[Desktop_architects] Applications and pre-installed machines

Segedunum segedunum at actuaria.co.uk
Fri Jan 27 11:49:51 PST 2006


On Friday 27 January 2006 17:14, John Cherry wrote:
> This list is generally concerned with (1) ISV porting issues, including
> things such as the Portland project, (2) open driver issues, including
> IHV and proprietary API issues, and (3) desktop developer issues,
> including the cooperation of desktop organizations in solving these
> issues.

Hmmmm. You might want to summarise these three points and what these actual 
'issues' are, because from where I sit ISVs don't have any issues because 
they have no reason to port to (or write software for) desktop Linux - as 
shocking as that might be for people. Before those three points above are 
addressed there's probably a couple of extra points to go in before them, 
with words to the effect of actually attracting ISVs to your platform in 
enough numbers for any desktop investment to be viable. This means making 
your platform attractive enough to start off with, which means supporting 
your open source developers first and foremost - within reason.

You're simply not going to see Quickbooks running on a Linux-based desktop 
any time soon, regardless of how many surveys people do. I've had some 
informal discussions with some some ISVs and people from Sage here in the 
UK, and they just simply aren't interested in any platform other than 
Windows. Sage are a company that sells financial software, primarily to 
SMEs. Apart from the number one reason for total disinterest in Linux (or 
even Mac), which is that SageandMicrosoft are like that, a few other 
problems came out. These may be right or wrong, and these are of course 
totally off-the-record and don't represent any sort of official position 
for them, but this is the general perception:

1. No market share. Regardless of people jumping up and down telling us 
that Linux has more market share than Apple, they see more Macs in the 
world and in businesses than any Linux desktop - and they don't port to 
Macs either. Desktop Linux is also diluted between different 
distributions, and the market share that there is is not coherent enough 
to target. How to go about boosting market share into something 
recognisable is number one, and comes first.
2. Support. They do not want to support any more than one distribution of 
Linux, or at the very least, distributions that do not use the exact same 
binaries and OS set up. Support has wider issues, such as the installation 
of the software, how it acts when running, just what they guarantee the 
customer it will run on and the documentation they provide. That's why 
many organisations are even jittery about installing Windows service 
packs. The LSB guarantees them nothing in that regard. It's easier just to 
remove any doubt and have the same binaries and OS!
3. Installation. They definitely *will not* use or support the many methods 
of installation that currently happen on all distributions. The only 
reliable way of getting software is through a distributor's repository and 
doesn't provide the flexibility they need in getting it to their 
customers. There is nothing comparable to the Windows installation methods 
in widespread use.
4. Development. Apart from some Java development tools like Eclipse, which 
run far better on Windows, the state of Linux desktop development tools 
and frameworks and supporting infrastructure is not great.

David



More information about the Desktop_architects mailing list