Re: [Cooker] fdio
On Wed, 24 Jan 2001, Alberto Vorano wrote: > Hi to all > It seems to me that several guys, myself included, that have upgraded to rpm > 4.0.x have the problem of the missing fdio. As there were no clear reply, for > what I have seen, can please someone say if this is a known bug of whatever > package or, opposedly, we made some silly thing? I have back the same message > if I try to launch from console either rpmdrake or kpackage. The strange thing > for me is that redhat 7.0 has both rpm 4.0x and kde, and they run smoothly, so > my opinion is we have made something wrong. > Please, give us some help! I get this launching rpmdrake; MandrakeUpdate, meanwhile, has a nasty habit of segfaulting. (ALWAYS when looking at a local Cooker mirror on disk; often after fetching the list from sunsite.uio.no.) James.
Re: [Cooker] About the text-based version of MandrakeUpdate
On Tue, 23 Jan 2001, John Allen wrote: > A useful thing for Madrake update sites to do would be to put a TIMESTAMP > file in the update directory, that way the update program could tell very > quickly whether any updates were added since the last time it was run. If retrieving via HTTP rather than FTP, it could use an If-Modified-Since header to check if files have changed since the cached version. Nice and simple; if there are no new files, the index would just return "Not modified" and MU knows there's nothing new. James.
Re: [Cooker] XFree86-4.0.2-4mdk.i586.rpm
On 22 Jan 2001, Vox wrote: > > During the bombing raid on Sun, 21 Jan 2001 17:21:26 -0600, Ed Wilts was heard > mumbling in fear:: > > > > "Due to certain legal liabilities and for the protection of intellectual > > property, Matrox reserves licensing rights to the library and prohibits > > reverse engineering but allows free distribution under any operating system. > > Matrox encourages members of the open source community to freely distribute > > and assist in the further development of this driver." "prohibits reverse engineering". Uh, what about countries where that restriction is illegal? :-) > And they think that somebody in the open source/free software community > is actually going to help them improve a driver under this conditions? ha! fat > chance! Indeed. Giving some closed source binaries to the community is almost worse than giving nothing at all: OK, it allows those of us unlucky enough to have their proprietary cruft to use said cruft until we can get a replacement, but... > > Even though the driver they ship is definitely not GPL, neither is Netscape > > and Mandrake has no problems shipping it. > > I happen to *strongly* agree with mandrake in their position about > closed source drivers, even if it means I have to go grab the nvidia drivers > from the nvidia site to get my card going properly. Then again, I don't depend > on X to be able to go grab stuff off the net, since I can do it from console > with no prob at all. > > On the other hand, when they say that they prohibit reverse engineering > they are not only going against your freedom (according to the FSF/GNU > definition) but against the law in most (if not all) countries. Sof them. Indeed. Reverse engineer away, after checking your local legislation :-) James.
Re: [Cooker] Very bad cooker install
On Sun, 21 Jan 2001, Pierre Fortin wrote: > James Sutherland wrote: > > > > On Sun, 21 Jan 2001, Pierre Fortin wrote: > > > > > [EMAIL PROTECTED] wrote: > > > > > > > > That's not what I saw - it was jerking randomly around the screen, with a > > > > tendency towards the top right. Not confined anywhere, just not moving > > > > properly. > > > > > > > > > You can only escape with a hard-reboot, or switching to the console :( > > > > > This is however probably not related to the "jumpyness-problem" that > > > > > occurs when you change from ps/2 to imps/2... > > > > > > This may sound way out in left field; but when this happens, switch between > > > consoles and look at the rightmost LED on the keyboard... for each console > > > (don't forget F12), if the LED is on, try Ctl+Q... I recently discovered this > > > on a similar "magnetic mouse" problem recently. > > > > That's getting into the realms of reflexology, I think :-) > > > > I've only made this mistake once, and solved it by rebooting and > > installing with the generic PS/2 option which worked fine - I don't think > > I'll get a chance to try your solution! Thanks anyway, though. > > I hope no-one does... but, I've seen the mouse problem more than once, and this > is the closest I got to a clue to what's happening. BTW, I'm still running the > same instance (avoided a reboot). You're still running the install program?! Must be a hell of a slow download you've got there :-) (Actually, I did a 2 hour install over a direct 100Mbit/sec connection from another PC. Painful... For some reason, the HDD on the server kept cutting out for a couple of seconds every minute or so?!) James.
Re: [Cooker] FTP Install of cooker,
On 22 Jan 2001, Guillaume Cottenceau wrote: > Oliver Stieber <[EMAIL PROTECTED]> writes: > > > I tried an ftp install of cooker at the weekend, > > the only problems I had were, > > > > Couldn't specify a netmask when setting up the network card. > > Yes, it's automatically calculated if the magic word "expert" is not > provided as a kernel parameter, according to Internet's standards: > > if first IP number is below or equals 127, sets to 255.0.0.0 > else > if first IP number is below or equals 191, sets to 255.255.0.0 > else > sets to 255.255.255.0 > > I don't think it's wrong? Yes :-) The 131.111 and 134.36 blocks were *allocated* as class B's, but are both suballocated by the universities concerned into multiple subnets - /24 in the first case (University of Cambridge), a mixture of /23 and /24 in the latter (University of Dundee). You can't assume the netmask is even one of /8, /16 or /24 now, let alone that the netmask is the default for that network range: that system died a long time ago. DEFAULTING to the values above is OK, but ASSUMING them is not. I'm sitting here on 131.111.237.83/24 :-) > If your network does not respect the Internet standards, try with "expert" > option at boot time. It's a dead standard, so don't assume compliance with it! James.
Re: [Cooker] Very bad cooker install
On Sun, 21 Jan 2001, Pierre Fortin wrote: > [EMAIL PROTECTED] wrote: > > > > That's not what I saw - it was jerking randomly around the screen, with a > > tendency towards the top right. Not confined anywhere, just not moving > > properly. > > > > > You can only escape with a hard-reboot, or switching to the console :( > > > This is however probably not related to the "jumpyness-problem" that > > > occurs when you change from ps/2 to imps/2... > > This may sound way out in left field; but when this happens, switch between > consoles and look at the rightmost LED on the keyboard... for each console > (don't forget F12), if the LED is on, try Ctl+Q... I recently discovered this > on a similar "magnetic mouse" problem recently. That's getting into the realms of reflexology, I think :-) I've only made this mistake once, and solved it by rebooting and installing with the generic PS/2 option which worked fine - I don't think I'll get a chance to try your solution! Thanks anyway, though. James.
Re: [Cooker] How to convert a FS to journalling ?
On Sun, 5 Nov 2000, Franck Martin wrote: > How to convert the standard linux file system (ext2) to a journalling > filesystem? Upgrade from ext2 => ext3. > Which filesystem is recommended? If you want to convert a partition to add journalling, ext3 is the only option; ReiserFS requires a reformat. ext3 also has the advantage of being backwards compatible. James.
Re: [Cooker] Journalling
On Sat, 4 Nov 2000, Leon Brooks wrote: > Joakim Bodin wrote: > > [XFS is] mostly stable right now and > > is only lacking some minor details IMO (quota support etc) which is being worked > > on. > > ReaiserFS doesn't (at last count) have quota either. I saw a rumour that Hans > was planning to add this RealSoonNow. Last time I looked (which I do quite often, since I run the UK mirror of their site!) there was a patch to add quota support. I'm not sure what state it's in, though. There are problems with quotas on ReiserFS, since ReiserFS doesn't store data in clusters like ext2 etc - a 17 byte file takes less space than a 25 byte one, making quota tracking rather tricky... James.
Re: [Cooker] I am puzzled by people's behaviour...
On Thu, 2 Nov 2000, Jason Straight wrote: > On Thursday 02 November 2000 08:12, you wrote: > > James Sutherland <[EMAIL PROTECTED]> writes: > > > > > > [...] > > > > > confusion. Your implication that neither of them is the real 7.2 suggests > > > you're working for SuSE :-) > > > > Sorry, I give up. > > Right, accusing and name-calling isn't going to help anyone here, just piss > everyone off. The point has been made already. Mandrake knows the confusion > that having 2 versions released under the same version can cause and will > hopefully find a way to educate consumers before it happens again. That's the wrong approach. It isn't the consumer who needs educating, it's whoever puts the version numbers on boxes. > This could have all been solved by some note or version difference > visible on the package at the stores. It may be a problem of having > 7.20 at the store while 7.21 is on download but it would be a lot > easier to just put "Can be updated, online, to newest versions with > included Mandrake Update utility!" than dealing with all the > confusion. Which is precisely what I and others suggested - only to be vetoed with "but the marketdrones wouldn't like it". The argument amounted to "Walmart want to have a box with the latest version number on, even if the contents aren't what the label says." James.
Re: [Cooker] I am puzzled by people's behaviour...
On 1 Nov 2000, Guillaume Cottenceau wrote: > You are misunderstanding the reason/utility for Release Candidates. > > If you want to call the retail version RC1, please call the online version > RC2. > > Again, I have to explain that Release Candidates, are Candidates to become > gold releases, no more no less. OK, next question: when is the gold release of 7.2 going to happen? The packaging implied the box Walmart have is that release; posts here implied the same of the 7.2 ISOs for download. You've released the software, via Walmart, under the name 7.2. Releasing a different product under the same name online is designed to spread confusion. Your implication that neither of them is the real 7.2 suggests you're working for SuSE :-) > The retail version is the distrib we judge to be good enough for 7.2; the > online version is stricly the same except that it has the latest kde2 > packages and a bunch of other minor fixes. So don't call it 7.2. The thing Walmart has is a release, by definition: you released it. Call the online updated version "7.2.1" or similar: it can't be a release candidate for anything. > > > That's an unsolvable problem, I think. > > > > Not entirely. The product they are selling is NOT "7.2", and shouldn't be > > sold as such. Just print the truth... > > I'm sorry, it IS "7.2", difference is some small updates.. > > AFAIK, it's printed on the boxes that it does contain a pre-version of > final kde2.0. Hrm... you're trying to confuse people. Now, if you could just make sure the installer would stop crashing while square rooting negative numbers... oh, and installing a kernel and boot sector would be nice, too. (This is 7.1, BTW.) James.
Re: [Cooker] I am puzzled by people's behaviour...
On Wed, 1 Nov 2000, Jacques Le Marois wrote: > James Sutherland wrote: > > > > Not entirely. The product they are selling is NOT "7.2", and shouldn't be > > sold as such. Just print the truth... > > The product in the shop is the 7.2 final. The ISO on Internet is 7.2 > final +updates and could be call something like "7.2-27/10 edition". A new subversion might be helpful, perhaps - 7.2.01 or similar? (Or 7.2 Service Pack 1 ) > In x weeks will do a new ISO/tree with all the updates and it would be > either 7.2 with a new build date - we are not going to call "final" > each last build. All the updates for MandrakeUpdate are going to be > targeted to the 7.2 of the shops. It would be a lot clearer if you could increment a subversion number to make this clear - perhaps call it "7.2 update 1"? James.
Re: [Cooker] I am puzzled by people's behaviour...
On 1 Nov 2000, Guillaume Cottenceau wrote: > Alexander Skwar <[EMAIL PROTECTED]> writes: > > > So sprach Pixel am Wed, Nov 01, 2000 at 12:48:45PM +0100: > > > - name the updated-7.2 7.2.1, but marketing people don't accept this. > > > > Ah! Finally a reason why the download version is also dubbed 7.2. Although > > I disagree with the reason (I always disagree with marketing people :]), I > > can understand it! > > They are right, they say that if we use two different names, people in the > shops would require the newest one, which the shop would be unable to > provide. Not necessarily; label the box as "7.2RC1" or whatever, and add a little "sticker" on the cover art boasting "Includes free online upgrade to 7.2 Final!". > That's an unsolvable problem, I think. Not entirely. The product they are selling is NOT "7.2", and shouldn't be sold as such. Just print the truth... James.
Re: [Cooker] ssh
On Tue, 31 Oct 2000, Brian J. Murrell wrote: > from the quill of James Sutherland <[EMAIL PROTECTED]> on scroll > <[EMAIL PROTECTED]> > > > > They prohibit (ab)use of the connection for running public servers. > > Why is running a public server on bandwidth that I am paying for abuse? You are NOT paying for that kind of bandwidth. You are paying for restricted access to shared bandwidth. You want to run a server? Pay for the bandwidth. > > Running publicly accessable WWW or FTP sites IS abuse, IMO, and SHOULD > > be > > prohibited. > > Why is it abuse? I am paying for bandwidth. How can you determine that > one use of it is use and another is abuse considering all uses only move > bits? Traffic levels and nature. > > Completely different situation. > > Only in their definition of use and abuse. When visiting > www.playboy.com is determined to be abuse it won't be different. It's > all in what they decide is use and what is abuse. > > > Airlines ALL do it, > > They don't put rules in place to prevent you from flying when they do > oversubscribe and they compensate those who suffer due to it. (I can't > believe I am defending airlines -- I despise airlines). They DO set restrictions on tickets - often restricting you to a specific flight, on a "use it or lose it" basis. They overbook, and cases do occur when they cannot carry everyone. > > buffet restaurants do it; > > They do? I don't think I was ever limited in how I ate at a buffet. You really believe they had an infinite amount of food available? I've seen them run out of food before. (The restaurant in question did close the next week...) > > ISPs all do it; > > We are talking about ISPs here. You can't use something you are trying > to define in it's own definition. I am not trying to define anything. I am pointing out that all ISPs do this, not just my one. > > banks > > all do this with cash. > > I have never been limited in any way from getting my cash out of the > bank either. Interesting; every cash machine I have ever used has had a withdrawal limit. There are limits on counter transactions as well. If you have enough balance, try withdrawing gbp100k from a cash machine, and watch the response. James.
Re: [Cooker] ssh
On Tue, 31 Oct 2000, Brian J. Murrell wrote: > from the quill of James Sutherland <[EMAIL PROTECTED]> on scroll > <[EMAIL PROTECTED]> > > > > That's the policy my cable co runs, more or less: no "public" servers > > (no > > anon-FTP etc), but SSH, FTP (users only), password-protected WWW etc > > is > > fine. Seems reasonable, IMO. > > Why does it sound reasonable? You are paying for bandwidth. Why should > you tolerate being told what you can do with it? They prohibit (ab)use of the connection for running public servers. Running publicly accessable WWW or FTP sites IS abuse, IMO, and SHOULD be prohibited. > Next, when they start telling you what web sites you can and can't visit > with the bandwidth you are paying for, is that going to "sound > reasonable"? Completely different situation. > OK, maybe that one is far fetched. But this is not: When you try to > use your paid for bandwidth to receive constantly streaming content > (say streaming video) and they elminate that from the allowable use is > that going to "sound reasonable"? Constantly maxing out a *SHARED* Net connection is abuse. > The problem is that your cable co. has over-subscribed their customers > to the bandwidth they are providing and now they are back-peddling to > try to fix the problem. This over-subscription is called fraud in other > businesses. Interesting. Perhaps you'd like to tell every other company on earth about that? Airlines ALL do it, unless you are actually leasing the entire aircraft as a single unit; buffet restaurants do it; ISPs all do it; banks all do this with cash. Find me a company which DOESN'T! James.
Re: [Cooker] ssh
On Tue, 31 Oct 2000, Ron Stodden wrote: > Bryan Whitehead wrote: > > > It's much more simpler than doing a stupid port scan of all your > > customers... > > Agreed, but it is a form of censorship, so will/should attract > political interest. Should not any packet addressed to you be > delivered to you? As a matter of principle and ethics? > > But I still think low volume access to these servers should be > permitted. That's the policy my cable co runs, more or less: no "public" servers (no anon-FTP etc), but SSH, FTP (users only), password-protected WWW etc is fine. Seems reasonable, IMO. James.
Re: [Cooker] ssh: end of discussion: MS make the right decision ?
On Fri, 27 Oct 2000, Brian J. Murrell wrote: > from the quill of Ron Stodden <[EMAIL PROTECTED]> on scroll > <[EMAIL PROTECTED]> > > Franck Martin wrote: > > > > > 99% of the people buying Mandrake do not have access to cable > > modem... > > > > Would you mind sharing your authoritative source for this > > information? > > What does it matter? A more relevant figure would be the number who actually WANT a server running. > > Of more relevance would be the proportion of Mandrake users having > > cable modem access. > > Can we just agree that the majority of us should not suffer due to the > few who are signed up with cable modem providers that have draconian > AUPs and need badly to be put out of business for sheer stupidity? More to the point, the majority of us who do not want to find they have a new system administrator should not get their machines broken into because someone thought a service might be useful? Install it, yes. Enable it? Try it and your distro goes straight back to the shop. Do NOT enable ANYTHING by default. James.
Re: [Cooker] ssh: end of discussion: MS make the right decision ?
On Sat, 28 Oct 2000, Franck Martin wrote: > 99% of the people buying Mandrake do not have access to cable modem... What percentage have a home network, and would be able to SSH into their machines? > The 1% are already lucky to have access to high bandwidth service, so > DO NOT bother us, change your service provider! Not an option: I for one do not have a choice. If Mandrake comes with redundant crap enabled by default, Mandrake has a /dev/null to go to. > Silly discussion indeed... > > Cheers > Franck > > So it is sshd installed by default! Installed, but **NOT** enabled. I've seen far too many machines broken into thanks to the braindead "let's enable everything - after all, it might come in handy one day to one of the crackers who'll be controlling this box tomorrow" policy of Linux distros. James.
Re: [Cooker] ssh: end of discussion: MS make the right decision ?
On Sat, 28 Oct 2000, Ron Stodden wrote: > Franck Martin wrote: > > > > Why don't we stop this silly discussion indeed and Mandrakesoft agrees to > > install sshd by default on ALL machines, the same way as telnet is installed on > > ALL machines. > > > > For an extra 100k sshd has a very small footprint for so much benefits. > > One good reason is that to do so would violate the service contract > of people with machines connected to the internet via a residential > service cable modem. If the cable company discovers you are running > a server (and they do this by regular port scans, not by traffic > analysis) they will cancel your service contract. > > So it is NOT a silly discussion. Fortunateny, my cable modem provider has a more enlightened philosophy (despite being part-owned by MS...) However, I certainly wouldn't want any service running I hadn't specifically enabled. Having lost a machine to an FTP exploit last week (FTP? Eh? The machine wasn't an FTP server ... Oh. Bad Prat had decided to enable all the services it could find...) Personally, I think "Paranoid" mode is the right way to go (and should probably be the default): do NOT enable all the crap you can find. If I need it, I'll enable it; if I don't, I do NOT want it enabled. EVER. James.
Re: [Cooker] ide-scsi hack
On 27 Oct 2000, Guillaume Cottenceau wrote: > Joshua Jackson <[EMAIL PROTECTED]> writes: > > > So, that wouldn't cause any problems for when I don't have the burner? > > Only type it when it's there, right? > > either when the burner is not here you don't have /dev/hdd so nervermind. > or else you have another ide driver and you'll need another boot option. If I read you correctly, you (Joshua) always have a device on /dev/hdd - sometimes a CD burner, the rest of the time it's a CD ROM drive? If so, I'd just use ide-scsi all the time, and call the CD ROM drive /dev/sr0. James.
Re: [Cooker] still trying to get 2.4.0 anything to work
On Wed, 25 Oct 2000, Allen Bolderoff wrote: > > [EMAIL PROTECTED] said: > > Let me guess: You enabled "Explicit Congestion Notification" (ECN) > > under network options, didn't you? Disable it - either in the kernel > > config, or by writing a "0" to /proc/sys/net/ipv4/tcp_ecn. > > *actually* it is enabled by default, either Mandrake need to re-compile with > it defaulted to 0, or they need to put it into /etc/sysctl.conf by default Oh dear. Considering the amount of breakage it causes, I don't think that was a very bright idea... Supporting it, for experimentation etc, is fair enough, but making it the default?! James.
Re: [Cooker] still trying to get 2.4.0 anything to work
On Tue, 24 Oct 2000, Jason Straight wrote: > A little off topic since I compiled my own 2.4.0test9, but there were > webpages I couldn't access under 2.4.0 that I can under 2.2 ! Very weird. I > couldn't even telnet to port 80 on us.imdb.com or www.linuxtoday.com. Went > back to 2.2 and everything is back to normal, and I flushed the firewall out > on both to make sure no firewalling was messing stuff up. Let me guess: You enabled "Explicit Congestion Notification" (ECN) under network options, didn't you? Disable it - either in the kernel config, or by writing a "0" to /proc/sys/net/ipv4/tcp_ecn. James.
Re: [Cooker] reiserfs? anyone?
On Sat, 21 Oct 2000, Warren Doney wrote: > "Brian J. Murrell" wrote: > > > > Is anyone successfully using Reiserfs? > > /me puts hand up. > > works fine ReiserFS is fine for me. I just wish I could say the same for Mandrake (release, not beta), which missed out the small detail of installing a kernel... James.
