Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 28 Feb 1996 18:17:04 -0800 (PST)
From:      Jake Hamby <jehamby@lightside.com>
To:        Terry Lambert <terry@lambert.org>
Cc:        terry@lambert.org, jkh@time.cdrom.com, hasty@rah.star-gate.com, root@dihelix.com, hackers@FreeBSD.ORG
Subject:   Re: Quake's out, where's that Linux ELF emulation?
Message-ID:  <Pine.AUX.3.91.960228180304.25940A-100000@covina.lightside.com>
In-Reply-To: <199602290103.SAA09633@phaeton.artisoft.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, 28 Feb 1996, Terry Lambert wrote:

> Uh, OS/2 hasn't been very successful compared to DOS (or Win95).

Sure, but from what I understand, OS/2 sells a million copies a month, you
just don't hear about it from the media much.  Personally, I gave OS/2
(both 2.1 and "Warp") multiple chances but gave it up due to:  hardware
incompatibilities (on a PC that successfully ran EVERY OTHER OS couldn't
boot Warp because of a bad KEYBOARD driver!), lack of native software,
desire to run Win32 apps, desire not to have to do too much tweaking to
get DOS progs to work, desire not to have to deal with two different UI's
since Windows apps were essentially running on top of Win3.1 on top of a
DOS VDM, etc... 

> Because it will have been regression tested on Linux but not on BSD.

Still don't understand how you can realistically regression test your
program for Linux considering the hundreds of different distributions +
the likely modifications the user has made.  Let's see:  "This product
tested with Slackware Linux 3.0 (Linux 1.2.13 and 1.3.42 running libc
5.0.9, ld.so 1.6, gcc 2.7.2, ELF svgalib 4.2, etc..)"!!  Is it just me, or
is that slightly insane?  :-) I was tired of having to get, e.g. a new
libc or "experimental kernel" to run XYZ program, then find weird
anomalies (hey 'make' doesn't work, oops 'ppp-on' doesn't work, guess it
must be that new LIBC I installed!!).  FreeBSD is at least regression 
tested with its OWN COMPONENTS, how many Linux users can say that about 
their distribution (at least after they've upgraded various components)?

This is still my major gripe with Linux, and it is because Linus only
tracks the kernel, he has no influence over any of the other components
that make up a "distribution."  Things were much simpler back in the a.out
days, that's for sure..  Now, how does all that compare with "This product
tested with FreeBSD 2.1.0-RELEASE (using Linux libc 5.0.9, etc..)"  :-)

> Because a commercial software vendor does not typically offer support
> for an OS running their product in an emulation environment.

Granted.  But I repeat, I would rather have an unsupported product than 
no product at all.  I would rather have a vendor say their product works 
with FreeBSD, then find that they have 100 purchasers calling them, "This 
product works great!  When can we expect a FreeBSD native version?" then 
to have no product at all, and no user base to demand a FreeBSD port.  
Think about it...


---Jake



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.AUX.3.91.960228180304.25940A-100000>