>From: Alexander Skwar <[EMAIL PROTECTED]>
>>=20
>> 1) I warn poeple when FSF programs do make life harder.
>Why make tools with a greater ease of use life harder?
Sure? let's use something different from GNUtar so you cannot tell
that I am only pointing to GNU tar:
Why does GNU tail does not support the -b option?
-b is in the standard and the fact that GNU tar is
far way off from the standard makes life harder when I
am constantly switching between OSes as I need to write
portable software.
>> - Look at the GNU tar source and judge your own. The=20
>> GNU tar source is a big pile of spaghetti code. It is
>> hard to find bugs by people who are not used to work on it.
>That's right.
And it makes no sense to really further enhance GNU tar. This has been
aggreed on with the current GNU tar maintainer.
There already was a discussion whether GNUtar should be relpaced by star
because star is written much cleaner. The demand from FSF failed because
FSF had illegal demands in the Copyright of star.
>> - Next time you will send me a bug report for cdrecord,
>> I will tell you that it is _your_ task to find a patch
>> for the bug. So please stay resonable!
>I don't get it - if I have a problem with cdrecord, it's of course my task
>to either write a patch for it and tell you about it, or to at least inform
>you, the author, about it. But why should I tell the guy next door that I
>have a problem with cdrecord, so that he writes a patch to cdrecord? That's
>what you are suggesting wrt. tar - you've said that I should patch tar. I'm
>not the author of tar. So what I said was reasonable - maybe my choice of
>words was not.
NO: you don't get it.
I wrote everal time in the GNU tar discussion, that I _did_ write bug reports
to the GNU tar maintainer in 1994 & 1995 and nothing happened!
The files in ftp://ftp.fokus.gmd.de/pub/unix/star/testscripts/ partially
are a result of this discussion. The main Problem with GNU tar is that there
is no systematic testing done with GNU tar as I do with star. Everytime when
thought of adding a new feature in star that already exists in GNU tar,
I thought of possible problems. Then I did tests with GNU tar and found a bug.
I hope you now understand the problem: GNU tar is buggy, I did report the bugs
long long ago but the GNU tar maintainers are unwilling to fix bugs in GNU tar.
My conclusion is to warn people from using GNU tar.
Note that this does not only apply to interoperability bewteen GNU tar and other
tar programs. Most of the bugs I reported were present with GNU tar <-> GNU tar!
>Okay, but he used /bin/bash, didn't he? With this approach, it is very much
>portable, as /bin/bash is not the default shell. So both mine and your
>remarks are exaggerated.
But there is no guarantee that there is a /bin/bash. /bin/sh os present on al
UNIX flavors and I cannot believe that the scripts cannot be made working for
/bin/sh. ... of course you need to know what sh officially supports as Linux
unfortunately links /bin/sh to /bin/bash.
>Good question. To enhance the ease of use: Yes, you do need such a feature=
>.=20
>But he may write something like I suggeseted, together with:
># If your system does not provide a /proc/kcore file or the 'free' tool,
># please enter the amount of system memory manually.
Other applications have similar problems and did solve them.
It is always a good idea to look at other applications to learn good practice.
J�rg
EMail:[EMAIL PROTECTED] (home) J�rg Schilling D-13353 Berlin
[EMAIL PROTECTED] (uni) If you don't have iso-8859-1
[EMAIL PROTECTED] (work) chars I am J"org Schilling
URL: http://www.fokus.gmd.de/usr/schilling ftp://ftp.fokus.gmd.de/pub/unix
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
