gnumed-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Gnumed-devel] install instructions


From: Ian Haywood
Subject: Re: [Gnumed-devel] install instructions
Date: Sat, 8 Nov 2003 10:44:28 +1100

On Fri, 7 Nov 2003 21:03:50 +0100
Sebastian Hilbert <address@hidden> wrote:

> On Friday 07 November 2003 19:00, Hilmar Berger wrote:

> to see something like it for GNUmed. Given the current state of GNUmed I do 
> not think that a graphical installer has high priority. Unfortunately we will 
I would have agreed, but I think that Yudhvir Singh Sidhu's experiences show
that gnumed's installation is currently so difficult we are discouraging 
*developers*
(let alone users) from the project, I think we have to move forward on 
packaging, even
though what we are packaging doesn't do much.

I have also used NSIS, but I simply had it install a snapshot of my C:\Program 
Files\Gnumed,
your method seems a bit cleaner.

> BTW , on Windows we only need an installer for the client.
> So this can be accomplished by using nsis. No need for extra coding.
Installers will have to be OS specific, (as a Debian user, I expect a .deb, I 
don't want some strange executable)
 but we can have a set of Python scripts which are shared
(the same script can be called from a Debian postinst script as from NSIS)
 
> happen. Considering the time it takes to produce the image I do not see
> any considerable , end-user visible progress. This is partly due to my 
> failure 
> to code more during the last six month but it shows very clearly that GNUmed 
> is definetly not progressing the way we (at least I) had hoped. 
The biggest problem is still a clear consensus about the widget-to-backend 
pathway,
Richard's GUI design makes this programmatically more difficult than might
be on other projects (but well worth the effort)  
once this sorted (and I think we are nearing this point), you will see a lot of 
frontend
functionality fairly quickly.

More importantly, once done, we can construct a task list and modules can be 
farmed out to developers 
to work on (reasonably) independently, effectively allowing parallelisation of 
development,
and expanding the team beyond a small group of people with the undocumented 
development plan locked firmly 
away in our brains (as Syan has discovered to his frustration)

Ian


-- 
PGP public key E750652E at wwwkeys.pgp.net
9BF0 67B7 F84F F7EE 0C42  C063 28FC BC52 E750 652E




reply via email to

[Prev in Thread] Current Thread [Next in Thread]