Slashdot Mirror

pork
naim

Pork hasn't been updated in 4 years, but I've used it constantly since I found it ~5 years ago and it does everything I need to.

If I really want video chat I'll fire up iChat.

Go Screen!

Received: by 10.54.59.6 with SMTP id h6mr97856wra;
Mon, 07 Feb 2005 13:03:03 -0800 (PST)
Return-Path: <newman@newman.newman-grt.oscar.aol.com>
From: AOL Instant Messenger <newman@newman.newman-grt.oscar.aol.com>
To:
Sub ject: AIM Account suspended!
Mime-Version: 1.0
Date: Mon, 07 Feb 2005 13:03:03 -0800 (PST)

AIM User

We have detected that the AIM Screen Name sixtyfourcubits associated with
this email address has been used in a way that violates the terms and
conditions of the AOL Instant Messenger service. Because of these
violations, AOL has suspended the use of this screen name. If you
believe that this suspension has occurred in error, please fill out the
form located at www.aim.com/suspended_accout. By providing accurate
information, you will enable us to properly investigate the situation.

Sincerely,
The AOL Instant Messenger Team*

* Please note this e-mail is not a commercial e-mail and is intended
only to provide official notice about an AOL Instant Messenger account
identified with this e-mail address.

I came across a nice way to use IM as a primary means of communication. I run naim with GNU Screen on a server on which I have a shell account. This way, naim functions as an "answering machine" when I'm not online, and a normal IM client when I am. Enjoy.

I came across a nice way to use IM as a primary means of communication. I run naim with GNU Screen on a server on which I have a shell account. This way, naim functions as an "answering machine" when I'm not online, and a normal IM client when I am. Enjoy.

Look at naim (Does AIM, IRC, ICQ & CMC). It supports client queueing... at least on AIM & IRC. (I haven't used the other two it supports). Pretty nice, but text based, so that will turn a lot of people off.

I ran Debian on an equivilent laptop to that. It didn't run amazingly fast, and I used Links (in the console) and Dillo (when in X) for web browsing. Mozilla was absolutely hopeless, although Opera was almost usable.

For IM, there was naim, and when I felt fancy, Gaim ran pretty well too, but then I couldn't really do too many things due to X taking up resources.

Let's start with the shell--zsh is by far the best one I've used. It has everything.

Moving on, Links (web browser) and Naim (AIM/ICQ/IRC client) rock. The only issue with the former is that Links doesn't support cookies, so I have Lynx in case I want to post on /. or something.

I don't have a console mail client on my machine--I have other methods of getting my email. For accessing my email account with my uni, I ssh into my uni's shell account and use pine from there or I use Links to access the Squirrelmail setup on my web server (over HTTPS, of course). To access my fastmail.fm account, I just use Links to access their web interface (they support both web and IMAP access for free).

For downloading stuff, I use giFTcurs, the btdownloadcurses.py BitTorrent client, and the venerable wget, depending on what I'm looking for and where I'm downloading from.

And, for the part that will generate the most flamage, my text editor of choice: Joe! Its interface is just as simple as nano, but with more features, such as find/replace and decent copy/paste, using text selection. On a related note, I use most as my pager--coloured man pages are good.

And, finally, who could forget NetHack?

Hmm...now I have an urge to find out how to make live CDs, so I can make a ``CLI survival kit'' live CD. Well, maybe not, as I'm too bloody lazy, but it's an idea...

Combined with screen, Naim is really nice for idling on AIM (to avoid missing IMs from people who are only awake when you sleep and such), and for switching computers without having to disconnect. The same convenience can be achieved using screen with an IRC client for IRC (I use Irssi).

Its much more convenient than GUI stuff when you switch computers a lot during the day. I can leave naim and irssi running in screen while I drive home from work and people can still IM me if they need to for those 30 minutes.

I use naim a lot for one reason: I can run it inside screen, detach from one computer and re-attach on another without ever going offline (or missing messages while I'm walking somewhere else). If I'm moving around a lot, screen also lets me have multiple connections to the same session, so I can read & reply from wherever I happen to be at the moment.

The other reason is that next to my main desktop at home, I have a nice little text-based LCD terminal (actually a partially disassembled 486 laptop) that I IM on -- saves screen real estate and I don't have to get offline when I'm doing stuff like kernel driver debugging that requires me to shut down X...

naim is a great, free, GPL'd instant messaging client. Very featureful, intuitive, and in my opinion one of the best examples of ncurses programming out there.

Don't forget naim. It's great for us console junkies!

I've asked for this last year and again this year. I want a color serial console to use upstairs running to my iMac downstairs so I can use screen'ed naim/centericq for IM, pine or elm for e-mail, and of course lynx for web browsing.

Quiet, low power, and can be turned on or off at a moment's notice. Leaves you logged into all the IM services simultaneously too!

Anyone know where to get these things on the cheap so I can forward the link to my SO? :) I think the higher level VT terminals support colors much like linux's terminal does...

There is an UnOfficial OSCAR implementation. It is used by a few clients. Again, naim is the most common OSCAR client that i know of, thought cLAIM (link not handy) has been around for awhile. naim uses libfaim, which is a tolerable implementation of the OSCAR protocol.

Also, gaim has been released as an oscar client. I need to read the freshmeat newsletters more often ;^)

oh the joys of being OT
/jbm

There is no buffer overflow in AIM that AOL exploits as client verification. If there was, the Free OSCAR clients would not work. This does not include gaim, which uses TOC; TOC is an "open" wrapper to OSCAR, not a native AIM client per se. The Free OSCAR clients include cLAIM, gtkFAIM, and naim. I know that at least naim works, and mfaim (in development) works. None of these have the buffer overflow, yet they continue to work. Therefore, it is very unlikely that AOL is screening people out through a buffer overflow.

For future reference, could we please make a distinction between OSCAR and TOC? They are two totally different protocols. TOC stores all your settings on an AOL server, and the client just interfaces with that "proxying server," for lack of a better term. OSCAR stores all your settings locally and interfaces with the Real AIM Servers. AOL loves it when we use TOC, because it keeps all the power in their hands. Which is why i spend my time working on an OSCAR client ;^)

For more info on naim, check out http://naim.n.ml.org, and http://www.auk.cx/faim/protocol/ has good (and very incomplete) info on the AIM protocol. And, as a side note, there are preliminary steps for gaim to use OSCAR as well, but that's still in progress.

This is the first time i've seen the Community listen to blatant M$ hype, and quite frankly, i'm disappointed.

/jbm

Get hold of the pgcc-1.1.3 source rpm ( pgcc-1.1 .3-3mdk.src.rpm) from Mandrake, if you plan to use pgcc anyway :-). It provides (apart from pgcc, of course) a libstdc++ package, which includes libstc++-2.8.0 and 2.9.0 (and various others) for compatibility to older packages. Built it yesterday, works fine. (Hey, now there's a use for rpm -bb --short-circuit... Building pgcc takes 3 hours on my machine, and I had the change to spec file tons'o'times. :-( )

If I'm not mistaken, there is little pgcc code that rolls back into egcs. pgcc offers better optimisation for Pentium-class chips, although I'm not entirely sure about the details. By passing it -march and -mcpu options, you can get it to use Pentium-only (or whatever-only; I think it even does MMX now?) instructions. But the biggest problem with gcc was that it assumed that you had a lot of general purpose registers. This is fine for a lot of architectures, but not so hot for i386 and friends. pgcc I think addresses some of the i386 deficiencies, but I couldn't tell you exactly how it optimises for Pentium-class processors.

As for benchmarking, you'd best go to the pgcc homepage, but I've heard improvements from 2% to 30%. The places where you'll find a lot of improvement (10%-30%) would be stuff like bzip, gzip, tar, stuff like that...stuff that doesn't really do much besides compute.

I'll have a mirror up at ftp://redwood.ml.org/pub/quake3/test/. I'll up the user limit from 100 to 200 for the hell of it too. At the moment I have it about 50% downloaded.

Since then, config.guess of each and every source package using configure bails out. "Can't determine host type" or the likes.

I've been adding "--host=`uname -m`-`uname -s | tr [A-Z] [a-Z]`" to my *.spec files, but that's stupid. Any suggestions?

As soon as I get the trailer it will be up at ftp://redwood.ml.org/pub/StarW ars/trailers/trailer2/. Also I'll try to dload it myself, but would welcome somebody anonymously uploading it to my incoming directory. I'll probably up my user limit to 250 or so later. Yes, my ml.org name still works.

That should be http://ubercool.ml.org/n64.htm not .html :-) I tend to flip-flop between the two.

I submitted this one, and I think it's VERY relevant.

According to the Arcana Mailing List the culture in the U.S. Patent office calls for approving more and more patents, and they're already approving 80% of patent submissions. Here's a cool excerpt from these E-mails from people who've worked for the patent office:

> This needs to be made clear to more people. The Patent Office is heavily
>biased towards rewarding examiners who allow a lot, and it is creating a
>culture to match. Unfortunately, our 'customers' are collectively paying
>the higher costs of confusion and legal help to sort the whole mess out.

I submitted this one, and I think it's VERY relevant.

According to the Arcana Mailing List the culture in the U.S. Patent office calls for approving more and more patents, and they're already approving 80% of patent submissions. Here's a cool excerpt from these E-mails from people who've worked for the patent office:

> This needs to be made clear to more people. The Patent Office is heavily
>biased towards rewarding examiners who allow a lot, and it is creating a
>culture to match. Unfortunately, our 'customers' are collectively paying
>the higher costs of confusion and legal help to sort the whole mess out.

If so, check out http://www.amush.ml.org/~rhw/kern el.versions.html
This lists kernel releases dating back to 0.01 on September 17th, 1991.

A rather good collection of links is at Wearables Central .

They should pass some resolutions (not that their resolutions do any good--back in the early eighties they passed a resolution that corporations should not market infant formula in Third World nations) that no member country shall tax, censor, or otherwise regulate Internet access.

Granted, it would be about as respect as their resolutions to (for example) Milosovec, the government of Rwanda, or the government of Libya. But, at least it would get a point across.

I always take comfort in the fact that in 150 years, none of the creeps around today will still be alive. That's not so long to wait. <grin>

The X TrueType fontserver for OS/2 is still there. Don't be discouraged by slow response times.

Good to see support for the Hurd is coming. For those of you aren't familiar with the Hurd (definition below), it's a kernel that is designed to handle really heavy loads. It is entirely GPL'd. Hopefully Hurd will eventually be of Solaris quality (I know, I know, it can't be done with opensource, but we've all heard that before). It will probably fill the same niche FreeBSD fills now. Hmm, Hurd servers and Linux clients... sounds nice to me.

One great thing about Hurd is that if it do get very good features, that can be folded into Linux without a hitch.

The Hurd was originally started by GNU as part of their project to create a completely free Unix system. Linux did not exist back then, and Linus was still learning his multiplication tables. (Or maybe they are smarter in Finland by that age? I don't know.)

Linux came out of nowhere and was able to make good use of the GNU toolset. It was exactly the vision that GNU had. However, they had already started work on the Hurd, and the Hurd had (or was going to have) some features Linux didn't (and doesn't) have. I know it may seem stupid to have people working on different Unix kernels-- but this is about diversity, not one-size-fits-all. If we just wanted one nice huge kernel we would use NT.

Be warned that the Hurd is not for newbies. I tried it a few weeks ago. It was neat, although it crashed a few times (probably due to my hardware configuration). Interesting things are going to happen to it.

``Hurd: `Hird of Unix-Replacing Daemons'. (definition of `Hird' below)''

``Hird: `Hurd of Interfaces Representing Depth' (definition of `Hurd' above)''

Yup, it's the X TrueType fontserver for OS/2 again. Just how many OS/2 users browse these forums, anyway? Happy xtruetyping!

...CICS for Linux. I use DB2/2 and CICS/2 (on OS/2). CICS is nice because it takes care of the transactions for you--it's quite trivial to design a nice site like Slashdot with DB2 and CICS (CICS is a transaction monitor).

For those of you who don't know what a transaction monitor is, it takes care of making stuff happen. So, you can define a procedure like "post comment". Then you can make your webscript for posting comments call that procedure via CICS, and CICS will make sure it gets executed. CICS takes care of errors occurring and it plugs into DB2.

Think of CICS as handling code the way DB2 handles data. Microsoft Transaction Server does the same thing but needs a lot more resources. CICS has been in use for 28 years. It's become quite popular as of late for building large websites that run on big iron.

There, I guess you know what company I like now. grin

CICS for Linux along with DB2 would almost be nirvana. (For nirvana, IBM would need to port their Java VM as well.)

My clumsy port of the X TrueType fontserver (xfstt) to OS/2

I have no personal experience with a touchscreen specifically supported by X. However, I have used (for that matter, I'm using it right now-- the IBM 8516 is staring at me and the IBM TouchSelect panel (one of this clipon jobbies) is on the shelf five feet from me)) touchscreens with X. OS/2 has excellent touchscreen drivers (for IBM's touchscreen of course). I use the XFree86 for OS/2 Xserver, which runs just fine and uses the OSMOUSE mouse subsystem--in other words, OS/2 managers the mouse.

Thus, I get a touchscreen and then run Unix programs with it. I especially like using touchscreens while using the Web.

This isn't really using a touchscreen with Linux--it's using an Xserver that has a touchscreen. But, it works (and besides, I'm an OS/2 junkie. It's a love-hate thing with me but I keep using OS/2 anyways.)

TrueType X fontserver ported to OS/2 (I told you I was an OS/2 junkie)

I put up a very ugly and extremely limited demo of some streaming audio done with non-proprietary software (I'm using OS/2 to be specific, but it would work fine with Linux. Linux would actually be a lot easier, faster, and better.) Take a peek at http://iiinet.ml.org:8/streaming-au dio-example/

I know it doesn't seem like the audio is streaming because it isn't from a live source, but it is. It's streaming from a program that simply plays those boring clips.

Yes, I realize the first one is one second long. I'm two weary to fix it right now.

Don't ask me for the software behind it; I'm using a non-opensource webserver (of alpha quality IMO) and it isn't doing any compression on the fly. You'll have to figure out how to do that for yourself. My own extremely experimental testing shows it can be done.

Here's what I've learned:

• 1. Streaming is possible over HTTP. Quality seems decent although the recovery isn't as good as RealAudio's (not that RealAudio's is good either). Quality is passable over 28800 baud modems, equivilent to mono AM radio. You can get stereo AM radio or mono FM radio over 56K modems (provided those 56K modems are getting connections of at least 40000 baud). It's easy to do things with named pipes. That's how I did because some of my tools were closed source (yuck). MP3 files are tileable; that at is, they can be split apart without getting troubles due to missing headers. I don't understand the MP3 format, but the blocks that my encoder put it out were seperate, independent entities.

I am no MP3 wizard nor audio streaming wizard. Do your own research or ask someone who knows. I just want to let you know it's possible!

HCP is another interesting project in the works. It lets you simply plug in existing protocols...