Re: Emacs command line option
Adam Lazur <[EMAIL PROTECTED]> writes: > > The second point is that I don't want emacs to always use > > auto-fill-mode but only if it is invoked by mutt. > > But I noticed that the auto-fill-mode breaks always at 70 chars. Even > > if I try to override the value. Strange thing. > > here's what I have in my ~/.muttrc to accomplish the auto-fill-mode > when emacs is invoked from mutt. > > set editor="emacs -nw %s --eval '(setq make-backup-files nil) > (auto-fill-mode)'" > > though all you really need is: > > set editor="emacs %s --eval '(auto-fill-mode)'" You can set emacs to auto-fill at any column number any time you want by evaluating: (auto-fill-mode 70)or any number you choose to wrap at. In emacs press M-: you will see a prompt saying Eval: Type (auto-fill-mode it will be evaluated and used on that buffer. This is what the muttrc line is doing too. -- Harry Putnam [EMAIL PROTECTED] Running Redhat Linux-5.1
from `zip' to sid via network - A general outline needed
Hopefully I will already have accumulated enough knowledge and skill to do this with minimum help: Summary: Beginning with no Debian OS at all, and over the network, get a minimal install, and finally a full blown installation of `Sid'. Details: Its not so easy to see how the above can be done by looking over the various docu stuff at www.debian.org. For example: Under the network install heading one is suggested to get a `minimal unofficial' cd first and proceed from there. Addresses to get them are given, but all I see are ones with names like `woody ' So its not clear if one can go to `sid' from there. I suspect doing what I want is mostly described on www.debian.org but without actual experience doing it a time or two, it just isn't apparent. Once one has a minimal install that is capable of internet connection, then what are the steps to upgrade a system all the way to absolute latest `sid'? I'm not a developer, nor will I be of much help to developers, so that isn't my angle here for wanting `sid'. I run one machine (of several) at home, that has nothing but experimental installs of various OS on it. Currently: Sol 8 (intel), openBSD-3.1, `Null' (The beta version of Redhat). I'm a retired heavy construction worker so have plenty of time to mess around with stuff I `have no business' with. (hehe). I like to tinker with stuff that is mostly over my head, so in some ways I may become quite a pest here over time. The description above isn't much of lead into getting help, but it probably comes sort of close to describing quite a few posters here, maybe mostly in a somewhat younger age group, but still people who like to tinker with stuff they don't fully understand. Anyway, what I'm asking for is a general outline of steps required to go from zero to a full install of `sid', over the network. I'm equipped to burn any CDs required. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: from `zip' to sid via network - A general outline needed
"Donald R. Spoon" <[EMAIL PROTECTED]> writes: [...] >> I'm asking for is a general outline of steps required to >> go from zero to a full install of `sid', over the network. I'm >> equipped to burn any CDs required. >> > > Hi Harry! > > Us "retirees" have to stick to gether . Hehe ... yes indeed. Nice blow by blow Donald. To posters in this thread: Thanks folks, these posts should have me setup. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: from `zip' to sid via network - A general outline needed
"Donald R. Spoon" <[EMAIL PROTECTED]> writes: > To get to SID, all you have to do is edit the /etc/apt/sources.list > file and change all the instances of "stable" (or woody) to > "unstable". Do an "apt-get update" and then do a "apt-get > dselect-upgrade" and sit back and watch(probably about 2-4 > hours??). I installed from one of the unofficial minimal cds out there. I selected only the security updates when asked about sources so my list only contains that one link: cat /etc/apt/sources.list deb http://security.debian.org/ stable/updates main What does a normal or working sources.list look like? Then I can edit as needed. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Fresh install no network
I had hoped the install routine would find my nic and know what to load but as I went thru the initial install, it wasn't found. Any attempt to select something from the module list for net, failed. I'm not real sure what card is in there but its one of the commaon 3com cards. I had hoped some kind of probe would find it and I wouldn't have to tear things up to find my original box etc. or open the case to get it. This case is in a inconvenient spot. I'm at the stage now where the initial system is intalled and I've rebooted ... setting at the config screen for basic stuff. What is the best way, with least hair pulling to get the nic setup? I'll check dmesg once bootup is complete. Oh and I lied a bit about being at the config screen, I will be shortly but I choose to make a floppy and boot that way since I have solaris installed on First Master disk, and wasn't sure I wouldn't scrub its boot loader or something by tinkering with lilo. I'll get to that later, but I notice the boot floppy takes an awfull long time, like 4-7 minutes. Seems an awkful lot even for a floppy boot. Would that be an indication of some kind of trouble? Eventhough I've installed several flavors of linux and 2 bsds, as well as a couple of regular unix's, I could use a little hand holding thru the rest of the install. Especially getting network enabled. I can dig up the exact info on my hardware but it will be some serious digging and I'd sooner avoid spending the time on that. The machine is an Athlon Tbird 1.3 ghz and a common 3com nic.
Neighbor table overflow.. what does it mean
After initial install of woody 3.0, at the beginning of basic configuration with nothing but the basic system installed. Upon aswering yes that I am ready to configure this message is printed to screen: Neighbor overflow table Sounds rather ominous. Dmesg shows this output: [...] PCI routing table Version 1.0 at Oxf1690 Intel ISA/PCI/Cardbus PCIC Probe no bridges found ds: No Socket drivers loaded Neighbor overflow table What is the problem here?
Configuring locales
During the second phase of woody install, I see a screen that asks me to configure locales. Seems to be no help on this screen and it isn't mentioned in the install instructions. How can I know what to do here? Will basic stuff be setup if I skip this? I am somewhat experienced in linux and installs of various flavors but I have no clue about this. I've never been asked to setup Locales manually. I have no special foreign language requirements and am a native enlglish speaker. Do I need to select the english ones?
Enough time wasted, moving on
Anyone who wants the official 6 CD set of woody 3.0 Can send me an address privately. First one I get in my inbox with snailmail address will get the cds mailed to them. I've burned up 2 full days, gotten no help whatever from the list. Found the install routines to be unusable. All I hear from Woody is `can't won't or doesn't' And this on a machine that has had 2 flavors of linux, 2 of the BSDs and Solaris (intel) install successfully. Hardware recognized etc. My nic is not found, X doesn't run, I get a message after the install that there were errors during install. Please fix the errors mentioned above... none are mentioned. I've tried the simplicity of Tasksel and the labor instensive dselect hell, 3 fullbore tries at installing. I have a non-functional system at the end of 2 days. When I see things so basic like the names referred to when the different cds are requested during package install. The names the routine asks for are nowhere on my CDs. Not that hard to see what is really meant but only because I'm already familiar with some of the lingo. Maybe I got hoodwinked getting them from `The sphere' When I see things like that gauranteed to cause confusion, I know that at the Debian organization someone is asleep at the switch. And can rest assured that I will be in for lots of unnecessary grief caused by `loose' organization and homeboy packaging. One has to think, what else is poorly done here. But then again for someone who enjoys major unnecessary tinkering. Relentless `fixes' and the like, Here is a chance to get a setup that will keep you busy for a long time, and absolutely free. I'll pay the postage too. Maybe I should have tried the released version, but something about the name `potato' puts one in mind of a large overweight weak vegetable lying on a couch. At least `woody' sounded.. well erect :-)
Re: Configuring locales
"Sean 'Shaleh' Perry" <[EMAIL PROTECTED]> writes: >> >> I have no special foreign language requirements and am a native >> enlglish speaker. Do I need to select the english ones? >> > > Everything will work fine for English speakers. If you want to add a locale > later look at /etc/locale.gen and related docs (it will point you to them). Thanks Sean, for your helpful reply. I'm sorry to say that I've found so many problems with my three attempts at installing woody 3.0, and found a number of things that seem to indicate a serious waste of time in store for me, that I've thrown in the towel and am offering anyone who wants them the 6 cd set of woody 3.0.
Re: Enough time wasted, moving on
PROTECTED]> writes: > Are we to understand that you bought cds of a version that is still > changeing constantly? Why? From whom? Why? >From the links on Debian home pages followed in this order: http://www.debian.org/CD/ http://www.debian.org/CD/vendors/ United States Vendor: Antonio Rodriguez URL: http://www.the-sphere.org/ URL for Debian Page: http://www.the-sphere.org/Linux/Debian.html Allows Contribution to Debian: Yes Country: United States Ship International: To some areas email: [EMAIL PROTECTED] CD Type: Official CD ; non-US ; Development Snapshot Architectures: i386 Note the part that says `Official CD' Why you ask? I didn't want potato, and foolishly thought that if CDS were available they would not be totally useless. And I'm no stranger to development tools. Alan Shutko <[EMAIL PROTECTED]> writes: > Harry Putnam <[EMAIL PROTECTED]> writes: [...] > Harry, woody hasn't been released yet, so there are no official CDs > yet. Since those CDs were pressed, a bunch of packages have been > updated and fixed, and there's more fixes that will need to be done > before release. > >> Maybe I got hoodwinked getting them from `The sphere' > > Probably, if they're passing woody off as done. Technically they aren't passing them off as `done' but you and I are both running lots of stuff that isn't `done' but can be used. Keith Parkansky <[EMAIL PROTECTED]> writes: > Harry Putnam wrote: >> >> I've burned up 2 full days, gotten no help whatever from the list. >> Found the install routines to be unusable. > > Same here on the last two counts, except > it's been two weeks with Potato r5. Sounds like it may not have been all that big of an improvement to get `potato'. "Liam Black" <[EMAIL PROTECTED]> writes: > Before you give up on Debian, I really would suggest that you try > potato. Though the imagery conjured up by the name may not be > pleasant, I've found it to be quite useable, and I'm a relative > newbie to Linux (my experience lies in AIX). I guess its a thought.. > While sometimes the list is unhelpful, I've been very impressed with the > attitudes and intelligence of many of the people present on it, and even > more impressed by those who don't have much time to parse it -- the people > actually working on bringing Woody to a stable release. I don't really doubt anyones intelligence at debian. I'm sure they are all vastly more talented than me. Probably have a lot more energy too... he he. But I did see what seemed like several glaring shortcommings. May have squawked a little harder than necessary. > But, really, give Potato a try -- it won't take more than another 2 hours > to do a full-on installation of it, and if it works -- well, that's just I'd have to get it first. And I did already spend $40 on this adventure. Not a forture to be sure, but I'd rather not throw away more.
Re: Enough time wasted, moving on
Geordie Birch <[EMAIL PROTECTED]> writes: >> At least `woody' sounded.. well erect :- > why don't you lose all those cd's, do a network install of stable, > then, before installing _anything_ beyond base, apt-get dist-upgrade to > testing or unstable? dman <[EMAIL PROTECTED]> writes: [...] > You can try installing a base potato system, then dist-upgrade to > woody. This has worked for me on several systems. A couple of posters have mentioned a network install. Where are the details spelled out? Are we talking installing from a running machine? Already network enabled? My situation is that I have 2 desktops before me. A full well used redhat 7.1 on one machine and A solaris x86 on another. The solaris (intel) has a second disk that had a skeleton install of Redhat on it before I began my adventures. I wanted to install Woody in its place. So with a network install how would I proceed? I mean in general. Not a detailed account. I'll find and read any documentation I need. Pointers would be nice though. I wouldn't mind the install being very basic for quite some time. Long as it is network ready. Don't really care about X either. At least not immediately.
Re: Enough time wasted, moving on
"Karl E. Jorgensen" <[EMAIL PROTECTED]> writes: > How much did you post to the list? I can only find the following > threads: > Fresh install no network > Neighbor table overflow.. what does it mean > Configuring locales > Enough time wasted, moving on You may have a point there. There does seem to be someting of a lag in turnaround. I didn't see any of this until sometime today. Except a helpfull response on the locales subject. I answered all the above that came in. But still haven't seen my answers appearing, as of 5:20pm pacific time.
Retry.. no more crybaby bs
OK, list posters, you've advised me to coninue my efforts and step back to potato. (See thread suject `Enough time wasted.. moving on') A coulpe of posters have suggested a network install. Thing is I have $40 dollars worth of cds setting here. Surely I can at least use them to get the base install going and maybe get my nic setup. Then do the rest via network? In the last three attempts my nic is not recognized and any attempt to install one of the 3com drivers is rejected. I didn't capture any error info but will this time if it comes up. My card is: 3Com 3C905B-TX (PCI)
Re: Fresh install no network
Sebastiaan <[EMAIL PROTECTED]> writes: > High, > > On Wed, 27 Feb 2002, Harry Putnam wrote: > >> I had hoped the install routine would find my nic and know what to >> load but as I went thru the initial install, it wasn't found. Any >> attempt to select something from the module list for net, failed. >> >> I'm not real sure what card is in there but its one of the commaon >> 3com cards. I had hoped some kind of probe would find it and I >> wouldn't have to tear things up to find my original box etc. or open >> the case to get it. This case is in a inconvenient spot. >> > Most 3com cards are supported, but you have to load the driver yourself. > Best way is to look what the card says. Then start 'modconf' as root, and > browse to the network card interfaces. > > Note: the 3c905 series need the 3c59x driver to get to work. Haa.. Thanks Sebastian that worked, and was not rejected. But the way modconf was not available during install, but I saw that choice on the modules sect screen. One would never know to choose it though, without being told.
Exim config question
Just got a basic install running and during the course of setting mail with the base-config tool, I think I answered something wrong but not sure what would have been right. Or where to hand edit it. I have a fair bit of experience with sendmail, but my first go around with exim. The details of my setup are: Single user machine dsl connected to internet thru local isp (dslextreme.com) and telco Using local isp for smart_host But pop server is newsguy.com not my local ip. I think I got most of this set alright except when I send a mail to myself at [EMAIL PROTECTED], exim tells me it is delivering locally and I have it back instantly. That is it never goes to newsguy, or thru my smart_host. I probably told exim to do that, but not sure what to change or where. My user is reader on this machine. But newsguy.com is a commercial online news/mail nntp/smtp/pop server. With sendmail I've always masqueraded as newsguy.com and it all worked. I think it is actually working now except sending mail to myself a [EMAIL PROTECTED] is just put directly into /var/spool/mail/reader
Re: Retry.. no more crybaby bs
> OK, list posters, you've advised me to coninue my efforts and step > back to potato. (See thread suject `Enough time wasted.. moving on') You all will be sick of me soon.. Just a little note of apology to the list and for being a crybaby. (Its not my fault ... its genetic .. :-) ) Also a little underhanded way to see if I got the mail sort of close to right. There is one little glitch .. already posted. but I think the fact that I know nothing about exim and was able to get mail working just with `base-config' speaks volumes. > A coulpe of posters have suggested a network install. Thing is I have > $40 dollars worth of cds setting here. Surely I can at least use them > to get the base install going and maybe get my nic setup. Then do the > rest via network? Now this item is kind of funny really. I accidently started this going after making a forth try at installing, and after the base install I opted to add ftp to the media possibilities. Scads of files are being brought down the pipe as I write. > In the last three attempts my nic is not recognized and any attempt to > install one of the 3com drivers is rejected. I didn't capture any > error info but will this time if it comes up. My card is: > 3Com 3C905B-TX (PCI) A kind poster named Sebastiaan, sent me a note that said to use the 3c59x driver. The install bought it and I am network enabled now. This is something that needs to be patched though. 1) The install screen where you pick out drivers and other modules is unecessarily cumbersom. Is there really a reason why one must scroll back thru the whole thing with each choice. I see no chance to mark or otherwise do things better than one by one with a complete rescroll at each stop. 2) The driver listed for the 3Com 3c905b is not the right one and in my case the OS rejects it. The right one is a couple lines further along. One would never know it unless told in advance as I was.
Re: Retry.. no more crybaby bs
Bob Thibodeau <[EMAIL PROTECTED]> writes: > My guess is an irq setting isn't being found correctly. What resources was > the card successfully using before? > Thanks Bob, I've posted the problem with it already but in brief I hadn't tried to install the correct driver. I used the base-config tool to load modules etc and at that screen it lists the wrong driver for my card. Another list poster told me the right one and it works now. p <[EMAIL PROTECTED]> writes: > // > > those cd's sound suspicious/flaky. [...] I managed to get fired up here by allowing only the basic install from the cds and going to ftp from there, with a working network, all things become possible.. Downloading and installing now from `stable'. Not sure I really understand what is happening though. Apparently apt-get is getting some stuff over ftp. Over 200mb was downloaded, but still askes me to install various of the cds at different points. And gets stuff from there too. > "...$40.00"? ouch! and it's > _un-debian! I got to it direct from the debian home page. It was the first one listed under us vendors `The Sphere'
no mouse in console
Managed to get a fresh install of woody running but have no mouse in console mode. Mouse works in X. I remember being asked where the mouse was and choosing psaux. It was just a guess. How can I debug this? The gpm man page isn't particularly helpfull. I do see /var/run/gpm.pid. But have no mouse.
The battle for mbr (lilo)
My installation is a dual boot setup with Solaris 8 (intel) on first master and Debian on second master. Solaris is very particular about its boot sector I think. So I'm booting from the floppy I made during install. Each boot takes 4-5 minutes, which is very slow. Seems it might mean some kind of problem. I want to set up booting from disk but don't want to jackup the solaris bootabililty. Anyone here familiar with dual booting with solaris.
No KDE for me
On a recent install of woody 3.X, if I say startx I get kde, not a choice window or something but just straight to kde. On boot up I get a kind of selection window That allows a gnome choice or two. I don't want anything to do with kde but its been so long since I had to choose between, I've forgotten how its done. I want to set the default to gnome and really I want gnome/sawfish. I'm accustomed to that combo and know how to get what I want from it with out much fumbling around. I guess the sawfish part is as easy as `apt-get install sawfish' But I think there are more than one model. Anyone here using that combo that can tell me how to get the default X session to be gnome/sawfish
dpkg - the letter prefixes.. what do they mean
dpkg -l '*desk*' shows a series of hits, all are preceded by `un' dpkg -l 'gnome*' shows a number of things some are proceeded by ii and some by un. Where does on look to find out what these letters mean. I went thru man dpkg several times looking for some referrence. A grep with `ii' gets no hits. dpkg --help doesn't appear to make any mention.
Re: dpkg - the letter prefixes.. what do they mean
Mark Janssen <[EMAIL PROTECTED]> writes: > dpkg -l | head > > The first 3 lines of the dpkg output list what the letters mean. > > [EMAIL PROTECTED]:/etc/courier/smtpaccess$ dpkg -l | head > 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 VersionDescription > +++-==-==- > ii a2ps 4.13b-15 GNU a2ps 'Anything to PostScript' > converter > ii acct 6.3.5-30 The GNU Accounting utilities. > ii adduser3.45 Add and remove users and groups Thanks. I saw that stuff and couldn't make heads or tails of it. My feeling is that it explains nothing and is downright confusing. Some of it, one can guess. It'd make some sense to have something in the man page for dpkg, especially a note at the section on `-l' that tells you point blank what each column means. Or points to a section that does that. I may be a slow learner or intellectually challenged or whatever but it would be a cold day before I got much meaning out of those lines. I'll submit a small patch when I get my new install squared away a little.
Avoid booting direct to X.. I want my starx back
Well, I got X working in my fresh woody install, but there are a few rinkles. 1) I want to boot to console mode and call x with startx. Currently, I get popped right into X. I thought this could be contolled by setting the defalult run level /etc/inittab But I see nothing in there that looks likely. In fact it claims the default is runlevel 2. Or is runlevel 2 X. I don't think so. In my redhat dealings it was possible to set default to runlevel 5 which force boot to bring up X. setting runlevel 3 gave you a console login. Where is this choice made on debian?
Re: Avoid booting direct to X.. I want my starx back
Harry Putnam <[EMAIL PROTECTED]> writes: > Well, I got X working in my fresh woody install, but there are a few > rinkles. > 1) I want to boot to console mode and call x with startx. Whee, got lots of help on that one. And a couple easy fixes. Thank you all. Now I can startx but now I get put into a kde session. I'd sooner set my hair on fire, and I'm mostly bald. I guess I should uninstall all the kde stuff but I seem to recall there being a problem with some of them like the libs or base, that allow some other things to work. All I really want is to use gnome/sawfish if in X. When I say `startx' thats what I want to start. I vaguely remember there being some place where you put a desktop choice and then its settled. Does debian have something like that too?
Re: dpkg - the letter prefixes.. what do they mean
dman <[EMAIL PROTECTED]> writes: > On Fri, Mar 01, 2002 at 07:48:59AM -0800, Harry Putnam wrote: > | Mark Janssen <[EMAIL PROTECTED]> writes: > | > | > dpkg -l | head > | > > | > The first 3 lines of the dpkg output list what the letters mean. > | > > | > [EMAIL PROTECTED]:/etc/courier/smtpaccess$ dpkg -l | head > | > 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 VersionDescription > | > > +++-==-==- > | > | Thanks. > | > | I saw that stuff and couldn't make heads or tails of it. > | My feeling is that it explains nothing and is downright confusing. > > It is confusing, until you actually stop and read it and follow the > ascii-art. The ascii art is arrows pointing from the description line > to the column the description refers to. The letters are just the > first letter of the word they represent. It seems rather intuitive to > me (once I decided to actually read that output). dman, must be gifted in some way I'm not. Not to press the point here, because I got good and thorough answers thanks to generous helpers here. But... It is confusing before, during, and after reading it closely and several times. In fact I dare say that even now that I know what it means I still see nothing in those lines to clue me in. For example, can you show me what, in there tells, what `ii' means? Yes, it would be a good guess that it means installed, but then what does the second i mean? Ditto with the others. So let me have my little fun when I get lined out here and submit a patch that even dopes like me can follow.
Re: Enough time wasted, moving on
dman <[EMAIL PROTECTED]> writes: [...] > | > A couple of posters have mentioned a network install. Where are the > | > details spelled out? > > http://www.debian.org/releases/stable/i386/install I read that over a bit before asking the above question. I'll admit I didn't read all of it word for word but the section entitled: Installing Debian GNU/Linux 2.2 For Intel x86 Chapter 5 Methods for Installing Debian Seemed like it would be the place, but it doesn't really deal with networking in any way other than NFS. Not to be a total PITA here, but can you point out some detailed description of how this is done there? Its no longer an urgent thing, so no need to bother really but I think it really isn't covered well and probably not a good place to point someone who needs to know how its done. > > | > Are we talking installing from a running machine? > > Nope. > > | > Already network enabled? > > Depends on your definition of "enabled". If there's no network > wiring, then you can't do a "network install". You must have a NIC > and a Cat[35] cable or a modem and it must be connected to a working > network. The machine itself might not even have any software on it. I pointed out in some previous posts that my NIC was not being found and that my attempts to install some of the various drivers at the screen where you select modules and such were rejected. So with no working nic, we could safely say the machine is not network enabled. This is all something of a moot point now since someone told me the driver I needed and it installed ok. My forty dollar disks were good enough to get the base system, then of course with a network, the skies the limit. I pulled down the rest in record time and got a good taste of the power of apt-get. My point earlier was that without a network enabled machine all bets are off.
Re: Enough time wasted, moving on
Manoj Srivastava <[EMAIL PROTECTED]> writes: > Considering you are asking about pre-release, unstable, wildly > unusable software on a users list, rather than the developers one, it > should come as no surprise no one would be able to offer any help. Horse pucky. Manoj, you do a disservice here. That software in woody is known to be usable and is widely used already. Your dead wrong about the help too. I just hadn't waited long enough. Not everyone here has your `hard case' attitude. Rubbish. You did not get these from Debian. There is no ``official'' CD set on any Debian archivce site. You run down to som e no name local shop and get an `official'' XP CD, guess what'll happen? Why should things be different for a a shoddy CD set not from Debian? Rubbish? I think you've been reading those `Winnie the Pooh' novels again. :-) I told exactly where I got them. And it wasn't on a street corner from a guy in a trench coat, they are not from Debian to be sure, nor are they `Official', it turned out, but the `No Name' shop is linked right into debain home. The very first US vender linked to debain home page. I don't really think there is any thing wrong technically with them either. My jab about someone asleep at the switch was poorly aimed.
Re: Enough time wasted, moving on
Carel Fellinger <[EMAIL PROTECTED]> writes: > received valuable help and turns out to be okee afterall. It appeared > that in the install he was adviced to use a certain driver for his > nic, turned out to be the wrong driver. Thanks to the list, he's back Close, but it was actually a helpful poster here who told me the RIGHT driver. The install itself lists the wrong driver. I can't find the notes I made at that step but on the screen it lists the NICs in one column and drivers across from them: nic1 somedriver nic2 someotherdriver nic3 yettaanotherdriver 3c509 WRONG ONE HERE The right one is a few lines away and not apparent what it goes to. I tried several and all were rejected. After Sabastiaan tipped me off what driver to get. It worked. That section of the install should probably be looked at, and I think some said they were doing that very thing.
Re: Enough time wasted, moving on
Manoj Srivastava <[EMAIL PROTECTED]> writes: > > All I hear from Woody is `can't won't or doesn't' And this on a > > machine that has had 2 flavors of linux, 2 of the BSDs and Solaris > > (intel) install successfully. Hardware recognized etc. > > I do have a hard time being patient with people maligning the > distribution, saying things about people being asleep at the switch, > while the problem was not with the project. The high traffic here is mostly about how well everything works. Its not really dozens, even hundreds of `project problems' being discussed here daily. No one is finding any difficulties with the install routines. And if they do, its probably their own doing, and not a `project problem'. Right about here is where the water starts running uphill. You can keep right on with the `hard guy' routine, but I think its only fair to warn you that my whole body is a weapon. Hands registered on 3 continents. The last guy that called me a juvenile banterer is pushing up daisies on a vacant lot in Chicago. :-) And you think `asleep at the switch' is too harsh I'm sure you will find none of that amusing: But: I've tried to detail some of the problems I had. The one with the nic was probably the most serious since it rules out any further install, or at least with apt-get. I'm clearly no expert, but then on the other hand I do have 5 yrs experience with linux BSD and some unixes. There are major holes in my knowledge and skill base, partly because my education ended at 9th grade. And I've made a living the past 35 years as a heavy construction worker. But I think it would be fair to say, if I have lots of trouble with an install, after installing one or another linux, bsd etc dozens of times on at least 20 machine in the last 3 or so years. That there are some actual problems there. I don't exaclty follow the reasoning I've seen presented as to why there isn't a more developed probing mechanism in place. Apparently it is felt that isn't something to spend time on. That may be true. But, hardware finagling is really a pain in the butt. My feeling is, the more you put on the machine to do, the better it will be. The thing about a crappy install routine is that once you succeed at it, you quit worring about it, so it never again seems so important. I'm having trouble now, remembering what was so hard.
Re: Enough time wasted, moving on
dman <[EMAIL PROTECTED]> writes: > Sure. As I look it over again (it's been a while since I read it), > the first sentence of Chapter 5 is : > > You can install Debian from a variety of sources, both local (CD, > hard disk, floppies) and remote (FTP, NFS, PPP, HTTP). > > Note the FTP and HTTP. Those are what you would want. You have a strange idea of details. And thats about how far it goes in that section too. But in the threads here I've put together a whole plan of attack. Involving editing the sources.list. Knowing how to get going with the nic. A series of commands for apt-get that clean things up and get ready to install a more fleshed out system. I suspect there is a smallish outline of things to do that would make it pretty clear how to conduct a network install. Some of that needs to be in that section. I don't think its enough to say, you can do it by ftp or http.
Re: dpkg - the letter prefixes.. what do they mean
dman <[EMAIL PROTECTED]> writes: > Ok, I'll work on my teaching skills a little more :-). I think you should be promoted to `oracle' or something. That was good. It seems so easy once explained, but I'll be damned if I could see that in it. Thanks.. Thats another one of those things that once you're onto it, it recedes into the background and doesn't seem like a problem any more. Having too much fun installing stuff with apt-get to worry about that now. .. hehe Oh that reminds me, I'm seeing some kind of bug in the rsync that is in woody testing now. Whats the proceedure for reports?
Re: Avoid booting direct to X.. I want my starx back
Chris Jenks <[EMAIL PROTECTED]> writes: [...] > I've noticed that it will take a couple of hours to get this to you, but... > what your looking for is .xsession or the .xinitrc Mark Zimmerman <[EMAIL PROTECTED]> writes: [...] > One (of several) ways to do this is (as root): > > update-alternatives --config x-window-manager > > You will be presented with a list of alternatives that are currently > installed on your system. Pick one. > > If the one you want isn't there, install it and run update- > alternatives again. > > This will set the system default. Individual users can override this > by writing a .xsession file. k l u r t <[EMAIL PROTECTED]> writes: [...] > it can be set in your home directory's .xsession file > > echo "exec gnome" > ~/.xsession > > ...now you can startx and gnome will be your window manager. Bob Thibodeau <[EMAIL PROTECTED]> writes: [...] > startx looks for ~/.xinitrc > you can set up your options there > > you probably have a system-wide xinitrc that calls kde More good help.. thanks again. Its kind of coming back a little now seeing the postings here. I've been doing upgrades but maintaining a set of configs for a couple of years now and haven't had to really fuss with X stuff too much. I've been able to put my rc files aside and just plop them back in many times. I caught hell when redhats defaut window manager changed from fvwm2 to gnome a couple of years ago. At first I cursed them for having made the change, and hated gnome. But if any of you have ever messed with the acres of config involved with fvwm, it was really a blessing. Just now trying to follow the maze of sourced variables and directeries with a gob of scritps in them uner /etc/X11 I never really did see what is calling kde. But I can set it as described above to over ride what ever is happine in that maze. You need a road map to follow all the twists and turns in there. I noticed at the default debian login (gnome) that the mouse is really sluggish. I know how to set it with xset m and of course once inside the session with the control panel settings. But to speed up the sloth at the actual login probably needs to be put into xintrc. Is it just a staight command in there like: xset m 2 14 or the like?
Terminfo and emacs
I may have jacked up my terminfo database. I built emacs from sources today. A very recent beta of emacs-21. I used no special flags with ./configure since emacs has quite a smart configure and make setup. Just ran it like: ./configure make make install There were no build errors or even warning messages that I saw. My first attempt to run it gives me the error: `can't find termcap data base' Being a new user of debian I hadn't notice until then that debian uses the newer terminfo. I found it at /etc/terminfo. I'm pretty sure emacs is smart enought to find the database but thougt maybe it wants it to be in /usr/share. At this point stupidity started to rear its oh so human head, and I decided to copy it to /usr/share. I took a hasty glance at ls /usr/share, but didn't see anthing to give me pause. cp -a /etc/terminfo /usr/share That blew up with a batch of messages like file such and such was the same as file so and so. Turns out there is already a /usr/share/terminfo. Now I'm not sure what I've done. /usr/share/terminfo is quite a lot bigger than /etc/terminfo But maybe a result of my dumb move. Are there really two terminfo directories? In addition to solving the emacs problem, now I want to know where to read up on the commands that give system package info. I see quite a few in `man dpkg', that give basic infomation but I think there are more. Things that will tell what package a file belongs to. In fact I'm pretty sure there are a full range of such commands that are similar to the rpm package manager in that they can pull up all kinds of detailed info about files and packages. Maybe something that can verify if a stock directroy has been screwed up. Something like the rpm commands `rpm -qf ' will tell exactly what package a file belongs to. Then with that info you can say rpm -V to find out if the package or its files are screwed. dpgk -l 'terminfo*' doesn't know anything about terminfo People have told me the debian package manager is more sophisticated yet, and can pull up more detailed info. What do I need to read to get on top of those kind of commands? Any hints on what is causing the emacs problem.
Re: The battle for mbr (lilo)
Hans Ekbrand <[EMAIL PROTECTED]> writes: > On Thu, Feb 28, 2002 at 11:12:16PM -0800, Harry Putnam wrote: >> My installation is a dual boot setup with Solaris 8 (intel) on first >> master and Debian on second master. Solaris is very particular about >> its boot sector I think. So I'm booting from the floppy I made during >> install. Each boot takes 4-5 minutes, which is very slow. Seems it >> might mean some kind of problem. >> >> I want to set up booting from disk but don't want to jackup the >> solaris bootabililty. Anyone here familiar with dual booting with >> solaris. > > No need to. Use lilo to set install a boot block on a floppy wich > starts a kernel on the /boot on second master. It's fast enough for > you. > > The significant line in /etc/lilo.conf is > > boot=/dev/fd0 Thanks that works just right. I've use lilo for several years and somehow had missed that one could do that so easily.
Re: Avoid booting direct to X.. I want my starx back
Mark Zimmerman <[EMAIL PROTECTED]> writes: > One (of several) ways to do this is (as root): > > update-alternatives --config x-window-manager > > You will be presented with a list of alternatives that are currently > installed on your system. Pick one. > > If the one you want isn't there, install it and run update- > alternatives again. > > This will set the system default. Individual users can override this > by writing a .xsession file. That doesn't seem to do the job here. Running that command gives me two choices twm and sawfish. No mention of kde or gnome. Both of which are installed. I do want to use sawfish so set that as default but startx still starts kde.
Re: Terminfo and emacs
Nic Ferrier <[EMAIL PROTECTED]> writes: > You can fix emacs by using > >apt-get install termcap-compat > > > It's about time that emacs supported terminfo but unfortunately there > is no one with the time to do the work. Thanks for the usage tips for dpkg. And the above install did get things working. However I ran into other problems. Emacs is not finding certain include files for X11 and won't compile with X11 support. I didn't save the error output since I don't really want to debug it right now. Too busy getting this new install working. I use emacs a lot so I just installed the debian version of emacs-21 instead. Or I should say I'm downloading it now. I'll get back to this at some point though, because I tinker with emacs a lot.
Re: Terminfo and emacs
Matijs van Zuijlen <[EMAIL PROTECTED]> writes:
> On Sat, Mar 02, 2002 at 02:56:02AM -0800, Harry Putnam wrote:
>> People have told me the debian package manager is more sophisticated
>> yet, and can pull up more detailed info. What do I need to read to get
>> on top of those kind of commands?
>
> I've found
> http://qref.sourceforge.net/quick/ch-package.en.html
> to be very helpful.
Oh thanks, this looks like the stuff I was after.
Point 5.3.1 Information on a file
To find the package to which a particular file belongs:
Is just the ticket I think but my trials with it only get an error
message:
A usage statement and "Bash --search command not found" with the
command line indicated there:
dpkg {-S|--search} pattern # search package from installed filename
My actual command line attempts were:
dpkg {-S|--search} /bin/more
dpkg {-S|--search} more
And with single quotes around the curley brackets:
dpkg '{-S|--search}' /bin/more
Maybe it wants a packagename
But seems unlikely since it starts by saying:
To find the package to which a particular file belongs:
dpkg {-S|--search} emacs21
Still no useful output.
Its pretty clear that I'm using it wrong but I did try to follow the
example listed there. What is expected at `pattern'
Re: Terminfo and emacs
Alan Shutko <[EMAIL PROTECTED]> writes: >> Emacs is not finding certain include files for X11 and won't compile >> with X11 support. > > Like redhat's -devel programs, most of the include files in Debian are > in lib*-dev packages. For the X libraries, I currently have installed Ahh cool.. Thanks Alan. I was glad to see your name popping up in the responses recently too. Seems debian has some handy help for many situations. Maybe you'll know about this one: In redhat package manager there is a rpm --replacepkg command that will replace an installed package all over again. I see `apt-get --reinstall' is kind of close. But barfs if files are missing or something. Even with `apt-get install --force-yes package' It barfs if you've scrubbed some of the files it expects to find. That --replacepkg thing in rpm is just for that kind of thing. I suspect apt-get or dpkg or a combination of them can do this too. Specifically the situation was that in shuffling stuff around after trying the emacs install from source, and installing emacs21 with apt-get I mistakenly removed some of the debian files in /usr/share/emacs/site-lisp (I'd mixed in a bunch of my own stuff and lost track of what was from the install and what was mine). Rather than piece it togetther with file date and such, I thought to just reinstall emacs21. But it barfs on the missing files. with --reinstall or --force-yes. I got around it by apt-get remove emacs21 the apt-get install emacs21 but suspect there is a more direct way.
Re: Enough time wasted, moving on
Michel Loos <[EMAIL PROTECTED]> writes: > Mmm, I am quit used to install, 3com NICs and I humbly think you > mixed 509 with 905, there is nothing in the description of the 3c509 > that says it should work with a 3c905. > The 3c509(B/TX) are old 10Mb ISA NICs, while you wher looking for the > 3c905B > a 10/100 PCI NIC. The description of the 3c59x (your driver) talks > somewhere of "newer" drivers, which includes the 3c59x and the 3c9xx > (not cited, this is a real problem). I think it was just a typo on my part. My nic is 3c905. I believe the Modules selections page has it listed wrong. But I haven't been willing to rerun base-config up to that section, to find out for sure. Afraid I will mess up the install I now have running. Maybe there is some way to page thru it with no risk, but some of the screens seem to demand an action be taken and won't allow you to proceed without taking some action. To recap this here: What I'm suggesting is that the screen presented during install that allows you to select a nic driver, has the wrong driver listed adjacent to 3c905. At this point I'm confused enough that I'm not really sure that is even true.
Re: Enough time wasted, moving on
Manoj Srivastava <[EMAIL PROTECTED]> writes: > Harry> You can keep right on with the `hard guy' routine, but I think its > Harry> only fair to warn you that my whole body is a weapon. Hands > Harry> registered on 3 continents. The last guy that called me a juvenile > Harry> banterer is pushing up daisies on a vacant lot in Chicago. :-) > > I see. You are not only juvenile, you are a moron. Only a > moron brings in juvenile machismo on a mailing list, based on some > kinda of idiotic ``my body is a weapon'' bullshit. Even us morons are usually capable of noticing when something is presented as comic relief. But apparently that escaped you.
Re: Terminfo and emacs
Manoj Srivastava <[EMAIL PROTECTED]> writes: > Can you expand on this? Which files are missing? As far as I > can see, when I reinstall emacs21, it re-downloads the .deb for me, > and installs it. (a script/log of the failure would be helpful too) I'd have to reruin the insall now to typescript the whole thing. Not sure I'm competent enough to do that without screwing it up again. But here is the sequence and a simulation of the errors seen. 1) It all began with an attempt to build emacs from source. I ran into troubles first with terminfo not being found (fixed by downloading termcap-compat) and then with missing X files. (Some of the X header files I think. But didn't make a record of it) But since I am more interested in getting this install running in other ways I didn't pursue the source build. I'll do that some other time. Some one has posted all the xstuff needed to make a source build work. 2) Instead, I opted to download and install emacs21. It went off without a hitch. `apt-get' brought down emacsen-common and emac21, Both were installed with no problems. 3) The troubles were all of my own creation. I moved some of my own files into /usr/share/emacs/site-lisp. But then decided to keep them separate from the official install by putting them in /usr/local/share/emacs/site-lisp. 4) Somehow in doing this, I scrubbed: cxref debian-startup.el noweb-mode.el sawfish.el 5) I wasn't sure what all had gotten nailed so I thought to just reinstall to fix it... So ran `apt-get install --reinstall emacs21' It quit with error code 2 and complained about the missing files noted above. But replaced all but the cxref and contents. 6) I then tried apt-get remove emacs21 followed by apt-get install emacs21. Again it quit with code2 mentioning the files above. 7) Tried apt-get remove emacs21 and emacsen-common followed by apt-get install emacs21 and emacsen-common. Again the errors indicated missing cxref especially. 8) Finally I looked around carefully for cxref (a directory) which hadn't really been rm'ed, just moved to the wrong location. Replaced it in /usr/share/emacs/site-lisp. 9) Ran apt-get remove emacs21 and emacsen-common followd by apt-get install emacs21 and emacsen-common. Which worked with no errors. > Umm, unless you have removed some files belonging to dpkg > itself, or packages marked essential (that is the definition of > essential -- packages without which dpkg fails). Apparently cxref/cflow cxref/cxref.el, were the cause of the problem laid out above. I don't think any of the files mentined here are part of emacs source package, but are debian additions.
Re: dpkg -S [was Re: Terminfo and emacs]
Hans Ekbrand <[EMAIL PROTECTED]> writes:
>> Its pretty clear that I'm using it wrong but I did try to follow the
>> example listed there. What is expected at `pattern'
>
> {foo|bar} means foo OR bar
>
> ~>dpkg -S emacs21
>
> will work, as will
>
> ~>dpkg --search emacs21
>
> or
>
> ~>dpkg -S /usr/bin/emacs21
> emacs21: /usr/bin/emacs21
>
> which means that /usr/bin/emacs21 belongs to package emacs21
Egad, maybe I should just take a header off the nearest pier...
I've been looking at too many awk scripts where the curley braces
contain the `program'.
CPAN and apt-get
I wondered if there is a perferred way to handle aquiring perl modules? The CPAN interface shell that gets installed with perl provides a nifty way to search, download, install most of what is on CPAN and can be enhanced with WAIT (a WAIS based search interface) All this is pretty easy to use and takes care of some of the work of finding and applying perl modules and packages. Seems like this would interfere to some extent with apt-get concepts of how to do things. Whats the normal way to handle this?
emacs21
A fairly long-time emacs user, I'm accustomed to using a site-start.el file as a place to do general things. In my single user system it can be quite a lot since no one else is stuck with it. Having installed the emac21 package, it seems my site-start.el file from other installs elsewhere, is not recognized as an init file and is not loaded eventhough it is in the `path' and its directory /usr/share/emacs/site-lisp shows up in C-h v load-path However changing its name to default.el causes it to be read as an init file. I don't see this behavior on several other installations of emacs-21 (from source) on other machines. I know default.el has long been a name that emacs recognizes as init but so has site-start.el, far as I know. I wondered if this change is something to do with my installation or if it is something debian has setup. Also curious about the the way files get installed with emacs21 and emacsen-common packages. I see it sets up /usr/share/emacs with subs of site-lisp and emacs-21 when the default is /usr/local/share/emacs/ . It also creates the ones under /usr/local but they are empty. Why both? Seems like, if the official deb emacs21 is going to use /usr/share then it would leave /usr/local/share for possible source built emacs for developement or what ever. That would provide a handy way to keep the separate and head off cluttered path type problems. And the files that get created under /etc/emacs/ site-start.el and site-start.d/ Yet another site-lisp sort of area. Giving a total of (5) site-lisp kind of directories 4 actual and 1 symlink. Probably good reasons for all this, I'm just curious what they are?
What shows what an uninstalled package does?
If I want to find out about a package. I mean, like what it actually does, what is the right command, when the package is not installed but apt-cache knows about it? `apt-cache showpkg ' gives some information but doesn't tell you what the package does. If its not installed then dpkg -I doesn't know about it. Combing thru `man apt-cache' I found the -f (full) operator too but `apt-cache -f showpkg still doesn't really tell what a package does. I don't see anymore choices in `man apt-cache' that look likely. `man apt-get' doesn't indicate a such an option either. In case any former redhat users see this, I'm looking for something that works like `rpm -qip ' (which also works on ftp sites that allow it) It gives a general blurb that tells what the package does. I guess the next thing after learning what packages do that one has in cache would be to be able to learn what packages do by name, where ever they are. That is, an http or ftp interface that pulls down that info.
Re: What shows what an uninstalled package does?
Martin Hermanowski <[EMAIL PROTECTED]> writes: >> `apt-cache showpkg ' gives some information but doesn't tell you >> what the package does. If its not installed then dpkg -I doesn't know >> about it. >> >> Combing thru `man apt-cache' I found the -f (full) operator too but >> `apt-cache -f showpkg still doesn't really tell what a package >> does. > have you tried apt-cache show? Yikes.. I guess the name is kind of a giveaway but the definition in the man page doesn't come across as giving that kind of information: show show performs a function similar to dpkg --print- avail, it displays the package records for the named packages. When I saw the part about `records' I just assumed it was stuff like size version etc. One expects a term like `summary' or the like for that kind of information. Unless you already know what dpkg --print-avail does. Not a good assumption in a man page. Thanks, I should have tried it I guess but that discription just made me keep looking.
Re: What shows what an uninstalled package does?
Manoj Srivastava <[EMAIL PROTECTED]> writes: > > Harry> Yikes.. I guess the name is kind of a giveaway but the definition in > Harry> the man page doesn't come across as giving that kind of information: > > Incidentally, if you happen to use dselect (I use dselect with > apt as a backend), looking through /var/lib/dpkg/available is quite > informative (especially when I am unsure of what I am looking for and > just use the regexp search in emacs or less). My one go around with dselect during my first installation attempt caused me to want to shoot myself after a few minutes. But once you start getting used to the keybindings it starts to work. What does it mean to use dselect with apt as backend?
dh-make-perl
What is the trick to running dh-make-perl so that it works: Here I said mkdir /usr/local/cpan cd there and said dh-make-perl --cpan File::Find Now the manpage gives this syntax: dh-make-perl [module_source_dir|--cpan module] Is there special meaning to [module_source_dir| Am I supposed to be a specific directory? I looked at the files installed for perl-base with dpkg -L and none of the directories look any more a module_source_dir than any other. I guess it just means where ever I decide to build things so that was /usr/local/cpan When I ran that command a hefty perl.tar.gz was downloaded with out me asking for it. After it was unpacked I get this message: [...] snipped massive list of unpacked files perl-5.6.1/x2p/walk.c perl-5.6.1/xsutils.c Cannot open /usr/local/cpan/build/perl-5.6.1/Makefile.PL ls /usr/local/cpan/build/perl-5.6.1/M* shows that sure enough its not there. But there is a /usr/local/cpan/build/perl-5.6.1/Makefile.SH It looks suspiciously like thats what need to be run but before doing so I wanted to see if any one knew what the deal was with this.
Re: dpkg - the letter prefixes.. what do they mean
"Mark S. Reglewski" <[EMAIL PROTECTED]> writes: >> I think that would *really* clear things up a lot. Maybe the >> explanations could even contain some more spaces. > > Excellent. Easier to read and understand than any of the ascii art > versions. dselect will still strike terror into the hearts of > newbies everywhere, but this is better. Just a comment from the peanut gallery here. I'm the newbie who started the thread, and after it was explained by dman I liked the ascii art. The number based one doesn't seem as good to me. What I think is missing is a brief explanatoin of how it works. The the ascii works if its explained. I guess the numbers would too, but I don't thing you can really escape an explanation. AT least at that secion (-l I think), in the manpage. There should be a full explanation of it.
Small bug in dh-make-perl? [was dh-make-perl]
Harry Putnam <[EMAIL PROTECTED]> writes: > When I ran that command a hefty perl.tar.gz was downloaded with out me > asking for it. After it was unpacked I get this message: > [...] snipped massive list of unpacked files > perl-5.6.1/x2p/walk.c > perl-5.6.1/xsutils.c > Cannot open /usr/local/cpan/build/perl-5.6.1/Makefile.PL > ls /usr/local/cpan/build/perl-5.6.1/M* shows that sure enough its not > there. But there is a /usr/local/cpan/build/perl-5.6.1/Makefile.SH > > It looks suspiciously like thats what need to be run but before doing so > I wanted to see if any one knew what the deal was with this. This is beginning to look like a small bug. Here is why: Calling this command: dh-make-perl --cpan Getopt::Std >From the cpan directory explained in previous post. Ends in the error mentioned: [...] snipped massive list that gets scanned perl-5.6.1/x2p/walk.c perl-5.6.1/xsutils.c Removing previously used /usr/local/cpan/build/perl-5.6.1 Cannot open /usr/local/cpan/build/perl-5.6.1/Makefile.PL Now invoking the CPAN shell directly: perl -MCPAN -e shell then install Getopt::Std Does the same thing but installs it with no errors. So apparently the dh-make-perl overlay is causing this error some how. I need some of the perl modules installed right now for a school project so am installing them directly with the CPAN shell, until I figure out if its my local technique that is the problem. I wonder if doing this without dh-make-perl, sort of behind debians back, will cause me some grief in some way later?
Re: testing symbolic links
Angus D Madden <[EMAIL PROTECTED]> writes:
> Does anyone know a of utility to check if a symbolic link is valid?
>
> I used a program called vpopbull (part of vpopmail package) to send a
> notice to a bunch of users. To my surprise, vpopbull ignored the
> complete path I gave it and created links to a message that doesn't
> exist. Now, when users try to open the message, it gives them an error.
>
> If there's a utitlity I can use with find to test the validity of links
> I could clear the problem up right away.
I'm new here so not sure if there is a deb package but there is a
program called symlinks. It does exactly what you want to do. You
just aim it at a directory and it recurses thru and gives various
reports depending on the flags you give it. Probably easy to find out
if there is a debian package but I didn't know the technique yet.
That aside you can use the -type l flag to find for this too.
find /path/to/tree -type l -exec ls -l {} \;
will give you a long ls on each of the links it finds.
(thats a lower case el (l) in there)
Re: What shows what an uninstalled package does?
Phillip Deackes <[EMAIL PROTECTED]> writes: > Try installing Storm Package Manager (stormpkg) - a GUI to apt. Very nice > - I always use it to find info about packages. It is also very good at > highlighting optional packages which the command-line apt-get does not do. OK, thanks. Sounds like a nice tool. Most of my work or at least a lot of it is done over ssh. I'm probably more comfortable at the command line than anywhere else from working like that a lot. But it would come in handy to have a nice gui tool at times too.
Re: How do I turn off the Virtual Desktop?
Peak Allan <[EMAIL PROTECTED]> writes: > I want the whole screen visible on my monitor. > > Please cc me since I'm not currently subscribed to debian-user. This won't be much help to you but I want just the reverse. I'm used to having a huge virtual desktop (1600 1200) but being a newby to debian I haven't figured out how to edit /etc/X11/XF86Config-4 to get that result. I know my monitor is capable of it since I've had it working under redhat for a couple of years I added something like this from a config on an old redhat install but it refused to startx then. I'm pretty busy configuring everthing so I just went back to a config that works and haven't worked on it since. But I do want my big desktop back. Subsection "Display" Depth 16 Virtual 1600 1200 Modes "1152x864" "1280x1024" "1400x1050" "1600x1200" "1024x768" Maybe you could use most of my config since it does not give the virtual desktop. Let me know if you want a copy.
Re: emacs21
Gary Turner <[EMAIL PROTECTED]> writes: > Hi Harry, try here [EMAIL PROTECTED] I've seen all > kinds of discussions on just this type of question. Some serious emacs > dev types hang out there. If you can't get an answer there, then the > only answer is "42" Oh, thanks, I didn't know about that list. But your reference to "42" sailed right over my head...
Re: How do I turn off the Virtual Desktop?
John Foster <[EMAIL PROTECTED]> writes: > SubSection "Display" > Depth 24 > Modes "1024x768" "1280x1024" "640x480" "800x600" > Virtual 1024 768 > Viewport0 0 I hope that helped Peak as much as it did me. I got my big desktop by using your technique but adjusting the numbers thanks
The quest for rodent power
Running woody (testing) I've been ignoring this issue and working on other things but I really miss having my little rodent in text mode. Its a logitec 3 button ps2, critter that works in X just fine. I can't remember how to tell with no doubt where the mouse is connected. At which device. I remember saying /dev/psaux during install but that was just a guess. Boot messages indicate gpm services are being started. But I have no functioning mouse in text mode. A grep of ps waux shows" ps waux|grep 'gpm' turns up nothing The devices directory itself shows ls -l /dev/ |grep 'mouse\|psaux\|gpm' prw-r--r--1 root root 0 Mar 1 15:26 gpmdata lrwxrwxrwx1 root root 7 Feb 28 21:41 mouse -> gpmdata crw---1 root root 10, 1 Nov 4 14:52 psaux crw-rw1 root root 10, 32 Nov 4 14:52 usbmouse >From posts here, I got the idea that the XF86Config mouse section is related to text mode mouse settings in some way. That section of mine looks like this: Section "InputDevice" Identifier "Configured Mouse" Driver "mouse" Option "CorePointer" Option "Device""/dev/psaux" Option "Protocol" "PS/2" Option "Emulate3Buttons" "true" Option "ZAxisMapping" "4 5" EndSection Section "InputDevice" Identifier "Generic Mouse" Driver "mouse" Option "SendCoreEvents""true" Option "Device""/dev/input/mice" Option "Protocol" "ImPS/2" Option "Emulate3Buttons" "true" Option "ZAxisMapping" "4 5" EndSection How can I tell beyond doubt which device the mouse is on. And get this thing configured and working.
Re: Small bug in dh-make-perl? [was dh-make-perl]
"Sean 'Shaleh' Perry" <[EMAIL PROTECTED]> writes: >> >> So apparently the dh-make-perl overlay is causing this error some how. >> >> I need some of the perl modules installed right now for a school >> project so am installing them directly with the CPAN shell, until I >> figure out if its my local technique that is the problem. >> >> I wonder if doing this without dh-make-perl, sort of behind debians >> back, will cause me some grief in some way later? >> > Put everything in /usr/local/lib/site_perl. Once the immediacy of your need > has gone away, you can spend time looking for a better solution. I would have > sworn File::Find was packaged. On my install, even perldoc was not available. Once added, calling perldoc File::Find was the test I used to see if it was available. It wasn't. But maybe more a shortcoming of my install technique than anything else. > The biggest issue you will face is if another package depends on say > File::Find > it will install it and you will have two possibly differing versions of a perl > module installed. Ok, that could be easily resolved I think so I guess its no biggie.
Re: procmail rule for html mail
Rob VanFleet <[EMAIL PROTECTED]> writes: > It seems like this has come up before, but I couldn't turn anything up > from searching. Basically, I am looking for a procmail rule that will > detect html mail, and pipe it to a script to strip the tags from it, > preferably before the other rules are applied, so it still ends up in > the proper mailbox, but that is not a necessity. One way to do it: 1) Scan a few of the kind of mails you want to do this to. Look for a `Contents-Type': header. Here is how one of the might look: Content-Type: text/html; charset=us-ascii. There are others This one has both a text and html version of the message. Once you have determined what content headers you need to grab: (Assuming you've already worked out the details of a functioning .procmailrc) :0 f * ^Content-Type: | my_stipper_script The `f' indicates that you want this to be a filter only. Not a delivery recipe. The message will be filtered thru your script then sent on to what ever else you want
Re: Enough time wasted, moving on
Petro <[EMAIL PROTECTED]> writes: > On Fri, Mar 01, 2002 at 11:48:18PM -0800, Harry Putnam wrote: >> You can keep right on with the `hard guy' routine, but I think its >> only fair to warn you that my whole body is a weapon. Hands >> registered on 3 continents. The last guy that called me a juvenile >> banterer is pushing up daisies on a vacant lot in Chicago. :-) > > Crap. There are no vacant lots in Chicago with daisies in them. > Hehe .. good one. Somebody realized it was a joke.
Re: The quest for rodent power
"Karl E. Jorgensen" <[EMAIL PROTECTED]> writes: > On Mon, Mar 04, 2002 at 07:03:14AM -0800, Harry Putnam wrote: >> Running woody (testing) >> >> I've been ignoring this issue and working on other things but I really >> miss having my little rodent in text mode. >> >> Its a logitec 3 button ps2, critter that works in X just fine. >> >> I can't remember how to tell with no doubt where the mouse is >> connected. At which device. I remember saying /dev/psaux during install >> but that was just a guess. >> >> Boot messages indicate gpm services are being started. But I have no >> functioning mouse in text mode. >> >> A grep of ps waux shows" >> ps waux|grep 'gpm' turns up nothing > > Odd. Perhaps there is a bug (gasp!) in gpm? > > To get more diagnostic info, try these after *stopping* your X server > (switching to a text-mode VT ain't enough): > # /etc/init.d/gpm stop > # /etc/init.d/gpm start > > and keep an eye on what gets added /var/log/syslog. `Stop' caused this to be printed: "DATE HOST /usr/bin/gpm[1837]: Removing stale pid file \ /var/run/gpm.pid" `Start' printed nothing whatever in log but did say it was started at the command line command prompt. > > Posting /etc/gpm.conf will probably help too. > > But since the mouse works in X, I assume that that X reads from > /dev/psaux. Edited gpm.conf: device=/dev/psaux responsiveness= repeat_type=raw type=ps2 append="" sample_rate= [...] Thanks for the tutorial overview of how it works >> How can I tell beyond doubt which device the mouse is on. Is there a difinitive answer to that... some canonical way to know for sure where the mouse is connected? Is that likely to be a source of problems here? > > /etc/gpm.conf: > device=/dev/psaux > repeat_type=raw > > /etc/X11/XF86Config-4 > Section "InputDevice" > Identifier"gpm-mouse" # human-readable ID > Driver"mouse" > Option"Device""/dev/gpmdata" > Option"Protocol" "PS/2" > # Add other options according to personal taste... > EndSection Edited version now is: Section "InputDevice" Identifier "Configured Mouse" Driver "mouse" Option "CorePointer" Option "Device""/dev/gpmdata" Option "Protocol" "PS/2" Option "Emulate3Buttons" "true" Option "ZAxisMapping" "4 5" EndSection What does the ZAxisMapping thing mean? > Section "ServerLayout" > ... > InputDevice "gpm-mouse" "CorePointer" > ... > EndSection In that ServerLayout section, did you mean to only have one line devoted to mouse? And others as they are? My current setup there is: Section "ServerLayout" Identifier "Default Layout" Screen "Default Screen" InputDevice "Generic Keyboard" InputDevice "Configured Mouse" InputDevice "Generic Mouse" EndSection Now changed to: Section "ServerLayout" Identifier "Default Layout" Screen "Default Screen" InputDevice "Generic Keyboard" InputDevice "gpm-mouse" "CorePointer" # InputDevice "Configured Mouse" # InputDevice "Generic Mouse" EndSection > I.e. stop X from reading from /dev/psaux, and tell it to read from > /dev/gpmdata instead (use the same protocol & options as those that work > for you now, no need to change them). No X running, and made the edits above. Rebooted and still have no mouse in text console, and now X crashes magnificiently with `startx' Hopefully there are the most relevent messages: Module Loader present (==) Log file: "/var/log/XFree86.0.log", Time: Mon Mar 4 19:20:31 2002 (==) Using config file: "/etc/X11/XF86Config-4" Data incomplete in file /etc/X11/XF86Config-4 Undefined InputDevice "gpm-mouse" referenced by ServerLayout "Default Layout". (EE) Problem parsing the config file (EE) Error from xf86HandleConfigFile() Fatal server error: no screens found [...] I'll guess that the `screens' thing is really brought on by the first error since that section seems to be in tact. Of course I saved the original files so can return to square one easily.
Generic file finder question ..what tool or database
If I have the name of a file like xpm.h and want to know what package
contains it. And I find that `apt-cache search xpm.h' doesn't know
about it.
And dpkg -S xpm.h doesn't either
I'm assuming that means it isn't here but might be in a
package in the distro on line somewhere.
Where is the database with this kind of information?
search.debian.org isn't the right place I don't think
One can compile a (rough) list of all file names in the install with:
Something like:
dpkg -L $(dpkg -l|awk '{gsub(/^/,"",$0);print $1}')
And with a little script, print the package name on each line too.
But where is a database that contains that kind of information for all
packages in a distribution? That is, a list of all the file names in
all the packages of a distribution.
Re: The quest for rodent power
John Foster <[EMAIL PROTECTED]> writes: > Harry Putnam wrote: > > >> > But since the mouse works in X, I assume that that X reads from >> > /dev/psaux. >> >> Edited gpm.conf: >> >> device=/dev/psaux >> responsiveness= >> repeat_type=raw >> type=ps2 >> append="" >> sample_rate= >> >> [...] Thanks for the tutorial overview of how it works >> >> >> How can I tell beyond doubt which device the mouse is on. > - > see output of "dmesg" Thought I had already mentioned that neither dmesg or /var/log/messages has any info about gpm. grep gpm /var/log/messages only shows the stale pid being removed dmesg|grep gpm What I'm asking for here is how to know what theh mouse is supposed to be on if all the easy ways are not telling me. >> Edited version now is: >> >>Section "InputDevice" >>Identifier "Configured Mouse" >>Driver "mouse" >>Option "CorePointer" >>Option "Device""/dev/gpmdata" I completely screwed up the advice given in the above edit: It should have been like below to follow the advice given: Section "InputDevice" Identifier "gpm-mouse" Driver "mouse" Option "CorePointer" Option "Device""/dev/gpmdata" Option "Protocol" "PS/2" Option "Emulate3Buttons" "true" Option "ZAxisMapping" "4 5" EndSection Which I did try. The result was no gpm mouse. X starts but has no mouse. > > the line above should be "/dev/mouse" /dev/mouse is a symlink to gpmdata > and you should have a hard link in the /dev dir for @mouse that points > to /dev/psaux > or depending on your version of gpm you may also have /dev/@gpmdata as > the pointer to /dev/psaux...either one should work. /etc/gpm.conf is set to /dev/psaux. So the X mouse cannot be ... right? Here is the current situation in /dev ls -l /dev |grep 'mouse\|psaux\|gpm' prw-r--r--1 root root0 Mar 4 23:56 gpmdata lrwxrwxrwx1 root root7 Feb 28 21:41 mouse -> gpmdata crw---1 root root 10, 1 Mar 4 23:45 psaux crw-rw1 root root 10, 32 Nov 4 14:52 usbmouse How should it look by your technique? >> What does the ZAxisMapping thing mean?
Re: The quest for rodent power
Carel Fellinger <[EMAIL PROTECTED]> writes: [...] >> > see output of "dmesg" >> >> Thought I had already mentioned that neither dmesg or >> /var/log/messages has any info about gpm. >> grep gpm /var/log/messages only shows the stale pid being removed >> dmesg|grep gpm > > Just curious, but what does `ps ax | grep gpm' tell, is there a running > gpm after you started it? (and be aware its gpm not gmp:) The message that opened this thread contained that data: Message-Id: <[EMAIL PROTECTED]> > A grep of ps waux shows" > ps waux|grep 'gpm' turns up nothing That was true then and is still true after the edits discussed.
Re: The quest for rodent power
Vineet Kumar <[EMAIL PROTECTED]> writes: > * Harry Putnam ([EMAIL PROTECTED]) [020305 00:29]: >> /etc/gpm.conf is set to /dev/psaux. So the X mouse cannot be >> .. right? > > Right. Also, as I've said before, adding a symlink in the mix can only > confuse things further. Forget about a /dev/mouse symlink. Tell gpm to > use /dev/psaux and tell X to use /dev/gpmdata. See the links below. Not sure I see what confusion it causes but it does seem extraneous. Are suggesting it is the problem here? >> Here is the current situation in /dev >> ls -l /dev |grep 'mouse\|psaux\|gpm' >> prw-r--r--1 root root0 Mar 4 23:56 gpmdata >> lrwxrwxrwx1 root root7 Feb 28 21:41 mouse -> gpmdata >> crw---1 root root 10, 1 Mar 4 23:45 psaux >> crw-rw1 root root 10, 32 Nov 4 14:52 usbmouse >> >> How should it look by your technique? > > I explained this in an email a few weeks back with subject line "mouse > freezes" back on Feb 13; check the archives. Actually, with a little > googling, it seems I've explained this more than once already =) > > See these, please: > http://groups.google.com/groups?selm=20010627135905.A10128%40doorstop.net > http://groups.google.com/groups?selm=20020214082809.GA947%40doorstop.net > ([EMAIL PROTECTED] seems to be missing from > groups.google.com, but it's quoted (albeit rather rudely) here:) > http://groups.google.com/groups?selm=20020213220924.GN16216%40ufies.org After reading thru those, it seems the setup I posted conforms to those suggestions. You say "(Start by reading Joost's message regarding mouse hardware.)" Joost says if the hardware connector is small and round it is PS/2. Thats what I've used. Everyone seems to agree if it is ps2 /dev/psaux should be the device: My current (non-working) gpm.conf: device=/dev/psaux responsiveness= repeat_type=raw type=ps2 append="" sample_rate= My current XF86Config (in part) Section "InputDevice" Identifier "gpm-mouse" Driver "mouse" Option "CorePointer" Option "Device""/dev/gpmdata" Option "Protocol" "PS/2" Option "Emulate3Buttons" "true" # Option "ZAxisMapping""4 5" (This is now commented) EndSection Two possible places where they might diverge. 1) I didn't rm /dev/mouse (The symlink) But I don't think you are suggesting that is the main problem. 2) More important probably. You talk about both configs (gpm.conf and XF86Config-4) Referencing the same mouse protocol but never really say what that is. For example: The gpm.conf I posted says `raw' for the repeat option and as you note in those posts is alright. But you never say what the X config should say... only that it should be the same. But don't think you mean `raw' in the X config. >From your post:Message-ID: <[EMAIL PROTECTED]> > Alternatively, you can specify any protocol you > like in your gpm repeater config, as long as you tell X to use the > same protocol. As someone else suggested, you can tell gpm to repeat > as type 'raw' in which case you should configure X to use the actual > protocol that your mouse is speaking. By `actual protocol' you mean PS/2 .. right? Can you say point blank what you see as wrong with the above? It seems to conform to your outline. But, I've been staring at it too long and probably missing some important difference. The final piece of the puzzle: Section "ServerLayout" Identifier "Default Layout" Screen "Default Screen" InputDevice "Generic Keyboard" InputDevice "gpm-mouse" "CorePointer" # InputDevice "Generic Mouse" EndSection Note Generic Mouse is commented out along with that entire section from further up in the config. Like this: ## Section "InputDevice" ##Identifier "Generic Mouse" ##Driver "mouse" ##Option "SendCoreEvents" "true" ##Option "Device""/dev/input/mice" ##Option "Protocol" "ImPS/2" ##Option "Emulate3Buttons" "true" ##Option "ZAxisMapping""4 5" ## EndSection With that setup, and in console mode running this command: # /etc/init.d/gpm force-reload Stopping mouse interface server: gpm. Starting mouse interface server: gpm. # ps wuax|grep gpm Appears as if gpm is started but nothing shows in ps output and mouse doesn't work. Mouse doesn't work in X with these settings either. >> >> What does the ZAxisMapping thing mean? > > ZAxisMapping is used to tell X which buttons you want to use for your [...] Thanks, Note above its commented out now.
Re: The quest for rodent power
Carel Fellinger <[EMAIL PROTECTED]> writes: > On Tue, Mar 05, 2002 at 08:38:38AM -0800, Harry Putnam wrote: > ... >> The message that opened this thread contained that data: > > sorry, missed that one. You see, I never got mutt to behave like > tin with respect to automatically deleting old stuff and espescially > hiding/unhiding read messages. So I basicly delete all messages > read or deemed uninteresting thinking I can get them from the > archives at any time. Unfortunately that anytime has a long initial > delay:( He He I bet it does. But this thread had gotten quite long so I've reposted some of the pertinent data. No one wants to back all the way up and read all that. > As an alternative route, did you try gpmconfig and its testing mode? > Maybe you'll get better error messages there. Not sure what you mean by testing mode. I did try increasing the verbosity level with -V+9 syntax but couldn't make heads or tails of what showed up in syslog output. The -D switch is mentioned as being helpfull for debugging but I get no output at all with /usr/sbin/gpm -D just a hung session
Re: Generic file finder question ..what tool or database
Johann Spies <[EMAIL PROTECTED]> writes: > On Mon, Mar 04, 2002 at 11:25:06PM -0800, Harry Putnam wrote: >> >> If I have the name of a file like xpm.h and want to know what package >> contains it. And I find that `apt-cache search xpm.h' doesn't know >> about it. > > Install auto-apt, do auto-apt update and then you can search like this: > > $ auto-apt search -f xpm\.h > usr/X11R6/include/X11/xpm.h devel/xlibs-dev > usr/share/doc/gnome-dev-doc/html/C/gnome-stock-xpm.html > doc/gnome-dev-doc > usr/share/doc/phpdoc/html/function.imagecreatefromxpm.html doc/phpdoc > usr/share/fsviewer/icons/xpm/home.xpm x11/fsviewer-icons > usr/share/gtktalog/xpm/help.xpm > misc/gtktalog > usr/share/solfege/xpm/happyface-32x32.xpm sound/solfege > usr/share/solfege/xpm/harmonicintervall24.xpm sound/solfege > usr/share/solfege/xpm/harmonicprogression24.xpm > sound/solfege > usr/share/solfege/xpm/holder.xpm sound/solfege > usr/X11R6/include/X11/xpm.h devel/xlibs-dev I haven't actually tried the above, but my question was about finding a file that is not on the system. There are several ways to find those that are. Does that technique off that option?
Re: Generic file finder question ..what tool or database
Johann Spies <[EMAIL PROTECTED]> writes: > > Install auto-apt, do auto-apt update and then you can search like this:$ Thanks Johann. Thats cool. I posted a dumb reply moments ago without having tried this. Please ignore it.
Re: The quest for rodent power
Carel Fellinger <[EMAIL PROTECTED]> writes: >> With that setup, and in console mode running this command: >> # /etc/init.d/gpm force-reload >> Stopping mouse interface server: gpm. >> Starting mouse interface server: gpm. >> # ps wuax|grep gpm >> >> Appears as if gpm is started but nothing shows in ps output and mouse >> doesn't work. >> >> Mouse doesn't work in X with these settings either. > > ofcourse not, because X depends on gpm to repeat the mouse events to > /dev/gpmdata. Without gpm there won't be anything to read from it:) I'm getting really confused here. With the most recent setup I posted I get no mouse in either console or X. You say that is right since gpm isn't running. But in my original setup (posted againg below), X mouse does work but there is no gpm running there either. Sorry to be so dense here, but I think I'm still missing some fundemental aspect of this. Settings that do work in X only: First we establish that gpm is not running: root # /etc/init.d/gpm force-reload Stopping mouse interface server: gpm. Starting mouse interface server: gpm. root # ps waux|grep gpm Gpm isn't running /etc/gpm.conf device=/dev/psaux responsiveness= repeat_type=ms3 type=ps2 append="" sample_rate= /etc/X11/XF86Config-4 [...] Section "InputDevice" Identifier "Configured Mouse" Driver "mouse" Option "CorePointer" Option "Device""/dev/psaux" Option "Protocol" "PS/2" Option "Emulate3Buttons" "true" Option "ZAxisMapping""4 5" EndSection Section "InputDevice" Identifier "Generic Mouse" Driver "mouse" Option "SendCoreEvents" "true" Option "Device""/dev/input/mice" Option "Protocol" "ImPS/2" Option "Emulate3Buttons" "true" Option "ZAxisMapping""4 5" EndSection [...] Section "ServerLayout" Identifier "Default Layout" Screen "Default Screen" InputDevice "Generic Keyboard" InputDevice "Configured Mouse" InputDevice "Generic Mouse" EndSection [...] Mouse works in X with above settings.
Re: [Harry Putnam ] Re: Terminfo and emacs
Manoj Srivastava <[EMAIL PROTECTED]> writes: > If this analysis is correct, there is no bug in emacs (well, > there is nothing emacs21 packages can do to avoid this). Just don't > go removing dirs ;-). I generally put things in /usr/share/emacs/site-lisp, > which takes precedence over paths in /usr/share/emacs; and let the > files that belong to packages be (or remove the debian package, and > use the local package only). I just finished reading the debian-policy thing on emacs. So have a little better idea what the game plan is now. I think there are some problems with it, but not really prepared to give a full analysis. One thing you might consider is that a regular emacs package from source would have had no problems with the directory having been moved. It would just have recreated it and gone on about its business. I noticed that my site-start.el file that has built up over a few years and worked in many places would not load as an init file even when placed in /usr/share/emacs/site-lisp. Which is contrary to emacs defaults. I had to change the name to `default.el' before it would load on init. It still doesn't load first, which again is not defaut and contrary to the emacs documentation (I think, but haven't researched it fully) While not major issues it seems a bad move to change stuff that is documented in the info files. Unless of course there is a good reason. A further thing I noticed in the policy statement was that some things are done by debian even if emacs is started -q -no-site-file. Which again breaks those commands and runs contrary to the documentation. That is supposed to give a fully vanilla emacs. Sometimes needed.
Re: The quest for rodent power
Simon Hepburn <[EMAIL PROTECTED]> writes: > Looks reasonable.try this from console as root: > > #gpm-mouse-test > > Anything useful ? If not, re-run gpmconfig, when you get to mouse type, type > help. You should see a list of all the supported mice types including some > odd ps2 ones. Suck 'em and see. I get this output: This program is designed to help you in detecting what type your mouse is. Please follow the instructions of this program. If you're bored before it is done, you can always press your 'Interrupt' key (usually Ctrl-C) *** Remember: don't run any software which reads the mouse device *** while making this test. This includes "gpm","selection", "X" Note that this program is by no means complete, and its main role is to detect how does the middle button work on serial mice /dev/atibm: No such device /dev/inportbm: No such device /dev/jbm: No such device /dev/logibm: No such device /dev/usbmouse: No such device Trying with 1200 baud The possible device nodes are: /dev/agpgart At that point it just hangs. The message does say it si for testing the middle button on serial mice.
Re: The quest for rodent power
Vineet Kumar <[EMAIL PROTECTED]> writes: > /usr/sbin/gpm -D -m /dev/psaux -t ps2 -Rraw > > The -D should be very helpful here; it should keep gpm running in the > foreground and log messages to stderr. Leave that going on a console to > see what's going wrong. If gpm just won't start at all, that's a > problem. That gives me some almost laughable output: root # /usr/sbin/gpm -D -m /dev/psaux -t ps2 -Rraw Alarm clock But still not running: root # ps waxu|grep gpm Also to clear the record here. Some are suggesting /etc/init.d/gpm is not present or other such basic things. I think the fact that this command has any output at all indicates that all is well in that quarter: root # /etc/init.d/gpm force-reload Stopping mouse interface server: gpm. Starting mouse interface server: gpm. Its just that nothing actually happens.
Re: The quest for rodent power
Carel Fellinger <[EMAIL PROTECTED]> writes: > On Tue, Mar 05, 2002 at 01:29:20PM -0600, Gary Turner wrote: > ... > > Just to make this one stick out amidst all advice given sofar! > >> Be sure that /etc/init.d contains gpm. Check the script that it hasn't >> been disabled, eg. an exit 0 thrown in to stop the script. > > Indeed check that /etc/init.d/gpm is sane. > > Do I remeber correctly that without gpm and X config pointing to > /dev/psaux you had a working mouse in X? > Yes, both configs the ones that work for X and the ones that don't have been posted in the last 2 hours or so. Maybe a little longer by now. One further point here: Concerning the validity of the hardware itself. Where this debian install now resides, there was a redhat 7.1 basic system (no X) installed. Using all the same hardware, that system came up with a working mouse as soon as the install was complete. This is not to compare OS's or say who's better or any of that bunk. But it should indicate the hardware isn't at fault here.
Re: AVI/DivX -> MPEG1 converter?
will trillich <[EMAIL PROTECTED]> writes: > there's http://packages.debian.org/ which > may help. see newbieDoc.sf.net/system/apt-get-intro.html (items > 7 and 12, specifically) for more ideas... Probably blind but I don't see anything there about finding files that are not installed. (not packages... files)
Re: Generic file finder question ..what tool or database
Simon Hepburn <[EMAIL PROTECTED]> writes: > On Tuesday 05 Mar 2002 7:25 am, Harry Putnam wrote: >> If I have the name of a file like xpm.h and want to know what package >> contains it. And I find that `apt-cache search xpm.h' doesn't know >> about it. > > apt-cache search only checks package names and descriptions > >> And dpkg -S xpm.h doesn't either > > It does if you have the relevant package installed I'm not talking about installed packages > #dpkg -S xpm.h > #xlibs-dev: /usr/X11R6/include/X11/xpm.h > > Otherwise try here: > > http://www.debian.org/distrib/packages#search_contents Thanks .. thats the baby. What about listing packages in cache that are not installed Browsing over man apt-get apt-cache dbkg I didn't notice a command that lists all the packages onboard that are not installed.
Re: AVI/DivX -> MPEG1 converter?
Harry Putnam <[EMAIL PROTECTED]> writes: > will trillich <[EMAIL PROTECTED]> writes: > >> there's http://packages.debian.org/ which >> may help. see newbieDoc.sf.net/system/apt-get-intro.html (items >> 7 and 12, specifically) for more ideas... > > Probably blind but I don't see anything there about finding files that > are not installed. (not packages... files) Wow ... not sure how the mangling happened but this message began life as a reply to Will Trillich From: will trillich <[EMAIL PROTECTED]> Subject: Re: Generic file finder question ..what tool or database To: debian-user@lists.debian.org Date: Tue, 5 Mar 2002 14:28:03 -0600 Message-ID: <[EMAIL PROTECTED]>
More cheese nibbler questions
Setup: Stock woody install from testing Kernel 2.2.20 Comments: X is installed but not running for these tests Mouse works in X but not in console If I say: # /etc/init.d/gpm force-reload Stopping mouse interface server: gpm. Starting mouse interface server: gpm. at the command line then `ps waux|grep gpm ' should show some thing with gpm in it ...right? Should lsmod also show a module running that is mouse realated? Neither of these happen here. Is it likely that upgrading to a newer kernel or reconfiguring the present one will help this situation? cat /etc/gpm.conf device=/dev/psaux responsiveness= repeat_type=ms3 type=ps2 append="" sample_rate=
Re: unsubscribe
"Karsten M. Self" writes: > shocked The fix in this case is to manually delete the > offending mail. If you don't have shell access to the mail server > (as the admin, I ;-), you'll have to scrap together enough > info on the POP mail protocol to list and delete messages. I've You can get quick example dialog for telnetting to lots of things by typing This into a google search box: "I sit in siberia"
Re: Enough time wasted, moving on
Corrin Lakeland <[EMAIL PROTECTED]> writes: > I agree, and if Harry had criticised the CDs then I would have refuted that. > But, looking at Harry's problems, none of them were caused by the > installation CD which (by luck more than anything) appeared to work > perfectly. So, I don't think it is far to complain he used unofficial, > unreleased CDs. Looking back at it now that I've successfully done the install and moved on. I think the biggest single stumbling block was the failure to find the nic. And the confusing (possibly inaccurate) section where you pick nic drivers. As Corrin and others have pointed out, without a network things are much more difficult. Once someone steered me to the correct nic driver my CDs were fine to get the basic system. Then editing sources.list and installing woody from testing was almost a snap. :-). Problems that followed were minor, but again driver or hardware related. The gpm issue which seems to crop up quite a lot. Still haven't resolved that but it is less significant. Since it does work in X. But here again its something that other linuxs and BSDS have found without major difficulty. Both the mouse setup and the nic. Once a user gets his hands on apt-get and its friends they will soon see what debian is all about. And why some people swear by it. There is truly a vast array of tools here for managing the OS. But as is always the case with a large arsenal, it takes some learning. But back to the problems for a moment: Its fine to say `Debian is not aimed at the same crowd' or similar. That is a fine thing to do, but still things like gpm are so simple a part of other installs, it must mean something is different there. It has apparently been a problem for quite some time. Not earthshaking but there. A final note here. I see some posters are still reposting things said in this thread, maybe some of you missed my outright appology to the list: Message-Id: <[EMAIL PROTECTED]> Please note too that since getting an install running I've documented and filed 2 official bug reports and a third semi-official one. These are not lifechanging issues but do reflect an attemp to pitch into the general pool here.
Ick.. More fresh Crow
More fresh Crow to be eaten it seems: I've reported in several theads that I thought the section of install where you pick modules and drivers was inaccurate in its listing for my nic `3com 905'. I argued that it had the wrong driver listed adjacent to it. I was half scared to run base-conf all the way to that section for fear of scrubbing something so hadn't checked for sure. I imagine there is a text representation of that stuff some where too. Dman posted that one can look at it with modconf any time so finally I did that and took a close look. It is mildly confusing to browse thru but not really that bad. And the driver is listed exactly as it should be. I got confused over those dumb 3com names that are so similar. The user isn't really told this but drivers are on the left and cards on the right. But it is fairly easy to see what is what. The listing looks like this: 3c95x +3c59x/3x90x/3c575_cardbus series boomarang/C There's my card right in the middle. There may still be an argument to be made that the install routine should have just found my nic, but my claims of inaccuracy are unfounded. So, I'll have at Todays serving of crow.
Re: More cheese nibbler questions
Carel Fellinger <[EMAIL PROTECTED]> writes: > On Wed, Mar 06, 2002 at 10:53:31AM -0800, Harry Putnam wrote: >> Setup: Stock woody install from testing >>Kernel 2.2.20 >> >> Comments: X is installed but not running for these tests >> Mouse works in X but not in console >> >> >> If I say: >> >> # /etc/init.d/gpm force-reload >> Stopping mouse interface server: gpm. >> Starting mouse interface server: gpm. >> >> at the command line then `ps waux|grep gpm ' >> should show some thing with gpm in it ...right? > > Yes, if all's well. But as I stated in an earlier posting, we still > don't know for sure that your /etc/init.d/gpm is correct. Any reason > you're not willing to share it's contents with us just to eliminate > the off change that it *is* mangled? Err no, but isn't that what is printing those messages above? I did look at it. Not an expert script writer but it looks pretty boiler plate.
Re: More cheese nibbler questions
Carel Fellinger <[EMAIL PROTECTED]> writes: > That depends. Is your ps/2 driver compiled in or compiled as a > seperate module? Look through /boot/config- for the awnser. Its not clear what any of that means and the only entry seems to be related to mouse is this: CONFIG_BLK_DEV_PS2=y I'm guessing the `y' means its compiled in. An `m' would have meant it was a module.
Re: The quest for rodent power
Carel Fellinger <[EMAIL PROTECTED]> writes: [...] Carel, Somehow I missed this post of yours Here are some answers: > e.g. I.m curious whether the above long gpm command > ># /usr/sbin/gpm -D -m /dev/psaux -t ps2 -Rraw > > did return immediately or kept running and that you did the > ># ps waxul|grep gpm That is a typo should be # ps waxu|grep gpm > in a different xterm/console > > It's the way you worded the events that made me wonder. > Probably it's just that you wanted to make sure that gpm > didn't put itself in the background although the -D option > told it not to, but I'm not sure. Its not clear what you want to know here. I just ran the same test again. There is about a 1 second delay before the prompt comes back and prints the odd message. Alarm Clock Any of these after the -t flag: ps2 imps2 autops2 will provoke the Alarm clock message. All pause about a second. [...] > > So we really still don't know if this file is okee, do we:) Now posted >> Its just that nothing actually happens. > > yep, that's why I think you should stick to calling gpm directly and > not use /etc/init.d/gpm OK, will do. Any further tests are command line only
Re: The quest for rodent power
"Karl E. Jorgensen" <[EMAIL PROTECTED]> writes:
[...]
> I had a sudden thought: perhaps it is time to experiment with the mouse
> type for gpm (completely ignoring X for now).
>
> You've established beyond any doubt that it works in X as a PS/2 mouse
> - and hence the kernel, connection, physical mouse etc is OK. And that gpm
> does *not* like to interpret the mouse as a vanilla ps/2.
>
> The ps/2 code in gpm is probably not identical to the ps/2 code in
> the X server. And even when gpm is told to repeat in "raw" mode it will
> still try to interpret the incoming mouse events. I guess that it doesn't
> understand them - hence your "Error in protocl" in the gpm debug you
> posted earlier.
>
> In other words: X and gpm may well have different interpretations of
> "PS/2". (my guess/conclusion, not necessarily fact, but I think it makes
> sense).
>
> Try
> # gpm -t help
>
> to get a listing of mice that gpm understands.
> In this list, these look like candidates for your logitec 3-button ps2:
> autops2 # if you're lucky :-)
> mman
> logim
> fups2
> imps2 # only 'cause it was mentioned in your XF86Config-4 at some point
> fuimps2 # same reason
>
> (and any other mouse type that catches your eye in gpm's list)
>
> So a few simple tests by running:
>
> # gpm -D -m /dev/psaux -t {some gpm mouse type}
>
> should reveal what mouse type does (not) work for gpm. If one of them
> works, then you know what to put in /etc/gpm.conf. Tweaking the X config
> afterwards should then be easy: replacing /dev/psaux with /dev/gpmdata.
I may not have understood the task correctly. Not sure what would
indicate working or not. So assuming you mean fully work as in
having a mouse ... hehe.
I went thru the list you show above adding only ps2 and exps2 for
good measure.
# gpm -D -m /dev/psaux -t {some gpm mouse type}
With mman logim fups2 exps2 all give a hung terminal that appears
to be waiting for standard in. Nothing is written to syslog except
the initial message (only once at the first invocation, not the others):
Removing stale pid file /var/run/gpm.pid
Using these types: ps2 imps2 autops2 and fiups2
After about a one second pause gives the message:
Alarm Clock
And then the terminal prompt reappears
No syslog is produced
This may be a further diagnostic observation:
# gpm
(just the bare command)
Will cause 3 lines in the center of my tty (console) to appear in
reverse video as if they had been hilighted by a mouse.
And ps waux|grep gpm will show that command running, until I kill it.
No syslog output is produced.
Re: The quest for rodent power
Carel Fellinger <[EMAIL PROTECTED]> writes: > On Wed, Mar 06, 2002 at 08:24:22PM -0800, Harry Putnam wrote: >> Carel Fellinger <[EMAIL PROTECTED]> writes: > ... >> Its not clear what you want to know here. I just ran the same test >> again. There is about a 1 second delay before the prompt comes back >> and prints the odd message. >> Alarm Clock > > It was the one sec delay that I was after:) So it's similar to what > happens here including that `Alarm clock' message. BUT I only get > that behaviour when an other gpm instance is running. Weird. Can you see the other gpm in ps output before running the test? Is it possible for gpm to be running in the background and not show up in ps output? Does the fact that running the bare gpm command produces 3 lines of reverse video in the center of my console, as if hilighted by a mouse indicate anything special to you?
cfdisk extended partitions
How does one make an extended partition with cfdisk? This would be the cfdisk used during and install from CD (woody 3.0) I'm familiar with fdisk and know how to switch to vt F2 during install to get a shell where I can use fdisk. And will probably do that, but just curious now how its done with cfdisk. The help option offers no clues.
sources.list post install
During install if you elect to add sources to the /etc/apt/sources.list it gets added and a nifty test is performed. How can that be done after install is over. I mean beside just using apt-get after adding something.
apt-get build-deps or -b switch
How can I capture the output of ./configure and make during a source build with apt-get -b source I think there might be a small bug in the building of emacs21 that way. The info flew by so quick but I noticed a `no' during ./configure that I think was prompted by a missing library `libungif4-dev'. I think that was supposed to be grabbed with apt-get build-deps emacs21 But it wasn't maybe. I'd like to rerun the ./configure and trace it down.
debian exim4 smarthost and relay
I hope this is the correct group for debian exim4 questions. I'm sure someone will point me elsewhere if not - Setup: New install of debian 10, in VBox vm attempting to get exim4 working. - Internet host with no recognized or dns resolvable name, or domain - Receives dhcp address, but made mostly permanent by router capabilities When editing a smarthost line in udpate-exim4.conf.conf How does one enter a server that requires ssl 465. Is adding 465 port all that is necessary? ..like dc_smarthost='smtp.server.com:465' I've also seen two colons used while googling on this: dc_smarthost='smtp.server.com::465' Which if any is correct? --- --- ---=--- --- --- If one uses the port number in update-exim4.conf.conf is it also required in passwd.client? smtp.server.com:465:user@some.domain:SECRETpw And again .. one or two colons? Or is this not required? --- --- ---=--- --- --- My googling has also caused some confusion to me about editing something for smtp transport: Will it also be necessary to edit something into smtp transport regarding the use of ssl 465? If so, can anyone show an example of how that is done?
Re: debian exim4 smarthost and relay
> Reco writes: > David Wright writes: > Brian writes: Thanks to all for the succinct and helpful answers Sometimes googling excessively can confuse me a bit.
about apt and aptitude update
I use both apt and cmdline aptitude. Mostly cmdline aptitude I'm curios if when I call `apt update' or `aptitude update', if they are refreshing the same database files... wondering if I could do just one update on either tool and that would do both.
What program is capturing key press on root window of X?
cat /etc/os-release PRETTY_NAME="Debian GNU/Linux 9 (stretch)" NAME="Debian GNU/Linux" VERSION_ID="9" VERSION="9 (stretch)" ID=debian (running LXDE desktop) Somewhere in the last few months my Debian OS has acquired an input box on upper right of base window (in X) that appears to grab any key presses aimed at the base window and print them in that small input window. Since I have run the `xbindkeys' program (An keybinding program that takes its intput from key presses in base window and calls programs thru a control script ~/.xbindkeys.scm), this new key grabbing program that seems to run at startup is ruining that whole process. I wasn't able to tell which program it is by examining the autostart programs ... can anyone tell me what that is that is running so I can turn it off?
Not finding openssl during a pkg build (mailsend)
cat /etc/debian_version 9.4 Summary: Trying to compile a tool called `mailsend' (https://github.com/muquit/mailsend/releases/) (Which does not appear to be available in the regular repo). But cannot get it to find openssl. I do have openssl and libssl-dev installed; --- --- ---=--- --- --- Details A check of apt-get search openssl A quick check of ./configure file suggests using: grep [\-]-with.*openssl ./configure --with-openssl=PATH where the root of OpenSSL is installed --with-openssl-includes where the OpenSSL includes are. --with-openssl-libraris where the OpenSSL library is installed. I tried --with-openssl=/usr/bin/openssl and --with-openssl=/usr/lib/ssl and --with-openssl-includes=/usr/include/openssl using dpkg -L and `grep -v' (to keep down the amount of data) shows: dpkg -L openssl |grep -v share /etc /etc/ssl /etc/ssl/certs /etc/ssl/openssl.cnf /etc/ssl/private /usr /usr/bin /usr/bin/c_rehash /usr/bin/openssl /usr/lib /usr/lib/ssl /usr/lib/ssl/misc /usr/lib/ssl/misc/CA.pl /usr/lib/ssl/misc/tsget /usr/lib/ssl/certs /usr/lib/ssl/openssl.cnf /usr/lib/ssl/private and dpkg -L libssl-dev (95 lines of output omitted) Are there other parts of openssl I need to install? I'm not figuring out how to direct this build toward openssl, but the configure process continues to show: [...] checking for OpenSSL... no [...]
Re: Not finding openssl during a pkg build (mailsend)
Greg Wooledge writes: > On Mon, Oct 01, 2018 at 12:47:01PM -0400, Harry Putnam wrote: >> cat /etc/debian_version >> 9.4 > >> I do have openssl and libssl-dev installed; > > In stretch, libssl-dev is the development package for OpenSSL version 1.1. > This is a relatively new version of OpenSSL, and it is not compatible with > version 1.0. > > Most programs that use OpenSSL will be expecting to use the version 1.0 API. > I don't know your particular program, but it's a reasonable guess that > it wants the older version. > > Try installing libssl1.0-dev (which will remove libssl-dev) and see if that > works. You can switch back and forth as needed when compiling various > programs, but you can't have both installed at the same time. Thanks for the helpful and precise input. So far haven't managed to make it work Something else must be a problem. After some more testingh/tinkering I may have a new query. Your response was helpful information anyway.
Re: uefi boot install and disk partitions
Bonno Bloksma writes: > Hi, > > > > I have been creating a small (300MB) primary /boot partition at the > beginning of the disk for as long as I can remember... That is after > disks got to be too big for the BIOS to reach all of the disk to be > able to boot from a file anywhere on the disk. > > So far so good, that still works but do I still need that > partition when I create an EFI System Partition (ESP) to boot using > UEFI? > [...] First off, I'm pretty sure that Pascal is far more knowledgeable than I but I just had some experience with setting up Uefi. Still working on it so not tested yet. It is on a zfs-on-root-linux install so that might be different. But the part of preparing the disk for booting seems to be the same. I used `gsdisk' which is in the package `gdisk' in the debian repo. There are detailed direction by Richard Laager: I'm quoting a small bit from URL: https://github.com/zfsonlinux/zfs/wiki/Debian-Buster-Root-on-ZFS 2.3 Partition your disk(s): [...] section on non efi boot snipped Run this for UEFI booting (for use now or in the future): sgdisk -n2:1M:+512M -t2:EF00 $DISK Run this for the boot pool: sgdisk -n3:0:+1G -t3:BF01 $DISK He's taking about a zpool setup but I think Its the same far as preparing the disk for uefi booting. There are more details further on concerning actually installing the boot code on the disk where booting will occur. You probably want to read this part: 4.8b Install GRUB for UEFI booting
VirtualBox on lenny
Googling about virtualbox on debian.. I see only older information. Looks like the wiki at http://wiki.debian.org/VirtualBox Appear to be somewhat old too and only goes up to etch. Other stuff shows its from 2007. I know VirtualBox is under very active development so wondered if things may have changed since that wiki was posted. aptitude search virtualbox shows a bewildering array of packages with virutalbox in the name... but I see no version numbers. Anyway, is there a recent outline about how to do this somewhere? Or any posters here have VirtualBox running on lenny that can say how they did it? -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: VirtualBox on lenny
Paul Fraser writes: > On Thu, Feb 12, 2009 at 07:21, Harry Putnam wrote: >> Or any posters here have VirtualBox running on lenny that can say how >> they did it? > > I'd recommend installing VirtualBox from experimental -- 2.1.2 is the > latest version, as opposed to 1.6.6 available elsewhere. Sorry to be so thick, but can you point me to something that explains how to access `experimental' And further... what packages are required? I noticed quite a long list with aptitude search -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
What do people do about large clock scews
I have debian lenny in several vmware applications on different windows machines so some of them don't get run too often. I noticed firing up one that hasn't been run for a few weeks that the time is off a by several of hours. To get it up to speed, looks like I'd have to shut ntpd down, run ntpdate then restart ntp. I'm thinking this might be a bit of a problem in other situations than my experimental setup. I'm guessing people are writing their own init script to run ntpdate on boot ahead of ntp, but I'm not yet conversant enough with the init script maze to know how to do that handily. I can script something to run ntpdate alright but getting it timed to run ahead of ntp may be a bit more daunting. But first, is there already a defacto way of doing this? -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: What do people do about large clock scews
Ron Johnson writes: > On 02/22/2009 12:30 AM, Harry Putnam wrote: >> I have debian lenny in several vmware applications on different >> windows machines so some of them don't get run too often. >> >> I noticed firing up one that hasn't been run for a few weeks that the >> time is off a by several of hours. >> >> To get it up to speed, looks like I'd have to shut ntpd down, run >> ntpdate then restart ntp. >> >> I'm thinking this might be a bit of a problem in other situations >> than my experimental setup. >> >> I'm guessing people are writing their own init script to run ntpdate >> on boot ahead of ntp, but I'm not yet conversant enough with the init >> script maze to know how to do that handily. >> >> I can script something to run ntpdate alright but getting it timed to run >> ahead of ntp may be a bit more daunting. >> >> But first, is there already a defacto way of doing this? > > AFAICT, ntpdate is run when networking is started, before ntp is run. Well, that should rule out a large clock skew then. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: VirtualBox on lenny
Umarzuki Mochlis writes: >> And further... what packages are required? I noticed quite a long >> list with aptitude search > > Try installing from virtualbox own repo to get the latest (2.1.4). Check > http://www.virtualbox.org/wiki/Linux_Downloads You mean to circumvent the apt tools right? Done that way, does it matter which compiler you have installed? I ask because I know on gentoo the VirtualBox package gives warning to use something earlier than gcc-4.3. Just made me wonder if other versions may be a problem too. I'm going to try that install sometime today, but do you have any other advice from experience with installing VirtualBox? -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: VirtualBox on lenny
"Christofer C. Bell" writes: > To add to this, I encourage using /etc/apt/sources.list.d for adding 3rd > party repositories. My apt pointer to the VirtualBox repository is in the > file /etc/apt/sources.list.d/virtualbox.list > > Which contains the following: > > # Sun xVM VirtualBox > deb http://download.virtualbox.org/virtualbox/debian etch non-free > > Obviously I'm still on etch. For lenny your file would be slightly > different. This allows you to keep logical separation between between > official Debian repositories and third party repositories. I'd probably > call it a "best practice." After doing that, I guess then I'm supposed to run apt-get update? When I do that I get this output at the end: Fetched 2283B in 3s (573B/s) Reading package lists... Done W: GPG error: http://download.virtualbox.org lenny Release: The following signatures couldn't be verified because the public key is not available: NO_PUBKEY DCF9F87B6DFBCBAE W: You may want to run apt-get update to correct these problems But of course when I run `apt-get update' it does NOT fix whatever is causing the error. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
