Return to BSD News archive
Path: sserve!newshost.anu.edu.au!harbinger.cc.monash.edu.au!simtel!news.sprintlink.net!howland.reston.ans.net!vixen.cso.uiuc.edu!newsrelay.iastate.edu!news.iastate.edu!wakko.cce.iastate.edu!marcus From: marcus@ccelab.iastate.edu (Marcus I. Ryan) Newsgroups: comp.unix.bsd.freebsd.misc Subject: Re: The Future of FreeBSD... Date: Mon, 24 Jul 1995 11:08:47 LOCAL Organization: Iowa State University Lines: 315 Message-ID: <marcus.202.00C5B93F@ccelab.iastate.edu> References: <3uktse$d9c@hal.nt.tuwien.ac.at> <3ulsro$ssl@agate.berkeley.edu> <3umkok$de2@nntpd.lkg.dec.com> <marcus.197.009F3034@ccelab.iastate.edu> <3us0rg$7ph@nntpd.lkg.dec.com> NNTP-Posting-Host: wakko.cce.iastate.edu X-Newsreader: Trumpet for Windows [Version 1.0 Rev B final beta #4] In article <3us0rg$7ph@nntpd.lkg.dec.com> Jon Jenkins <jenkinsj@ozy.dec.com> writes: >From: Jon Jenkins <jenkinsj@ozy.dec.com> >Subject: Re: The Future of FreeBSD... >Date: 22 Jul 1995 23:16:00 GMT >Hi Marcus, >thanks for replying to my post. You're welcome :) >marcus@ccelab.iastate.edu (Marcus I. Ryan) wrote: >>In article <3umkok$de2@nntpd.lkg.dec.com> Jon Jenkins <jenkinsj@ozy.dec.com> >wri >tes: >Actually Windows/386 was the first and the compatilbility problems >killed the other efforts. I'm sure that this compatibilty problem >was not all by "accident". I tried Windows/386 and Windows/286 before windows v3.0 - it was horrendous. 3.0 was a dramatic improvement - so much so I tend not to acknowledge the earlier versions ;) [stuff about VXP -- Visual X windows Programming Interface and LessTif] >Well this would be a good start as long as it will work on the LessTif >base. Motif being a commercial and relatively expensive is not >a consideration for "FreeBSD" at least in my view. Well, that's LessTif's intent - 100% compatibility with Motif... >I'm hoping to get together (in the electronic sense) with Terry >Lambert to have a look at this area and we will certainly look >at this option. Interested in helping out ?? I'd love to, but unfortunately I've already promised some of my time to the documentation project and with my schedule that's already more than I can handle :) (don't let the .edu fool you, I work for a living too :) >> There are already other X-Windows >>applications to do as you suggest. Though the ones I can think of off the top >>of my head are commercial, I'd be amazed if there weren't shareware ones out >>there somewhere... >The reason perhaps you cant think of them is that they are scarce! >How many Windows apps builders can you think of immediately ? >VB, MSC, BC++ and the first of the complete object orientated >paradigm generation and the best of all: Borlands Delphi Well, that and I've not looked - I was raised in the command-line world, so I run after my text tools first. :) >Did you know at last count that there are 10 million OS/2 users. Yes >thats correct. Most of them are in the commercial sector >so we dont hear the hype in the more academic world but >OS/2 has a very solid and fanatical user base. Dont be Don't forget Government - there are probably about as many government agencies running OS/2 as commercial companies (though that's just my overall impression - not based on any facts or figures). >surprised if OS/2 becomes a major player over the the next >few years. Did you know that OS/2 is being ported to >other workstation platforms ?? I know it's being ported to the PowerPC :) Other than that I was not aware. However, I don't think that will have much of an affect. What makes an operating system popular is who developes what apps for it, or at least what will run on it. I'll be interested in seeing what happens with FreeBSD and its kin when the WINE project fully implements windows in X. I may be completely nuts, but I seem to recall reading the newsgroup a while ago where people were already running MicroSoft Word under X. >> OS/2 Warp managed to catch up by quite a bit, simply by including the >>Internet BonusPak. Do you think that "You on-ramp to the information >>superhighway" and "Enhances You Exisiting DOS and Windows(TM)" on the side of >>the box did nothing for sales? If FreeBSD had the resources to do the kind of >>advertising they have, I bet that FreeBSD would be much more popular. >Again I think we will have to disagree. I don't think we have to disagree completely. I use OS/2 and like it, but can you honestly and with complete confidence say that despite the fact that they based entire commercials around that bonus pack it had no significant impact on sales? >Linux was never advertised and look at its poularity. >In Europe its user base is 10x that of FreeBSD. Advertising is not necessarily commercials. Advertising is also done by word of mouth, getting magazines to review the products, having booths at shows, etc. FreeBSD is capable of competing with Linux and its level of advertising to some degree (6 or so distributions of linux being sold by commercial vendors, vs. 1 distribution of FreeBSD being sold by Walnut Creek). I don't think either will have advertisements on TV like IBM, MicroSoft, etc., anytime soon which is more of what I was referring to. >>UNIX is not any harder to use than DOS. >Whoa there, surely you jest! I wont go into a >diatribe here. Just compare setting up two simple >applications (forgetting all the nasty kernel configuration) >compare setting up Eudora Mail on Windows with /etc/sendmail.cf >on UNIX!!!! and then compare setting up a dot matix >printer in Windows versus printcap etc on UNIX: Nuff said. Windows != DOS! Don't make this mistake. I was not comparing X with Windows, I was comparing command-line, text-based UNIX with command-line, text-based DOS. On top of that, sendmail.cf is something I've never needed to play with. It's default settings work well for my system. If you want to compare readers/viewers, comapre Pegasus Mail for DOS, and mh (because it's what I've set up). 1) FTP them to your local machine. 2) Extract the archives (PKZip and pkg_add) 3) run PConfig for PMail, run mh-install for MH. 4) If a different editor for PMAIL is prefered, you're SOL. If a different editor for MH is prefered, edit .mh_profile and add EDITOR: uemacs :) 5) Use them. mh takes the same number of steps to install, unless you want to use something other than the systems default editor, an option not available in any DOS mail programs I know of. Remember, lets compare apples with apples. >> People talk about having to reconfig >>the kernel, and edit some files in /etc, but look at all of the time and >>tweaking it takes in DOS just to get some device drivers going, >I have never had a problem with this but I have >had plenty of problems setting up FreeBSD and I work >inside the kernel of OSF/1 and ULTRIX everyday!! 2.0.5 is a little easier when it comes to PS/2 mice, but other than that, at most I've had to open up LINT and GENERIC together and get the proper line from LINT. If I REALLY want to get fancy (which I usually do for speed's sake), I read through and figure out which devices I don't need, and then remove them (though I must admit it annoys me to get compile errors if I remove the PCI line even on a non-PCI machine *Shrug*). In MicroEmacs, it's a matter of a few ^K (cut) and ^Y (paste) commands here and there, then config GENERIC && cd ../../compile/GENERIC && make depend && make && install\ && reboot and I'm done. Comparing setups for my machine at home: 486DX/2 66MHz Intel, EISA ATI Mach32 Graphics Ultra Pro 2MB PC Logic Modem on Com3 IRQ 5 SMC Elite16 Soundblaster Pro Adaptec 1740 Quantum 1G hard drive Mitsumi CD-ROM Colorado Jumbo 250 UNIX/X (1 CD, or FTP connection): Run the standard FreeBSD install program and check the various options. Installing X, I have to answer a few questions that I pull straight out of my manuals about my monitor and video card, but for the most part it's just checking boxes and answering yes/no questions. Add a few packages with the pkg_manage, do a quick kernel reconfig (for a faster bootup, and to add the sb device, copied over from LINT). Type in the command above, and in a little bit (20-30 minutes, unfortunately), I have a freshly optimized system. I admit that the time needed for the reconfig is irritating, but as a trade-off for the advantages, it's acceptable. DOS: (17 floppies for DOS, Windows, and drivers, no apps) Install DOS. (3 disks) Install Mitsumi CD-ROM driver (1) Install Windows (7) Install ATI Mach32 driver (2) Install SMC Ethernet drivers (1) Install Soundblaster drivers (2) Install Colorado Tape software (1) tweak memory, clean up autoexec.bat and config.sys so everything is readable, (so I don't have 3 PATH=%PATH%;... statements, and to add a SHELL=C:\DOS\COMMAND.COM /E:1024 /P so that all of the new environment variables won't run me out of environment space). To get enough memory to run my applications, I have to run PCTools Ramboost, undo some of its changes, and then run MemMaker. Several applications still won't run, so I'll probably have to get Multimedia Cloaking utilities, or set up multiple boot configurations for my various games and large DOS programs. That says nothing of the addtional setups I'll have to do to actually efficiently run on my NetWare network (editing net.cfg so that lsl and vlms don't take up most of my conventional memory remaining). Of course, all of that assumes that those installs do both DOS and Windows at the same time. Otherwise, I have to do even more installs and configs to get the things to work in Windows. None of the installs look or operate the same. >> let alone to >>get the thing on the network. >Windsock was trivial, perhaps a few minutes. /etc/hosts >/etc/hosts.conf, bind, NIS, SLIP, PPP, routing, device >slattach, etc etc .... There really is no comparison Again, Winsock is Windows, but I'll acccept it. Winsock: Install, add Trumpet's directory to path, odipkt and winpkt commands to startnet.bat, and reboot. Run windows, and then trumpets setup. Add your IP#, domain name, netmask, default gateway, name server, Packet Vector, and if you want, play with a few other settings, though their not necessary. Exit and restart for changes to take effect. FreeBSD: During install tell it IP#, domain name, hostname, netmask, default gateway, and name server. When you reboot to load the system the first time, it's there and ready. You keep comparing by saying everything you can tweak about UNIX. Just because it's there doesn't mean you have to mess with it. Look at all of the settings you can add to WIN.INI and SYSTEM.INI. They'll make it run faster, but you don't have to touch them in most cases. The same is true with Autoexec.bat and config.sys, but those you likely will need to play with. The main problem is that no devices come pre-installed with DOS, and Windows at minimum requires you to go in after it's done (with the exception of printers that they already have drivers for) and feed it a few disks. On top of that, you don't have the problem of keeping track of all of your driver disks, keeping each up-to-date seperately, etc. It all comes on one CD, and when one thing is updated, they're all updated. Everything that runs under a standard UNIX command prompt should run under an xterm in X, but not one of my favorite games (Doom, X-wing, Wing Commander 3, etc.) will run under a dos shell in any of OS/2, NT, or Windows 3.11. OS/2 does offer the best chance though with all of the settings you can change. [snip] >As for "ls" well here is perfect example: if we had a good file manager >for X then "ls" would be available for your use if you wanted but >bascially obsolete for everyday use. Yes I do have xfm but we miss the point: X is not officially part of FreeBSD, so why is it on FreeBSD's shoulders to develop it? On top of that, any X-windows program should compile on any UNIX if programmed correctly, which means if the linux camp developes one, as long as the source code is available, FreeBSD can use it to. That's the beauty of UNIX, free software, and an OS that comes with compilers built-in :) [snip - GUI builder app stuff] >Which is exactly what I am saying: >If FreeBSD is to put effort into anything over the next few years >then a object based GUI builder app should be high on the list >of priorities. I agree with the need for such a thing, but I disagree that the FreeBSD core team should be the ones to do this. Say things were a little different in the DOS world. MicroSoft decides to sell DOS to Netware, so that DOS and Windows are no longer owned by one company. What you are are saying is that Novell (owners of DOS) should develop a nice GUI builder app. for Windows. [snip] >Just as an interesting aside did you know >that the US goverment has accepted NT as an Open >System. If you dont get the significance of this >then it effectively allows NT to replace any >UNIX operating system in the Goverment services. >Two large military organisations (I'm not sure >I can tell you who so they will remain nameless) >have already announced they will swap from UNIX >to NT. You have got to ask yourself why ? Okay, why? Especially when Novell's UnixWare is available, and completely manageable using the Motif-based manager tools. It's still UNIX, but is as easily configured as NT. That's a different conversation though. I also use NT and love it. The only reason I keep DOS around is because NT won't run my games, and because Reachout and my scanned don't have NT version yet (they still rely on DOS device drivers) [snip] >Unless the UNIX community gets together >and provides the common toolsets to develop >fast cheap GUI applications including >system admin and configuration and >mulitmedia in an object based paradigm >then I predict that within 10 years >UNIX will be relagated to >academic circles with NT and OS/2 being >the predominat OS for both Goverment and >commercial systems. I disagree, but I won't argue the point, as it has been argued repeatedly. I do agree that FreeBSD, and UNIX in general will not replace NT, OS/2, etc., for quite some time to come without a lot more applications, just like OS/2 won't outdo windows until there are a lot more native OS/2 applications (although OS/2 will run Windows apps, you have to buy a copy of windows to do this meaning that for each person running a windows app under OS/2, one version of each was sold. OS/2 may make again against it's previous sales, but it doesn't close the gap any faster). Regardless, you state directly that this is not a problem of just FreeBSD, but of UNIX as a whole. Rather than saying FreeBSD will die if this doesn't happen, and pinning all the responsibility on the team, and freebsd users, pin it on EVERYONE. Get the Linux, NetBSD, Mach, and 386BSD, and WINE project programmers involved too. Personally I think all of the UNIX teams need to work together to get people using UNIX, to develop the tools needed to make it work, to get applications ported to UNIX, and to make people notice it in general. Then we can fight and bicker about whose dad's can beat each other up :) ---------------------------------------------------------------- Marcus I. Ryan |*Diplomacy is the art of saying 'Nice Asst. SysAdmin, CCE Labs | doggie!' until you can find a rock. Iowa State University |-------------------------------------- shad@iastate.edu |*Diplomacy - Letting someone else have (515) 294-0715 | things your way. ----------------------------------------------------------------