*BSD News Article 88772


Return to BSD News archive

Path: euryale.cc.adfa.oz.au!newshost.carno.net.au!harbinger.cc.monash.edu.au!munnari.OZ.AU!news.ecn.uoknor.edu!feed1.news.erols.com!newsfeeds.sol.net!newspump.sol.net!ddsw1!news.mcs.net!ddsw1!not-for-mail
From: les@MCS.COM (Leslie Mikesell)
Newsgroups: comp.os.linux.misc,comp.unix.bsd.misc
Subject: Re: Linux vs BSD
Date: 8 Feb 1997 14:24:22 -0600
Organization: /usr/lib/news/organi[sz]ation
Lines: 85
Message-ID: <5dindm$42i$1@Venus.mcs.net>
References: <32DFFEAB.7704@usa.net> <5dd0p1$72n@innocence.interface-business.de> <5dekt3$aiu$1@venus.mcs.net> <5dfpah$cre@innocence.interface-business.de>
NNTP-Posting-Host: venus.mcs.com
Xref: euryale.cc.adfa.oz.au comp.os.linux.misc:157211 comp.unix.bsd.misc:2416

In article <5dfpah$cre@innocence.interface-business.de>,
J Wunsch <joerg_wunsch@interface-business.de> wrote:
>les@MCS.COM (Leslie Mikesell) wrote:
>
>> No, it was an AIC-7850 on an IWILL motherboard.
>
>Hmm, i don't know much about it, although i know of AIC-7850s that
>work.

I'm sure they do now, but at the time I had Linux and FreeBSD CD's
and neither would install as shipped.  Linux emitted the remarkably
philosphical diagnostic: 'an idle task may not sleep' and freeBSD
wasn't any better.  I found the Linux solution in a couple of days
(and perhaps within days of when it was completed) but it was
several months before I saw a freeBSD solution.

>> As for not using the official support forums, why are they so well
>> hidden?
>
>Why are you not reading supplied READMEs?  The support forums aren't
>hidden, but even if your boot disk fails to see the SCSI controller,
>it doesn't fail to show you the documentation.

The controller was being detected as a supported chip. The documentation
I found implied it should work.  The *bsd newsgroup messages I found said
(only) that it didn't work for other people too.

>> These days it's downright unsociable to not show up in a search
>> using altavista, hotbot, dejanews, etc.  I can understand a
>
>We don't run them, so don't blame us for what they're considering
>to include or not.

I'm not 'blaming' you in the sense of demanding that you should do what
I want for free, but I am suggesting that if you want people to find
and use your product the way to do it is to make the answers to
their questions easy to find.  And these days, http://www.altavista.com
and http://www.hotbot.com, along with dejanews provide the easy way to
find things.  If someone could feed the mailing list archives to
monharc or something similar and point the search engines at it
the world would be a nicer place.

>> preference for mailing lists over newsgroups, but if you want
>
>Huh?  Did you even hear about comp.unix.bsd.freebsd.misc?  That ain't
>a mailinglist, mind you, and there's a lot of traffic in it.  (I
>should know, incidentally i'm answering many questions there.)

Yes, but as I said, a dejanews search for drivers found only questions
on the *bsd groups at the time, no answers.  There may very well
have been answers earlier on the mailing lists, but how was I supposed
to find them?
 
>> While I'm at it, another question: does freeBSD (or it's Linux 
>> emulation) support sysV style ioctl's on tty devices (#include 
>> termio.h  instead of termios.h)?
>
>I don't know about its Linux emulation, but i would suspect it does it
>by a layer in the library.  If so, the emulation should automatically
>pick up them.

I'd expect this to work for binaries, but you need termio.h to
compile.  Linux has one.

>FreeBSD itself doesn't, and certainly won't (unless you feel to
>contribute this to our libcompat ;-).  However, it's usually a matter
>of five minutes to convert a SysV-termio program into at least
>something roughly resembling a working termios program (even if you
>still do it by ioctl's, as opposed to use the Posix-official tcfoo()
>and cffoo() functions).  Practically, the only difference between both
>is that SysV termio doesn't know about split input and output line
>speed, while Posix termios does (in theory -- the PC hardware doesn't
>support it anyway).

Agreed, it wouldn't be an unreasonable task to change, but I have several
programs (phone smdr, a bunch of wire services including real time commodity
exchange data that get turned into screen updates and multiplexed out on a
satellite uplink) and need to keep them working under sysV also for a
while.  It probably saved a couple of days work to avoid #ifdef'ing
and testing everything that would have to compile both ways.  (In other
words, more time than I'd ever expect to save even if freeBSD does
have faster network code...).

Les Mikesell
  les@mcs.com