OPN status update, assistance request

Posted 1 Mar 2001 at 08:49 UTC by lilo Share This

Open Projects Net is an extremely small infrastructure project which, in its present form, has provided services to the community since early 1998. OPN IRC provides interactive services to projects and groups involved with "Open Source, Open Technology, Open Information". We are about to embark on a program of accelerated growth. This article describes the state of the project and mentions areas where we could use assistance.

The problem.

Open Projects Net has existed in its current form since early 1998. At its inception it had about 100 users and some small number of channels. It currently has peak load of nearly 3,000 clients and 1,000 channels. To a certain extent, we're a victim of our own success. We've had a fairly steady growth rate, doubling every 5-7 months. Some of the affiliated groups are listed here. In addition, we have architecture channels relating to MIPS and PowerPC and various small development channels, as well as publications channels and support groups.

Various internal projects must come to usability for us to keep running and growing. We are converting to a modified hybrid server which should reduce the impact of various denial attacks, but we have no coders dedicated to finishing the project. We emphasize friendly communication and project coordination, but staff training is essential and, as the founder, I have little time to write and test coursework. We've begun a process to develop a long-term replacement for IRC, but this is the early design stage and recruitment of developers with cryptographic expertise to complete the design and begin coding takes time which, again, is lacking.

I should be working on OPN fulltime, managing the projects, recruiting developers, working on training and operations issues. To accomplish this, we need funding, so I've spent time talking to businesses in the community. To date, while I've had very good access to upper management in several large Linux companies, I can report that none of them see a business advantage in funding a community project as "out of the box" as ours, with as little direct revenue opportunity for them. And if we were to change our goals to produce that direct revenue, we would be a different project, one a lot less useful to the community.

The solution.

We need at least one full-time participant to grow well, and we need to grow to be able to fund a fulltime employee. So I've constituted a new internal project to resolve the situation. I call the project Mister Toad's Wild Ride.

Its goals are to publicize the network and attract new projects and users, to scale up in bumpy but serviceable fashion to the point where funding via voluntary contributions becomes feasible. To this end, we've redesigned the OPN website and placed it on ibiblio, where it should be able to handle a fair number of accesses. We've reworked the priorities in the TODO lists for the dancer-hybrid-6 and dancer-hybserv conversions in an effort to make the individual items a bit easier to tackle. We've commissioned some banner ads to provide to organizations that are willing to display them and the process begins.

Help we're looking for.

Primarily we need coding support, in the form of developers experienced with the hybrid-6 ircd tree who are willing and able to bang out solid patches for our tree, some of which will be a bit unusual. If you have professional graphic arts experience, we could use more banner ads. And if you're in the media, we need publicity. Contact me (lilo at openprojects dot net) and I'll be glad to talk to you about the network and how you can help. You can also look at the web site or find me online at irc.openprojects.net on channel #mistertoad.

Thanks in advance for any assistance you can provide.

Random Thoughts, posted 1 Mar 2001 at 20:35 UTC by Ankh » (Master)

Not sure I'd start with hybrid, but it's probably a better code base than ircu or bahamut right now.

We've talked about open sourcing SorceryNet services for a while; like a lot of IRC projects, they were written by someone who was learning C, and then gradually cleaned up, but the good side is, at least they weren't written by someone who was learning C++ or LISP :-)

I think there are a number of directions a project like OPN could take; stronger web/habber/IM/mail integration, for example, getting rid of the min span tree, and maybe using LDAP or soemthing for distributed services.

The hardest part of running an IRC network is having motivated staff who are there to help users, not just to sneer at them and feel elite.

I don't know, though, of a good business model for IRC. There are IRC clients that show adverts to users (I keep a list of IRC clients and some of them are pretty odd), but I don't think that would go over well.

I haven't had time to do much with irc++, but I would like to see at least some of the ideas used, so feel free to steal them.

hybrid-6 is nearly out of date :-), posted 2 Mar 2001 at 16:57 UTC by k » (Journeyer)

As some people know, I'm one of the ircd-hybrid hackers and even though I'm not active right now, I was part of the small group of people who went through the hybrid-7 codebase a couple of months ago and redesign most of its internals from the ground up.

Its nowhere near completely stable, but with just a little more work it could become stable. hybrid-7 is flexible enough to adapt to quite a lot of funky new-ircd concepts people have been toying around with over the last few years but have never quite implemented. (Which, once I've finshed some other stuff, I'm going to come back to.)

In any case, how about migrating some of your patches to hybrid-7? Hell, I'd even help you migrate some of your patches over. (But I'm not volunteering for anything else, lord knows I have too much going as it is.) Hybrid-7 should also fix a few on-irc-DoSses, but it obviously won't fix the IP layer DDoSing that is popular today :-P

IP DOS attacks, posted 3 Mar 2001 at 00:58 UTC by Ankh » (Master)

WIthin about an hour of introducing IP masking, whereby IRC commands such as whois no longer show user's hostnames or IP addresses, almost all IP DoS attacks stopped on SorceryNet. The scheme is not perfect by any means, but we had about 200 sub7 bots online before, and now we sometimes get one or two.

We also check for open SOCKS servers, but I presume hybrid7 does that too.

We cannot, of course, stop DoS attacks against the servers, as people have to be able to connect to those to get onto the network.

DoS, mask, posted 3 Mar 2001 at 07:54 UTC by Malx » (Journeyer)

Are you cheking for other proxies also? (Squid-ssl-connect?)
Is IP masked in /whowas ?
(Irc.ForestNet.org also have IP masking for Non-Oper users).

Re: Comments, posted 5 Mar 2001 at 00:11 UTC by lilo » (Master)

Thanks to everyone for your comments. Agreed, Ankh, that motivated and friendly staff is key. We aren't doing badly in that area, but there is always room for improvement, and one of the internal projects has as its purpose to do just that.

k, I think hybrid-7 would do the job for us with some modifications, if it were in production. It's been a disappointment to me that we have had so much trouble finding coding help from hybrid-6 developers, so soon after it was released. Everybody's getting excited about working on -7. It's understandable, I guess, but makes things difficult.

Malx, thanks for your comments. We will end up doing some proxy checking. Our server hiding does make the appropriate changes to WHOWAS output as well.

To anyone who might have skills we could use, please consider putting in some time. Thanks to everyone who has spent time on OPN in whatever capacity, and those of you who have expressed interest.

Business/Community Interface?, posted 5 Mar 2001 at 12:10 UTC by davidw » (Master)

Would packaging openprojects as a way for businesses to get involved in open source be workable in some fashion? Create a set of tools to get businesses on line, and provide assistance in setting them up with what they need?

Just a random thought...

whats your goal?, posted 5 Mar 2001 at 22:44 UTC by Netdancer » (Journeyer)

When I read your request for help, I wondered what your goal is. Is it to become a professional infrastructure/support organization? If yes, why not become a business? If thats not your goal, in which ways are you going to differ?

re: whats your goal?, posted 7 Mar 2001 at 08:49 UTC by lilo » (Master)

Netdancer wrote:

When I read your request for help, I wondered what your goal is. Is it to become a professional infrastructure/support organization? If yes, why not become a business? If thats not your goal, in which ways are you going to differ?

Hmm. I guess OPN is already doing most of what I want it to do, at least in some basic form. We host interactive discussion areas for projects, we provide avenues for online support. We've been doing this for free for a long time, and I think I would rather we continue to do it that way. Most of what we do is done with donated hardware and bandwidth and money never changes hands at all. That seems to work well, because it encourages participation from people who care about the result rather than just the payment. The opposite attitude seems to be a common thing these days.

On the other hand, I would like to manage the network full time, and I think it would be a beneficial thing. Paying salaries would be of the few things that really does require some sort of budget.

Would putting up banner ads in some way damage the network? I'm inclined to think so. It would change the atmosphere a bit more than would be healthy. So I'm asking for contributions to the operating fund. We've gotten some so far and if the network keeps growing, it'll eventually get to the point where those contributions will be respectable in size. Maybe enough to pay a few salaries.

The thing is, I'm not sure we fit any of the conventional models for a professional infrastructure/support company and I think I'd be unwilling to get VC to do what we'd do, because the VC would almost inevitably change what we do into something quite different. It's possible we'll stay something like our current amorphous structure, it's possible we'll form a non-profit to get the idea across that we really are just trying to do some fairly simple things without embellishment. It just remains to be seen as we continue to grow.

Business/Community Interface?, posted 7 Mar 2001 at 09:00 UTC by lilo » (Master)

davidw wrote:

Would packaging openprojects as a way for businesses to get involved in open source be workable in some fashion? Create a set of tools to get businesses on line, and provide assistance in setting them up with what they need?

Just a random thought...

There are other people trying to do similar things in this space. Collabnet comes to mind. They're both of the community and of the business world and their business model seems to be to help foster that business-to-community connection. They seem to be prospering at it and I have no doubt they're doing good work.

I guess at this point I am a little bit more interested in the community-community connection. There are going to continue to be plenty of jobs for people working with free software, take that as a given. But the community itself needs to grow, not just to pay salaries, but to achieve the kinds of things it can achieve. Maybe that seems too introspective, I don't know. But it seems to me that everybody is interested in using the community as a resource, and maybe it's a resource that ought to be given a little watering for its own sake.

New Advogato Features

New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.

Keep up with the latest Advogato features by reading the Advogato status blog.

If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!

Share this page