Re: fairy nuff (fwd)

[Brandon Long <bl...@uiuc.edu>]

Beth asked me to forward this information to you, not sure if it'll help
much now . . .

>From our "OneStep" downloader, the following statistics for what was
downloaded apply.  This doesn't include our ftp server, and since the
OneStep downloader is a hand-holder as it is, I expect the source 
percentage is lower then otherwise.

Total:			4707	100
source:			683	 14.5
aix 3.2.5: 		225	  4.8
hpux 9.0.5: 		356	  7.6
irix 4.0.5: 		51	  1.1
irix 5.3: 		242	  5.1
linux 1.2.13: 		1441	 30.6
osf 3.0: 		121	  2.6
solaris2.4_sparc: 	660	 14.0 
solaris2.4_x86: 	51	  1.1
solaris2.3_sparc: 	182	  3.9
sunos 4.1.3: 		610	 13.6 
ultrix 4.0: 		85        1.8

Now, we don't have access to some machines which could be fairly
high downloaders, like BSD/OS, FreeBSD, and NetBSD.  I know a lot of
the source downloaders are from random sysv compatible machines as well.

Brandon

----- Forwarded message from Beth Frank -----

Hi Brandon,
	If we have any stats on how many binaries vs.
source are picked up would you please send it to the
apache list?

	-Beth

Forwarded message:
> From: Rob Hartill <ha...@ooo.lanl.gov>
> Subject: Re: fairy nuff
> Date: Thu, 30 Nov 95 9:37:04 MST
> 
> > Ok.
> > What's the problem that is holding us up?
> > The system binaries are a complete shambles. We should give up on them:
> 
> No, but I agree with you that they needn't hold up the release of the
> source.
> 
> > 1. Release of a 1.0.0 source distribution to go ahead now.
> > 2. Discussion of a binary release to proceed separately, with agreement on
> >    binary distribution to be achieved by 1.1 at the latest.
> > 
> > Thus I amend Rob's timetable as follows:
> > 1. Everyone stop building binaries now.
> > 2. We build a final source distribution (including manual) by tomorrow.
> >    (Preferably by end of today.)
> > 3. We allow three working days for the distribution to be propogated to
> >    mirror and archive (NCSA, sunsite) sites, and for the online documentation
> >    to be finalised.
> > 4. We announce the new release.
> > 
> > We shall do it my way, unless someone else can convince me their way 
> > is better.
> 
> I second this.   +1
> 
> Back to the binaries. My boss asked me the other day how popular
> Apache was (thinking 1.0 had been released weeks ago {hmm, I wonder
> where he got that idea from :-)  }). He thought that NCSA would
> continue to dominate the server "market" because they offered binaries
> (which he knew we didn't (yet)).
> 
> Binaries are needed. I've had direct requests for binaries. People do
> want them. Some bozo at HP was threatening to bad mouth Apache
> in his "small company" because he couldn't compile it and there wasn't
> a binary available for him to pick up. I shut that clown up, but if
> someone from HP can't use their own compiler, or find local help to
> compile the thing, you can imagine the trouble other bozos^H^H^H^H^Husers
> will face.
> 
> It would be very interesting to find out how many people use the
> NCSA binaries. I doubt if there are any figures for this, and there's
> no way to check from Andy's (Netcraft's) server survey. Some time ago
> I suggested we name the binaries differently so that we could spot them
> using HEAD. Can we do this please?
> 
> Apache 1.0.0*
> Apache 1.0.0x
> Apache 1.0.0(precompiled)
> Apache 1.0.0(binary)
> 
> or something similar.
> 
> 
> rob
> 


-- 
		Elizabeth(Beth) Frank
		NCSA Server Development Team
		efrank@ncsa.uiuc.edu
----- End of forwarded message from Beth Frank -----

-- 
 Brandon Long         "I think, therefore, I am confused." -- RAW
 HTTPd/SDG/NCSA	  "Of all the things I've lost, I miss my mind the most."
 ECE/UIUC	    blong@uiuc.edu   http://www.uiuc.edu/ph/www/blong
 N9WUC		Don't worry, these aren't even my views.