Subject: Linux-Misc Digest #399
From: Digestifier <Linux-Misc-Request@senator-bedfellow.MIT.EDU>
To: Linux-Misc@senator-bedfellow.MIT.EDU
Reply-To: Linux-Misc@senator-bedfellow.MIT.EDU
Date:     Thu, 9 Dec 93 14:13:20 EST

Linux-Misc Digest #399, Volume #1                 Thu, 9 Dec 93 14:13:20 EST

Contents:
  Who is the typical Linux user? (Phil Hughes LJ Editor)
  Re: Security (Winfried Truemper)
  Re: Linux Consortium (The Outlander)
  Re: Linux Consortium (The Outlander)
  Re: New Yggdrasil LGX boot floppy (Christian Seyb)
  Re: how fast is linux? (Curt L. Olson (Admin))
  Re: Yet another benchmark results.. (J. D. McDonald)
  Re: Boot Linux with OS/2 Boot ManagerHi, (Jens Wilken)
  Re: Yet another benchmark results.. (J. D. McDonald)
  Re: Linux-activists list berzerk? (Christopher Wingert)
  OGI Speech Tools for Linux (Just a fellow traveller...)
  Help fixing Lyre for OGI Speech Tools (Just a fellow traveller...)

----------------------------------------------------------------------------

From: phil@fylz.com (Phil Hughes LJ Editor)
Subject: Who is the typical Linux user?
Date: Wed, 8 Dec 1993 07:57:39 GMT

Our publisher is attempting to put together a profile of a typical
Linux user.  I keep attempting to tell him there is no such thing
as a typical one.  Below is my guess at what the Linux community
looks like today and what it will look like in a year or two.  I would
be real interested in either hearing a confirmation of my guess or
any ideas why I am on the wrong track.  In particular, if you are
way out of my profile guess I would like to hear about you.

Today I see the majority of Linux users as people who have at least
some Unix experience.  Most are fairly technical and ranging in
background from students to Unix systems programmers.  Many of
these people are work with a "real" Unix system at work and have
a Linux box at home.  And some of those people are attempting to
convince people at work that Linux would be a reasonable alternative
to a commercial system for some work-related tasks.  The remainder
of this population is the MS-DOS hacker that has seen Linux as
a chance to run a real operating system without even having to
change computers.

In one to two years I think the population will be much less technical
with tens of thousands of people moving from MS-DOS to Linux
because Linux offers more functionality and at a lower cost.
Some of these people won't really know much about Linux itself but
will pick it because it will do their job.  I also see a large
increase in the number of commercial uses of Linux systems.  This
will include imbedded applications as well as a platform for
scientific and commercial applications.

-- 
Phil Hughes, Editor, Linux Journal, P.O. Box 85867, Seattle, WA 98145-1867 USA
E-mail: phil@fylz.com   Phone: +1 206 524 8338 FAX: +1 206 526 0803

------------------------------

From: truemper@fileserv1.MI.Uni-Koeln.DE (Winfried Truemper)
Subject: Re: Security
Date: 9 Dec 1993 16:03:58 GMT

In article <2e53si$hbk@jacobs.jacobs.mn.org>, root@jacobs.mn.org (Mike Horwath) writes:
|> Mathias Koerber (mathias@solomon.technet.sg) wrote:
|> : Ok, that might work for some users. I am thinking of installations, where
|> :    a) many people should have access to that PC as normal users
|> :    b) they are encouraged to use floppies for backup etc.
|> 
|> : I want to protect the system against booting from floppy (not actually bad in
|> : itself) or moreover against other prople mounting the root filesystem
|> : from other OS'es they boot.
|> 
|> : Mathias Koerber
|> : email: mathias@solomon.technet.sg
|> : swispl@solomon.technet.sg
|> 
|> this can be accomplished with current AMI bios' by setting a password to
|> get into the bios and YOU set the bios to boot off of C: and ignore A:
|> for booting.  Should be in the Advanced CMOS settings.
|> 
|> --
|> Mike Horwath    IRC: Drechsau   BBS: Drechsau   LIFE: lover
|> root@jacobs.mn.org  drechsau@jacobs.mn.org
|> Jacob's Ladder  612-588-0201  UUCP, UseNet, Linux files, BBS


Hi folks,

its really useless to upgrade the bios in order to get the bios protected
since there is a DOS program called "amisetup". The program allows reading 
and changing every entry in the CMOS-RAM very comfortabel, even the 
password.

Its availiable via ftp, for example on "ftp.Uni-Koeln.DE" under 
"pc/msdos/dosutils/amise210.zip".

Winfried

------------------------------

From: yvain@microsup14 (The Outlander)
Subject: Re: Linux Consortium
Date: Wed, 8 Dec 1993 12:45:00 GMT

Andreas Klemm (andreas@knobel.knirsch.de) wrote:
: markline@henson.cc.wwu.edu (Mark Line) writes:

: >andreas@knobel.knirsch.de (Andreas Klemm) writes:

: >>Magnus Y Alvestad <magnus@ii.uib.no> writes:

: >>>Linux Consortium Principles v0.2

: >>>o Our 'public' are non-experienced users that want to buy Linux. We
: >>>  will judge distributions and whether they are suited for novices. We

: >>Better they run good at all. Don't know what suited for novices
: >>should mean ... There is good info material out there, public domain, too.

: >Anybody who reads your comments below will realize that you haven't a
: >clue what 'suited for novices' means -- you didn't have to spell it
: >out. If you're not interested in novices, say so and leave it at that.
: >Those of us whose livelihood involves catering to novices (yes, they
: >even qualify as human beings and *everything*) *are* interested, and will
: >do whatever it takes to keep them out of hot water. For most computer
: >users, solving a glitch by recompiling the kernel is not an option.

: No I have nothing against guidance of end users if they really
: have the occasion to come to an own conclusion what to buy and 
: what not !

Don't they anyway? They can choose to buy bsd386, NetBSD, Coherent, SCO,
UnixWare, etc etc etc. Buying Linux in any case is an act of free choise
in most situations.

: I think a consortium makes this process too easy ... You
: give a product the label "Consortium proofed" or whatever
: and everybody buy's _blindly_ "Consortium proofed" packages.
: Although there might be better products in means of 
: functionality.

If there are a number of distributions with a mark of approval frm the
Consortium, then the end user will still have freedom of choise, and will
still have to think about what they want. But "making it too easy"? Maybe
your thought didnt translate that well into english, but i think making
a decision easier for an end user is a good thing! Aside, the LC would be
allowing a consumer to have some idea as to the chances of the prospective
system working, and working relatively easily. Look at it this way, there
will always be those people that do blindly leap into things, be it Linux
distributions, religion, politics, choise of music. If we were to apply your
argument generally, we would have to do something about removing religion,
politics, music and all else so that people didnt jump into it blindly. I am
not convinced that this would be such a good idea

: I fear, that a consortium could possibly be not neutral enough.
: Where are the criterions that everbody can accept for being
: recommended by a consortium ? 
: I know either this nor the persons, who will judge over
: user friebdlyness.

True, while a goodly proportion of any decision that the consortium would
make would be subjective, i feel that there are ways of judging user 
frendliness in a qualitative sort of way.

Is the documentation written in relatively clear English (or whatever the
distribution's language is)?

Is the documentation correct?

Does the documentation require an in-depth knowledge of Unix, C, mail or
the like? Does it have a lot of jargon that is not otherwise explained?

Similar sorts of heuristics can be used for software...

Are the options relatively clear to understand?

Is the user interface relatively easy to grasp?

How steep is the learning curve?

And for distributions...

Is there sufficient documentation describing the installation options,
installation process, as well as describing common errors and fixes to them ?

Is the distribution organised in such a way that the various packages do not
conflict with each other when loaded?

Is the distribution organised in such a way that the various depenancies 
between packages in the distribution are respected?

Does the distribution contain utilities to make as much of the initial
configuration as possible easy to do?

Is the distribution structured such that standard installs have sufficient
utilities to make a fully workable system, with the option of adding more
interesting or esoteric packages on top? Does it provide an easy way of
adding/removing packages after the initial setup?

Does the damn thing work?

: Financial interests could perhaps manipulate decisions.
: Why I say this ? Some things said by Magnus sound so strange,

Perhaps they can... Heh, but it could hardly be any worse than the real world
could it?

: Another thing is, that the need for a consortium like that
: with certain rules weren't discussed in the public.

What's happening now?

: Instead of this you present finished and in my eyes questionable
: concepts, that everybody should accept.

You arent accepting it... I am not sure that they have sent hitmen out
after you..

: Where was the public discussion about that - similar to the
: democratic process of creating a new newsgroup ?!!

What's happening now?

: Another example that makes me thoughtful is the reply from
: adam@yggdrasil.com (Adam J. Richter) writes:
: >       As I understand it, the Linux Consortium is not attempting
: >to restrict distribution of Linux.  The Linux Consortium is an
: >attempt to provide some reasonably unbiased information about
: >installing Linux.  A lot of people who are interested in Linux could
: >use this information.

: He is someone from a Linux on CD producer who is for such
: a consortium ... What I don't understand is, why someone
: could be interested in such a consortium, when the own
: product has so many flaws. Read the thread about

One of the problems with a person producing a distribution is that they
have to coordinate a heck of a lot of packages etc. A consortium could
be useful in providing information to the producer before the product is
released as to where the major nasty bugs with the distribution is... This
could save the producer substantial time, and provide one more level of
quality control.


: Best wishes for freedom of distribution 

People will allways be free to distribute.


chris.

------------------------------

From: yvain@microsup14 (The Outlander)
Subject: Re: Linux Consortium
Date: Wed, 8 Dec 1993 12:53:26 GMT

Andreas Klemm (andreas@knobel.knirsch.de) wrote:
: markline@henson.cc.wwu.edu (Mark Line) writes:
: >andreas@knobel.knirsch.de (Andreas Klemm) writes:

: [...]


: >documentation, because nobody writes documentation for free; and you
: >have implied (c) Linux is for Hackers. In addition, we have (d) the
:                         ^^^^^^^^^^^^^^
: Is mainly for people that can live with the fact, that it's alpha stuff.

Actually, I think we must make a distinction between the status of the
kernel, and the status of the individual components. Is the kernel still
considered to be alpha? It is true that dosemu is alpha, that some of the
other packages are beta. But what about GNU shellutils? fileutils etc?
What about Xfree? What about ghostscript etc? Are all these things alpha 
as well? No.

I am not overly sure who linux is "for" as such. Perhaps it is "for" anyone
that wishes to use it.


chris

------------------------------

From: cs@gold.muc.de (Christian Seyb)
Subject: Re: New Yggdrasil LGX boot floppy
Date: Thu, 9 Dec 1993 15:48:24 GMT
Reply-To: cs@gold.muc.de

In <CHqM0s.Lqr@khijol.yggdrasil.com> adam@adam.yggdrasil.com (Adam J. Richter) writes:

>       You probably used a high density floppy instead of a low
>density floppy.

Hi,

no, the disk is damaged. There is no /dev directory and there are
perhaps more things missing. You see it if you mount the floppy.
But you can still copy the new kernel to the old boot disk and run
lilo on it.

regards Christian
-- 
Christian Seyb  |  cs@gold.muc.de  |  Mailbox/uucp/Fax: 08106-34593

------------------------------

From: clolson@me.umn.edu (Curt L. Olson (Admin))
Subject: Re: how fast is linux?
Date: Thu, 9 Dec 1993 16:46:12 GMT

trucken@hecto.cs.umn.edu (Dave Truckenmiller) writes:

>Yes, it is much slower than my Sun IPX here at school.  I measured this
>morning, and the IPX is two inches to the left, while my PC at home
>running Linux hasn't moved a bit.

Yes, but everyone knows your Sun IPX is on steroids since it was disqualified
from the world championships last year and had all its medals revoked.

Curt.

------------------------------

From: mcdonald@aries.scs.uiuc.edu (J. D. McDonald)
Crossposted-To: comp.sys.ibm.pc.hardware
Subject: Re: Yet another benchmark results..
Date: Thu, 9 Dec 1993 16:54:06 GMT

In article <1993Dec09.025654.12591@ksmith.com> keith@ksmith.com (Keith Smith) writes:

>     Computer                             Time spent

>486DX2-66 EISA/VL 16Mb RAM
>running Linux (Slackware 1.1.0).
>gcc compiler.
>Single user                               27 sec.


486DX2-66 ISA/VL 16Mb RAM 256K Cache       
MS-DOS
MicroWay NDPC 4.30 -n2 -n3 -OLM -exp       25 sec


same machine, compiler
Windows, this program only                 27 sec
Windows, plus Trumpet and Telnet           44 sec

SGI Crimson, Unix, totally unknown load    14 sec




Doug McDonald


------------------------------

From: Jens.Wilken@arbi.informatik.uni-oldenburg.de (Jens Wilken)
Subject: Re: Boot Linux with OS/2 Boot ManagerHi,
Date: Thu, 9 Dec 1993 16:26:38 GMT

stalher@uni-muenster.de (Frank Stalherm) writes:

>I seem to have a problem with booting linux from disk.

>On my site there are 2 harddisks. On the first disk there is one
>primary partition with OS/2 and Boot Manager on it and a extended
>partition with a logical drive, HPFS formated.
>On the second drive there are 2 more logical drives, the first
>HPFS formated and the second is the Linux partition.

>Now is there a chance I can get the OS/2 Boot Manager to boot Linux ?

>thanks, Frank

>Keywords: 

I have my system running with 2 HDs:
hda:
pri          01 MB BootMgr
pri bootable 88 MB OS/2
pri bootable 08 MB DOS5
pri          08 MB Linux Swap

hdb:
pri bootable 130 MB Linux
pri          100 MB DOS data

So I can boot OS/2, DOS or Linux via Bootmgr.
For Linux I installed LILO, with a LILO.config that saied:
boot partition /dev/hdbx
root partition /dev/hdbx
with dev/hdbx as the partition where Linux is on.
So LILO installs itself on /dev/hdbx and when booting, first the Bootmanager
is invoked and after choosing Linux, LILO is started and boots up Linux.

I hope this helped a bit.

        Jens

==============================================================================
E-Mail  : Jens.Wilken@arbi.Informatik.Uni-Oldenburg.DE
Inhouse : snoppy

------------------------------

From: mcdonald@aries.scs.uiuc.edu (J. D. McDonald)
Crossposted-To: comp.sys.ibm.pc.hardware,comp.os.vms
Subject: Re: Yet another benchmark results..
Date: Thu, 9 Dec 1993 17:03:27 GMT

In article <CHqIHq.9pw@dscomsa.desy.de> kruse@zow.desy.de (Andres Kruse (NIKHEF)) writes:


>Please, please!! Don't continue running this program on your machines
>and post the results.

Please DO do so. It's interesting.


>   Look at the code first and decide if it makes any sense to do it. 
>There are several oddities:

>- It is using the time(2) function... check your man pages to 
>  see what that means...


Yes, indeed . This means that it measures the ACTUAL time the program takes.
This is what ACTUALLY MATTERS to the user.


>- It is putting a lot of emphasis on trigonometric functions.

Sort of .. sqrt and log are not trig functions.. So what it tells you is, 
how fast programs that are dominated with transcendtal functions
will ACTUALLY run on your system under real conditions.


>- You have to quote the CPU type, the cache size, the memory
>  size, the compiler options etc. All this has a big influence!
Yes, of course.


>Seeing so many people blindly taking this source and wasting their
>CPU cycles and bandwidth in the NET I think that it's good to
>have SPEC around. SPEC *does* give a quite good estimate on how
>the performance compares. 


Well, yes and no. SPEC is a good set of representative programs,
but it does not normally measure the ACTUAL time to run a program.


Both types of benchmarks are useful. But the ACTUAL clock times
are, in truth, more important. If a computer does well
on SPEC and poorly on this, it is telling you something ... mainly
that SPEC is not really truly going to tell you how long your
program will take to run, because it does not normally measure
actual clock time. Also, this benchmark will generate a BIG difference
in numbers on similar hardware depending on how that hardware is used.
Knowing the RANGE is a VERY important fact. OF course, the actual
code in SPEC could be changed to give the same information .. and THAT
would, I argue, be an excellent, if for each machine a histogram
of such times was published. 

Doug McDonald

------------------------------

From: crw@TorreyPinesCA.ncr.com (Christopher Wingert)
Subject: Re: Linux-activists list berzerk?
Date: Thu, 9 Dec 93 16:16:56 GMT

In article <1993Dec9.062033.19447@unlv.edu> maniac@unlv.edu (Eric J. Schwertfeger) writes:
>Am I the only one that has received over 500 (mostly corrupt) messages
>from the linux activists  mailing list?  I'm trying to figure out
>which end has the problem, so I can uncork my mail.
>
>-- 
>Eric J. Schwertfeger, maniac@cs.unlv.edu
Actually No.  I have been receiving mailings....not on that scale though...
I usually receive about two to three per day.  The most recent one being

MAIL>Return-Path: <owner-linux-activists@joker.cs.hut.fi>
MAIL>Sender: owner-linux-activists@niksula.hut.fi
MAIL>To: linux-activists@niksula.hut.fi
MAIL>Subject: HELP
MAIL>Date: Wed, 8 Dec 1993 21:04:28 +0200
MAIL>X-Mn-Key: ADMIN
MAIL>X-Note1: Remember to put 'X-Mn-Key: ADMIN' to your mail body or header
MAIL>
MAIL>
MAIL>-----------
MAIL>
MAIL>list channels
MAIL>show channels
MAIL>help
MAIL>send help


There is the possiblity that I could be receiving that many a day.  I am
on the other side of a major bottleneck, more commonly named firewall.  I
have received mail delayed for 45 days at one point (almost 5x slower than
USPS)..  Sigh.....All in the name of security.  Anyway if whoever can't fix
it...Please remove me from the list.  I have the feeling that I would just
get the message back, if I sent a leave request.  Thanks......


-- 

                                    Regards,
                                        Christopher

 ------------------------------------------------------------------------------
| That secret you've been guarding, isn't.                                     |
|                 --fortune .signature Shell Version 1.0                       |
 ------------------------------------------------------------------------------
|         Christopher R. Wingert                   Programmer / Analyst        |
| Christopher.Wingert@TorreyPinesCA.ncr.com    Direct Phone (619) 597-3533     |
 ------------------------------------------------------------------------------

------------------------------

From: jedubins@unix.amherst.edu (Just a fellow traveller...)
Subject: OGI Speech Tools for Linux
Date: 9 Dec 1993 12:27:39 -0500

[Crossposted manually from comp.os.linux.announce as an afterthought.
Apologies given to those with threaded news readers.]

Here is the LSM for OGI Speech Tools, followed by the original post on
the port to Linux by Tilo Schuerer in Germany.  The original post
lists his site in Germany with files, but I strongly urge you to get
the files from sunsite.unc.edu(or tsx-11.mit.edu as soon as they become
available there).  There are two good reasons for this.  First the
link to Tilo's machine is extremely slow, at least from the US.  I am
not exagerating when I tell you that you could be waiting several minutes
before you even get to log in via ftp.  The second reason is that his
original files didn't have much of a directory structure, so I unpacked
all of them in the right places(except for the man pages which I left alone
as there really was no need to do anything with them).  So now you
can get the file from sunsite.unc.edu, stick it where you like, and it
will unpack into repespective directories under the top directory 'ogi-speech'.

Install the man pages seperately.  This is really a neat package.  I've
been looking forward to a port of it since the source for the package
was made public in July.  After talking with people at the Center for
Spoken Language Understanding at the Oregon Graduate Institute(OGI),
the head of the project said they had plans to port it to a free *nix
eventually, but they wanted to use 386bsd instead of Linux.

A note of warning: there still are a few bugs in the package, but they
are mostly in the X-graphical tool Lyre(and auto_lyre which is a scripted
version of Lyre).  You'll often get a floating point exception error
when trying to use Lyre.  I'm not sure what's going on here either.  If
you would like to help email Tilo at tilo@cs.TU-Berlin.DE.  I'm sure he'll be
able to send you his diffs or a list of his patches.  The general *nix source
release is on speech.cse.ogi.edu in /pub/tools/ogitools.v1.0.tar.Z

This package is potentially useful in many areas, especially for the
computational linguistics types in the Linux community.  Among the uses are:
phonetic, phonemic, and adaptive word recognition.  This can be accomplished
through the building of database files to run witha neural net which is
trained by the neural net trainer which comes in the package.  I've been
made to understand that OGI has this running with a multi-lingual
database.  Also this package would be useful in voice signal analysis
which could then be used to write voice compression/decompression
algorithms.  This package comes with Linear Predictive Coding analysis
tools.  It can disaply the sound file in wave form and spectrogram form
among others.  Have fun.  If you have any other questions I can be reached
at this address (jedubins@unix.amherst.edu) as wells as my address as
listed in the LSM entry.  Here's the LSM entry:

Begin2
Title        =     OGI Speech Tools
Version      =     1.0
Desc1        =     Speech data manipulation research tools:
Desc2        =     signal manipulation, phonetic, phonemic and word analysis 
Desc3        =     has the capability to build audio databases which can
Desc4        =     be used to train neural networks with a NN trainer
Desc5        =     in the package.      
Author       =     Center for Spoken Language Understanding (CSLU-OGI)  
AuthorEmail  =
Maintainer   =     Tilo Schuerer        
MaintEmail   =     tilo@cs.TU-Berlin.DE
Site1        =     sunsite.unc.edu      
Path1        =
File1        =     ogi-speech.tar.gz    
FileSize1    =     1616K
Site2        =     tsx-11.mit.edu       
Path2        =
File2        =     ogi-speech.tar.gz
FileSize2    =     1616K        
Site3        =     ftp.cs.TU-Berlin.DE  
Path3        =     /pub/sci/speech/ogi
File3        =     linux-bin.tar.gz  
FileSize3    =     431K (binaries only, other files available separately)
Site4        =
Path4        =
File4        =
FileSize4    =
Required1    =     Needed some sort of sound card that can digitize
Required2    =     voice input under Linux.  A sound format conversion
Required3    =     utility like Sox would be extremely useful as well.
Required4    =     X-window environment
CopyPolicy1  =     Freely distributable.
CopyPolicy2  =
Keywords     =     If you are into computational linguistics, or are
Comment1     =     interested in speech compression, recognition, or
Comment2     =     applications of neural networks this is one great
Comment3     =     research package to check out.  Still a few bugs     
Comment4     =     in the X-window tool Lyre.  Tilo needs help with this.
Comment5     =     Non-Linux ported source at speech.cse.ogi.edu in
Comment6     =     "/pub/tools"
RelFiles1    =     ogi-man.tar.gz       
RelFiles2    =     
RelFiles3    =     
Entered      =     09DEC93
EnteredBy    =     Jim Dubinsky
CheckedEmail =     jed1@cec.wustl.edu 
End

As appeared in Comp.speech:

/*****************************************************************************/

     ANNOUNCING THE BINARY LINUX VERSION OF THE OGI SPEECH TOOLS (Version 1.0)

/*****************************************************************************/

The binary version of the OGI Speech Tools (the best speech related tools I ever
saw - congratulations to all the people from OGI!!!) together with the changes 
I had to make to get the stuff compiled under LINUX are now on the following 
ftp-server:

        ftp.cs.TU-Berlin.DE under /pub/sci/speech/ogi

The code is compiled with the gcc version 2.4.5 and the libc version 4.4.4
(Slackware 1.1.0).

There are the following files:
        
        437689 Dec  4 00:18 linux-bin.tar.gz     ---->  including the binaries
         54351 Dec  4 00:03 linux-lib.tar.gz     ---->  including the libraries 
                                                    and include files
          4287 Dec  4 00:44 linux-compile-diff   ---->  including the changes to 
                                                    the soure code I I had to 
                                                    make to compile the code
    352937 Dec  4 09:31 ogi-doc.tar.gz       ---->  the OGI docu
    808882 Dec  4 09:32 ogi-examples.tar.gz  ---->  some example data
     36352 Dec  4 09:33 ogi-man.tar.gz       ---->  the OGI manuals
   1546417 Oct 25 13:19 ogitools.v1.0.tar.gz ---->  the original OGI-code 
                                                    including docu, examples
                                                    and manuals
                                                    

The soure code itself can be found:

        US:        speech.ogi.cse.edu under /pub/tools
        Europe:    ftp.cs.TU-Berlin.DE under /pub/sci/speech/ogi 
               (mirrored from OGI)

Tilo Schuerer
tilo@cs.TU-Berlin.DE


Now a short part of the original ANNOUNCE from OGI about the Speech Tools:

A. What are the OGI Speech Tools
================================
The OGI Speech Tools are set of speech data manipulation tools
developed at the Center for Spoken Language Understanding (CSLU) at
the Oregon Graduate Institute of Science and Technology (Portland
Oregon). The tools can be used to compute and display signal
representations, label speech at different levels (e.g., phonetic,
phonemic and word), train neural network classifiers, and display the
output of classification or recognition algorithms time-aligned with
the speech.

A Software Capitalization grant from NSF has allowed to enhance,
document and distribute the tools. Continued maintenance and
improvement of the tools is supported by CSLU.

These include

1. An X windows display tool (LYRE) for displaying data in a time
   synchronous fashion. These mainly constitute the following data
   representations:
   
     a. the speech signal
     b. the spectrogram
     c. the phoneme labels
 
   The data displayed may, however, also include any other type of
   time synchronous information, provided the data is stored in the
   required format.

2. A Neural Network (NOPT) training package, which uses Conjugate
   Gradient descent optimization.

3. An extensive set of C library routines (LIBNSPEECH) for the
   manipulation of speech data. These also include the following signal
   processing routines.

    a. PLP Analysis
    b. Rasta PLP Analysis
    c. Linear Predictive Coding
    d. Mel Cepstrum Coding
    e. Fast Fourier Transform

4. An extensive set of utilities for converting from files
   of different format. These include routines such as:

   a. Converting from ADC to NIST type headers
   b. Converting to mu-law encoded files
   c. Converting from mu-law encoded files back to NIST headered files
   d. Filtering the speech waveform.
   e. Converting to from binary to an ascii data stream.

5. A database utility (find_phone) to efficiently automate speech data
   related queries.  This utility creates a database based on a listing
   of speech files and there corresponding location and label files
   (lola).  It allows the user to specify a particular label or set of
   labels in a given context, display all occurrences of the label, and
   relabel the occurrences if desired.

6. A Vector-Quantizer based on the Linde Buzo and Gray (LBG) algorithm.

7. An set of PEARL Scripts which have been used mainly to automate
   the use of the OGI Speech Tools. 

8. MAN Pages for all routines and programs developed, as well as
   a user manual in both in postscript and {\bf tex} format.




------------------------------

From: jedubins@unix.amherst.edu (Just a fellow traveller...)
Subject: Help fixing Lyre for OGI Speech Tools
Date: 9 Dec 1993 12:34:27 -0500

Read the other OGI announcement that was posted in c.o.l.m and c.o.l.a if
you already haven't.  I talked about this a little in the announcement,
there is a runtime bug in, Lyre, the x-graphical display and manipulation tool
which comes with the package.  It often comes up with a 
floating-point-excetion error.  Please contact tilo@cs.TU-Berlin.DE, or
my self, and I'll relay the information to tilo or try it myself.
Run the programs yourself and take a look.  They are extremely powerful
tools for computational linguistics analysis and research.

                                Jim



------------------------------


** FOR YOUR REFERENCE **

The service address, to which questions about the list itself and requests
to be added to or deleted from it should be directed, is:

    Internet: Linux-Misc-Request@NEWS-DIGESTS.MIT.EDU

You can send mail to the entire list (and comp.os.linux.misc) via:

    Internet: Linux-Misc@NEWS-DIGESTS.MIT.EDU

Linux may be obtained via one of these FTP sites:
    nic.funet.fi				pub/OS/Linux
    tsx-11.mit.edu				pub/linux
    sunsite.unc.edu				pub/Linux

End of Linux-Misc Digest
******************************
