Re: ＫＤＥ日本語パッケージについて
Ben Lutgens wrote:- > On Sat, Aug 21, 1999 at 02:28:22PM +0900, Shinichi Miyazawa wrote: > > Ｄｅｂｉａｎ用のＫＤＥの日本語化されてパッケージで、ｓｌｉｎｋ版は有りま > > せんか？ > > 以下ののｆｔｐサイトにｓｌｉｎｋのフォルダーは有りますが、空で実態が有り > > ません。 > > ftp://kde.tdyc.com/pub/kde/debian/dists > Content-Description: Shinichi Miyazawa のカード > > Wow! wtf is this! lol!!! It's a standard Japanese message, and I can read it perfectly in Mutt . He's asking if there's a Debian Slink version of Japanese KDE, mentioning that at the above ftp site there is a directory for it, but the directory is empty. Any ideas? Neil.
Re: Please help a poor gaijin!
Stephen Pitts wrote:- > Hi! I'm trying to figure out what level of Japanese input support exists > in Debian. Not much to my knowledge, unless you install Debian-jp. > Ideally, I'd like to be able to receive messages written in > Unicode with mutt (via the mutt-ja package), and type messages/documents > in kana/kanji with vim and have them stored as Unicode. I don't think you mean Unicode - I don't think mail is sent in Unicode. Most use EUC encoding with iso-220-jp. I have mutt set up to read and send Japanese. To do this, you need at a minimum the following (there may be something I've omitted): a) Japanese console fonts, and a Japanese console. konfont is the package for the fonts, and kon / kterm (say) for the virtual console / X terminal emulator. b) Japanese patches for mutt. See http://home.sprintmail.com/~kikutani/mutt-e.html These patches apply almost cleanly to the debian slink source (this is the route I took). c) Setting up Japanese input is quite tricky. I use canna, with standard slink emacs20. The various packages for this come with Debian-jp. I haven't got the printing working yet, but I've hardly tried. I think it's much easier than the setting up any of the above. A Japanese Debian book I've picked up points to a package called escpf at http://www.debian.or.jp/prospective-packages.html which I haven't downloaded yet. > Being able to print > Japanese text would also be really nice. I'm a bit overwhelmed by how > all of the packages work together. Do I need kinput2, canna, mutt-ja, > and jvim-canna? Will I need some type of kana/kanji dictionary? Do I > need to 'enable' kinput somehow, perhaps by editing my XF86Config? With canna, one way to do it is to select a key which when pressed with the CTRL key, toggles English / Japanese input. I use the Windows key right by the CTRL key on my standard English keyboard. xmodmap -e 'keycode 115 = Kanji' in .xsession achieves this. emacs20 has its own interface to canna, using CTRL + o. Editing files in Japanese works beautifully, and it's even possible to give files Japanese names. > The English documentation is a little confusing and my Japanese fluency > isn't enough to begin to try and understand the stuff written in > Japanese. Thanks to any and all who can shed light on this for me. I found Craig Oda's stuff very informative about the issues behind a Japanese setup:- http://tlug.linux.or.jp/~craigoda/writings/linux-nihongo/ but the bit about emacs and canna didn't work for me; luckily what was detailed in my book was a tad simpler and did work. If you get close, I may be able to help with the finishing touches. 頑張ってね。 Neil.
[NeilB@earthling.net: Re: Please help a poor gaijin!]
Miles Bader wrote:- > > Thing's are quite a bit simpler if you just use emacs-20's built-in > input methods -- they basically just work out of the box (as long as > your emacs was compiled with the leim support enabled). Just do: > > (set-language-environment "Japanese") > > Then type C-\ (toggle-input-method) and begin typing! > > Newer versions of (stable-release) gnus actually are harder to make work > correctly than older versions, but I have a small set of hacks that I > put in my .gnus file, which make everything work pretty flawlessly. Hi Miles, Thanks for the reminder, but I had tried that (and forgotten about it) because I found it so frustrating. It doesn't accept "standard" input methods - to type in "Nihongo" I habitually type a double n like so:- nihonngo which is the way kanna and everything else I've used accept it, in addition to accepting the single n. Emacs insists on a single n. Then trying to undo mistakes gets even more frustrating - it seems to remember single keystrokes that have no representation onscreen when you backspace, and only appear when you type the next character. I found it a constant battle to use, whereas canna is just very smooth. Also, Emacs's dictionary seems very limited - nihongo is not recognised as a word, but gets split as Nihon and hiragana go! I may be using it wrongly, of course. Neil.
Two ethernet cards
I have 2 PCI Planex ethernet cards in my machine, the first uses "tulip.c" as its driver, and the second uses "de4x5.c". When I inserted the second card today and rebuilt my kernel (no modules), during boot-time initialization the tulip driver initializes successfully as eth0, but the second driver fails with:- eth1: region already allocated at 0xe400 because that I/O port was grabbed by the tulip driver. Is there a way around this, or can I not use these two cards in the same machine? I can't see any jumpers or switches that might change either I/O port on the cards. Cheers, Neil.
Re: Two ethernet cards
Peter Iannarelli wrote:- > Hello Neil: > > via the card configuration tool from the manufacturer, > you should be able to change the IRQ and I/O mapping > for the card(s) in question. > > Peter Hi Peter, Literally all that came in the box was a manual, a cable for Wake-On-LAN, and the card itself. Could you elaborate? Neil.
Re: Two ethernet cards
George Bonser wrote:- > > If both cards are PCI there really should not be a problem. Try swapping the > order of the cards in the PCI bus and see if that helps. Hmmm, yes they are both PCI - swapping the card order made no difference - the same card gets recognised first, and the second one ignored. Does this indicate a hardware problem? > Just curious, why are you using the different drivers for the two cards? Can't > they both use the same one? Yes, I now think (from the manufacturer's website) that they can both use the tulip driver, but it seems to just ignore the second one (presumably because they share I/O address). Using the other Dec driver, with which it seems compatible too, was what actually threw up an error message, making me think that was the true driver. Neil.
Forcing a core dump?
I have found a bug in some software that causes it to segfault. However, there is no core file to use with gdb. Why is a core file not generated / how do I force one? Thanks, Neil.
libslang problem after upgrade to potato
After doing a distribution upgrade from slink to potato, I keep getting errors about /usr/lib/libslang.so not existing. Doing "ldconfig /usr/lib" results in ldconfig: warning: can't open /usr/lib/libslang.so (No such file or directory), skipping ldconfig: warning: can't open /usr/lib/libslang.so (No such file or directory), skipping The libslang contents of /usr/lib are as follows:- lrwxrwxrwx 1 root root 20 Nov 3 00:31 libslang-ja.so.1 -> libslang-ja.so.1.2.2 -rw-r--r-- 1 root root 242144 Jun 1 04:37 libslang-ja.so.1.2.2 -rw-r--r-- 1 root root 335618 Aug 13 13:20 libslang.a lrwxrwxrwx 1 root root 22 Aug 13 13:23 libslang.so -> /lib/libslang.so.1.2.2 The problem is that I have a dangling symlink here. 1) Should I have an English libslang here? Or should I delete the symlink? 2) I have one in /lib as so lrwxrwxrwx 1 root root 17 Nov 2 23:57 libslang.so.1 -> libslang.so.1.3.9 -rw-r--r-- 1 root root 328428 Oct 14 08:59 libslang.so.1.3.9 Should this be here? Should it be in /usr/lib instead, or symlinked from there? If I try an "apt-get install slang1" it tells me I already have the newest version, so I'm not quite sure where to go from here. I daren't reboot without sorting this out as libslang is in the base distribution, and quite a few upgrades did not go smoothly because of this libslang problem. Any advice appreciated. Neil.
Soundblaster installation problem.
Hi, I've been trying to set up a Soundblaster ISA card under Debian. The card (AWE64 I think) works fine under Redhat with sndconfig, albeit with slightly different hardware. I have 4 IDE devices installed, 2 PCI NICs and a PCI ATI graphics card. I'm using a modular 2.2.13 kernel. If I do the following:- modprobe sound insmod uart401 insmod sb io=0x220 irq=5 dma=1 dma16=5 mpu_io=0x330 the kernel logs give Nov 27 23:04:39 monkey kernel: Soundblaster audio driver Copyright (C) by Hannu Savolainen 1993-1996 Nov 27 23:04:39 monkey kernel: SB 170.170 detected OK (220) Nov 27 23:04:39 monkey kernel: sb: Interrupt test on IRQ5 failed - Probable IRQ conflict so, as it says, there may be an IRQ conflict. However, I don't think so. /proc/interrupts gives 0: 60622 XT-PIC timer 1: 3295 XT-PIC keyboard 2: 0 XT-PIC cascade 5: 0 XT-PIC soundblaster 10: 20 XT-PIC eth1 11:104 XT-PIC eth0 12: 6070 XT-PIC PS/2 Mouse 13: 1 XT-PIC fpu 14: 5341 XT-PIC ide0 15: 34 XT-PIC ide1 so all seems fine. What am I missing? Thanks in advance for any hints. Neil.
Re: apt-get dist-upgrade -u on testing wants to uninstall 189 packages
Anthony Towns wrote:- > The problem here is some weird bug in apt's analysis of which packages are > mor important: it's deciding to remove perl-5.005 and keep data-dumper, > instead of the other way around, for some reason. > > Easiest way to avoid this nastiness is to apt-get remove data-dumper > first. Is there an easy way here to see which package is at the root of the chain of dependencies and causing problems? I found it by trial-and-error, but it can be terribly opaque. Neil.
Re: wvdial doesn't connect
Henning Otte wrote:- > > What kind of modem? Winmodems aren't supported. (In > > general). > > How do I find out, if I've got a winmodem? > (It's called 'ASKEY 56k Plug&Play Modem', it supports Rockwell K56flex, > but not V.90) Is it internal or external? You should have no probs with an external one. Internal ones are more likely to give problems. Neil.
Re: telnet doesn't find localhost
Jeronimo Pellegrini wrote:- > socrates tmp$ telnet localhost 80 > telnet: could not resolve localhost/80: Temporary failure in name resolution FWIW I have exactly the same problem. Just started recently. I have a dial-up connection too. The problem only exists when I'm not connected to the internet. As soon as pppd is up, the problem vanishes, and it reappears when I disconnect. I haven't a clue either. Neil.
Perl's default include path
I get these messages trying to install recent versions of debconf: monkey:/usr/src/gcc/build# dpkg -i /var/cache/apt/archives/debconf_0.5.22_i386.deb (Reading database ... 36741 files and directories currently installed.) Preparing to replace debconf 0.5.25 (using .../debconf_0.5.22_i386.deb) ... Unpacking replacement debconf ... Setting up debconf (0.5.22) ... Can't locate lib.pm in @INC (@INC contains: /usr/lib/perl5/5.005/i386-linux /usr/lib/perl5/5.005 /usr/local/lib/site_perl/i386-linux /usr/local/lib/site_perl /usr/lib/perl5 .) at /usr/share/debconf/frontend line 21. I have 5.6 installed as the default, but as you can see it's still looking in the 5.005 directory for stuff. How do I change the @INC default path to the correct one? Neil.
Re: Neighbor table overflow
Hi Alex, This is almost always caused by not having a loopback interface. If you haven't, make sure the first interface you configure is lo. Neil. Alex McCool wrote:- > Helpp > > What is neighor table overflow?? > > I'm getting this after changing the IP addr of a machine > > I changed /etc/network/interface from xxx.xxx.xxx.17 to xxx.xxx.xxx.16 > > I had seen this before when I switched a diff machine to a 192.168.100 addr, > but I never found a solution > > Alex
Digital cameras and Linux
Hi, I'm considering buying a digital camera or videocam, but am concerned about being able to download the JPEG images to Linux. Of course, the cameras come with a serial cable and software for downloading the images to Windows. Does anyone know if these cameras simply pass the JPEG data down the serial line, or is there some special camera-specific protocol they use (rendering it useless without special software)? In the former case, how would I capture the data coming down the serial line? Thanks for any information or advice, Neil.
Re: Digital cameras and Linux
Hi all, Thanks for the very useful info. With this, and having found gphoto.org and noticing gphoto comes with Debian, I have everything I need to buy a camera that works with Linux. Cheers, Neil.
debconf error
I'm running the latest potato, but if upgrade something to the latest and greatest, I tend to get the following error: debconf: failed to initialize Dialog frontend debconf: falling back to Text frontend How can I fix this? Neil.
Package pools and Japanese Debian problems
I use Japanese Debian, and ever since pools were introduced I've not been able to update many packages at all without download errors, and new revisions seem to be far less frequent. Here is a typical apt-get example:- monkey:/home/neil# apt-get install modutils Reading Package Lists... Done Building Dependency Tree... Done 1 packages upgraded, 0 newly installed, 0 to remove and 19 not upgraded. Need to get 173kB of archives. After unpacking 12.3kB will be used. Get:1 ftp://ftp.debian.or.jp woody/main modutils 2.3.23-2 [173kB] Err ftp://ftp.debian.or.jp woody/main modutils 2.3.23-2 Unable to fetch file, server said '/debian-jp/pool/main/m/modutils/modutils_2.3.23-2_i386.deb: No such file or directory ' Get:2 ftp://ftp.debian.or.jp woody/main modutils 2.3.23-2 [173kB] Err ftp://ftp.debian.or.jp woody/main modutils 2.3.23-2 Unable to fetch file, server said '/debian-jp/pool/main/m/modutils/modutils_2.3.23-2_i386.deb: No such file or directory ' Failed to fetch ftp://ftp.debian.or.jp/debian-jp/pool/main/m/modutils/modutils_2.3.23-2_i386.deb Unable to fetch file, server said '/debian-jp/pool/main/m/modutils/modutils_2.3.23-2_i386.deb: No such file or directory ' E: Unable to fetch some archives, maybe try with --fix-missing? monkey:/home/neil# Note that apt-get update works fine. Here's my sources.list:- deb ftp://ftp.debian.or.jp/debian-jp woody main contrib non-free deb ftp://ftp.debian.or.jp/debian-jp woody main contrib non-free deb ftp://ftp.debian.or.jp/debian-non-US woody/non-US main contrib non-free deb-src ftp://ftp.debian.or.jp/debian-jp woody-jp main contrib non-free deb-src ftp://ftp.debian.or.jp/debian-jp woody main contrib non-free deb-src ftp://ftp.debian.or.jp/debian-non-US woody/non-US main contrib non-free What's wrong? TIA, Neil.
Mutt and viewing attachments simultaneously
When I run Mutt and I receive a mail with 3 attachments, I see something like: [-- Attachment #1 --] [-- Type: text/plain, Encoding: 7bit, Size: 0.4K --] [blah blah blah] [-- Attachment #2: test-cpp-if-2.patch --] [-- Type: text/x-patch, Encoding: 7bit, Size: 0.8K --] [-- text/x-patch is unsupported (use 'v' to view this part) --] [-- Attachment #3 --] [-- Type: text/plain, Encoding: 7bit, Size: 0.3K --] How can I view all 3 attachments at once on the screen? The above is particularly annoying when replying to a message - I want to quote from the attachments too. I couldn't find anything in the manual apart from auto_view in the muttrc, but that seems to be about converting non-text to text. Here I just want to see all the text attachments at once. Thanks, Neil.
Installing xlibs wants to remove xdm and other stuff
This is what apt-get -s gives:- monkey:/home/neil# apt-get -s dist-upgrade Reading Package Lists... Done Building Dependency Tree... Done Calculating Upgrade... Done The following packages will be REMOVED: intlfonts-japanese intlfonts-japanese-big xbase-clients xdm xfonts-100dpi xfonts-75dpi xfonts-base xfonts-cjk xfonts-scalable 2 packages upgraded, 0 newly installed, 9 to remove and 0 not upgraded. Remv intlfonts-japanese Remv intlfonts-japanese-big Remv xfonts-scalable Remv xfonts-cjk Remv xfonts-100dpi Remv xdm Remv xfonts-base Remv xfonts-75dpi Remv xbase-clients Inst xlibs [shellutils on hurd] Inst xlibs-dev [shellutils on hurd] Conf xlibs Conf xlibs-dev monkey:/home/neil Why does it want to remove this stuff, including xdm? I need to keep the various fonts; how can I resolve this? Thanks, Neil.
Re: 質問です 。
If you have a (fast) internet connection you should be OK. You can download two or three floppies from Debian's web site, and that is all you need to install. The rest of the distribution is downloaded from the internet. If you haven't got a fast connection, it can take a while. But I did it on a 56K modem... Neil. [Sorry, my Japanese input is broken at present] Nakatani Yoshitaka wrote:- > 古いパソコンにlinuxをインストールしようとしている者です。 > 私の古いパソコンにはCD-ROMドライブが無く、フロッピーディスクドライブしか無いのです。 > debianはフロッピーのみでインストールできるという話を耳にしたのですが、 > それは本当なのでしょうか。 > その古いパソコンというのは DynaBook GT475 031CS です。 > > よろしくお願いします。 > > > > > -- > To UNSUBSCRIBE, email to [EMAIL PROTECTED] > with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] >
Gnome panel + sawfish = very slow ?
I installed sawfish about 3 wks ago, and have installed a couple of revisions of gnome panel since then. However, if I bring up panel after starting sawfish, dragging a window becomes jerky and slow. I don't remember this behaviour in WindowMaker. Somtimes, but not always, panel complains about Sawfish not being Gnome compliant. So I suspect it is trying to make up for the supposed non-compliance by hooking into a lot of events it wouldn't otherwise. Other panel-related things are sluggish too, like menu drawing and tracing them with the mouse. Without panel, everything is snappy, but Sawfish is a bit limited :-( I know Sawfish is compliant, so what could be wrong? I'm running the latest debian unstable, XFree 4.1. Neil.
Re: Gnome panel + sawfish = very slow ?
First, thanks to Johnny and Stephan. > You can start Gnome in two ways in your .xsession: > > start the panel > start a windowmanager > > This will work, but will not be session managed. > > The smarter way is: > > gnome-session > > Gnome-session (and do NOT start a window manager manually before or after) > > Gnome-session will automatically start the window manager noted in the > environment variable WINDOW_MANAGER (note the underscore) > And this way you will recieve session management. > > I suspect that the sluggishness of your setup comes from 2 sawfish > biting at eachother. > - The one you started manually > - The one gnome-session starts > > They probably don't like eachother. I don't have gnome-session installed at all. I was just running panel from an rxvt window. I'll install gnome-session and see if it helps. If it does, I tend to think it is a bug that running panel from an rxvt causes such sluggishness; it really is unusable. If it doesn't help, well something else is wrong :-) Neil.
Re: Gnome panel + sawfish = very slow ?
Jim McCloskey wrote:- > It works well for me. > > I'm running X 4.0.3 from testing, and version 0.99 of sawfish from > testing. Version 1.0.55 of gnome-panel, also from testing, Hi Jim, It's a little weird. If I start panel in my .xsession before sawfish, everything is snappy like you'd expect. [My test of snappiness is dragging a window across the screen: whether it jerks across or drags smoothly]. If I instead start panel from an xterm, it seems to make window dragging jerky and things are generally sluggish. I don't know why... So, the only remaining issue is where does Gnome get its window manager list from? It only lists Windowmaker, and if I try and add sawfish it segfaults. I would have thought this should be handled by the sawfish install scripts, but maybe not. Thanks, Neil.
Re: bug: "neighbor table overflow" message
Josef Dalcolmo wrote:- > Hello, > > I am getting a "neighbor table overflow message" during installation of > woody from 2001-11-09 and also afterwards > Does anyone know where this message comes from? This usually means you are not configuring the loopback interface, or not configuring it first. Neil.
tty1-6 screwed up
I just finished playing xmame in full screen mode, and coming out of full screen mode messed something up. X died, and gdm was respawned. However, interestingly, the virtual consoles 1-6 are now screwed; if I switch to them I see a mess of mostly green and black in graphics mode. Everything still works; I can log in and out there, but I can't see anything useful. What's gone wrong here, and how can I reset them without rebooting? Thanks, Neil.
Re: tty1-6 screwed up
Shaul Karl wrote:- > I do not know what gone wrong. > > As for having a vt reset without rebooting, try issuing a reset command > (reset) after logging in. Ignore the fact that you might not see this > command while you keying it in. Thanks, but I'd already tried that. No luck; it stays in graphics mode. Like someone else said, it seems to confuse the graphics card (or some part of the kernel, because I noticed when rebooting 8-) that the kernel manages to cure it before shutdown is fully complete). It happens when I come out of full-screen XMAME under the sawfish WM; sawfish (I presume) dies; I get thrown back to gdm, and the 6 ttys are hosed. Can anyone else reproduce this? It's fully reproducable for me. Neil.
Re: t0rn v8
Stephen Gran wrote:- > Hello all, > While running chkrootkit, I got this message (among a bunch of others > saying nothing found): > > Searching for t0rn's default files and dirs... nothing found > Searching for t0rn's v8 defaults... Possible t0rn v8 (or variation) > rootkit installed > > and > > Searching for suspicious files and dirs, it may take a while... > /usr/lib/xemacs-21.4.1/lisp/.cvsignore Have you looked inside this file? It's on my system too. It's a harmless text file. Neil.
Re: t0rn v8
Stephen Gran wrote:- > Sorry, bad form to have to reply rather than include the info in th > original message, but hindsight and all that. A few things I have > done to try to see if t0rn is in fact present: > lsof|grep LISTEN: > portmap273 root4u IPv4303 TCP *:sunrpc > (LISTEN) > rpc.statd 277 root5u IPv4418 TCP *:32768 > (LISTEN) > inetd 286 root6u IPv4424 TCP *:smtp > (LISTEN) > inetd 286 root7u IPv4425 TCP *:auth > (LISTEN) > cupsd 289 root0u IPv4692 TCP *:ipp > (LISTEN) > sshd 306 root3u IPv4566 TCP *:ssh > (LISTEN) > Sorry about the bad wrap ; ) > > and lsof|grep -i t0rn: > No results. > > nmap localhost: > Starting nmap V. 2.54BETA30 ( www.insecure.org/nmap/ ) > Interesting ports on localhost (127.0.0.1): > (The 1544 ports scanned but not shown below are in state: closed) > Port State Service > 22/tcp openssh > 25/tcp opensmtp > 111/tcpopensunrpc > 113/tcpopenauth > 631/tcpopencups But what about to external hosts? Are they open or closed by your firewall? I'd be particularly concerned about sunrps and cups, and only allow access to and from specific IP addresses. If they are visible externally, you should investigate further. If you don't already, I'd suggest you run one of those scripts that filters and mails your logs to you every 1 hour or so. Reducing the background noise from legitimate stuff is the most tedious thing there though. For my machine I have: Starting nmap V. 2.54BETA30 ( www.insecure.org/nmap/ ) Interesting ports on localhost (127.0.0.1): (The 1540 ports scanned but not shown below are in state: closed) Port State Service 13/tcp opendaytime 22/tcp openssh 25/tcp opensmtp 37/tcp opentime 110/tcpopenpop-3 139/tcpopennetbios-ssn 631/tcpopencups 2401/tcp opencvspserver 22273/tcp openwnn6 but only SSH and SMTP are visible outside my LAN (as verified by various firewall testing web sites). Neil.
Unavailable packages question
This is the output of a couple of dpkg commands on my system:- monkey:/home/neil# dpkg -l "mutt*" Desired=Unknown/Install/Remove/Purge/Hold | Status=Not/Installed/Config-files/Unpacked/Failed-config/Half-installed |/ Err?=(none)/Hold/Reinst-required/X=both-problems (Status,Err: uppercase=bad) ||/ Name Version Description +++--- ii mutt 1.3.15-2 Text-based mailreader supporting MIME, GPG, PGP and threading. un mutt-i (no description available) rc mutt-ja 0.95.4i.jp2-2.1 Text-based mailreader for Japanese. un muttzilla (no description available) monkey:/home/neil# dpkg -p mutt-i Package `mutt-i' is not available. Use dpkg --info (= dpkg-deb --info) to examine archive files, and dpkg --contents (= dpkg-deb --contents) to list their contents. So, why is "mutt-i" still listed if it is not available? How can I flush these "unavailable" packages so they don't clutter up the packaging system? dpkg --forget-old-unavail appears to be what I want, but it does nothing from what I can surmise. Thanks, Neil.
Re: Unavailable packages question
Sean 'Shaleh' Perry wrote:- > # dpkg --get-selections|sed -e 's/deinstall$/purge/'|dpkg --set-selections > # apt-get dselect-upgrade > > what this does is find packages that were removed but not purged and purges > them. That doesn't appear to do what I want. The mutt-i package is not flagged "deinstall"; in fact only 2 or 3 unrelated packages are. It has never been installed on my system. Neil.
Re: GCC packages fighting amongst themselves?
> (Ironically, the reason I want to get rid of gcc-2.95 entirely and use > only gcc-3.x is that I'm a GCC maintainer, and want to do lots of testing > with the 3.x series. If the solution to this dependancy mess involves > destroying gcc-doc, that's fine with me; I have many copies of the GCC > manual already. :-) Takes two to tango 8-). I just let CVS GCC install itself in usr/local after a "make install", and then it's first on the path anyway. Neil.
C locale and error messages
If I open an rxvt window or xterm, which for me defaults to LANG=C, and do echo < foo instead of getting bash: foo: No such file or directory I get bash: foo: ?? If I set LANG to en_GB, it doesn't help. If I set LC_MESSAGES to en_GB, I then get the proper error message. If I then unset LC_MESSAGES, I continue to get a proper error message (so maybe a caching issue?) At the Linux console, I get a proper error message regardless. What is going on here? I'm using Debian unstable, but this issue has existed for a long time and is started to annoy me. Thanks, Neil. FWIW a typical "env" when getting the '?' is as follows: PWD=/home/neil COLORFGBG=15;0 WINDOWID=35651586 PAGER=jless HOSTNAME=monkey LESSCLOSE=/usr/bin/lesspipe '%s' '%s' LESSOPEN=| /usr/bin/lesspipe '%s' LV=-Or LESS=-M USER=neil MACHTYPE=i386-pc-linux-gnu [EMAIL PROTECTED] EDITOR=vi LANG=C COLORTERM=rxvt DISPLAY=:0.0 SSH_AGENT_PID=9701 SHLVL=1 LOGNAME=neil PERL_BADLANG=0 SHELL=/bin/bash HOSTTYPE=i386 OSTYPE=linux-gnu SSH_AUTH_SOCK=/tmp/ssh-XX0QUFpi/agent.9667 TERM=rxvt HOME=/home/neil PATH=/home/neil/bin:/usr/local/bin:/usr/bin:/bin:/usr/bin/X11:/usr/games JLESSCHARSET=latin1 OUTPUT_CHARSET=ISO8859-1 _=/usr/bin/env
Re: line numbers in code
Romain Lerallut wrote:- > You can run *any* text through cpp (not just C program sources, I use it > for my Fortran codes:) That's not true, certainly in GCC 3.0. Neil.
Re: line numbers in code
Romain Lerallut wrote:- > > Romain Lerallut wrote:- > > > > > You can run *any* text through cpp (not just C program sources, I use it > > > for my Fortran codes:) > > > > That's not true, certainly in GCC 3.0. > > > > Neil. > > > h: > > echo '__LINE__' | cpp-3.0 -P > 1 > > looks like you can ( with my Fortran codes too!) > :-) I don't understand what your example proves. We made an effort to help Fortran and assembler. Although in general you cannot preprocess even Fortran and assembler. With Python, for example, you'd lose all your tabs. Basically, if you don't care about whitespace preservation, and your file lexes like C (in that, e.g. all charconsts and strings are closed), you can probably get away with it. You're advised not to rely too heavily on it though - 3.0 is more strict than 2.95 and will not accept some stuff that 2.95 did. Neil.
