services.xo and services.us

General talk about EFnet

Moderators: Website/Forum Admins, EFnet/General Moderators

-wassup-
Posts: 103
Joined: Wed Aug 13, 2003 8:25 pm
Location: Middle East

Postby -wassup- » Fri Sep 05, 2003 11:37 pm

didnt know u mean cpu cycles :oops:
User avatar
dubkat
Posts: 40
Joined: Fri Sep 19, 2003 4:59 am
Location: Colorado, USA
Contact:

Postby dubkat » Mon Apr 19, 2004 8:02 am

seiki wrote:
-wassup- wrote:is there any reason on why services.int is closed source?
that's a very good question which more and more admins are asking these days..

-seiki
I know this post is a little bit old, but i noticed no one put a little update. While services.xo/int is closed source, there is Open Chanfix which can be found at http://www.garion.org/ocf/ which is, as the name implies, open souce.
Hardy
Site Admin
Posts: 394
Joined: Wed Jul 02, 2003 4:54 pm
Location: Oslo, Norway
Contact:

Postby Hardy » Mon Apr 19, 2004 8:25 am

dubkat wrote:
seiki wrote:
-wassup- wrote:is there any reason on why services.int is closed source?
that's a very good question which more and more admins are asking these days..

-seiki
I know this post is a little bit old, but i noticed no one put a little update. While services.xo/int is closed source, there is Open Chanfix which can be found at http://www.garion.org/ocf/ which is, as the name implies, open souce.
Old thread it is.. however the current services.int are finally open source also after some preasure.. http://www.codestud.com/services/
-- Hardy
Administrator: irc.underworld.no
Services Administrator
http://www.efnet.org admin/staff
-wassup-
Posts: 103
Joined: Wed Aug 13, 2003 8:25 pm
Location: Middle East

Postby -wassup- » Sat Apr 24, 2004 11:28 am

is there any chance the source for services.xo will be released?
leeh
ircd-ratbox coder
Posts: 48
Joined: Wed Jul 02, 2003 5:43 pm
Location: UK

Postby leeh » Mon Apr 26, 2004 12:03 pm

-wassup- wrote:is there any chance the source for services.xo will be released?
Id doubt it - the features were supposed to have moved to services.int with services.xo dissapearing a long time ago..
Hardy
Site Admin
Posts: 394
Joined: Wed Jul 02, 2003 4:54 pm
Location: Oslo, Norway
Contact:

Postby Hardy » Mon Apr 26, 2004 1:10 pm

But i belive Sentinel IRC Services should cover most of the needs if you need a "services.xo lookalike". from what i understand Sentinel IRC Services is built based on the code services.xo has.

Then again, i could be wrong :)
-- Hardy
Administrator: irc.underworld.no
Services Administrator
http://www.efnet.org admin/staff
-wassup-
Posts: 103
Joined: Wed Aug 13, 2003 8:25 pm
Location: Middle East

Postby -wassup- » Tue Apr 27, 2004 8:08 pm

yes that is what i currently use on my internet and i would like to extend a great thank you to Hwy for writing and maintaining it. i am pretty sure that it doesnt use any services.xo. it was based on geostats and extended far beyond that.

http://ircd.botbay.net/sentinel/ <--for anyone interested
Quai
Posts: 5
Joined: Fri Feb 06, 2004 8:43 am
Location: Drammen, Norway
Contact:

No need for services.xo anymore

Postby Quai » Sat May 01, 2004 9:43 am

With the TESTMASK command in ratbox (and other ircds?) there are no need for tcm/oomon to query xo anymore. Its an oper-only command to see how many clients matching a u@h mask.
ie;

/quote TESTMASK *@avalonworks.ca

Right now, it shows;

-!- *@avalonworks.ca 15 0 Local/remote clients match

and nothing more :)

And the DRONES command on .xo is preatty useless.
--
Kay Sindre Bærulfsen
// Quai * irc.avalonworks.ca
Yeah, my english sucks!
Hwy
Posts: 66
Joined: Wed Jul 16, 2003 12:27 pm

Postby Hwy » Sat May 01, 2004 2:14 pm

Sentinel has no services.xo code, I have not seen services.xo's code and 99% likely never will. It was made to work with TCM/oomon (with /msg services@services.blah clones 4) and to provide a functional DRONES command. The functionality in DroneMon and current OOmon blows Sentinel's DRONES command away, however.

Also, TESTMASK is included with ratbox 1.4 and higher only. The original module worked with earlier ratbox and hybrid-7 releases (though probably not with 7.0/7.01 and certainly not 7-CURRENT) and is available from http://ircd.botbay.net/pub/hwy-mods/m_testmask.c. Note that the original module uses server NOTICE's and the ratbox module uses proper numerics.
-wassup-
Posts: 103
Joined: Wed Aug 13, 2003 8:25 pm
Location: Middle East

Postby -wassup- » Mon May 03, 2004 6:52 pm

anyone know if tcm/oomon/dronemon support testmask yet?
Hwy
Posts: 66
Joined: Wed Jul 16, 2003 12:27 pm

Postby Hwy » Mon May 03, 2004 7:22 pm

TESTMASK is primarily something for opers to use. A bot probably doesn't need to use it.

Also, it appears TESTMASK is under a bunch of political bullshit, so it may end up going away (don't quote me on that)
Hardy
Site Admin
Posts: 394
Joined: Wed Jul 02, 2003 4:54 pm
Location: Oslo, Norway
Contact:

Postby Hardy » Mon May 03, 2004 8:15 pm

Hwy wrote:TESTMASK is primarily something for opers to use. A bot probably doesn't need to use it.

Also, it appears TESTMASK is under a bunch of political bullshit, so it may end up going away (don't quote me on that)
/me pushed the quote button :D

Yea, the discussion is if our ircd should be allowed to gather information about how many clients there is globaly on the network. Some would like to continue having global stats on a global service instead of the ircd, while others want the functionality locally..
-- Hardy
Administrator: irc.underworld.no
Services Administrator
http://www.efnet.org admin/staff
Hwy
Posts: 66
Joined: Wed Jul 16, 2003 12:27 pm

Postby Hwy » Mon May 03, 2004 8:51 pm

Well, TESTMASK is open source, services.xo is not. That's a plus there.

It could be easily made to OPERSPY or be logged. That can be done with 1-3 lines of code.

TESTMASK came about after I got bitched out on the EU EFnet mailing list about seconing a GLINE that some said was too broad. To devise a solution to the problem of knowing if a mask was right or too broad, I wrote TESTMASK first in Sentinel, as I felt that's where it should go. Shortly after, I included an implementation in my TCM-like ircd searching module, as it's easy to do and not TOO CPU intensive, as well as more available to operators. This sat for about a year until ratbox 1.4-2 included an improved implementation.

It was INTENDED to just help opers choose a proper GLINE mask, so that no innocents are caught. In this age of OPERSPY, I don't see why it's even questioned...as it's simply counts. TESTMASK can't show the users or what server's they're on, and it's all stuff that you can do with the OPERSPY variant of WHO and other, less efficient, methods (which are not logged).
Hwy
Posts: 66
Joined: Wed Jul 16, 2003 12:27 pm

Postby Hwy » Mon May 03, 2004 9:00 pm

And also, TCM and oomon query services.xo for the CLONES command. There is no easy way to do an equivalent in ircd, and TESTMASK can't really do it either, since you need to know the user@host you wish to test.

CLONES #

Lists all clients services.xo considers clones, where the number you send is the number of clones that are not allowed. The current default for services.xo appears to be 6, it lists any user@host that has 6 or more connections to EFnet. TCM and oomon use 4 by default, allowing users to have 3 clones networkwide, but if they get 4, they'll be klined.

(and yes, non-opers are still permitted to use the CLONES command)
Hardy
Site Admin
Posts: 394
Joined: Wed Jul 02, 2003 4:54 pm
Location: Oslo, Norway
Contact:

Postby Hardy » Tue May 04, 2004 7:47 am

Hwy wrote:Well, TESTMASK is open source, services.xo is not. That's a plus there.

It could be easily made to OPERSPY or be logged. That can be done with 1-3 lines of code.

TESTMASK came about after I got bitched out on the EU EFnet mailing list about seconing a GLINE that some said was too broad. To devise a solution to the problem of knowing if a mask was right or too broad, I wrote TESTMASK first in Sentinel, as I felt that's where it should go. Shortly after, I included an implementation in my TCM-like ircd searching module, as it's easy to do and not TOO CPU intensive, as well as more available to operators. This sat for about a year until ratbox 1.4-2 included an improved implementation.

It was INTENDED to just help opers choose a proper GLINE mask, so that no innocents are caught. In this age of OPERSPY, I don't see why it's even questioned...as it's simply counts. TESTMASK can't show the users or what server's they're on, and it's all stuff that you can do with the OPERSPY variant of WHO and other, less efficient, methods (which are not logged).
I belive the discussion is where do you draw the limit on what you hide into a ircd version. If it just gave the output of local users no one would have said anything, but most draw their limits when it comes to performing global searches of hosts on ircd with no accountability.

And even tho it "is the age of OPERSPY" , operspy sends a global broadcast to all opers when used, testmask dont.

The discussion isnt about the testmask itself, its where you draw the limit of what should be a local command vs. a global services command. Not every admin of this network has the same common sense when picking modules that the rest have..
-- Hardy
Administrator: irc.underworld.no
Services Administrator
http://www.efnet.org admin/staff

Who is online

Users browsing this forum: No registered users and 14 guests