Re: compiling with gettext
Yesterday I reported that I was no longer able to compile the package 'sp'. Santiago Vila and Tom Lees both kindly offered suggestions, but both of these still leave sp in an uncompilable state. Santiago pointed out that the problem might be intrinsic to the existing gettext-0.26, and suggested I wait til gettext-0.28 becomes available. This might be a very long time, since gettext-0.26 came out in December. Tom pointed out that my include line: #include /usr/share/gettext/intl/libgettext.h referred to libc6 headers, which must surely have caused a problem since I was linking with libc5. So I changed this to #include /usr/include/libintl.h. But this produces the same result as if I had no include line at all: ../lib/libsp.a(MessageTable.o): In function `GettextMessageTable::GettextMessageTable(void)': MessageTable.o(.text+0x35): undefined reference to `bindtextdomain' ../lib/libsp.a(MessageTable.o): In function `GettextMessageTable::getText(MessageFragment const , Stringchar ) const': MessageTable.o(.text+0x69): undefined reference to `dgettext' If anyone has any other suggestions, I'd be very grateful. Susan -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
compiling with gettext
I have been trying for some time to solve Bug #8882 against the 'sp' package, which says that in order to make it buildable under glibc, I need to call libintl as well as libnls in order to accommodate glibc, and to define LINUX_TYPES_H for glibc. I made those changes and could no longer get the program to compile. After a while I realized that the changes were unrelated to my inability to compile the program; which I tested by going back to the original sources and trying recompilation without the fixes. Sure enough, I got the same error messages: 1. First there's a complaint about an undefined dgettext in lib/MessageTable.cxx. I think I bludgeoned this one by inserting the line #include /usr/share/gettext/intl/libgettext.h in the top of that file. 2. When I then recompile the program I get the message: MessageTable.cxx:102: parse error before `const' MessageTable.cxx:103: parse error before `const' MessageTable.cxx:104: parse error before `while' MessageTable.cxx:105: parse error before `while' where the four lines in question are: extern char *dgettext(const char *, const char *); extern char *gettext(const char *); extern char *textdomain(const char *); extern char *bindtextdomain(const char *, const char *); I thought I saw a message in debian-devel that gettext was now outmoded. Perhaps this is related to my problem? I notice that Ulrich Drepper has version 0.27 of gettext on his server, but he still has the generic gettext package pointing to 0.26, which is what we have in the distribution and what I have installed. Anyway, it would seem pretty meaningless to me to have a package in the Debian distribution which can't even be compiled. If 'sp' is removed, then so should a lot of other programs that depend on it. If anyone could provide any pointers or advice, I'd try to upload a fixed version very quickly. Susan -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Bug#4623: threading in exmh
Package: exmh
Version: 1.6.9-2
Recently, in an attempt to thread messages with exmh, I wrote to the author
of thread.tcl (Ignacio Martinez [EMAIL PROTECTED]) and asked for
some help, which he has graciously given.
He said that the version of thread.tcl which was in 1.3.9 was outmoded,
and sent me a new version. I've attached it below.
He also said that I'd need to patch folder.tcl in order to use the
revised thread.tcl. I've attached that below also.
He instructed me to run auto_mkindex in /usr/lib/exmh.
Finally, he said I'd have to add some buttons to my .exmh_defaults file.
*Fops.ubuttonlist: thread
*Fops.thread.text: Thread
*Fops.thread.command: Thread_DisplayAll
Two problems with the threader are:
a) every time one executes a Commit after marking a message for deletion,
the whole folder is automatically re-threaded. This takes a lot of time.
b) threading doesn't work on folders longer than about 2000 files.
In any case, I believe it would be valuable to make this updates to
the exmh package.
Regards,
Susan Kleinmann
==new version of thread.tcl===
# thread.tcl
#
#
# Display FTOC messages in a threaded manner
#
# Ignacio Martinez[EMAIL PROTECTED]
# Fundesco
# Madrid, April 1996
#
proc Thread_PrintReplies { msg minfo off mark {indent -1} } {
upvar $minfo msginfo
global exwin ftoc
if {$indent 0} {
set indent 0
set blank
} else {
incr indent [expr [string length $mark] + 1]
set blank [format %*s $indent ]
}
set maxoff [expr $ftoc(scanWidth) - 2] ;# newline counted as well
foreach m $msginfo(refs,$msg) {
if {[lsearch $msginfo(out) $m] 0} {
set text $msginfo(text,$m)
set tmplist [list [string range $text 0 $off] $blank $mark \
[string range $text [expr $off + 1] end]]
set newtext [join $tmplist ]
if {[string length $newtext] $maxoff} {
set newtext [string range $newtext 0 $maxoff]
}
$exwin(ftext) insert end $newtext\n
lappend msginfo(out) $m
Thread_PrintReplies $m msginfo $off $mark $indent
}
}
}
proc Thread_IsRel { minfo msg } {
upvar $minfo msginfo
if {[lsearch $msginfo(selm) $msg] = 0} {
return 1
}
foreach m $msginfo(refs,$msg) {
if [Thread_IsRel msginfo $m] {
return 1
}
}
return 0
}
proc Thread_Scan { folder minfo } {
upvar $minfo msginfo
#
# We only care about what is currently displayed into the FTOC.
# New messages are ignored.
#
set maxlines $msginfo(maxl)
set firstmsg [Ftoc_MsgNumber 1]
set lastmsg[Ftoc_MsgNumber $maxlines]
set scan_fmt %(msg)%{message-id}%{in-reply-to}%{references}
set scan_cmd [list scan +$folder $firstmsg-$lastmsg \
-noheader -noclear -width -format $scan_fmt]
if [catch {open |$scan_cmd} pipe] {
Exmh_Status scan failed: $pipe purple
return 1
}
set numline 0
set status Scanning $folder for cross-references ...
set pass [expr int($maxlines/10)]
set msginfo(hits) 0
set msginfo(tref) 0
Exmh_Status $status blue
while {[gets $pipe line] 0} {
if ![regexp {^ *([0-9]+)([^]*)(.*)} $line x num mid newline] {
# no message-id?
regexp {^ *([0-9]+)} $line x num
set mid {}
set newline {}
}
if {$num != [lindex $msginfo(msgs) $numline]} {
Exmh_Status thread/scan message mismatch. Rescan? purple
return 1
}
incr numline
if {$maxlines 250 [expr $numline%$pass] == 0} {
set done [expr 10*$numline/$pass]
Exmh_Status $status $done% done blue
}
set msginfo(refs,$num) {}
set msginfo(isref,$num) 0
set msgnum($mid) $num
set line $newline
while {[regexp {([^]*)(.*)} $line x mid newline] == 1} {
if [info exists msgnum($mid)] {
set ref $msgnum($mid)
lappend msginfo(refs,$ref) $num
set msginfo(isref,$num) 1
incr msginfo(hits)
} else {
if ![info exists unres($num)] {
set unres($num) {}
}
lappend unres($num) $mid
}
set line $newline
incr msginfo(tref)
}
}
close $pipe
#
# Second round. Disordered messages (i.e. replies received BEFORE their
# originals)
#
foreach res [array names unres] {
foreach mid $unres($res) {
if [info exists msgnum($mid)] {
set ref $msgnum($mid)
lappend msginfo(refs,$ref) $res
set msginfo(isref,$res) 1
incr msginfo(hits)
}
}
}
return 0
}
proc Thread_Display { {breakoff 20} {mark
Re: Can't contact Bill Mitchell
Hi Dale -- When I try to send e-mail to Bill at: [EMAIL PROTECTED] I get a no such user error. Does anyone have a good contact address for Bill? Bill sent a message to debian-devel a few days ago from this address: [EMAIL PROTECTED] Better yet, does anyone have a copy of the patch we did on less to fix the problem reading proc files? I think this is the message you are looking for is the one attached below. Cheers, Susan == Date: Thu, 30 May 96 17:30 CDT From: Mike Coleman [EMAIL PROTECTED] To: [EMAIL PROTECTED] CC: [EMAIL PROTECTED], debian-devel@lists.debian.org, [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Subject: Re: psutils and pipes bug Date: Thu, 30 May 1996 12:29:03 -0500 From: Lukas Nellen [EMAIL PROTECTED] Unfortunately, that doesn't solve the problem. Could the problem be related to the libraries you link psutils with? The version of libc on my system is: libc5 5.2.18-6 which is the only shared library used by psnup and pstops. Or could there be another difference between our systems which might be relevant? The problem seems to be one or both of: (1) a change in the way psutils decides whether a file is seekable or not, and (2) libc's (possibly changed) optimization of fseeks. Specifically, the current version of psutils decides whether a file is seekable by SEEK_CUR with 0 as the arg (which is a no-op seek to the current position), checking to see whether this generates an error. Libc, however, optimizes this fseek away (i.e., without calling lseek), and doesn't generate an error. The attached patch seems like a reasonable fix. It tests seekability by seeking to EOF and back to the current point, checking for errors. (I'm also Cc:'ing Angus (author of psutils) on this.) --Mike P.S. One of the debian.* files mentions [EMAIL PROTECTED] as the package maintainer. Is this really the same as [EMAIL PROTECTED] or a typo? --- psutil.c~ Mon Feb 19 11:27:19 1996 +++ psutil.cThu May 30 17:15:16 1996 @@ -82,6 +82,7 @@ FILE *ft; long r, w; char *p; + long fp_pos; #if defined(WINNT) struct _stat fs ; #endif @@ -90,8 +91,9 @@ if (_fstat(fileno(fp), fs) == 0 (fs.st_mode_S_IFREG) != 0) return (fp); #else - if (!fseek(fp, 0L, SEEK_CUR)) - return (fp); + if ((fp_pos = ftell(fp)) = 0) + if (!fseek(fp, 0L, SEEK_END) !fseek(fp, fp_pos, SEEK_SET)) +return (fp); #endif #if defined(MSDOS)
Re: where are guidelines?
Hi Joost -- I think there may be 2 reasons why you can't find the package containing the guidelines: 1. Guidelines doesn't exist any more. It has evolved into the policy manual and the programmer manual, and 2. The PP manuals aren't in a package (that I could find). They are uploaded seperately into doc/package-developer; the newest ones are in /private/project/Incoming. They are called policy.ps.gz, policy.html.gz, programmer.ps.gz, and programmer.html.gz. HTH, Susan
Bug#4548: tset man page appears to be aborted
Package: ncurses-bin Version: 1.1.9e-1 The last line of the tset man page ends in a comma: SEE ALSO csh(1), sh(1), stty(1), tty(4), termcap(5), ttys(5), envi- ron(7), I do not know what ending the author intended, but: -- the comma implies _something_ is missing; -- it seems that terminfo(5) should be among the SEE ALSO's mentioned; and -- it would be nice to know the author of the page (I was curious because it's one of the more clearly written man pages). Susan Kleinmann
Re: Can I get a debiandoc package?
Hi Syrus -- You said: I would like to generate ps versions of the new policy and programmer manuals which are formatted for letter size. I notice that sgml2latex does not work on the .sgml ... If you'd just like to convert the existing postscript files (which were formatted for A4 paper) to letter paper, then you can use: psresize -Pa4 -pletter policy.ps policy.letter.ps Ditto for programmer.ps. To use the .sgml files, you'll need to fetch and install the DTD that Ian used, then the conversion files he used. Good luck, Susan Kleinmann
dpkg and dependencies
Ian Jackson and I have had several exchanges by private email having to do with the dependencies in dpkg, following my bug reports: #4262 (dpkg-source requires cpio) and #4263 (dpkg-source requires patch). We have now come to an impasse. Ian suggested if I were to bring this to debian-devel perhaps someone there will set you straight. Here are the arguments: I cpio is only required by dpkg-source, and not to install packages. I I've added it to Suggests and mentioned it in the Description. This I will be in 1.3.10. (ditto for patch.) S Now I'm confused. I would have thought that since a command within the S dpkg package requires cpio in order to operate, that the whole dpkg package S would be said to Depend on cpio. I No. S This is not my reading of the programmer's manual, which says: S Suggests: S This is used to declare that one package may be more useful with one S or more others. Using this field tells the packaging system and the S user that the listed packages are be [sic] related to this one and can S perhaps enhance its usefulness, but that installing this one without S them is perfectly reasonable. I The manual is correct. I believe dpkg is in accordance with it now. S It is unreasonable to install software (dpkg-source) which won't S function at all without additional files. I So are you suggesting that I should split dpkg up into half a dozen I packages ? Since debian-changelog-mode.el won't work without Emacs I I have to make dpkg depend on Emacs ? Since you can't use I dpkg --print-architecture to determine the build architecture unless I you have GCC and libc installed I should make dpkg depend on those ? I Since dpkg's NFS installation method won't work without NFS loaded I into the kernel I should refuse to allow dpkg to install properly if I you don't have that configured in ? S Yes. S In any case, what is the mechanism to warn the user that only 1 S (or perhaps 0, 2, or some other number) of components of a package S he's got installed ought to be expected to work? I Suggests, plus the Description, and the manuals for the package I components if this is thought necessary. S I find this logic absurd. If it is reasonable to install software that S won't work, then what is the reason? Just to remind oneself of its S existence? S Debian users should be told during the initial installation they will S have 2 choices: S -- realize that some, most, or perhaps all of the software they just Sthought they installed won't work at all, and that they'll need to read Severy doc and every man page and every info page associated with the Spackage in order to identify the components that won't work, or S -- follow religiously every suggestion made by every package. S I doubt this is the common understanding of the word suggests, except S perhaps in the world of organized crime. S The package dependency mechanism does little good if there is not a common S understanding and use of it. OK, will someone please set me staight? Susan Kleinmann
Bug#4305: metmail uses non-existent flag in postinst
Michael Meskes writes:
Susan G. Kleinmann writes:
Package: metamail
Version: 2.7-6
Oops, wrong depends line. Please upgrade your mime-support package.
(Perhaps this should be a seperate bug report; if you want one, please
let me know.)
I upgraded mime-support to 2.02-1 sometime yesterday.
Today I re-ran dpkg --configure metamail. This produced some odd
queries:
# dpkg --configure metamail
Setting up metamail (2.7-6) ...
New action 'view' for MIME type 'image/*'...
-- package=metamailview=showpicture -viewer xloadimage -view
-quiet %s
1) package=metamailview=showpicture -viewer xv %s
Place at what priority? (1-2) --1
^^^--- Notice that these indices do not match
the choices above [-- and 1)]
New action 'view' for MIME type 'text/richtext'...
-- package=metamailview=shownonascii iso-8859-1 -e richtext -p %s;
test=test `echo %{charset} | tr [A-Z] [a-z]` = iso-8859-1 -a $DISPLAY
!= ; copiousoutput
1) package=metamailview=shownonascii iso-8859-8 -e richtext -p %s;
test=test `echo %{charset} | tr [A-Z] [a-z]` = iso-8859-8 -a $DISPLAY
!= ; copiousoutput
Place at what priority? (1-2) --1
^^^--- Notice that these indices do not match
the choices above [-- and 1)]
New action 'view' for MIME type 'text/richtext'...
-- package=metamailview=richtext %s; copiousoutput
1) package=metamailview=shownonascii iso-8859-1 -e richtext -p %s;
test=test `echo %{charset} | tr [A-Z] [a-z]` = iso-8859-1 -a $DISPLAY
!= ; copiousoutput
2) package=metamailview=shownonascii iso-8859-8 -e richtext -p %s;
test=test `echo %{charset} | tr [A-Z] [a-z]` = iso-8859-8 -a $DISPLAY
!= ; copiousoutput
Place at what priority? (1-3) --1
^^^--- Notice that these indices do not match
the choices above [1) and 2)]
This seems to be a problem with install-mime; or perhaps I did
something wrong. ??
Susan
Re: Bruce - fiat required to end discussion on lyx/copyright ?
Ian Jackson wrote: Dale Scheetz writes (Re: Bruce - fiat required to end discussion on lyx/copy right ?): ... Pine is in non-free because it's copyright places restrictions on the distribution of source. Xforms has more severe restrictions on the distribution of source than pine does. It is my understanding that this source distribution restriction is what makes Xforms' proper location to be non-free. Please read chapter 2 of the new policy manual. This is my synopsis of the relevant parts of Chapter 2: Packages go into contrib if their copyrights or patents require that they: a. allow distribution of no source code b. allow distribution of only some source code, but not all the source code needed to compile the program (even given the existence of other sources in the Debian distribution). c. depend on a non-free or contrib package in order to be used d. allow use only for a trial period e. lack vital functionality f. are installer packages g. fail to meet some other policy requirement Packages go into non-free if their copyrights or patents require that they: h. disallow distribution for profit i. disallow distribution on certain media j. disallow distribution except if special permission is obtained k. have any other onerous conditions. My reactions: Condition (a) is redundant, given condition (b). It is not clear either what is meant by condition (k), nor how condition (k) differs from condition (g). Without such a distinction, non-free and contrib overlap. The word onerous in condition (k) would seem inconsistent with the Debian objective to be a base upon which value-added GNU/Linux distributions can be built. Susan Kleinmann
Bug#4294: gcc -bi486-linux can't find cc1
Package: aout-gcc Version: 2.6.3-13 FYI, the aout-gcc file is in: /usr/lib/gcc-lib/i486-linuxaout/2.6.3/cc1 Susan Kleinmann
Bug#4305: metmail uses non-existent flag in postinst
Package: metamail Version: 2.7-6 # dpkg --configure metamail Setting up metamail (2.7-6) ... /usr/sbin/install-mime: Unknown option '--quiet' /usr/sbin/install-mime: Unknown option '--quiet' /usr/sbin/install-mime: Unknown option '--quiet' /usr/sbin/install-mime: Unknown option '--quiet' /usr/sbin/install-mime: Unknown option '--quiet' /usr/sbin/install-mime: Unknown option '--quiet' /usr/sbin/install-mime: Unknown option '--quiet' /usr/sbin/install-mime: Unknown option '--quiet' ... dpkg: error processing metamail (--configure): subprocess post-installation script returned error exit status 2 Errors were encountered while processing: metamail Susan Kleinmann
Re: Bruce - fiat required to end discussion on lyx/copyright ?
Ian Jackson writes: Susan G. Kleinmann writes (Re: Bruce - fiat required to end discussion on ly x/copyright ? ): ... This is my synopsis of the relevant parts of Chapter 2: Packages go into contrib if their copyrights or patents require that they: a. allow distribution of no source code b. allow distribution of only some source code, but not all the source cod e needed to compile the program (even given the existence of other source s in the Debian distribution). c. depend on a non-free or contrib package in order to be used d. allow use only for a trial period e. lack vital functionality f. are installer packages g. fail to meet some other policy requirement Packages go into non-free if their copyrights or patents require that they: h. disallow distribution for profit i. disallow distribution on certain media j. disallow distribution except if special permission is obtained k. have any other onerous conditions. My reactions: Condition (a) is redundant, given condition (b). Yes, if you think about them like that. I haven't expressed it quite that way. It is not clear either what is meant by condition (k), nor how condition (k) differs from condition (g). Without such a distinction, non-free and contrib overlap. (k) is there as a catch-all, in case someone comes up with another example of a bad thing in a copyright. non-free and contrib do overlap - they are intended to. The way I have phrased it makes it clear that if a package meets the bad criteria for needing to be in non-free, and those for contrib, it must go in non-free. The word onerous in condition (k) would seem inconsistent with the Debian objective to be a base upon which value-added GNU/Linux distributions can be built. I don't understand this at all. The above several paragraphs actually echo a common theme. The word onerous is commonly taken to be perjorative; certainly the phrase bad criteria is perjorative. Therefore one interprets the category non-free as perjorative, rather than simply being a statement of fact that copyright restrictions exist. The use of such language is unnecessary and inconsistent with Debian's purported objective of being a base for value-added distributions. To be plain about it, you don't normally go around telling people you think they're bad or their ideas are bad, and then expect to attract them to the notion of using your software. Susan Kleinmann
Bug#4262: dpkg-source requires cpio
Package: dpkg Version: 1.3.8 In unpacking the sgmlspm dsc file on a fairly fresh system, I found: # dpkg-source -x sgmlspm_1.03ii-2.dsc dpkg-source: failure: exec cpio: No such file or directory dpkg-source: failure: cpio gave error exit status 2 # type -a cpio type: cpio: not found Installing cpio cured the problem, but suggests that the newest versions of dpkg need to depend on cpio, which (I think) means that cpio would have to be added to base. Susan Kleinmann
Bug#4263: dpkg-source requires patch
Package: dpkg Version: 1.3.8 In unpacking the sgmlspm dsc file on a fairly fresh system, I found: # dpkg-source -x sgmlspm_1.03ii-2.dsc dpkg-source: extracting sgmlspm in sgmlspm-1.03ii dpkg-source: failure: exec patch: No such file or directory dpkg-source: failure: patch gave error exit status 2 Once patch was installed, dpkg-source completed successfully. Since patch isn't in base, I assume this means that the dpkg Depends field has to be amended to add 'patch'. Susan Kleinmann
Re: Bug#4164: Ferret extended description has blank lines
I agree with Brian that micro-bugs should probably be reported to the maintainer, just because it takes more work to contact the reporter and close the bug than to fix it. However, in this case the reporter has to keep his _own_ bug tracking system, if he wants to make sure that the maintainer actually fixed the bug. This may or may not be a big deal, depending on the responsiveness (read, speed and verbal skills) of the maintainer. Also, the reporter then takes the moral responsibility of (a) having wasted a third person's time who may chance across the bug (it takes time to identify a bug), and (b) having wasted someone else's time reporting a bug. I agree with Lars that the bug tracking system is not a shame pole. I agree with Dale that trivial (or incorrect) bugs give a bad and unbalanced impression of the distribution. I agree with Ian that we can't afford to let bugs go unreported because they might be publicly visible (that's actually a lot of help!). And I also agree with Ian that debian-user is a bad, bad place to report bugs because of its traffic. Having to do with psychology, it seems to me there is another advantage of using the bug tracking system, and that has to do with the less-than-perfect side of humans: all of Debian's developers are busy folks who need to save time. If someone mails a developer a message regarding a trivial (or maybe not so trivial) bug fix, and he deems it unimportant, he might be tempted to tell the reporter to forget it. This leaves these possibilities: A. the reporter agrees it was silly and defers to the maintainer, and no one else is ever bothered by this. B. the reporter agrees it was silly and defers to the maintainer, and someone else does find the same bug, thereby wasting his time and _really_ giving the Debian distribution a bad name because of real bugs. C. the reporter doesn't agree with the maintainer, takes the bug to the bug tracking system, thereby deliberately turning up the heat on the maintainer, and potentially causing a mini-war. So, except where the ratio of triviality-of-bug / responsiveness-of-maintainer is near 0, or (better) where the reporter realizes the maintainer has his own motivations for fixing the bug right away, I agree with Ian that the right thing to do is to log the bug. I believe the bug tracking system may be the single most important aspect of the Debian system. Even when a reporter isn't sure that an event is a bug, he should report it, since this eventually leads to FAQ's which can then be aired and resolved :-). (I now feel guilty for not having always used the BTS myself in the past.) Best regards, Susan Kleinmann
Bug#4238: mirror requires perl
Package: mirror Version: 2.8-6 The control file for mirror should be changed to show that it requires the installation of the full Perl package, not just the small package included in the base system (what's missing is a timezone.pl file). Alternatively, the perl package in the base system needs to have the timezone.pl file added to it. Susan Kleinmann
Bug#4234: errors in /usr/man/man8/make-kpkg.8
Package: kernel-package Version: 2.03 The man page for make-kpkg has a few typos, some of which refer to file names and might therefore cause confusion. A diff file is attached. Susan Kleinmann *** make-kpkg.8 Thu Aug 22 10:40:19 1996 --- make-kpkg.8.rev Thu Aug 22 10:56:23 1996 *** *** 28,34 file using PGP. This option will override the builtin default and the site wide customizations stored in the file ! .B /etc/kernal-package.conf. This option is only relevant to the .B dist target, since that is the only target that uses --- 28,34 file using PGP. This option will override the builtin default and the site wide customizations stored in the file ! .B /etc/kernel-package.conf. This option is only relevant to the .B dist target, since that is the only target that uses *** *** 128,138 If there is no .I .config file in the kernel source directory, a default configuration is provided ! similair to the one used to create the .B Debian boot\-floppies. The package produced updates the .I /boot/psdatabase ! file at install time, it also update symbolic links in the root directory to point to the new kernel image in .I /boot. It also offers to run the --- 128,138 If there is no .I .config file in the kernel source directory, a default configuration is provided ! similar to the one used to create the .B Debian boot\-floppies. The package produced updates the .I /boot/psdatabase ! file at install time, it also updates symbolic links in the root directory to point to the new kernel image in .I /boot. It also offers to run the *** *** 186,202 file run by .B make\-kpkg also looks for site\-wide defaults in the file ! .I /etc/kernel-packge.conf. The default configuration allows there to be a site wide override for the full name and email address of the person responsible for maintaining the kernel packages on the site, but the ! .I /etc/kernel-packge.conf file is actually a Makefile snippet, and any legal make directives may be included in there. .B Note: Caution is urged with this file, since you can totally change the way that the ! make is run by suitable editing this file. .SH SEE ALSO .BR dchanges (1), .BR make (1), --- 186,202 file run by .B make\-kpkg also looks for site\-wide defaults in the file ! .I /etc/kernel-package.conf. The default configuration allows there to be a site wide override for the full name and email address of the person responsible for maintaining the kernel packages on the site, but the ! .I /etc/kernel-package.conf file is actually a Makefile snippet, and any legal make directives may be included in there. .B Note: Caution is urged with this file, since you can totally change the way that the ! make is run by suitably editing this file. .SH SEE ALSO .BR dchanges (1), .BR make (1),
Bug#4189: dselect highlights too many lines when TERM=linux
Lars Wirzenius wrote: someone else who is now trying to hide her identity wrote:: Wishing to see the colored version of dselect, I tried this from a color xterm: # export TERM=linux # dselect linux is, of course, the wrong terminal type. For color xterm's, the correct type is xterm-color. (dselect doesn't seem to be using colors in the main menu, but it does in the selection screen.) Not that it works too well for me, anyway (color-aware programs seem to want to prefer the background to be black, and will set it to back when exiting). Oops. Right you are. BTW, whether TERM=linux or TERM=xterm-color, dselect DTRT when it is invoked with one of the menu options, e.g., dselect access brings up a nicely colored screen. (Susan -- needing very small font here.)
Bug#4189: dselect highlights too many lines when TERM=linux
Package: dpkg Version: 1.3.5 Wishing to see the colored version of dselect, I tried this from a color xterm: # export TERM=linux # dselect I didn't see any colors, which was disappointing but hardly fatal. However, when I used the arrow keys to scroll up or down, dselect highlighted _all_ the menu options I scrolled over, not just one. Susan Kleinmann
Bug#4106: mime-support doc files are installed into /usr/doc
Package: mime-support Version: 2.0-1 mime-rfc-1542.txt.gz and mime-install.HOWTO.gz are installed into /usr/doc, rather than into /usr/doc/mime-support. Susan Kleinmann
Re: Documentation formats
Rob Browning writes: Lars Wirzenius [EMAIL PROTECTED] writes: I don't think we have free software packaged to do full text searches. We have glimpse and ferret, neither of which is free. There's something that is part of freeWais, but I haven't looked at it yet. Someone with the time should find and package one. It should be usable from the command line so that it can easily be integrated into Debiandoc. It'd be nice to have something like altavista, that could generate a page containing a (heirarchical) list of the references found. The search program ht://Dig (which has an admittedly odd name) comes with the GNU general public license. Its home page is: http://htdig.sdsu.edu/ Also, there's been a Javascript on comp.infosystems.www.authoring.cgi that uses the Altavista search engine to search only your own site. (Seems like exploiting rather than excluding robots.) Regards, Susan Kleinmann
Bug#4019: Decompression failure on Install Boot Disk
Hi Chris -- I don't know if you have gotten Linux booted yet. If you haven't, please let me know. I'll be happy to make a boot disk for you and mail it to you by real, old-fashioned physical US Mail. Just let me know what kernel you need to have on it, or if you need a special one. Cheers, Susan Kleinmann [EMAIL PROTECTED] Package: Boot1440.bin Version: Debian 1.1 I have been unable to boot from the boot1440.bin image. I have tried the following: - downloaded the image from a variety of sources. - used several different floppies - erased (using dd if=/dev/zero ...) the floppies before writing the image to them. - formatted the floppies using a variety of tools When I attempt to boot, I get the following: invalid compressed format (err=2) Then it tells me that the system is halted. I've tried entering boot options (such as 'linux ether=10,0x320,eth0'). When I do this, I get a slightly different error message: incomplete literal tree Where do I go from here? Chris Hertel -)- [EMAIL PROTECTED] . -- Christopher R. Hertel -)- University of Minnesota [EMAIL PROTECTED] Networking and Telecommunications Services
Bug#4017: boot disk doesn't recognize NCR scsi
Hi Richard -- I've made a couple of NCR-aware kernels for friends. If you'd like one of them, you might try: ftp://sgk.phast.umass.edu/pub/boot1440.bin.gz If you have some special requirement and can't make your own, let me know and I'll make one for you. Cheers, Susan Kleinmann [EMAIL PROTECTED] Package: bootdisk Version: 1.1.1 Attempted to install Debian 1.1.1 from PHT cd rom. Bootdisk doesn't see my SCSI system, based on NCR 53c8xx PCI Rev 2.0 1993 This device worked find under Debian .93R6/ Linux 1.2.13
Bug#3953: Tk 4.1's bell flashes the display.
I don't get the Visual Bell when I run the wish script you mentioned. Perhaps you have vb (visual bell) set in your terminfo? Susan Kleinmann
Bug#3901: dotlock should be setgid mail
A couple of days ago, Lars mentioned I suggest staying with rwxrwxrwt. with respect to the permissions for /var/spool/mail. Mine is: drwxrwsr-t 2 mail mail 1024 Jul 31 13:08 /var/spool/mail/ I couldn't find the debian package that installed this directory, so I couldn't figure out if these permissions were right, whatever that means. So, a) can someone tell me what package/file/script does install /var/spool/mail? (I'd use this info to find out why my search method failed.) b) Are those permissions right? c) Lars, what did you mean by saying staying with rwxrwxrwt? Regards, Susan Kleinmann [EMAIL PROTECTED]
Revised Debian FAQ
A copy of the revised version of the Debian FAQ has been placed on my WWW server at: http://sgk.phast.umass.edu/FAQ/debian-faq.html Other formats for this document are available by anonymous FTP at: ftp://sgk.phast.umass.edu/pub/debian-faq.txt.gz (plain text) ftp://sgk.phast.umass.edu/pub/debian-faq.ps.gz(Postscript) ftp://sgk.phast.umass.edu/pub/debian-faq.tex.gz (LaTeX) These documents should be thoroughly reviewed by Debian developers before they are publicized. Please alter the authors of the faq of required or recommended changes by sending mail directly to us. (We have a mailing list set up for the debian faq, but it hasn't been working this last week.) After a week, I will forward the documents to Sven Rudolph for inclusion in a revised debian-doc package. Thanks in advance for your comments, Susan Kleinmann
the Search system
I just found the Debian search system on http://www.debian.org/ghindex.html. This seems to me like a big step forward, and am sorry I didn't see it before. Just yesterday I remarked to someone (for whom I had answered a user-question by specifying the appropriate HOWTO) that the HOWTO's desperately needed a search interface. Questions and comments: 1. In the line Files in this directory, what does the word 'this' refer to? 2. Would it be possible (i.e., would the motivation/pain ratio exceed 1) to add the HOWTO's to the list of indexed files, and then automate the rebuild of the glimpse index as new packages appear for doc-linux? 3. Going further, would it possible (same as above) to simply add all of /usr/doc and all of /usr/info to the list of indexed files? If so, the text above the search engine form could be greatly streamlined. 4. Comment: it would be nice to put the selectable directories in a SELECTOPTION /SELECT grouping, where the default was ALL. I would have been happy to send this directly to the author of the ghindex.html, but I don't know who that is. It might be useful to add a mailto: at the bottom of the page. Cheers, Susan Kleinmann [EMAIL PROTECTED]
a reconnaissance of Debian mirrors
In the process of updating my WWW pages last night, I decided to check all the Debian mirrors; given the new release, I thought this might be a good time. As a spot check on how uptodate a mirror was, I looked for/at unstable/binary/base/base-*.deb. In particular, I was checking the version number on base. I found several surprises. a) One site (ftp.usyd.edu.au) was months out of date only because the administrator (Andrew Howell) didn't have a mirror account. b) Most of the international debian.org aliases seem to be broken. The only exception was ftp.debian.org.au - ftp.tower.net.au c) Many sites (including Simon Shapiro's) encourage folks to get more info from [EMAIL PROTECTED] I saw this in at least one .message file at the top of the unstable directory. d) Many sites seem to be missing the base package. I suspect this is because of a bug with a link. e) A number of mirrors listed on previous emails or in the www.debian.org Debian FTP sites list seem to be surprisingly out of date or incomplete. This may be due to aborted mirror runs. I don't know. I've written to all (I think) of the administrators about these problems, but thought you all might want to be aware of the situation. Also, attached below please find my updated WWW page. Feel free to use/incorporate/edit it as you wish. I think it would be useful to have a script or two, to check automatically whether all of these sites (or at least a reasonable number on one's continent) are on-line and current. It would be good to have the output easily accessible. Also, folks might want to have a script that ping'd the sites within some radius of their site, as an indicator of the fastest route to a host at any particular time. Perhaps such scripts already exist. Comments? Susan Kleinmann [EMAIL PROTECTED] = HTML HEADTITLEDebian Mirrors/TITLE/HEAD BODY H3 ALIGN=CENTERDebian Mirrors/A/H3 The official list of Debian mirrors is listed in debian/README.mirrors. The list below was gleaned from the debian-user mailing list, and is larger than the official list. P Sites marked with an asterisk (*) keep source code. P TABLE TRTH ALIGN=LEFTCountry/TH TH ALIGN=LEFTHost:/Directory/TH /TR TRTDAustralia/TD TD A HREF=ftp://ftp.debian.org.au/pub/linux/debian; ftp.debian.org.au:/pub/linux/debian/A*, a.k.a. ftp.tower.net.au /TD /TR TRTDAustralia/TD TD A HREF=ftp://ftp.it.com.au/mirrors/linux/debian; ftp.it.com.au:/mirrors/linux/debian/A* /TD /TR TR TDBelgium/TD TD A HREF=ftp://ftp.kulnet.kuleuven.ac.be/pub/mirror/ftp.debian.org/debian; ftp.kulnet.kuleuven.ac.be:/pub/mirror/ftp.debian.org/debian/A* /TD /TR TR TDBelgium/TD TD A HREF=ftp://ftp.belgium.eunet.net/pub/linux/debian; ftp.belgium.eunet.net:/pub/linux/debian/A* /TD /TR TR TDBrazil/TD TDA HREF=ftp://linux.if.usp.br/debian; linux.if.usp.br:/debian/A*, /TD /TR TR TDCosta Rica/TD TD A HREF=ftp://simula.efis.ucr.ac.cr/pub/debian; simula.efis.ucr.ac.cr:/pub/debian/A* /TD /TR TR TDFinland/TD TD A HREF=ftp://ftp.funet.fi/pub/Linux/mirrors/debian/; ftp.funet.fi:/pub/Linux/mirrors/debian/A* /TD /TR TR TDFrance/TD TD A HREF=ftp://ftp.ibp.fr/pub/linux/distributions/debian; ftp.ibp.fr:/pub/linux/distributions/debian/A* /TD /TR TR TDGermany/TD TD A HREF=ftp://ftp.inf.tu-dresden.de/pub/os/linux/debian; ftp.inf.tu-dresden.de:/pub/os/linux/debian/A* /TD /TR TR TDGermany/TD TD A HREF=ftp://ftp.inka.de/debian/;ftp.inka.de:/debian/A* /TD /TR TR TDNetherlands/TD TD A HREF=ftp://ftp.leidenuniv.nl/pub/linux/debian; ftp.leidenuniv.nl:/pub/linux/debian/A* /TD /TR TR TDNetherlands/TD TD A HREF=ftp://ftp.nluug.nl/pub/os/Linux/distr/Debian; ftp.nluug.nl:/pub/os/Linux/distr/Debian/A* /TD /TR TR TD Sweden/TD TD A HREF=ftp://ftp.lh.umu.se/debian; ftp.lh.umu.se:/debian/A* /TD /TR TR TDSwitzerland/TD TD A HREF=ftp://nic.switch.ch/linux/distributions/debian; nic.switch.ch:/linux/distributions/debian/A* /TD /TR TR TDUnited Kingdom/TD TD A HREF=ftp://ftp.mcc.ac.uk/pub/linux/distributions/Debian; ftp.mcc.ac.uk:/pub/linux/distributions/Debian/A /TD /TR TR TDU.S. (was Michigan)/TD TD A HREF=ftp://ftp.debian.org/debian; ftp.debian.org:/debian/A*. Currently, this site has been moved to master.debian.org. /TD /TR TR TDU.S. Florida/TD TD A HREF=ftp://debian.med.miami.edu/debian; debian.med.miami.edu:/debian/A* /TD /TR TR TDU.S. (Michigan?)/TD TD A HREF=ftp://ftp.microworld.net/debian; ftp.microworld.net:/debian/A* /TD /TR TRTDU.S. Oregon/TD TD A HREF=ftp://master.debian.org/debian; master.debian.org:/debian/A* /TD /TR TR TDU.S. New York/TD TD A
Unidentified subject!
According to the Ftape-HOWTO, Travan drives are supported. Hope this helps. Susan Kleinmann [EMAIL PROTECTED]
origin and pronunciation of Debian
I'm working on the FAQ now, and believe it or not, these questions come up (at least in the back of people's minds) fairly often. So, is it Deee'-bian (long e) or Deb'-ian (short e), or something else? Also, where did the word come from and why does the project have that name? TIA, Susan Kleinmann [EMAIL PROTECTED]
pointer to a definition of the debian format
Can someone specify for me a document (either within /usr/doc or on the WWW somewhere) that says exactly what the internals of a .deb file look like? I need (want?) this for the FAQ. Though users might not _need_ to know what the format is, I believe that people would like at least to think that they know what they're dealing with. TIA, Susan Kleinmann [EMAIL PROTECTED]
Bug#3312: installation and startup problems with mathpad
Package: mathpad Version: 0.60-0 1. During the postinst script, the following message appears: Making font directory /usr/X11R6/lib/X11/fonts/100dpi... even though (on my system) that directory already exists. Also, the postinst script reported no error message indicating that it didn't need to make the diretory. 2. When I started up mathpad for the first time, a dialog box appeared asking whether I wanted to make the directory ~/mathpad. Though I clicked on the 'create' button, no such directory was made, and the dialog box didn't go away. These problems may somehow be system-specific, and Erick Branderhorst (the current maintainer) did not see this behaviour on his system. Susan Kleinmann [EMAIL PROTECTED] ii base1.1.0-4Debian Base System Miscellaneous Files ii dpkg1.1.5elf Package maintenance system for Debian GNU/Li
Re: Doubts building a new package...
On Mon, 17 Jun 1996, Rob Browning wrote: [EMAIL PROTECTED] (Emilio Lopes) writes: 2- The logo itself is a gif file. Is it non-free? I can convert it to a jpeg file if needed. Don't know about the other stuff, but for this I'd recommend png, not jpeg. PNG is a bandwidth-conservative, patent-free replacement for GIF (as well as many uses of TIFF). If you want to use it, you might be interested in looking at: http://www.boutell.com/boutell/png/ Cheers, Susan Kleinmann [EMAIL PROTECTED]
Bug#2097: Problem building dvips
Package: dvipsk Version: 5.58f I obtained the sources from ftp.debian.org within the 0.93R6 directory tree, and tried to recompile simply using the command: debian.rules build The first thing that had to be fixed was to go and fetch the kpathsea sources, since one can't compile dvipsk unless one also has those sources on disk, in a particular directory sequence. But the more fundamental problem is that these sources seem to be looking for the new version of the C libraries: ed: can't load library 'libc.so.5' whereas the stock 0.93R6 distribution includes only libc.so.4. (This whole problem came about because I'm trying (but failing) to get color output from dvips. In trying to make it work, I had downloaded several version of color.pro, colordvi.sty, etc., etc., and since none of these worked, I wanted to replace my dvips files with the original ones that came with the dvipsk distribution. At the same time, I wanted to look into the sources to see if there was any info in there to help.) I'm not sure this is a bug that's worth fixing, since ELF is well on its way, but I thought it might at least be worth reporting. Susan Kleinmann [EMAIL PROTECTED]
Bug#2087: ppmforge seg faults
Package: pbmplus Version: 10dec91 Rev 2 ppmforge -stars produces an image which appears to be truncated, then stops with a Segmentation fault error. Example: % ppmforge -stars 0 junk ppmforge: planet: -seed 1357751156 -dimension 2.40 -power 1.20 -mesh 256 ppmforge: -inclination 18 -hour 12 -ice 0.40 -glaciers 0.75 ppmforge: -stars 0 -saturation 125. Segmentation fault This error appears to be independent of the numeric option given to stars. That is, the program also segfaults with -stars 100. If one then attempts to show the output image with 'xv', the result is an error message junk: file appears to be truncated. Despite this, xv succeeds in showing that part of the image which is not truncated (which is 98% of it). Two other options fail in the same way: ppmforge -clouds ppmforge -planet The portion of the image which is being truncated can be illustrated fairly dramatically using: ppmforge -clouds | xv -root -quit - On the other hand, this succeeds: ppmforge -stars Interestingly, this is the only one of the options which produces only black and white output. This system has: === ii base0.93.6 13 Debian Base System Miscellaneous Files ii image 1.2.13 5Linux kernel binary image. ii ldso1.7.10 1The Linux dynamic linker, library and utilities ii libc4.6.27 6The Linux C library. ii pbmplus10dec91 2An extended portable bitmap toolkit. ii xlib 3.1.2 2XFree86 3.1.2 shared libraries. ii xs3 3.1.2 1XFree86 3.1.2 S3 server. ii xv 3.10a 1Interactive image display for the X Window Syst === Susan Kleinmann [EMAIL PROTECTED]
Bug#2072: MANOPT not parsed well?
Package: man
Version: 2.3.10-6
I can't seem to find a way to set MANOPT to change the default pager for
displaying man pages to less -s.
I want to do this:
MANOPT='-P less -s'
But man apparently doesn't parse the double quotes, and treats the -s as a ne
w
man option. The man page for man hints that embedded spaces should be escaped
with \, so I tried:
MANOPT=-P less\\ -s
With this, running man I get:
sh: less -s: command not found
man: command exited with status 32512: /bin/gzip -dc
'/var/catman/cat1/man.1.gz' | { export MAN_PN LESS; MAN_PN='man(1)';
LESS=$LESS\$-Pm\:\$ix8mPm Manual page $MAN_PN ?ltline %lt?L/%L.:byte
%bB?s/%s..?e (END):?pB %pB\\%..; less\ -s; }
Apparently man correctly includes -s as part of the option, but doesn't strip
the \ so it gets passed down into the resulting shell command above. I've
tried many variations on quoting and backslashing, and none seem to give the
desired results. Any ideas, or can this be classified as a bug?
--
Robert Leslie
[EMAIL PROTECTED]
I'm having a problem which may or may not be related to this:
I've been trying to get MH to use less with the '-c' option. That is,
I put these lines into my .mh_profile:
mproc: /usr/bin/less -c
lproc: /usr/bin/less -c
Like Robert Leslie, I've tried many variations of quotes and backslashes.
I get a similar error message:
unable to exec /usr/bin/less -c: No such file or directory
I've also tried using an alias, e.g.,
alias les=/usr/bin/less -c
with the lines in .mh_profile set to:
mproc: les
lproc: les
This produced the same error message.
I had presumed this was an MH problem, and should be sent to mh-users.
Now I'm not so sure. Any suggestions would be appreciated.
Susan Kleinmann
[EMAIL PROTECTED]
Bug#2072: MANOPT not parsed well?
less(1) shows that the program takes options from the $LESS. Thanks for the hint. Setting LESS to -c makes less work the way I wanted within MH. (As always it pays to RTM, just couldn't resist :) Yes, you're right. But the problem (and I think this is an interesting problem) is to figure out what M to R. For example, at first I thought the problem I had was with MH (which involves a ton of reading); then I thought it was with bash (another ton). It just didn't occur to me that the solution was within less, whose manual page is itself about 20 reasonably dense pages long. What's needed is a smart search engine, i.e., one that's matched to the various documentation resources that exist. Hope springs eternal. Thanks again! Susan Kleinmann [EMAIL PROTECTED]
Bug#2036: Printer stuck #2.
Eddie Maddox wrote: 1. I didn't have lp installed since the installation parameter screen for this said line printer, not just printer. Having been a computer operator in times past I KNOW my little HP Deskjet 500 is NO line printer. I agree that the use of the term line printer is an anachronism. It would probably be impossible to eradicate it from *nix internal use, but if it's possible to remove it from any installation scripts, I think that would be a good idea. Susan Kleinmann [EMAIL PROTECTED]
Bug#2002: Missing manpages
Before I knew about the Free-BSD man pages, I started to build man pages for the diff utilities from the diff info file and from their usage statements. In particular, I had made a man page for 'cmp' that way. When you (Jeff) pointed out the existence of the FreeBSD page, I incorporated information that was in it but which had previously been unclear (at least to me) from the info and usage statements. The FreeBSD page for 'cmp' is now outdated, since 'cmp' supports more flags and handles some arguments in different ways; it would not be surprising to find out that the same was true for the other diff utilities. Therefore, my current plan is to finish building revised man pages for diff, sdiff, and diff3 using all of the info mentioned by everyone; then I plan to upload it to Paul Eggert, to distribute with the code if that's what he'd like to do. Paul is absolutely right that it is a pain to maintain several information resources, so I'm not s sure he'll want to incorporate this in a release. Meantime, I'm writing to Jordan Hubbard, to determine whether it's OK to use any text they'd written in their pages. Susan Kleinmann [EMAIL PROTECTED]
Bug#1790: latex2rtf can't find fonts.cfg
Package: latex2rtf Version: 1.1-0 This command line: latex2rtf file.tex file.rtf yields this error message: latex2rtf: ERROR: cannot open file 'fonts.cfg' even when fonts.cfg is in the right place: bash# ls -l /usr/lib/latex2rtf total 8 -rw-r--r-- 1 root root 3491 Nov 1 14:23 direct.cfg -rw-r--r-- 1 root root 1201 Nov 1 14:23 fonts.cfg -rw-r--r-- 1 root root 1349 Nov 1 14:23 ignore.cfg This is probably a bug only in the sense that it doesn't force the new user to do the right thing, whatever that is. Susan Kleinmann [EMAIL PROTECTED] ii dpkg 1.0.5 0Package maintenance system for Debian GNU/Linux ii source 1.2.13 5Linux kernel source. ii base0.93.6 10 Debian Base System Miscellaneous Files ===
man pages for accouting package
I am planning on adapting the information in the /usr/doc/accounting texinfo page to a series of `man' pages in the acct package. Some might argue that this is redundant and therefore wastes disk space and introduces the chance for errors. Others might argue that this is a useful complement to the remainder of the man page documentation. If you have a strong preference, please express it. Susan Kleinmann [EMAIL PROTECTED]
Bug#1773: duplicate Keystroke docs
Package: doc Version: 0.93-5 There are nearly duplicate entries for the Keystroke HOWTO in /usr/doc/HOWTO and /usr/doc/HOWTO/mini. The file in /usr/doc/HOWTO has HOWTO mis-spelled in its first line (HOWTU), whereas the file in /usr/doc/HOWTO/mini has no such problem. Two fixes are needed: HOWTO should be spelled correctly in the document /usr/doc/HOWTO/Keystroke-HOWTO, and the file /usr/doc/HOWTO/mini/Keystroke should be eliminated. Susan Kleinmann (P.S. I wish there were a mailing list for micro-bugs.) ii source 1.2.13 4Linux kernel source. ii dpkg 1.0.5 0Package maintenance system for Debian GNU/Linux
Bug#1736: `atrun' man page: -l is optional
Pacakage: at Version: 2.8a-2 In atrun's manpage, the SYNOPSIS line reads: atrun [-l load_avg] [-d] I believe it should read: atrun [[-l] load_avg] [-d] I am using dpkg 1.0.3-0, kernel 1.2.13-4, and libc 4.6.27-6. Susan Kleinmann [EMAIL PROTECTED]
