Speak Freely To Be Withdrawn January 15

wrenhunt writes "The Speak Freely site has this: 'On January 15th, 2004, Speak Freely will be discontinued and removed from this Web site. Existing users may continue to use the program as long as they wish, but no further releases will be forthcoming. For details and the reasons why Speak Freely is being discontinued, please see the full end of life announcement.'" The reasons are various and interesting; it's graceful of the author to provide an explanation of why a piece of software is going away. Update: 01/11 19:22 GMT by T : As reader pi_rules points out, this story is a duplicate -- my apologies.

Comments

Cheap routers..

[Aliencow]

Why isn't it easier for people to open up ports on their cheap routers ? Tell someone to "Just forward your port 4893 to your computer" and they'll look at you like you're an alien, so why not include an application to do it that goes in their start menu (in addition to the web based interface) that would detect software trying to listen, and then asking if you want these to be open ? A bit like ZoneAlarm but controlling the router...

Re:Cheap routers..

[uradu]

> He's referring to ISPs NATing off their customers, not customers being restricted by their own routers

His rant gives no indication either way, I don't know how you draw that conclusion. Your own experience (and mine, and most others') tells you that you've never heard of ISP-level NAT, so why would he mean that? He's just bitter about NAT for whatever reasons and venting by the most dramatic means he has: EOL-ing a fairly popular piece of software. Well, I know why he hates NAT, but that's hardly NAT's fault, that's similar to getting angry at the color Yellow for being so bright. Instead of pouting, he could think about or work on some generic method to overcome NAT's inherent weaknesses.

In fact, since--as he himself puts it--NAT will be with us for a long time, even after switching to IPv6, it might be very worthwhile for him to think about methods of addressing private computers below the transport level, but above the application level. A universal method of sub-addressing machines would be very useful, since not all machines will ever be on the public internet, whether for security or address limitation reasons. Port mapping works well enough for some things but has inherent limitations (16 bit, many apps assume fixed ranges etc.), and ports were really meant to identify applications on a single machine, not machines on a network. It's really a hack, and you don't build future technologies on hacks.

Re:Cheap routers..

[TheSpoom]

Then it's time for a paradigm shift, since I've obviously been misunderstanding.

Admittedly, NAT can stop inbound connections from reaching a computer that otherwise would receive all connections had it not been behind a NAT router. But my computer is no longer a peer on the internet; my NETWORK is now a peer on the internet, with ports opened and forwarded to multiple machines as I see fit. In one way of thinking, it allows me to use the computers in my home more as I would had I been running a corporate perimeter network, with different machines running web servers, FTP servers, and the like.

Admittedly, Joe Sixpack has no idea why his computer won't allow inbound connections anymore after he's put a router on his network, but here's the thing: Joe Sixpack has no idea what an inbound connection is, nor, likely, does he even know SpeakFreely even exists. If Joe Sixpack doesn't want the feds snooping on his conversations, he'll find a way to forward his ports, like all decent home-level routers allow. If John Walker wanted to combat this NAT-related inability to use his software, why didn't he just post some documentation or links showing how users can forward the correct ports? The moment Joe Sixpack wants to use SpeakFreely, he could go to the site and see "hey, I have a Linksys router, and this link that says 'IF YOU HAVE A ROUTER CLICK HERE' shows me how to get around it!"

IMHO this whole end of life thing seems a bit much if it's based entirely around home-level routers, as this issue is largely avoidable.

1996 will be a very exciting year for the WWW.

[Doc Ruby]

How about a Slashdot search engine that accepts boolean operators and phrases? Or searching on a phrase plus other fields in the comment/story's DB record, like author, date, topic/section? A better search engine would use less server resources when searching, and members could search their own post history to link a new comment to an old, but still relevant, point. Slashdot's server seems to use something like the ancient "swish" freeware. This post is practically a quote of a similar email I sent to a customer back in 1995! These features are coded by Slashdot users every day. Who will help me add it to the Slashcode? Who at Slashdot is interested in rolling it out at Slashdot? I'd rather code than complain.

Re:Too bad -- design was obsolete

[uradu]

> As has been pointed out, what we really need are easier solutions such as port forwarding

What we really need is a generic method of sub-addressing machines. The public/private network paradigm is here to stay for various reasons, so we should shape our protocols to cope with that. We need another protocol between IP and TCP/UDP: IP addresses a point-of-presence on the internet, TCP/UDP a POP on a machine (i.e. an app), we need something that addresses a POP on an internal network. In fact, it could be a nestable protocol that replaces IP and allows for unlimited levels of private subdivision. That way a large company could have multiple internal NAT setups and you could still address a specific machine several levels down the hierarchy. I guess one could modify IP to be nestable, and IP stacks inside routers to be aware of it. Then you would address a private machine as a.b.c.d/e.f.g.h where a.b.c.d is the public IP address, and e.f.g.h the private one. The public NAT router would examine the next nested IP header (in this case e.f.g.h) and pass the packet to the correct internal machine (which could be another NAT box, ad infinitum).

The downside of course is that we're then back to the old UUCP days where you had to explicitly specify the route to the destination machine, making the network more fragile. Still, given that for the vast majority of setups it would be just a two-tiered setup (public internet and internal LAN), it should be workable.

Re:Too bad -- design was obsolete

[mysticalreaper]

Wow. I *completely* disagree with what you've just stated here. Allow me to explain why.

First off, the internet was BUILT as an end-to-end network. You cannot just sweep this fact aside by saying it's "outdated". This principle is what MADE the internet successful. Without end-to-end, the internet would have gone nowhere. Really.

We want the application to run end-to-end, because that is what make the application useful -- but folks have confused this with requiring the mechanism to be identical from end to end

But now, in the new system, it requires that the network be AWARE of the application, and configured EXPLICITLY to allow this certain type of data to be transferred. Now you have to ask permission from the people who control the network to run your application. Now you have to make configuration changes in the network itself before you can run any new application. Gone is the open development environment of the internet. Gone are new applications that pop up that anyone can use immediately. (This is how the web started. Your NAT support would have made the web so difficult that it wouldn't have gone anywhere. Imagine the millions who would have had to configure their NAT to work with a new system of doubious worth.)

You say that the network should be SEPERATE from the application, and then go on to promote the application being DEPENDANT on the specific configuration of the network.

"like in the days of the telegraph, the mechanism and the application were synonymous. That is an obsolete model, though. Our needs and demands have gotten more varied and complex from the point of view of the applications -- the mechanism (IPv4) needs to be separated out from the applications."

AND IT IS! That's the POINT, Bookwyrm. Currently, in the 'obsolete' model, the network is TRANSPARENT to the application. No specific configuration of the network is requried. The network is seperate from the application. However, NAT makes the application depend on the network, and thus makes the network and the application once again joined, like the telegraph, phone and cable TV networks of the past. That's a step BACKWARDS.

Even now, because of NAT, we can observe the harmful effects of new development. VoIP doesn't work properly. File sharing applications are suffering massively because people can't share, even when they want to. Running a server of any kind, (a game server for you and your budies to play on) requires additional configuration, making it harder. People in certain situations, like in university, for example, have no ability to influence the functionality of the NAT, and are stuck being internet consumers. And don't forget that it's even MORE arduous to have multiple computers doing the same thing, like being a webserver, behind the NAT. Now you have to specify to the CLIENTS to use different ports for different servers behind the NAT. It begins to get so ugly that people give up.

Your goals are noble, Bookwyrm, but your thoughts on the matter are misguided. This site might help shed some additional light on the situtation.

And finally, the people who invented the internet for real though that end-to-end addressing was the best idea, and from their efforts, we have the most advanced communcation system humans have ever seen. To say that they are utterly wrong requires some guts, and also a LOT of backing up. In other words, the proof is in the pudding. Where is YOUR all NAT internet?