum"...)

That is what the bloated mods are for.
If somebody wants the bells and whisles he/she can go using one of those.
Then there are people who prefer lean and mean clients, like me.

Quote

Alright then... (ZZUL + a very minimum set of features) is pretty much everything is needed ... I'll take you up on that.
For now.

Good. I am sure there are interested people also on the original board, otherwise you wouldn't have made this "experiment"...
Let me add this: if it took 30 minutes to add the CA to ZZUL (hoping you haven't screwed anything in the process) it looks like a little effort for a huge gain, at least for those users who need the CA added to a basic mod for releasing. So I guess we can agree it was a good bargain regardless. If you worked 30 minutes to add colors that would be wasted time.

This post has been edited by LorenzoC: 14 May 2008 - 05:11 PM

[gav616]

LorenzoC, on May 14 2008, 04:57 PM, said:

See, this is only my opinion but:

1. CiccioBastardo could stop developing the Bastard. He is missing somewhere after he said he was leaving for Sweden and he sounded a little de-motivated lately and/or not having time for modding.

noooo!! don't say that
but if it had to come to that, i hope when the person(s) merge it to .49 they don't change anything.
only CiccioBastardo can change/fix/remove stuff, its his philosophy on how a perfect eMule modification should be.

[cellax]

I didn't see the CA log, may I ask why?


I am testing it, it goes fine

This post has been edited by cellax: 14 May 2008 - 05:38 PM

[Tuxman]

What log?
CA output is moved to the Verbose window, like in beba.

[LorenzoC]

After several hours of running the only strangeness I noticed via task manager is the virtual memory usage.
The "standard" memory stays the same as ZZUL but the virtual memory grew to the double of the "standard".
On my sistem it is about 35M and about 60M, rispectively.

Re-testing ZZUL to compare.

This post has been edited by LorenzoC: 14 May 2008 - 06:10 PM

[aSceT]

I very like the idea of a minimalistic mod with CA and ZZUL features!
I think that there are room for implementing additional useful features (if development will evolve), that don't need any user interaction, like SUWT, for example. In fact, there are many of those tiny/small but useful features that could be implemented with minimalistic-style: that will just work, without spending hours on their fine-tuning from user-side...

[cellax]

Tuxman, on May 14 2008, 06:41 PM, said:

What log?
CA output is moved to the Verbose window, like in beba.

oh sorry, I was using sharkx and during the Analyzzul test the first thing that I looked for was the CA log like on sharkx.

you are right

@aSceT:
I agree with you

This post has been edited by cellax: 14 May 2008 - 07:29 PM

[Tuxman]

Adding a particular log window for everything that logs is ridiculous...

@ aSceT:
I can't remember any clear definition of "useful". Do you?
SUQWT... might be, but doesn't need to...

This post has been edited by Tuxman: 14 May 2008 - 07:35 PM

[LorenzoC]

Well, SUQWT should be present in the Official, so it looks "useful" in whatever mod, IMHO.
It seems fair to save the user's queue waiting time instead of resetting it upon closing the connection.
Of course in some situations, like NO CS, it is more critical. I usually disable credits on Emule, that is because I find SUQWT "useful".

[Tuxman]

LorenzoC, on May 14 2008, 09:41 PM, said:

I usually disable credits on Emule

And you dare to tell me sth about "fairness"?

(That's not the topic to discuss here, though...)

I think the ZZUL system itself doesn't require any SUQWT as the official client maybe does...

[LorenzoC]

"(That's not the topic to discuss here, though...)"
You are right but I am an annoying guy
Given the design of the network the only "fairness" clearly allowed is to serve requests by waiting time and file priority. Serving requests by received/exchanged data is trading. One may like trading more but it is not "fair" in this network, on the contrary, besides the "shaping" effects, fairness based on waiting time is incompatible with fairness based on trading, one steals the slots of the other. You can try to "balance" it but at the end it is an obvious nonsense. Just to explain my idea of "fairness" here and why I do belive the CS should be forbidden. Sorry for the OT.

Regards.

This post has been edited by LorenzoC: 14 May 2008 - 08:07 PM

[Tuxman]

So if you upload to someone, you don't want to be rewarded in any way?
That's, indeed, a way of thinking I didn't follow yet...

Quote

fairness based on waiting time is incompatible with fairness based on trading, one steals the slots of the other

That's even a good reason not to add SUQWT at any point in the future. Why should people who left the queue be able to push in later?

This post has been edited by Tuxman: 14 May 2008 - 08:23 PM

[LorenzoC]

See that is why I wrote about it. The reason why people should start the queue where they left (SUQWT) is because the waiting time, plus the file priority, is the only "allowed" (by the network design) criteria to handle the queue.
It is so obvious the CS was added on top of a network that is designed to work against trading and I don't get what is the point in making the waiting time and priority less relevant while using trading as main criteria for handling the queue.
It is like trying to correct potatoes to make them taste like apples.

Exactly, I don't want to be "rewarded" for my upload.
I prefer much more to be granted a slot depending on my waiting time.

The reasons for this are:
1. you can't be rewarded for complete files, the only effective CS usage is while exchanging parts of incomplete files. One of the consequences of this is "leeching", at least for complete files, is allowed in this network by design. The other consequence is people unsharing files upon completion.
2. I insist that file trading is contrary of the network design. People must be granted a slot when they make a request, regardless what they have to give you back. Otherwise, selecting "interesting" clients only is the logical "Dazzle-like" tactics.
3. Those who are exchanging parts push me down in the queue when I don't have anything to give, that makes very difficult to get a slot from somebody who is sharing complete files (given that they are actually sharing complete files that are useless for shaping) AND downloading other files in the same time, unless this one is giving the complete files a "release" priority or higher. It is easy to see that while using the official.

For the above reasons I would prefer a client without CS and with SUQWT instead, right because I belive the trading the "wrong" idea of fairness. Of course you can disagree but I am always puzzled by people who can't separate upload from the idea of being "rewarded". It is only good when you upload to other clients than the ones that are uploading to you, while it is wrong when you upload to the same clients you are downloading from. In the first case your queue is "open", you are uploading all files you are sharing, in the second your queue is "closed", you are uploading only files you are downloading.

This post has been edited by LorenzoC: 14 May 2008 - 09:05 PM