[Web4lib] What Sort of Library is Open Source Software For?

[Don Robertson]

> The big difference between commercial support for proprietary software and
> commercial support for Open Source software is that the former is
> obligatory and the latter is optional.  If you decide not to pay ongoing
> support for proprietary software, you lose the right to use
> it--totally.  If you decide not to pay for ongoing support for open source
> software, you still retain the right to keep using it.  That is one of the
> main reasons why support costs for proprietary software are so high in
> comparison to support costs for Open Source software.  This is what Mike
> was alluding to when he used the word "monopoly."  (My only disagreement
> with Mike on this point is that, realistically, there is not much choice
> for getting support.  In theory, anyone could offer commercial support for
> Index Data's Open Source software.  In practice, only Index Data does.)
>
> >  There are many libraries that don't have "root" on their own system,
> > which is reserved by the vendor for maintenance and repairs (and to keep
> > the library from getting itself into a mess that the vendor will then
> > need to fix). So library "system staff" may not be doing programming or
> > development currently, and the library may not have staff that can create
> > applications.
>
> Here again you assume that there is no commercial company supporting Open
> Source "turnkey" software.  A library that so wishes can treat Open Source
> software from a commercial company as a black box to be delved into only by
> the vendor.  The main reason that this is so rarely done is that most
> librarians share your misconception that Open Source software is only
> appropriate for libraries who have geeks to grok it.  By perpetuating this
> myth and saying that library Open Source software is appropriate only for
> libraries with lots of technical experts (with lots of time on their
> hands), you do a disservice to companies like Index Data and LibLime whose
> Open Source software is fully (and commercially) supported and is quite
> appropriate for any library, regardless of the level of involvement they
> want with their software.

You also assume vendor support for proprietary software.

I have been using ans supporting open source software for quite some time. I 
now find myself using proprietary LMS software. It has been a revelation to 
me how bad the support is. It reminds me when I was supporting Windows 95 in 
- er - the mid 90's. 

Most support comes from mailing lists hosted by a north American university, 
not by the vendor. theere are few Open Source projects without mailing lists 
and bug track databases.

The Vendor is trying - they have given access to all the bug reports, not just 
the ones you have filed. Unfortunately, they bought into a proprietary 
Knowledge Base thing - and it does not seem to work without MS IE. 

We run a Linux version of the software, and I use Linux nearly all the time. 
But I will need to reboot into MS to use Explorer to use a web based support 
site to search for documents relating to the Linux software the vendor ... 
okay - they donated it to us ... but still.

I have emailed to find out the price of upgrading to the latest version but 
have not heard back. 

Our version runs on Red Hat 7.1. I would like to install it on SUSE, as a 
Khmer version is being developed by KhmerOS.info (incidentally, all you Khmer 
speakers - KDE, OpenOffice, Thunderbird and maybe more are available in Khmer 
now - see KhmerOS. Microsoft - maybe the next version - maybe not. We will 
have to ask MS .... ) Anyway - I cannot get information about installing on 
SUSE, Fedora, RH 9 or whatever is new ...

> >>No.  There is no reason at all why maintaining an open source system
> >>should be more expensive (in either time or money) than a proprietary
> >>one: in fact, the converse is usually true, since _any_ suitably
> >>skilled/experienced programmer can make changes as required to an open
> >>source system, whereas with a proprietary system you are tied into the
> >>vendor's support arrangements or nothing, with all the usual negative
> >>(for the customer!) consequences of any other monopoly.

Exactly - Open Source usually uses existing published standards for data (or 
publish their own standards), and has published API's. Adding features or 
integrating other software  is much easier. Our LMS is almost impossible to 
understand - there is no published API's, no documentation of what variables 
are used - the WebOPAC seems to use the same variable - informativly named 
$0100 etc for multiple values depending on which part of the page you are on.

You do not seem to be able to take, say, the DDC number, and move it somewhere 
promenant.

Disgruntled of Phnom Penh
-- 
Don Robertson
Systems Administrator
Hun Sen Library
Royal University of Phnom Penh
Kingdom of Cambodia
don at robertson.net.nz 012 769 280