Re: [Gulp] 2 links na mesma maquina
Olá! Em Ter 14 Jan 2003 10:32, Rodrigo Gruppelli escreveu: Eu quero que o default route da maquina seja pelo link adsl (eth1). Mas setando a rota default pro eth1, tudo que entrar pela ppp0 nao consegue retornar, pq vai tentar retornar pelo eth1. http://www.lartc.org/ :) -- Fábio Olivé Leite ([EMAIL PROTECTED]) Pesquisa Desenvolvimento GoDigital Marketing de Precisão (http://www.godigital.com.br/) Rua Antônio das Chagas, nº 161 - Chácara Santo Antônio - São Paulo/SP 04714-000 - Fone: (11) 5187 1516 - Fax: (11) 5181 9482 Av Padre Cacique, 320 / Bloco A / 2º Andar - Porto Alegre/RS 90810-240 - Fone: (51) 3230 4400 (http://www.marketingdeprecisao.com.br/)-Visite!
Re: [gulp] Re: [OFF-TOPIC] (Pelo menos p/ a d-u-p :) SMB/CIFS sobre Ip Masquerade
Oi Fabiano, On Fri, Oct 05, 2001 at 09:14:02AM -0300, Fabiano F. Siqueira wrote: |o|- embora existam mais de quatro bilh?es de endere?os ipv4, sua distribui??o |o| em classes A, B, C e etc ? muito perdul?ria por uma s?rie de motivos |o| sobre os quais qualquer livro de redes poderia encher p?ginas, e ent?o |o| n?o h? endere?os IP suficientes para que qualquer empresa tenha em todas |o| as suas m?quinas IPs v?lidos; Mas quem deu essa ideia a voce? Qualquer aula de redes, e experiência própria com internet acadêmica e comercial. Não entendi exatamente se achas estranha a idéia da distribuição perdulária de IPs ou eu ter colocado isto como uma motivação por trás do masquerading. |o|- intranets podem querer ser apenas clientes da Internet, e portanto n?o |o| precisam de IPs v?lidos em todas as m?quinas se houver masquerading; E pq nao se pode usar uma classe A reservada pela RFC 1918 fazendo roteamento entre suas redes internas, E QDO FOR sair para internet faz um NAT Dinamico. MAIS CONHECIDO, VULGARMENTE, POR MASQUERADING Cara, foi exatamente isto que eu escrevi ali em cima, leia de novo. :) |o|- mesmo quando se tem servidores e etc para expor ? Internet, pode ser mais |o| barato comprar de um provedor apenas uma subnet /29 e usar masquerading e |o| redirecionamento de servi?os para IPs inv?lidos da intranet; Concordo com voce isso vai de caso a caso. Sim, de repente tu podes colocar centenas de máquinas e serviços atrás de um só IP válido, mas depende do tráfego. :) |o|- ? mais seguro, pois assim s? quem se exp?e diretamente ? Internet ? o |o| firewall que est? fazendo o masquerading; Discordo voce pode ter n maneiras para proteger sua rede. E daí? Eu apenas falei que esta era uma das vantagens do masquerade, que protege um pouco mais a rede interna. Ninguém falou que é a única. |o|:) Apenas roteamento se faz quando todo mundo tem IPs v?lidos e fixos. Aonde vc arrumou essa teoria? E onde tu leu que roteadores dos grandes backbones da internet roteiam 10.0.0.0/8 e 192.168.0.0/16? Note que eu não estava dizendo que não se faz roteamento dentro da rede interna, mas sim roteamento direto para a internet. O caso em questão era sobre masquerade, quando se ia colocar uma rede na internet. Aí escrevestes por que não usar apenas roteamento?, e eu respondi que neste caso não rola. Se vais fazer só roteamento para ligar uma intranet na internet, precisas de IPs válidos e registrados por toda a intranet. Tá explicado? Abraço! Fábio -- Fábio Olivé Leite http://www.conectiva.com.br/~olive I drowned in the universal pool of entropy Eris has saved me, and she has set me free A31F CCB3 6BC4 A875 27E7 C2B4 2A74 5690 F127 519B pgp5SK9epAfyn.pgp Description: PGP signature
Re: [gulp] [OFF-TOPIC] (Pelo menos p/ a d-u-p :) SMB/CIFS sobre Ip Masquerade
Olá! Espero que, ainda que não seja assinante, possa postar para a d-u-p. On Wed, Oct 03, 2001 at 10:13:21AM -0300, Rodrigo Gruppelli wrote: gente, voces sabe de alguma coisa sobre o protocolo SMB/CIFS se portar mal sobre Ip Masquerade? Digo isso pq tem um servidor NT squi na UnC de Canoinhas, cujo cliente perde a conexao com o servidor. O Cliente acessa uma base de dados cobol, atravez de mapeamento de rede windows etc.. Detalhe: entre o servidor e o cliente tem 3 roteadores Debian GNU/Linux :) sendo que 2 desses fazem mascaramento... Se ficar muito tempo sem mexer no programa ou fazer alguma coisa, a conexao com o servidfor NT cai.. Aih eh preciso fechar o programa, abrir de novo e se logar... Se demorar muito... isto está com cara de firewall com masquerading baseado nos kernels 2.2. Ele tem um timeout muito mais baixo que o timeout (e portanto o keepalive) do TCP, sendo assim quando uam conexão fica um pouco inativa ele cai direto. O masquerade abandona a conexão, e aí os pacotes não passam mais. Se não me engano o default é 20 minutos. Existem soluções em todos os lados: - diminuir o tcp keepalive no /proc/sys/net/ipv4/tcp_keepalive_time (pode atrapalhar o uso normal do protocolo, mas não deve); - aumentar o tempo de timeout no masquerade (pode ficar ingerenciável se o masquerade não faz connection tracking de verdade); - diminuir o tcp keepalive no servidor, mas aí vai afetar todo mundo (mas não deve, novamente, pois fará apenas com que alguns pacotes inúteis trafeguem com mais freqüência); e a melhor opção de todas: - trocar os firewalls com masquerade por firewalls baseados no kernel 2.4 e o Netfilter/iptables, usando connection tracking. iptables rege! Abraço! Fábio -- Fábio Olivé Leite http://www.conectiva.com.br/~olive I drowned in the universal pool of entropy Eris has saved me, and she has set me free A31F CCB3 6BC4 A875 27E7 C2B4 2A74 5690 F127 519B pgpofGz0PtoiE.pgp Description: PGP signature
Re: [gulp] Re: [OFF-TOPIC] (Pelo menos p/ a d-u-p :) SMB/CIFS sobre Ip Masquerade
Olá! On Wed, Oct 03, 2001 at 12:13:15PM -0300, Fabiano F. Siqueira wrote: Pq e' necessario fazer ip masquerade? Nao se pode fazer apenas roteamento! Sonho meu, sonho meu... ;) Existem muitas razões interessantes para o masquerade: - embora existam mais de quatro bilhões de endereços ipv4, sua distribuição em classes A, B, C e etc é muito perdulária por uma série de motivos sobre os quais qualquer livro de redes poderia encher páginas, e então não há endereços IP suficientes para que qualquer empresa tenha em todas as suas máquinas IPs válidos; - intranets podem querer ser apenas clientes da Internet, e portanto não precisam de IPs válidos em todas as máquinas se houver masquerading; - mesmo quando se tem servidores e etc para expor à Internet, pode ser mais barato comprar de um provedor apenas uma subnet /29 e usar masquerading e redirecionamento de serviços para IPs inválidos da intranet; - é mais seguro, pois assim só quem se expõe diretamente à Internet é o firewall que está fazendo o masquerading; - redes inteiras podem ser conectadas à Internet via linhas dial up graças ao masquerading; - ... :) Apenas roteamento se faz quando todo mundo tem IPs válidos e fixos. Abraço! Fábio -- Fábio Olivé Leite http://www.conectiva.com.br/~olive I drowned in the universal pool of entropy Eris has saved me, and she has set me free A31F CCB3 6BC4 A875 27E7 C2B4 2A74 5690 F127 519B pgpX831P4dYTY.pgp Description: PGP signature
Re: [gulp] GRUB Gráfico
Olá! Tô respondendo tarde assim porque tava de férias... :)) On Mon, Jul 02, 2001 at 10:25:57AM -0300, Rodrigo Gruppelli wrote: Gente. Aquele GRUB gráfico é hack da conectiva ou dá pra fazer com o GRUB normal mesmo? Pq eu tentei e nao deu... :\ O Paulo César, nosso desenvolvedor do XFree e que fuçou no GRUB também, preparou esta modificação para o GRUB suportar modos gráficos simples, porém o pessoal do GRUB não aceitou. Como nós achamos interessante, mantivemos no nosso GRUB. :) -- Fábio Olivé Leite · http://www.conectiva.com.br/~olive ·Never whiss while you pistle, or something like that·
Re: [gulp] GRUB Gráfico
Olá! On Thu, Jul 26, 2001 at 03:17:37PM -0300, Henrique de Moraes Holschuh wrote: Hmm... eu estou interessado. Vou dar uma olhada no patch, e dependendo do que ele implica, posso passar a manter um pacote não-oficial do Grub pro Debian com o patch de vocês. Ótimo! :) Deves encontrar o patch no nosso pacote mais atual, acessando http://snapshot.conectiva.com/ . Afinal de contas, como podemos ter um Grub com supore a japonês sem um fb gráfico? (e haja bloat...) ;-) Bom, aí o interessante vai ser como vocês vão fazer para colocar o menu em Japonês, hehehe. :) Abraço Fábio -- Fábio Olivé Leite · http://www.conectiva.com.br/~olive ·Never whiss while you pistle, or something like that·
filters package
Hi there, I'd like to know what happened to the filters package that was in hamm but is not in slink. It contains some funny text filters like jive. Please Cc: me as I'm not on the list. Fabio ( Fábio Olivé Leite[EMAIL PROTECTED] ) ( NEW -*- http://descartes.ucpel.tche.br/~olive -*- NEW ) ( Linux - Distributed Systems - Fault Tolerance - Security - /etc ) (BC 50 7F 7A B9 2E 0A 26 91 8A D1 C0 B1 E4 DA A4)
sys_ptrace
Hi there, I need help on understanding and using the ptrace syscall. I've been rtfming for some days, and still can't produce even a simple program that will print ops! everytime the traced process does a syscall. This is obviously not my objective, I actually need to code a special purpose debugger that will trace a process and automagically display a copy of it's state. gdb or another general purpose debuggers are out of the question, I just want to code something that will fork, TRACEME, exec the target program and stop it every once in a while to collect some info. Pointers anyone? Example code anyone? CC me, I'm not on the list. Thanks! Fabio ( Fábio Olivé Leite[EMAIL PROTECTED] ) ( NEW -*- http://descartes.ucpel.tche.br/~olive -*- NEW ) ( Linux - Distributed Systems - Fault Tolerance - Security - /etc ) (BC 50 7F 7A B9 2E 0A 26 91 8A D1 C0 B1 E4 DA A4)
Re: I'm not giving up: backpack CD-ROM drive.
Ola! ] 2) Use a null modem cable to connect to a desktop and piggy back ride on ] that systems CD-ROM. ] Q: I don't know if the Linux base system supports null modem connections. It does since a long time ago, as I installed myself my first debian system on an old notebook with no cdrom or any pcmcia cards (like net and modem). My only choice was to use a null modem cable to set up a PPP connection between my desktop (which used redhat at the time) and NFS export the /cdrom directory. The install took over 12 hours, but it worked. And it has survived upgrades from debian 1.2 to 1.3, to 2.0 and 2.1. Try it out and mail me if necessary. I'm not on the list anymore (too much traffic). Fabio ( Fábio Olivé Leite[EMAIL PROTECTED] ) ( NEW -*- http://descartes.ucpel.tche.br/~olive -*- NEW ) ( Linux - Distributed Systems - Fault Tolerance - Security - /etc ) (BC 50 7F 7A B9 2E 0A 26 91 8A D1 C0 B1 E4 DA A4)
Re: 3 button mouse
Ola! ] I just bought a new 3 button mouse for the convenience of better cut and ] paste, ] and find that I can't seem to make it work. When I first got it, I just Most likely your mouse is (like mine) a dual mode mouse. Check if it doesn't have PC and MS written somewhere around the ball hole. For sometime this kind of mouse came with a nice switch below it that would select the mode, but then god-knows-why they changed it to an insanely great approach: the mouse defaults to micro$oft mode, which uses only two buttons (don't ask); if you want it to work in MouseSystems mode, which has three buttons, you have to turn on your PC holding the left mouse button down. This ain't no joke, although it looks like a joke from those %#@@!#$ing mouse manufacturers. You don't _really_ have to turn your computer off and on again just 'cause of this fscking mouse, it suffices to unplug your mouse and plug it back again while holding down the left mouse button. Of course I allways forget this intuitive trick when I have to shutdown my box for maintenance/upgrade/etc. Just for the record, I've tried all kinds of tricks, like setting those weird X options to disable DTR or RTS on the serial port, but none worked. The only thing that makes this sick mouse use the middle button is the unplug/plug windows approach. Hope this helped! Fabio ( Fábio Olivé Leite[EMAIL PROTECTED] ) ( NEW -*- http://descartes.ucpel.tche.br/~olive -*- NEW ) ( Linux - Distributed Systems - Fault Tolerance - Security - /etc ) (BC 50 7F 7A B9 2E 0A 26 91 8A D1 C0 B1 E4 DA A4)
Re: WEIRD!
Ola! ] I have noticed lately when I look at my server when it has been idle there ] is multiple lines that say Unable To Load Interpreter ] ] Maybe something ran out of memory, or your processtable was full so bash (or ] another shell) couldn't get loaded. Maybe you have some script as a cron job or something alike that uses a #!/path/to/interpreter as the first line, and this interpreter is somewhere else or missing. Sometimes I do get errors when I forget to chmod +x a new script on /etc/cron.daily. Your system might even be trying to execute some corrupted binary, whose signature can't be understood and the kernel tries to load a module for a strange binary format. Fabio ( Fábio Olivé Leite[EMAIL PROTECTED] ) ( http://descartes.ucpel.tche.br/~olive ) ( Linux - Distributed Systems - Fault Tolerance - Security - /etc ) (BC 50 7F 7A B9 2E 0A 26 91 8A D1 C0 B1 E4 DA A4)
Re: Help! Fetchmail broke
Hi! ] I checked /var/spool/mail and all folders are either root/mail or srgore/mail ] with rwx rw rw. The last rw allows any user to read and write to the files (mailboxes), what makes it a horrible security hole. Quickly run chmod 1770 /var/spool/mail/* to get things back to normal. HTH Fabio ( Fábio Olivé Leite[EMAIL PROTECTED] ) ( http://descartes.ucpel.tche.br/~olive ) ( Linux - Distributed Systems - Fault Tolerance - Security - /etc ) (BC 50 7F 7A B9 2E 0A 26 91 8A D1 C0 B1 E4 DA A4)
ndisasm is full of bugs
Hi there! I've noticed some problems with ndisasm really long ago (maybe even as long as debian 1.2), but I never bothered to complain. It's just that when I try to disassemble something, it just doesn't recognise the automatic sync points it should (using -a). When I try to set fixed sync points it just stops disassembling at the point I tell it to sync. Today I also tried to skip a part of the binary I was trying to disassemble with -k, and it also stopped at the point I told it to ignore. I'm using the latest nasm.deb (it's the same on slink and potato, BTW). I did not find any open bugs on the subject on debian.org, nor was it mentioned on Nasm's homepage at www.cryogen.com/Nasm. Am I the only one having these problems? Cheers! Fabio ( Fábio Olivé Leite[EMAIL PROTECTED] ) ( http://descartes.ucpel.tche.br/~olive ) ( Linux - Distributed Systems - Fault Tolerance - Security - /etc ) (BC 50 7F 7A B9 2E 0A 26 91 8A D1 C0 B1 E4 DA A4)
Re: Boot floppy
Hi there! ] I am trying to make a boot floppy. I downloaded the boot-floppies* ] package. But during the installation it craps out due to dependency ] problems. The packages that dpkg complains as not been installed are ] actually there (eg teTeX) since I built those myself from sources without ] using dpkg. I have the custom kernel image in .deb format. Is there an ] alternate way to create a boot floppy? Well, if you did not install teTeX via dpkg, then for dpkg they're _not_there_. You can force install overriding dependencies, though, if you feel it should work. Just man dpkg. Anyway, if you feel like building your own packages, you can allways do so using debian source packages. You can then modify whatever you want and still have your dependencies right. HTH! Fábio ( Fábio Olivé Leite[EMAIL PROTECTED] ) ( http://descartes.ucpel.tche.br/~olive ) ( Linux - Distributed Systems - Fault Tolerance - Security - /etc ) (BC 50 7F 7A B9 2E 0A 26 91 8A D1 C0 B1 E4 DA A4)
Re: missing file (/etc/ld.so.preload)
Hi there, ] What I'm doing is running a pipe between two background processes. Opening ] pipe for reading and writing both fail illegal seek. Does this qualify ] as special requirements ? Not at all. Probably your program is trying some funny file-oriented syscall on the pipe, so it fails. You can't do lseek on a pipe, for example. ] In any event I'm curious now to know what this file does do (and perhaps ] some example of the sort of special requirements that might require it) ? man ld.so It will give you some insight on how the shared libraries get loaded. HTH! Fábio ( Fábio Olivé Leite[EMAIL PROTECTED] ) ( http://descartes.ucpel.tche.br/~olive ) ( Linux - Distributed Systems - Fault Tolerance - Security - /etc ) (BC 50 7F 7A B9 2E 0A 26 91 8A D1 C0 B1 E4 DA A4)
RE: Continuing saga...
Hi there, ] (the ttyS* and cua* devices are hardware-equivalent, but if I'm ] not mistaken using the /dev/cua* device is better for dialin, which ] you might want for a modem). Whoah! cua devices have been dropped from Debian (and Linux itself, so to speak) long ago and their use nowadays is _strongly_ discouraged. They don't even exist on a recent Debian install. There are docs about that somewhere, but I don't recall any URL at the moment. Just go shopping on your usual Linux docs site. :) Cheers! Fábio ( Fábio Olivé Leite[EMAIL PROTECTED] ) ( http://descartes.ucpel.tche.br/~olive ) ( Linux - Distributed Systems - Fault Tolerance - Security - /etc ) (BC 50 7F 7A B9 2E 0A 26 91 8A D1 C0 B1 E4 DA A4)
Re: kernel 2.2.5 symlinks
Hi!
] the README for Linux 2.0.36 kernel tells me, that
/usr/include/{asm,linux,scsi}
] should be symlinks to /usr/src/linux/include/{asm,linux,scsi}, respectively.
] They aren't. They are normal directories. That's wrong, isn't it? The
] installation of Slink has done so.
That's an old Linux issue. We were supposed to keep those symlinks
pointing straight to the kernel includes, so that when we compiled stuff
we would get the latest and greatest (and most accurate) definitions. The
problem is if you keep changing kernel versions, it's likely that each of
your programs will have been compiled with a different set of includes,
and things may not allways work as expected. At least that's my version of
it :).
For such a great big distribution like Debian, it's nice to coordinate
which headers the maintainers use to generate the binaries, so that the
chance of interaction problems between packages is minimized, if not
eliminated. That's why the Debian policy mandates that these headers not
point to the kernel, but rather to some known to work fixed set. Again,
that's how I understand it. :)
] I've upgraded to Linux 2.2.5 kernel. Its README doesn't say anything about
] those symlinks. But I've problems to compile some libraries. Someone
] told me, that this can be caused by different versions of installed
] includes or libraries, that don't fit together.
I think you should elaborate more on those problems, as this is too vague.
] Must I delete the directories /usr/include/{arm,linux,scsi} and replace
] them with those symlinks?
I would rather not. Make sure you really have to use the kernel's
includes, and if so, don't delete the directories, but rename them to
asm.old, linux.old, etc.
Hope this helps
Fábio
( Fábio Olivé Leite[EMAIL PROTECTED] )
( http://descartes.ucpel.tche.br/~olive )
( Linux - Distributed Systems - Fault Tolerance - Security - /etc )
(BC 50 7F 7A B9 2E 0A 26 91 8A D1 C0 B1 E4 DA A4)
Re: FW: Problem with debian-user-list?
Hi! I just sent a message to debian-user@lists.debian.org, and it made it through to the list straight away, but I also go this reply from somewhere else. I SENT the message to and only to debian-user, it was not a reply or else. What does this mean? This means this user from .hu is a subscriber and has its mailbox full. The proper thing to do, IMHO, would be to return this kind of message to the list owner, not the sender. Then the list-owner would be able to do something about it, like temporarily remove this temporarily faulty addresss from the list. HTH Fabio ( Fábio Olivé Leite[EMAIL PROTECTED] ) ( http://descartes.ucpel.tche.br/~olive ) ( Linux - Distributed Systems - Fault Tolerance - Security - /etc ) (BC 50 7F 7A B9 2E 0A 26 91 8A D1 C0 B1 E4 DA A4)
Re: Help Needed with install Choose Debian archive path
Ola! My question is: if the path is /dists/slink/main/binary-a/ do I use the first '/ ? do I use the last / ? Do I need to include the Package file inside binary-a ? IIRC, if you are installing from official CDs you don't have to set anything special on this path. The install script will find it. Basically, you have to set up the path to something that contains the dists directory, but it's in the root of the CD, so just set this path to nothing. The install script knows which directories to go to inside dists. HTH Fabio ( Fábio Olivé Leite[EMAIL PROTECTED] ) ( http://descartes.ucpel.tche.br/~olive ) ( Linux - Distributed Systems - Fault Tolerance - Security - /etc ) (BC 50 7F 7A B9 2E 0A 26 91 8A D1 C0 B1 E4 DA A4)
Re: PPP not finsihing connection for Earthlink
-BEGIN PGP SIGNED MESSAGE- Hi there, However, I get the LCP problem where they are being sent but not received. I put LCP-MAX-CONFIGURE 30 in my options file and have upped it all the way to 150, but I still don't get any LCP's received. Finally after 30 - 60 seconds the modem simply hangs up... I've seen this on a local ISP, apparently their router doesn't care about responding to echo requests, and it looks like pppd doesn't like it. I've read every man pages and HOWTOs about it, but did not find a way to tell pppd not to care about echo requests... It seems all we can do is switch ISPs in this case. I hope someone can prove me wrong on this and tell us how can it be done! []! Fábio ( Fábio Olivé Leite BSc in Computer Science, MSc on the way! ) ( leitinho @ akira.ucpel.tche.br http://akira.ucpel.tche.br/~leitinho/ ) ( Debian GNU/Linux: just what your clunker needs to fly off the ground ) -BEGIN PGP SIGNATURE- Version: 2.6.3ia Charset: noconv iQB1AwUBNLWNs+k9m1I0y+XNAQGdDgL/UQx3pVJghqjGJbucWnIwM5Z2JzIkqnqU O9VjBfAmq4p0zoqPwgl48xsdq2AR6r+f0XfZeobAEDNdi4SCo7W5p8qwx4YZJueO lM8xN/qRPFLjVp1Eze2NfM3S+09pu5WW =Jikv -END PGP SIGNATURE- -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: todos
-BEGIN PGP SIGNED MESSAGE- Hi there, In which debian package lurk the conversion utilities 'todos' and 'fromdos'. Thanks, Richard which todos: /usr/bin/todos which fromdos: /usr/bin/fromdos dpkg -S /usr/bin/todos: sysutils dpkg -S /usr/bin/fromdos: sysutils Install the sysutils package. ;) [] Fábio ( Fábio Olivé Leite BSc in Computer Science, MSc on the way! ) ( leitinho @ akira.ucpel.tche.br http://akira.ucpel.tche.br/~leitinho/ ) ( Debian GNU/Linux: just what your clunker needs to fly off the ground ) -BEGIN PGP SIGNATURE- Version: 2.6.3ia Charset: noconv iQB1AwUBNLJgb+k9m1I0y+XNAQGIugMAqPgP3sOU6QWPPLhEV2+DQsktaShIRtPa xs5BRI9G2tVLHViI/UkCBQiVs8fUGnJyOd/lxsAcCVsGfq3E+fAHSTDSI+5Narg6 26tgxKStMqnVjbkG8RY8fwzt7MafG55f =hoo4 -END PGP SIGNATURE- -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
pcmcia 2.9.6: broken?
-BEGIN PGP SIGNED MESSAGE- Hi there, I use a Debian notebook, and recently upgraded from pcmcia-* 2.9.5 to 2.9.6, recompiled the modules and was very confused when I noticed the card manager would allocate _only_ already used IRQs... It was real funny: The pcmcia controller sits on IRQ 11, and that was detected flawlessly, but then when I inserted one card, 11 was exactly the IRQ allocated to it! To make things worse, I inserted another card and it got IRQ 14 (my ide0 went nuts). Both cards (a modem and an Ethernet) get detected perfectly, but the IRQs seem to be a little bit screwed. Downgrading to 2.9.5 and recompiling the modules got everything back to normal. Is this old news? If so, sorry, I just subscribed to this list (I use debian since may/97 and linux since march/95, but this list since ten minutes ago :). []! Fábio ( Fábio Olivé Leite BSc in Computer Science, MSc on the way! ) ( leitinho @ akira.ucpel.tche.br http://akira.ucpel.tche.br/~leitinho/ ) ( Debian GNU/Linux: just what your clunker needs to fly off the ground ) -BEGIN PGP SIGNATURE- Version: 2.6.3ia Charset: noconv iQB1AwUBNLJi6ek9m1I0y+XNAQGTxwL+Mj054d96g1YP09/XrFXMdE0NpwKtgOPi P10TQ2UWZFToopd9Xo2wDp4jio9JhebZEEohdah0FHGCqD5GF0nGcrjbXskWjeBK bWXhyLc+E/OdytDFQ0VvdSVJ1gTNFJx3 =SgzR -END PGP SIGNATURE- -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: a count-down timer?
Hi there, ) bell or do something to alert me when the five minutes is up. There ) are many programs such as rclock that can alert me at a specific time ) on a specific day but I want to be able to say five minutes from now ) without a lot of hassle. I recall a program named leave that does exactly that. What I do not remember is wether there is a .deb pack for it. []! Fabio ( Fábio Olivé Leite BSc in Computer Science, MSc on the way! ) ( leitinho @ akira.ucpel.tche.br http://akira.ucpel.tche.br/~leitinho/ ) ( Debian GNU/Linux: just what your clunker needs to fly off the ground ) -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Lynx: proxy server
Hi there, ) how do I set up Lynx to use my ISP's www proxy server? There's nothing ) obvious in either the man page or the online help. ) ) Open up /etc/lynx.cfg (the system-wide configuration file) and search ) for proxy. I learned the hard way that you can also set the http_proxy environment variable for the URL of the proxy, like what follows. I use squid.ucpel.tche.br at the port 3128 as proxy, so I set the variable to http://squid.ucpel.tche.br:3128/ as if it were a URL just like anyother. []! Fabio ( Fábio Olivé Leite BSc in Computer Science, MSc on the way! ) ( leitinho @ akira.ucpel.tche.br http://akira.ucpel.tche.br/~leitinho/ ) ( Debian GNU/Linux: just what your clunker needs to fly off the ground ) -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
