Re: first install of RefDB .debs

"Markus Hoenicka" <[email protected]>
Newsgroups gmane.text.refdb.general
Message-ID <[email protected]>
Hi,

Dan O'Donnell writes:
 > One of the results of playing around with these many installations is
 > that I am understanding why commercial programs like Oxygen or Crossover
 > Office tend to skip the whole distro specific thing (or offer rpms only)
 > in favour of, as far as I can tell, building from source using a script.
 > Synapsen does it as well, and it is a joy to install, though a much more
 > limited program.
 > 
 > I wonder if something like this would be useful for refdb? Installation
 > is an issue for newbies, I think--perhaps even more so in the case of

I've had mixed experiences with these kinds of install scripts. I've
installed vmware on a Debian box the other day without a hitch -
everything just worked out of the tarballs. On the other hand I had to
watch a less ingenious install script of a tax software eat my
harddrive because of a minor version mismatch - the box just froze and
left the filesystem in a corrupted (but fixable, luckily) state. I
reckon the end-user experience is proportional to the manpower and the
care that is put into these scripts, and this makes it rather unlikely
for a project like RefDB to offer an install script that works
reliably on 99% of the computers out there. This is not meant to
discourage your attempts, or to critizise your abilities, but it is a
hard fact that the script you whipped up works reliably only for one
particular RefDB version on one particular version of one particular
operating system.

The basic idea of building the software on the end-user's machine
instead of providing binaries has been adopted by many operating
systems - most notably the *BSDs which provide "ports" in addition to
or instead of binary packages. These "ports" are essentially Makefiles
which do pretty much the same as you do in your install script: fetch
the tarball(s), resolve dependencies, build, and install. However, we're
looking at an integrated system here. Resolving dependencies can be
nested at an arbitrary level, whereas a script approach either has to
know about all dependencies of the dependencies of what it tries to
install, or has to live with the fact that it can only do so much.

David has already pointed out many of the problems in your script. You
may want to fix the script along these lines, but you should carefully
think about what audience you target. Is it 99% of the potential RefDB
users out there? How many OSes and distributions and how many versions
of these will you be able to cover and test? Or is it only the Ubuntu
users which use the same version as you do? In the latter case I'd
rather suggest to "port" David's Debian packages to Ubuntu. I don't
think that we manage to solve a problem (the app installer that works
on all OSes) that no one else has solved yet. But it takes only one
committed port or package maintainer to make RefDB available to yet
another OS or distribution.

regards,
Markus

-- 
Markus Hoenicka
[email protected]
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de



-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.