Linux-Development-Sys Digest #932, Volume #6 Sun, 4 Jul 99 09:14:30 EDT
Contents:
Re: libc and a released product. (david parsons)
Re: Why not C++ (Timo Tossavainen)
test (Jordan Aslett)
Re: Domain Names Limited to two or three Characters? (CodeWright)
Device Drivers (Matt Jervis)
Re: C++ templates: More than Turing Complete? (Nathan Myers)
Re: Why not C++ (Nathan Myers)
HELP me Please (Shawn Lang)
Help me PLEASE (Shawn Lang)
Re: Why we are still holding on to X Windows ("Adam JBC")
Re: Help me PLEASE (Thorsten Heit)
Re: Download freeBSD and Slackware 4 (Nir)
Re: HELP me Please (Scott Lanning)
----------------------------------------------------------------------------
From: o r c @ p e l l . p o r t l a n d . o r . u s (david parsons)
Crossposted-To: comp.os.linux.development.apps
Subject: Re: libc and a released product.
Date: 4 Jul 1999 00:40:29 -0700
In article <[EMAIL PROTECTED]>,
Andreas Jaeger <[EMAIL PROTECTED]> wrote:
>>>>>> Christopher Browne writes:
>
>Christopher> On 23 Jun 1999 13:23:27 -0700, david parsons <[EMAIL PROTECTED]>
>Christopher> wrote:
>>> What Linux frantically needs is (a) a regression test suite that
>>> can be run on new libraries to validate them and (b) to have the
>>> major distributions refuse to ship new libraries that don't
>>> pass the test suite.
>
>Christopher> Agreed.
>
>Christopher> There may already be something "internal" at Cygnus, whether as part
>Christopher> of EGCS, GLIBC, or otherwise; if there were a public test suite, that
>Christopher> would be, while not the most sexy thing in the world, a terribly
>Christopher> valuable thing to have.
>
>Christopher> The person *vastly* more likely than anyone else to know about this
>Christopher> would be Ulrich Drepper.
>
>glibc has a test suite. Just run `make check'. When errors get
>reported I try to make a test program out of them and add this test
>program to glibc so that we don't make this error again:-).
Test suites put out by the software vendor are useful, but, alas,
aren't very good for regression testing (and on a personal level, I
am all to familiar with the tendency to update the libraries and
the test suites at the same time, so a test suite that ships with
glibc is useless for regression testing.)
One simple regression test that can be done, which unfortunately
requires manpower and hardware, is to statically link the testsuite
and see if it works on machines that don't have the libc de jour.
____
david parsons \bi/ getpwuid() considered broken.
\/
------------------------------
From: Timo Tossavainen <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.development.apps
Subject: Re: Why not C++
Date: Sun, 04 Jul 1999 14:21:58 +0300
Nathan Myers wrote:
>
> C++ is _not_ an Object-Oriented Language. It supports procedural
> (top-down), object-oriented, generic, and functional styles, plus
> others which don't have names but which we all use at times. It is
> precisely this catholicity that makes it a good at decomposing
> problems -- that is, writing libraries -- because big problems call
> for different styles at different scales.
>
> I don't know of any other language designed with such close attention
> to the effect of design choices on library designers and users.
Common Lisp is also quite good at this, for instance CLOS is totally
built in Lisp on top of Lisp (some implementations may have a low
level implementation for increased efficiency) bringing a whole new
paradigm of programming to the language, the same has been done for
logic programming as well.
> >Also dynamic extendibility is
> >something not easy to achieve in C++ programs, many programs
> >don't need it but some benefit greatly (for instance, Emacs).
>
> One often achieves this in a C++ program by linking an interpreter.
> Python is specifically designed to be used that way; some Scheme
> interpreters are, also. Furthermore, the virtual function mechanism
> lends itself to dynamically linked libraries, so much so that
> Microsoft (probably unwisely) built their whole DLL architecture on
> top of it.
There are programs (not commercial, I think) that have plugins
unwisely locked to a certain compiler by requiring plugin dll's to
be compiled classes with certain virtual functions. This not the
fault of the language, but the developer, who has chosen the easy
way. It's still tempting to think that everyone is using the same
compiler... on linux this doesn't matter because egcs is free and
you can build the interface yourself, but where compiler is
proprietary it matters.
I've heard that Microsoft is planning another OO language called
COOL, which integrates OLE/COM with C++. I only hope that if
this is the case then the developers won't use it too much,
because it will destroy the portability of their applications.
> It is no Bad Thing to use two different languages in the same
> program to address radically different goals.
No it's not, it's actually a good solution.
This is an isolated incident, most will not run into it: Sometimes
the extensibility interface demands great speed and is difficult
to achive. Extensible software synthesizers are one example. You
may process vectors of data, which is fast enough but destroys the
possibility of using feedback in the signal processing networks, or
you may process one sample at a time which leads to the procedure
calls eating up too much processor power so you lose realtime
capability. As a hobby project I'm trying to see if it is possible
to create a network compiler in Lisp (the compiler that compiles
lisp is available at run-time), that produces fast enough code from
plugins (which are actually more like code templates in this case)
> Already there: C++ defines a subset of its types to be "POD types"
> which are compatible with C and with the standard C ABI for the
> target. Approximately, a POD type is anything that can appear in
> a C header. ("POD" is named by analogy with the telecomm jargon,
> "POTS", which is usually elaborated as "Plain Old Telephone System".
> I swear I am not making this up.)
Plain Old Data. Yes I know. Extensibility only requires people to
do a bit of extra work and they have to be quite skilled in C to
do it. Still doing it in a cross-platform way (dll, so's, etc)
requires some more work, which amounts to a minor extra nuisance.
A standard for crossplatform C/C++ dynamic libraries would be useful,
because a simple recompile would suffice.
> >I still run into the numeric conversions part in C++, the
> >compiler should be a bit more helpful in informing about them
> >especially signed/unsigned and float/int. I mostly program numerical
> >algorithms for work.
>
> Do you use '-Wall'? If you have suggestions for good warnings,
> by all means send them to the Egcs maintainer. The quality of
> a compiler's warnings makes a huge difference in its usefulness.
Yes I use -Wall, how about :
signed converted to unsigned
float converted to int
in both cases as a part of a computation, not an assignment.
> If you're doing numerics, do you know about the Blitz++ library
> and its kin? See http://www.oonumerics.org/ .
Yes I know blitz++, it seems to be quite usable. Actually I do
signal compression research and many of the algorithms don't
use arrays or matrices explicitly (like the FFT) although
matrix inversions and eigenvectors/eigenvalues are needed.
> The C ABI _is_ the lingua franca, and will be for a long time.
> Still, it's worth noting that some languages are taking on a
> degree of direct C++ compatibility, including Eiffel, Java,
> and Python.
Yes and these are object oriented languages. It's a lot easier
for them to usefully map the C++ classes to language primitives
than for other languages, but it's not impossible for other
languages. I'm curious though, that how is it possible to do
direct C++ compatibility with these languages if there
is no binary standard for C++ objects or do they use the C++
compiler as a final step ?
Timo
------------------------------
From: Jordan Aslett <[EMAIL PROTECTED]>
Subject: test
Date: Sat, 03 Jul 1999 22:49:49 -0400
Reply-To: [EMAIL PROTECTED]
Sorry about this guys but I have to test for a customer.
Thanks,
Jordan Aslett
======================================
ARACNET INTERNET SYSTEMS
------------------------------
From: CodeWright <[EMAIL PROTECTED]>
Subject: Re: Domain Names Limited to two or three Characters?
Date: Sat, 03 Jul 1999 22:02:37 -0400
Bj=F8rn Reese wrote:
> =
> JDJohnson wrote:
> =
> > Are domain names limited to two (.uk, .de) or three (.com, .org, .edu=
, .gov)
> > characters in length? If not is there a max number of characters?
> =
> Currently there longest domain name is .arpa, and others (.firm, .shop,=
> .info) may be comming soon.
> =
> I am not sure if there is a well-defined upper limit. There seem to
> be some contradition between RFC 952 and RFC 1123.
> =
> RFC 952 says
> =
> "A "name" (Net, Host, Gateway, or Domain name) is a text string up
> to 24 characters drawn from the alphabet (A-Z), digits (0-9), minus
> sign (-), and period (.)"
> =
> whereas RFC 1123 says (in section 6.1.3.5)
> =
> "The DNS defines domain name syntax very generally -- a
> string of labels each containing up to 63 8-bit octets,
> separated by dots, and with a maximum total of 255
> octets."
> =
> I do not know if any of these RFCs have been obsoleted by more
> recent RFCs.
> =
> If you really need an upper limit I guess 256 would be the most
> sane choice.
There's a difference between theory (as defined in the above quoted RFCs)=
and
practice (as defined by the IANA and/or ICANN, and in RFC 1591).
In theory, a domain name consists of parts that (according to RFC 952)
are "a text string up to 24 characters drawn from the alphabet (A-Z),
digits (0-9), minus sign (-), and period (.)". =
In practice, the Internet only permits Top Level Domains of two catagorie=
s
1) General TLDs (COM, NET, ORG, EDU, MIL, INT, GOV), and
2) Country TLDs (AU, CA, US, UK, etc.)
There has been discussion about expanding the General TLDs, but AFAIK, th=
is
has not yet happened. In any case, the TLD is not arbitrary; it must be o=
ne
approved by the IANA/ICANN
-- =
Lew Pitcher
Master Codewright and JOAT-in-training
------------------------------
From: Matt Jervis <[EMAIL PROTECTED]>
Subject: Device Drivers
Date: Sun, 04 Jul 1999 19:41:16 +1200
Reply-To: [EMAIL PROTECTED]
Hi,
I'm real new with Linux and need some help.
Can anyone tell me how to get my: Vibra 16 Sound Card
to work in Linux
Thanks if you can help !!
------------------------------
From: [EMAIL PROTECTED] (Nathan Myers)
Crossposted-To: comp.os.linux.development.apps,comp.os.linux.networking
Subject: Re: C++ templates: More than Turing Complete?
Date: 4 Jul 1999 02:56:21 -0700
Davin McCall <[EMAIL PROTECTED]> wrote:
> [EMAIL PROTECTED] (Nathan Myers) wrote:
>>Split the program into pieces, and then you have interfaces between
>>the pieces. Turing completeness says nothing about those interfaces.
>>
>>Getting those interfaces right, and getting the cut lines in the
>>right places to allow them to be right, is the domain of engineering.
>>C++ has a large variety of very strong tools to describe library
>>interfaces because it is an engineering language. C has many fewer
>>such tools.
>
>So you are saying that the choice of programming language is an
>engineering issue, not a computer science issue?
Of course. Science has nothing to say about construction methods.
>>Principles, divorced from experience, rot. If you want to understand
>>good principles, the only trustworthy source is good code.
>
>But why C++ code? The same principles found in the C++ language are
>applied in other languages also.
Some are. Many aren't. You don't know which ones before you even start.
You are starting from the assumption that you already know all
the important principles. If you actually want to learn anything,
you must begin by assuming you don't know everything yet.
>> If you
>>want to understand the principles behind the success of C++, you must
>>study good, real C++ programs and libraries. There are no shortcuts.
>
>I disagree. My reasons have been stated. You are saying something
>analogous to this:
>
> "if you want to know how a fuel injected engine works [-> the
>principles behind it], you must study the engine in car {X}. Although
>cars {Y} and {Z} have similar engines, studying them will not reveal
>*any* of the principles used to design the engine of car {X}."
If you need an analogy, how about:
"What are you looking for?"
"My keys. I dropped them over there".
"Why are you looking here?"
"The light's better here."
If you want to know why car {X} wins races, studying {Y} and {Z}
(which don't) is no substitute for a very careful study of {X}.
To assume fuel injection must be the key factor handicaps you before
you even start studying.
--
Nathan Myers
[EMAIL PROTECTED] http://www.cantrip.org/
------------------------------
From: [EMAIL PROTECTED] (Nathan Myers)
Crossposted-To: comp.os.linux.development.apps
Subject: Re: Why not C++
Date: 4 Jul 1999 03:18:33 -0700
Timo Tossavainen <[EMAIL PROTECTED]> wrote:
>Nathan Myers wrote:
>>
>Sometimes
>the extensibility interface demands great speed and is difficult
>to achieve. Extensible software synthesizers are one example. You
>may process vectors of data, which is fast enough but destroys the
>possibility of using feedback in the signal processing networks, or
>you may process one sample at a time which leads to the procedure
>calls eating up too much processor power so you lose realtime
>capability. As a hobby project I'm trying to see if it is possible
>to create a network compiler in Lisp (the compiler that compiles
>lisp is available at run-time), that produces fast enough code from
>plugins (which are actually more like code templates in this case)
There was an article in DDJ last year about a regular expression
package that works that way, except it generates machine code
directly. It's a good approach when nothing else will do, but
tends to be fragile because it's harder to spot errors by inspection
and harder to trace problems back to the (hand-written) code
responsible. It sure can be fast, though.
>A standard for crossplatform C/C++ dynamic libraries would be useful,
>because a simple recompile would suffice.
We have GNU Libtool.
>Actually I do
>signal compression research and many of the algorithms don't
>use arrays or matrices explicitly (like the FFT) although
>matrix inversions and eigenvectors/eigenvalues are needed.
I'm curious about anyone's experience with FFTW.
>> Still, it's worth noting that some languages are taking on a
>> degree of direct C++ compatibility, including Eiffel, Java,
>> and Python.
>
>... I'm curious though, that how is it possible to do
>direct C++ compatibility with these languages if there
>is no binary standard for C++ objects or do they use the C++
>compiler as a final step ?
They "know about" a few compilers. For example, Gjc, the GNU
Java compiler, is part of Gcc, and even shares exception
semantics with G++, so you can throw in C++ and catch it in
Java. I think Python understands the vtable formats of some
compilers. The whole prospect is pretty shaky still, and I
would avoid it and use a C interface.
--
Nathan Myers
[EMAIL PROTECTED] http://www.cantrip.org/
------------------------------
From: Shawn Lang <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Subject: HELP me Please
Date: Sun, 04 Jul 1999 04:24:05 -0700
i need help i need to make a symbolic link in my system
but i can make the dir usr/local/domains/***.com/Ops/
but i need to know how to make this dir's go to
/home/***/www/Ops/
please help me
------------------------------
From: Shawn Lang <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Subject: Help me PLEASE
Date: Sun, 04 Jul 1999 04:41:18 -0700
i need help i need to make a symbolic link
this is what i get when i use ln
ln -s /home/qudesttalk/www/Ops /usr/local/domains/questtalk.com/Ops/
ln: cannot create symbolic link `/usr/local/domains/questtalk.com/Ops/'
to `/home/qudesttalk/www/Ops': No such file or directory
what can i do to get this to work
i can make a dir in the local dir
help
------------------------------
From: "Adam JBC" <[EMAIL PROTECTED]>
Subject: Re: Why we are still holding on to X Windows
Crossposted-To: comp.os.linux.development.apps
Date: 4 Jul 1999 10:37:28 -0100
> >>Let's face it. X Windows is a really premitive base for modern GUI,
> >>terrible font support breaks GUI all the time, no sound capability,
...
> >>If Linux is going to desktops to compete with Microsoft, it got to come
> >>up with something much better then X.
> >>
> >>So, why don't we drop the X and innovate?
> >
> > "Write it and let us know when it's ready for testing"
>
> Provide a coherent and reasonably-complete design specification, and
> that would be enough to establish, at least in general terms, whether
> the would-be replacement actually "innovates" over what X offers.
How about an interface which meets the following criteria:
-Dosn't rely on screen size in any way (e.g. dialog boxes change size (they
can be tiny in Windoze with a large moniter/high resolution))
-Dosn't make a funny sound every time the mouse moves (can I stop this?)
-Can use multiple moniters (like Win98)
-Something like in Star Trek (seriously), their UI seems to be very
efficient, albeit full of meaningless numbers also good graphics
-I'm having a hard time coming up with ideas, but I think KDE at least is
crap, but it's all I have, before I can learn Linux and change it :-) I
espically hate the "Where do you want to go tomorrow", it's soooo tacky.
Okay, these are just wild ideas which popped into my head, feel free to
ignor them :-)
Adam.
------------------------------
From: Thorsten Heit <[EMAIL PROTECTED]>
Subject: Re: Help me PLEASE
Date: Sun, 04 Jul 1999 14:09:27 +0200
> ln -s /home/qudesttalk/www/Ops /usr/local/domains/questtalk.com/Ops/
>
> ln: cannot create symbolic link `/usr/local/domains/questtalk.com/Ops/'
> to `/home/qudesttalk/www/Ops': No such file or directory
That means that either the parent directory questtalk.com doesn't exist
or you don't have permissions to create a file/directoy/link etc. in it.
If it's the first case try "mkdir -p /usr/local/domains/questtalk.com",
otherwise you should contact your sysadmin ;-)
Thorsten
------------------------------
From: Nir <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc,comp.os.linux.setup
Subject: Re: Download freeBSD and Slackware 4
Date: Sun, 04 Jul 1999 14:06:38 +0300
vineet wrote:
> Can anyone tell me the site where I can download the full and latest
> version of FreeBSD and Slackware 4?
>
> ------------------ Posted via SearchLinux ------------------
> http://www.searchlinux.com
Try ftp://ftp.cdrom.com
anyway you can always go the the companies sites and try the site they
give you.
------------------------------
From: [EMAIL PROTECTED] (Scott Lanning)
Subject: Re: HELP me Please
Date: 4 Jul 1999 12:09:38 GMT
Shawn Lang ([EMAIL PROTECTED]) wrote:
: i need help
Yes....yes, you do... "development.system" doesn't mean
developing your skill at using the system. This would be
more appropriately asked....wait, you didn't even ask,
you just are begging... How pathetic.. <sigh>
: i need to make a symbolic link in my system
: but i can make the dir usr/local/domains/***.com/Ops/
: but i need to know how to make this dir's go to
: /home/***/www/Ops/
: please help me
man ln -- in particular the -s option.
Oh, alright, it's your birthday.. First, don't use '***'
obviously... (unless that's what you meant, in which case
use a "for" loop -- ask in comp.os.linux.misc)
ln -s dir_to dir_from
That's read, "symlink to dir_to from dir_from". I know, it's
confusing remembering the order because you're used to something
like "cp from_file to_file". Just pretend you're copying the
location of the file/directory to the new location, so it's
more like "copy (location of) dir_to to dir_from", but you
mentally (well, don't push it...) replace "copy" with "ln -s".
--
Scott Lanning: [EMAIL PROTECTED], http://physics.bu.edu/~slanning
"Besides a mathematical inclination, an exceptionally good mastery of
one's native tongue is the most vital asset of a competent programmer."
--Edsger Dijkstra
------------------------------
** 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: [EMAIL PROTECTED]
You can send mail to the entire list (and comp.os.linux.development.system) via:
Internet: [EMAIL PROTECTED]
Linux may be obtained via one of these FTP sites:
ftp.funet.fi pub/Linux
tsx-11.mit.edu pub/linux
sunsite.unc.edu pub/Linux
End of Linux-Development-System Digest
******************************
