From: Digestifier To: Linux-Misc@senator-bedfellow.mit.edu Reply-To: Linux-Misc@senator-bedfellow.mit.edu Date: Sat, 4 Sep 93 03:13:10 EDT Subject: Linux-Misc Digest #92 Linux-Misc Digest #92, Volume #1 Sat, 4 Sep 93 03:13:10 EDT Contents: Re: SLS considered harmful (wasRe: Bashing Peter MacDonald) (Stephen Harris) Re: help with SIGTSTP (more) (Vesa Ruokonen) Re: SLS considered harmful (wasRe: Bashing Peter MacDonald) (P. David Gardner) Re: Stacker-like Compression? (Andrew Goh Liu Kang) Re: Low Cost SLS 1.03 on Diskette (Stephen Balbach) Re: netguide.. PostScript? (Toh Guan Nge) Re: Ian Jackson (dan@oea.hobby.nl) Public Domain driver for SMC Ethernet Card (Mike Stein) Re: Bashing Peter MacDonald (Peter C Olsen) Re: BSD UNIX (Charles Hedrick) Re: SCCS for Linux (Jim Haynes) Re: NT versus Linux (Kevin Brown) ---------------------------------------------------------------------------- From: harris@teaching.physics.ox.ac.uk (Stephen Harris) Subject: Re: SLS considered harmful (wasRe: Bashing Peter MacDonald) Date: 4 Sep 93 0:44:38 BST Look, this whole slagging of distributions is totally irrelevant. Times change, distributions change. WHat was great yesterday is awful today. My experience: I started with 0.10 kernel (eee, so long ago!) with the root+boot floppies Linus made. Then MCC created a massive 5 disk setup complete with gcc. Wow! I couldn't imagine anything better. THEN SLS created a 30 disk setup! After a lot of phone calls (no direct connection) I grabbed SLS. 1 week later it changed. I grabbed the changes. 1 week later it changed again. I got the changes. 2 days later it changed again! At no time did I even CONSIDER slagging off any of the authors involved. My attempt at humour ("I hate Linus" postings) were read as such by some people (luckily those who did the work understand humour!) Now, through the Ygd... CD, and a fair bit of ftpmail, I have rebuilt 90% of the Linux setup I want with my own compiling. The whole point of this posting: the work done by Peter, Le Blanc et al is great. If you don't like it, DON'T USE IT! I have never downloaded a system from these people that has not worked, after a little common sense. OK messy it may have been (SLS around March was real culprit!) but it *worked*. To paraphrase: those who can, do; those who can't, moan. Stephen Harris harris@teaching.physics.ox.ac.uk Opinions are just opinions, and the facts are the facts. But what are what? ------------------------------ Crossposted-To: comp.os.linux.help From: ruokonen@lut.fi (Vesa Ruokonen) Subject: Re: help with SIGTSTP (more) Date: Fri, 3 Sep 1993 16:49:40 GMT Reply-To: Vesa.Ruokonen@lut.fi In article , c9108932@peach.newcastle.edu.au (Simon J Ferrett) writes: > jmorriso@rflab.ee.ubc.ca (John Paul Morrison) writes: > >Can someone explain the correct way to handle SIGTSTP in Linux? > >When I type ^Z, both programs go berserk; they go into a loop > >and the continuously update the screen. I dont know why talk hasnt > a friewnd of mine and I modified ytalk so that it would compile > properly under linux, and it too suffers the same symptoms... Try something like: suspend_func() { kill(getpid(), SIGSTOP); /* some terminal restoring required here? */ } main() { ... signal(SIGTSTP, suspend_func); ... } -- Vesa.Ruokonen@lut.fi HELLO! I'm a .signature virus! Join in and copy me into yours! ------------------------------ From: dgardner@netcom.com (P. David Gardner) Subject: Re: SLS considered harmful (wasRe: Bashing Peter MacDonald) Date: Sat, 4 Sep 1993 01:09:39 GMT Stephen Harris (harris@teaching.physics.ox.ac.uk) wrote: : I couldn't imagine anything better. THEN SLS created a 30 disk setup! After : a lot of phone calls (no direct connection) I grabbed SLS. 1 week later it : changed. I grabbed the changes. 1 week later it changed again. I got the : changes. 2 days later it changed again! I think this is what pisses most people off, at least those who don't have a direct connection and have to download through a modem. Of course, they forget this is still beta software and subject to change at a drop of a hat. I'm grateful to be doing Unix at home affordably, and Linux/SLS works just fine for me. -- -- Dave Gardner in The Animal House, South Pasadena, CA dgardner@netcom.com ------------------------------ From: eng10408@leonis.nus.sg (Andrew Goh Liu Kang) Subject: Re: Stacker-like Compression? Date: 2 Sep 1993 03:42:27 GMT I've got this message from Werner Almesberger . However, I've no idea when the proposed "compressed ext2fs" will be done. -- From almesber@bernina.ethz.ch Sun Aug 22 23:32:57 1993 Received: from bernina.ethz.ch (bernina.ethz.ch [129.132.1.170]) by leonis.nus.sg (8.2/8.2) with SMTP id XAA04953; Sun, 22 Aug 1993 23:32:54 +0800 Message-Id: <199308221532.XAA04953@leonis.nus.sg> Subject: Re: Loop devices, version 1 released To: eng10408@leonis.nus.sg (eng10408) Date: Sun, 22 Aug 93 17:32:41 MET DST In-Reply-To: <199308110113.JAA04734@leonis.nus.sg>; from "eng10408" at Aug 11, 93 9:13 am From: Werner Almesberger Sender: almesber@bernina.ethz.ch Status: RO > You could provide a loop device that passes data to `gzip` transparently > before writing to the loop-device. Then we will have some sort of a > dos "Stacker" equalivant on linux. Hmm, this wouldn't work too well. First, you would be limited by the rather small number of blocks you can look at for compression. Second, this would create unpleasent dependencies accross the entire FS, making any recovery after an error less probable and increasing the possible damage. Third, loop devices don't have any interface to user mode programs, so this would be something rather different (but there exists an (old) implementation of user-mode block devices by Lawrence Foard). Fourth, there would be no way to know whether the data being written is already compressed or not, thereby wasting CPU time and possibly also disk space compressing data that's already been compressed by a better method. Fifth, you'd get "interesting" semantics, e.g. an attempt to erase a file may result in a disk full condition. Sixth, and most importantly, Stephen Tweedie is already working on a compressing version of the Ext2 FS. I expect this to yield much better performance than a strictly block-oriented approach. A more interesting case migh be compression of swap files or partitions. NeXT do this with apparently quite good success. Something to think about for the future ... - Werner -- _________________________________________________________________________ / Werner Almesberger, ETH Zuerich, CH almesber@bernina.ethz.ch / /_IFW_A44________________________________________________________________/ -- Goh Liu Kang National University of Singapore, Singapore 0511 Internet: eng10408@leonis.nus.sg Run linux - the FREE! unix clone that runs on a 386 PC. See comp.os.linux, comp.os.linux.announce. & or ftp tsx-11.mit.edu ; cd /pub/linux ; get the FAQ. ------------------------------ From: stephen@clarknet.clark.net (Stephen Balbach) Subject: Re: Low Cost SLS 1.03 on Diskette Date: 3 Sep 1993 21:55:56 -0400 I don't think the problem is distributing disks for a charge. Most will agree this is a needed service. The problem is advertising on USENET. And the problem is when 10 diffrent companies all vie for the readers attention, suddenly c.o.l.* becomes a junk mail-box, somthing know one wants. One possible solution that was tried in the past but died for lack of support was a mail-order catalog in which all vendors would place adds and the catalog was posted periodically rather than haveing the group peppered by cheesy individual slogans and blatent attention getters. If I was the .announce moderator, a moderated newsgroup that can be controlled, I would require mail-order adds to be in a catalog form. And if no catalog exists it would be the vendors responsibility to create one. And when a new vendor tries to post refer them to the catalog maintainer. All adds would be listed alphabetically or some other fair way (not by order of arrival). The catalog, like the one on tsx-11 (or was it sunsite), would have a short description and a long description of x lines each. With short descriptions listed first then the long descriptions last. This would solve the problem on .announce Of course the other groups are really more public and can not be controlled beyond the forces of the market. If you are a vendor and would like to discuss this further please contact me, thanks, Stephen -- Stephen Balbach . Clark Internet Services . Washington D.C./Balt. metro area . mail info@clark.net . FAX 410-730-9765 . Corp. accounts . Linux on Disk . 31 disks $45 . stephen@clark.net . voice 410-740-1157 ------------------------------ From: egntoh@ntuix.ntu.ac.sg (Toh Guan Nge) Subject: Re: netguide.. PostScript? Date: Thu, 2 Sep 1993 09:01:03 GMT Darcy Boese (dboese@spartan.ac.BrockU.CA) wrote: : Did something go a little wrong? The postscript version of the netguide only : came out to be about 4K... The actual one is at the directory at one level higher. -- Toh Guan Nge, School of EEE, Nanyang Technological University, Singapore Internet Email: egntoh@ntuix.ntu.ac.sg Bitnet Email: egntoh@ntuvax.bitnet ------------------------------ From: dan@oea.hobby.nl Subject: Re: Ian Jackson Date: Fri, 3 Sep 1993 21:30:54 GMT guest@hpacv.com wrote: : If this Ian Jackson is such a wonderful and helpful sole, why is then in the : weeks we have carried comp.os.linux.all there has not been ONE post or : helpful response or anything from this guy except his AMAZING dialy : pointer, ie: "Dont bother me ... read these FAQs" : annoyed. He is too busy flaming Peter MacDonald and SLS. -- |< Dan Naas dan@oea.hobby.nl >| +---------------------------------+ ------------------------------ From: mikes@hgc.edu (Mike Stein) Subject: Public Domain driver for SMC Ethernet Card Date: Fri, 3 Sep 1993 18:35:22 GMT Are there any Public Domain drivers for the SMC3016 Ethernet card. What ftp sites are there for Public Domain drivers? Mike Stein Hartford Graduate Center ------------------------------ From: pcolsen@super.org (Peter C Olsen) Subject: Re: Bashing Peter MacDonald Date: Fri, 3 Sep 1993 11:49:49 GMT I would like to thank Peter MacDonald for distributing a workable, though not perfect, Linux system that make my life a good deal easier. I used to administer my own personal Unix System running on an NCR MiniTower. When it died, I switched to Linux on a '486, using the SLS installation. While there are things I would change, I was overjoyed to find that SLS provided me --- "out of the box" --- with a set of software that it had taken me months to download, configure, and install on my MiniTower, *even when I knew where to look for it.* I see SLS as a trade-off between "convenience" (the software is there) and "perfection" (Peter has installed it as I would have done), and I have found that SLS's advantages in convenience have far outweighed it's failings in perfection. Thanks, Peter. Peter Olsen, P.E. P.S. My strongest suggestion is that /usr/local be reserved for *truly* local material, so that I only have to back up one directory to insure that I'm ready for any upgrade. -- Peter Olsen, n2ell, pcolsen@super.super.org ...!uunet!super!pcolsen P.O. Box 410, Simpsonville, MD 21150-0410; 410-997-8584 "Engineering is the art of applying a professional knowledge of mathematics and the physical sciences to improve the quality of life" ------------------------------ From: hedrick@geneva.rutgers.edu (Charles Hedrick) Crossposted-To: comp.unix.bsd,comp.os.386bsd.misc Subject: Re: BSD UNIX Date: 4 Sep 93 04:23:44 GMT michaelv@iastate.edu (Michael L. VanLoon) writes: >It's kernel structure is not as clean, having had things piled on top >to add functionality many times. Actually I think this is not true. Linus is quite good about doing rewrites and redesign where needed to introduce new functionality. >They have recently adopted Net/2, but it is >still somewhat buggy. I wish they had used a different name, because lots of people have the same impression. What Linux calls net-2 is simply the second release of the Linux networking code. It has nothing to do with the Berkeley net2. It is certainly the weak point of Linux. While I find SLIP quite usable, based on the reports I see in both the public newsgroup and the mailing list, it's clear that when you start putting Linux on a busy network and having lots of simultaneous network connections, you're likely to have trouble. Someone said they were planning to integrate the Berkeley net2, but I don't know whether that project is really going anywhere. There's a certain wariness in the Linux community about using BSD code. I don't know other people's reasoning, but I think the USL suit provides a good reason. I personally think the USL suit is completely absurd, but courts have made ruling that I think are absurd before. Thus until that suit is settled, I believe the community needs an alternative that is completely independent of Berkeley. If it weren't for that, I'd say Linux would be better off just adopting BSD net2. >libraries so uses much less memory and disk space. Also, its rumored >to co-exist with DOS better, but this may be more opinion than fact. Linux uses DOS partitions, including "logical" partitions in an "extended" partition. Linux file systems and swap space are simply created in a DOS partition or logical partition. Since the DOS structure allows for up to 26 partitions (using logical partitions -- only 4 primary partitions), I think there's enough flexibility. The advantage of using the DOS structure is that Linux can read and write from DOS partitions using mtools or directly using the DOS file system code that's been in the kernel for the last 6 months or so. Linux has a version of fdisk that can be used to manipulate partitions of other operating systems. The Linux structure lets me move partitions and logical partitions back and forth between Linux and other operating systems. As I understand the BSD setup, you pretty much have to pick a single DOS partition, and once you've done that you're stuck with that amount of space unless you want to reinstall. The Linux approach seems like it's a bit more flexible, though there can be situations where it might be more dangerous. ------------------------------ From: haynes@cats.ucsc.edu (Jim Haynes) Subject: Re: SCCS for Linux Date: 4 Sep 93 05:38:43 GMT SCCS started out in life as AT&T licensed code. Hence the popularity of RCS (even at Sun installations) and the rarity of SCCS in the world. Now I may have this mixed up, but the Berkeley folk were or are using SCCS internally for their development; and I've seen some Makefiles come out of Berkeley that did SCCS operations. So there might be some non-licensed programs that do a very minimal subset of SCCS as a way to get you going if you get handed a file that is in SCCS form. -- haynes@cats.ucsc.edu haynes@cats.bitnet "Ya can talk all ya wanna, but it's dif'rent than it was!" "No it aint! But ya gotta know the territory!" Meredith Willson: "The Music Man" ------------------------------ Crossposted-To: comp.os.ms-windows.advocacy From: kevin@frobozz.sccsi.com (Kevin Brown) Subject: Re: NT versus Linux Date: Thu, 2 Sep 1993 16:13:47 GMT In article <930830.082811.4Z3.rusnews.w165w@mulvey.com> rich@mulvey.com (Rich Mulvey) writes: >muts@compi.hobby.nl (Peter Mutsaers) writes: > >>>> On Thu, 26 Aug 1993 02:27:00 EDT, rich@mulvey.com (Rich Mulvey) >>>> said: >> >> RM> folks? It's a for-profit company. They exist solely to make >> RM> money. What's the best way to make money? Kill your >> RM> competition, *especially* if they have a better product. Even >> RM> if they don't. What exactly does morality have to do with this? >> RM> Saying that they are 'evil' is basically saying that people >> RM> shouldn't strive to be successful. Gee, maybe we all should >> RM> spend the rest of our lives flipping burgers for each other. >> RM> But make sure that we avoid trying to provide a decent standard >> RM> of living for our families. >> >> RM> That would be moral by your logic, right? >> >> Yes, but they go too far. In the end it will damage themselves, like >> IBM was damaged too by the almost-monopoly they got in the 70s. First >> they get big profits, but the losses will be even bigger. > > If they damage themselves, that's *their* problem. New companies >will spring up to fill the void - just as it has always happened. And >*those* companies will probably be lean and mean at first, until *they* >kill themselves with bloat. That's the way that capitalism works. Which is true of the dynamics of companies within capitalism. But we're not really concerned about the companies so much as we are about the *products*. Let's take a look at two cases where the company in question was large, had an inferior product, and managed to have that product get real popular anyway due to the stupidity of the public. Case 1 is, of course, IBM. They came out with the PC. When it first came out, there wasn't much that was widely available that competed with it. Because it was an IBM product, *everyone* bought it. The PC gained most of its ground because people naturally thought IBM was good stuff. Little did they know that the product was actually quite badly designed, making use of a badly-designed CPU and a badly-designed bus (and a badly-designed DMA scheme, etc). Now look where we are today. What's the most popular type of computer? A PC clone. What kind of hardware is it? It's running a relatively broken CPU like the 486, something that has an instruction set and register set that leaves quite a bit to be desired, and running the *same* broken bus architecture that is responsible for the low performance of most of the hardware available for it. But because clones are so cheap, they're popular, and as a result people spend a great deal of time developing software for this broken hardware. Linux is a *perfect* example. I'm glad it happened, but I wish it had happened to better hardware. Case 2 is, of course, Microsoft. They came out with DOS for the IBM PC. Because people initially bought lots of IBM PC's, DOS became popular. Not on its technical merits (DOS has none), but on the IBM name. Because of this, people wrote programs for this broken "operating system", which caused it to become more popular, etc... Now look at where we are today. What's the most popular "operating system" available today? DOS, of course. Guess what needs it to run? Windows. We all know how badly broken and inferior DOS is. Yet, because it's so popular (due to the stupidity of the public), there are a lot of things that won't run under anything *but* DOS. Because people often do things that only one or two applications can provide the functionality for, they are often *forced* to run DOS, ugly as it is. Word processing is a perfect example. Lots of people have Wordperfect for DOS. Despite *its* brokenness, it's also one of the most popular things around. But to run it, you have to run DOS. Even if Microsoft were to go bankrupt *now*, DOS will be with us for a long, long time because it takes a long, long time to unwind the recursive dependency cycle which is behind the popularity of DOS. There are more examples, e.g. System Vr4, but I trust my point is clear: it doesn't matter *how* well a company may be doing in the future, the problems that they've left us with will cost us dearly, *exactly* as is happening now. An outside observer who wasn't really on the inside of all this might easily conclude that people in general *prefer* inferiority. This, of course, isn't really true. The biggest problem is that most people aren't bright enough or well-trained enough to recognize an inferior product when they see it, so they base their decision on things like how much it will cost or how much "support" they'll get (when, of course, few people realize that the promise of "support" for a product usually simply means that there will be someone on the other end of the telephone telling them that the bug they're reporting doesn't exist, hasn't been heard of before, and is being fixed even as they speak). >> Every company must fight its competitors, I agree; but they must also >> learn self-control. If you kill everyone you get an unhealthy >> situation, bad for everyone. > > But you just said that they are going to kill themselves as well, so >what's the problem if they will eliminate themselves as players in the >game? They'll leave us all with a legacy of inferiority. >> The former eastern-block was a >> monopolized economy, the results are clear. Microsoft is evil in a >> way, but more so, the buyers are stupid. They fall in a trap with >> opened eyes. Blinded by the lies and marketing stories of Microsoft, >> and too incompetent to judge the different choices on their technical >> merits. There are too many incompetent people on decision-making positions >> in companies, especially in automation departments. > > I have no sympathy for stupid/ignorant people. If they make decisions >about software and hardware without doing the proper research, they >deserve what they get. It is their responsibility. If a person is so >clueless that they think a salesman's sole interest isn't in padding his and >his company's pockets, and actually *believe* what the salesman tells them, >they shouldn't be making purchasing decisions. Yeah, but they usually *are* making purchasing decisions. >And if, as often happens, >the company fails because of poor decision-making, then we're only seeing >natural selection in action. Which, in the end, is a *good* thing. Agreed with respect to companies. But see above for why the process itself yields suboptimal results compared to making the correct choices to begin with. Additionally, you know what will happen, right? When Microsoft (or whoever) goes down, their product will eventually be replaced with another inferior product. >- Rich -- Kevin Brown kevin@frobozz.sccsi.com This is your .signature virus: < begin 644 .signature (9V]T8VAA(0K0z end > This is your .signature virus on drugs: <> Any questions? ------------------------------ ** FOR YOUR REFERENCE ** The service address, to which questions about the list itself and requests to be added to or deleted from it should be directed, is: Internet: Linux-Misc-Request@NEWS-DIGESTS.MIT.EDU You can send mail to the entire list (and comp.os.linux.misc) via: Internet: Linux-Misc@NEWS-DIGESTS.MIT.EDU Linux may be obtained via one of these FTP sites: nic.funet.fi pub/OS/Linux tsx-11.mit.edu pub/linux sunsite.unc.edu pub/Linux End of Linux-Misc Digest ******************************