dollar sign and ignore feature

default · Post by **default** » 2003-02-25 03:45

how come we can't use a dollar sign ($)?? is this due to the programming language for DC++?? :/

and how about an ignore feature on users in the chatroom?? =]

-default

qqzm · Post by **qqzm** » 2003-02-25 04:14

It's due to the fact that Jon Hess (the programmer of the original DC) is a moron.

aDe · Post by **aDe** » 2003-02-25 06:46

dollarsign ($) and pipe (|) are characters used by the protocol of dc and therefore you cant use them, but theres some work in progress to escape those characters so they can be used..

cologic · Post by **cologic** » 2003-02-25 09:05

The dollarsign doesn't have to be escaped at all in main chat and private chat messages, actually... The pipe still does, though.

Post by **GargoyleMT** » 2003-02-25 19:58

default wrote:how come we can't use a dollar sign ($)?? is this due to the programming language for DC++?? :/

and how about an ignore feature on users in the chatroom?? =]

Oh, and cologic's mod, BCDC++ has both features. You can use it or bug him or someone else to port them to DC++ proper though.

KLesK · Post by **KLesK** » 2003-02-26 04:29

oDC have this feature too ( $ and | )

Opera · Post by **Opera** » 2003-02-26 07:22

And have had for a loong time now, by simple converting a $ to [Ã‚Â§] and vice versa. The same for | is [Ã‚Â¦]

I guess I was the one setting this standard

Btw, I don't think we should mention mods or forks that contains fake- or leech features in this forum. Try to keep the quality a little bit higher than that, please.

Opera

cologic · Post by **cologic** » 2003-02-26 08:40

If you count four months (since 10-10-2002, when your release page indicates you added $ and | support) as a "looong time", I wonder what you'd call 9 months (since I added it to BCDC++ at the end of May 2002). So despite the sanctimonious suggestion in your post, you weren't "the one setting the standard." (BCDC++ contains contains no fakeshare features, and never has.)

Futher, you don't need to escape $ (in main chat and private messages, anyway, and in general any message without embedded $$'s to separate arguments), so your implementation is unncessarily incompatible with standard DC clients. The clients merely have to look for the message termination character |, which does have to be escaped.

Opera · Post by **Opera** » 2003-02-26 09:25

Okey, noone mentioned it to me before I implemented it. Maybe I should've checked it out before I said I was first

sorry.

About your mod:

allow a user-adjustable small file size
Can you make 4 gig a limit? Great, then you can enter hubs with maximum slot rules and still break the rule, since everyone can download from you anyway

allow a custom description string format
Great, now I can enter the kind of desc tag I want. Wanna hide "H:"? No problem.

allow upload limiting below 6kB/s
Okey, this implies download limiting aswell, but it still is an upload limit

better masquerade as other DC clients
So now I can enter a hub that thinks I'm nmdc, so they won't bother that I'm in 1000 hubs at the same time. Yeah!

remove the check for an already running instance
Do I need to comment this further?

learn user/IP correlations, and uses that information to resolve known search spy IPs into nicks
Big Brother is watching you... I did this feature myself 4 months ago. Didn't release it. And removed it after just a few days. This kind of spy should be used in the hub, if anywhere.

automatically grant slots to users to favorite users
I've already commented this

support ignoring users
I hope this is in chat/pm only, and not by connection. Not that I would be suprised if it was..

I think all of these by them selves fall under the category "bad clients" that many hubs ban. Not very strange dc++ got banned from most hubs like a year ago.
People just can't stay out of messing with it.

Opera

Opera · Post by **Opera** » 2003-02-26 09:27

Btw, on your homepage I only found the binary, not the source. If you didn't know, this is against the GPL license.

Opera

Opera · Post by **Opera** » 2003-02-26 09:28

*Stupid me*.. Didn't see the "related files" section... Or maybe it was ment that way

Opera

aDe · Post by **aDe** » 2003-02-26 11:09

BCDC++ gives move freedom to the user, its not an explicit faking program, however if i was cologic i'd just keep that client to me and other trusted people.
I like it though. also- cologic doesn't have a donate button leading to his own account.

Napro · Post by **Napro** » 2003-02-26 11:53

i use BCDC++ i use it because from what ive found its the only DC/DC++/Whatever client that alows u to find ogg filez when searching for audio, ogg media when searching for video, etc. i got no real complaints agaist the client itself expect for the ppl who...abuse its "features"

Napro · Post by **Napro** » 2003-02-26 11:53

and also about the custom tag...WOULDNT AN OP SOMEWHERE GET SUSPICOUS?!

cologic · Post by **cologic** » 2003-02-26 16:40

Can you make 4 gig a limit? Great, then you can enter hubs with maximum slot rules and still break the rule, since everyone can download from you anyway

4GB probably isn't a valid limit anyway, as it uses a 32 bit quantity as I recall, but, say, 2GB certainly is. However, that still doesn't allow one to circumvent maximum slot limitations, as there are a limited number of "extra" slots that can be used for small files and file lists. It opens up to 3 slots only.

Code: Select all

int getFreeExtraSlots() { return max(3 - getExtra(), 0); };

allow upload limiting below 6kB/s
Okey, this implies download limiting aswell, but it still is an upload limit

Er, if the person can't download faster than 5kB/s (or 2kB/s if they're capping at 2kB/s), they can't be much of a leech can they?

better masquerade as other DC clients
So now I can enter a hub that thinks I'm nmdc, so they won't bother that I'm in 1000 hubs at the same time. Yeah!

I'm annoyed with hubs that discriminate against certain clients.

remove the check for an already running instance
Do I need to comment this further?

I believe NMDC can already do this; in any event, each instance's upload slots and number of hubs can be checked, so this isn't a particularly effective leeching tool. Further, you could always run DC++ as another user, bypassing the single-instance limitiation.

learn user/IP correlations, and uses that information to resolve known search spy IPs into nicks
Big Brother is watching you... I did this feature myself 4 months ago. Didn't release it. And removed it after just a few days. This kind of spy should be used in the hub, if anywhere.

Users should assume even without BCDC++'s presence, their searches don't have any privacy, and this essentially ensures users who care know. I'd rather know about unpleasantries in the network than pretend for my comfort that because commonly available clients don't implement some pretty obvious feature, that the protocol doesn't reveal that information.

automatically grant slots to users to favorite users
I've already commented this

Link me to them please, as I haven't read them (unless you're referring to bypassing the max slots/hub)? Someone can grant slots to the entire hub with stock DC++, as well. This permanent slot granting also forms more of a bond between users within a hub.

support ignoring users
I hope this is in chat/pm only, and not by connection. Not that I would be suprised if it was..

This is only for main chat at the moment, though I may at some point extend it to private messages (it's not really a defense against a user annoyed at one specifically as of yet, so much as one spamming main chat).

So far, my reponse has been mostly a point-by-point rebuttal of how BCDC++ is more limited than you give it credit for; however, it also serves as an example of a non-hostile (if you're worried about something as relatively benign as BCDC++ harming DC, DC has no chance of surviving) as a viable filesharing network.

Post by **GargoyleMT** » 2003-02-26 16:42

Napro wrote:i use BCDC++ i use it because from what ive found its the only DC/DC++/Whatever client that alows u to find ogg filez when searching for audio, ogg media when searching for video, etc. i got no real complaints agaist the client itself expect for the ppl who...abuse its "features"

The search type, of adding the .ogg and .ogm extensions to your ShareManager, will only work on local files. If you search for Audio files, a BCDC++ user will return .OGG files, but a normal one won't. Of course it works, but I don't think it works the way you think it does.

Opera, I admire your work on your oDC++, are we going to see more bug fixes and features making their way into DC++? (The same question goes out to cologic and sarf, too.

)

smitty · Post by **smitty** » 2003-03-01 14:50

Open Direct Connect Has Been Able To post $ and | for over 2 years now so neither of you started it...lol

Napro · Post by **Napro** » 2003-03-03 17:22

GargoyleMT wrote:The search type, of adding the .ogg and .ogm extensions to your ShareManager, will only work on local files. If you search for Audio files, a BCDC++ user will return .OGG files, but a normal one won't. Of course it works, but I don't think it works the way you think it does.

i guess i was misunderstanding on that

sarf · Post by **sarf** » 2003-03-04 04:15

GargoyleMT wrote:[snip]
Opera, I admire your work on your oDC++, are we going to see more bug fixes and features making their way into DC++? (The same question goes out to cologic and sarf, too. )

Well, as I've released the source code to my latest client (unlike some people

) it's just for arne to cut'n'paste what he likes... I might put up a specific patch/diff/how-to for adding ed2k/sig2dat/dchub links though, as that info is quite useful, as well as putting up the code for the "ip" command (not that it is hard to do, it's just that it's a necessary feature for some fellas with changing IPs).

Sarf
---
Acting without thinking can be awfully entertaining.

aDe · Post by **aDe** » 2003-03-04 04:40

GargoyleMT already sent a patch for the IP command:

arnetheduck wrote:Anyway, the ip handling code needs to be updated to work on a per-hub basis (at least when nothing is entered in the ip field)...after that, I'll seriously consider the patch gmt sent me...

Anyway- if you like, you could update that code he is talking about.

sarf · Post by **sarf** » 2003-03-04 06:44

Oh. Alright. Well, the ip command is needed for those with pesky ISPs who change their IPs/force them to change their own IPs (by force reconnect every 24 hours or whatever).

Sarf
---
Submitting to the will of a higher power has never appealed to me

aDe · Post by **aDe** » 2003-03-04 06:55

yep it sure is, i am one of them myself.

for those wondering what this is,
here is the feature request,
here is the document,
and here is the topic.

Post by **GargoyleMT** » 2003-03-04 08:08

sarf wrote:Well, as I've released the source code to my latest client (unlike some people ) it's just for arne to cut'n'paste what he likes... I might put up a specific patch/diff/how-to for adding ed2k/sig2dat/dchub links though, as that info is quite useful, as well as putting up the code for the "ip" command (not that it is hard to do, it's just that it's a necessary feature for some fellas with changing IPs).

You obviously missed this memo: (from a thread in the resort)

arnetheduck wrote: Well, I'm not going to hunt a hundred modified ++'s, if they want a feature in the main release, they'll have to send a nicely done patch...

If you're adverse to that, I'll see if you have any good features (;-)) and make a diff and send it in myself (with appropriate credit).

aDe wrote:GargoyleMT already sent a patch for the IP command

And just to make sure credit goes where credit is due, I just took it from cologic's BCDC.

aDe · Post by **aDe** » 2003-03-04 08:24

GargoyleMT wrote:And just to make sure credit goes where credit is due, I just took it from cologic's BCDC.

Hehe, and I invented the command!

Seriously though, did you make this patch after the issues with setting to local ip and not the otherway around, and did you add that other internal IP range that was not included at first?

Post by **GargoyleMT** » 2003-03-04 08:47

aDe wrote:Hehe, and I invented the command!
Seriously though, did you make this patch after the issues with setting to local ip and not the otherway around, and did you add that other internal IP range that was not included at first?

Guilty, on both counts.

So, is there a uniform way for hubs to say that they support the $UserIP command, or is the best way to just send it and count on the hub to ignore it if it doesn't understand the command?

aDe · Post by **aDe** » 2003-03-04 08:53

GargoyleMT wrote:So, is there a uniform way for hubs to say that they support the $UserIP command, or is the best way to just send it and count on the hub to ignore it if it doesn't understand the command?

Since there is a script which adds support to the NMDC hub, there is no good way of knowing. A $Supports command is possible for new hubs, but as you know not possible with NMDCHub. (at a good level, in / before / just after the $Lock..)

So, yes, the client should try to send it blindly. The NMDC hub ignores it, and i believe other hubs do too...

Phantom · Post by **Phantom** » 2003-03-04 15:48

Is it really important to know if a hub supports it or not? You only need to connect to 1 hub that supports it for it to set your ip. Maybe just get one of your regular hubs to add support.

Post by **GargoyleMT** » 2003-03-04 20:19

Phantom wrote:Is it really important to know if a hub supports it or not? You only need to connect to 1 hub that supports it for it to set your ip. Maybe just get one of your regular hubs to add support.

Well, why send any command that the remote side might not understand if you can detect whether or not it will be understood beforehand? But, as aDe pointed out, just sending it seems to be the best way to handle it. Maybe at a later date when all hubs have a built in facility for showing their capabilities, it can be handled the "right" way.

Or... something.

DC++

dollar sign and ignore feature

dollar sign and ignore feature

Re: dollar sign and ignore feature

Open Direct Connect by Luke Jr.