instant-user — For user and developer info about Inst.ant

[Instant-user] FASD project: Online survey launched

[<ben...@id...>]

Dear Open Source developer

I am doing a research project on "Fun and Software Development" in which I kindly invite you to participate.
You will find the online survey under http://fasd.ethz.ch/qsf/. The questionnaire consists of 53 questions and you will need about 15 minutes to complete it.

With the FASD project (Fun and Software Development) we want to define the motivational significance of fun when software developers decide to engage in Open Source projects. What is special about our research project is that a similar survey is planned with software developers in commercial firms. This procedure allows the immediate comparison between the involved individuals and the conditions of production of these two development models. Thus we hope to obtain substantial new insights to the phenomenon of Open Source Development.


With many thanks for your participation,
Benno Luthiger


PS:
The results of the survey will be published under http://www.isu.unizh.ch/fuehrung/blprojects/FASD/.
We have set up the mailing list fa...@we... for this study. Please see http://fasd.ethz.ch/qsf/mailinglist_en.html for registration to this mailing list.

_______________________________________________________________________

Benno Luthiger
Swiss Federal Institute of Technology Zurich
8092 Zurich

Mail: benno.luthiger(at)id.ethz.ch
_______________________________________________________________________

[Instant-user] Re: Design and features

[Danno F. <sh...@ya...>]

> Danno,
> The current design seems to be a bit too tightly
> coupled to swing/awt with references in a number of
> the net.sf.inform classes. You know this is the
> classic case for the application of the bridge and
> factory method patterns. 


It's not the net.sf.instant.inform classes that are so bad, it's the 
references to the net.sf.instant.inform.swing references that we should 
worry about.  The classes in inform.* itself I designed to be generic 
and to merely hold information from the presentation layer, which is 
implemented in inform.console and inform.swing, so we would need an 
inform.swt.  If there is presentation specific classes in inform.* they 
should definatly be abstracted out.

What we need is some sort of a class to generically get the splash  and 
wizard screens and attach the to their respective models in inform.*. 
 Right now they are done in the task classes itself, and you right we 
should seperate it and place it inside inform as smething like a 
PresentationLayerFactory class.

The question this raises is how do we decide which presentation layer to 
trip?  I was initially using a magic property "inst.ant.console," if the 
awt classes failed to initialize or this property was set to true then I 
would do the console path.  We could stay with the magic property and 
change it to infom.ant.display and let ti take values of swing, swt, and 
console.  And we could add an attribte to InstSplash and Wizard to 
over-ride this (with console always being a back-up).


> Yes, I was asking specifically about what install
> related tasks you envisioned. Ant already supplies a
> great number of tasks that are useful during an
> install. However, some things I was thinking that are
> missing (and needed in during an install specifically
> in the Windows environment) are ways to interact with
> the registry and do things like add a shortcut to the
> desktop or a icon to the tooltray.
> 
> Ted



Those would go under net.sf.instant.propellant, and definatly need to be 
written at some point.  Single platform specific tasks would be ok, and 
I would think they should go under propellant.win, propellant.linux, 
propellant.macosx, etc.  Raw Registry interaction would go here.  If any 
of those tasks run on a platform that doesn't make sense for them the 
should quietly do nothing (perhaps logging that on the verbose level). 
 Tasks that do things that can have similar affect on multiple platforms 
can go jsut under propellant, such as adding desktop icons, application 
menu icons, and quick-start links.  These should deligate to classes 
inside of propellant.* based on the paltform ideally.  As for adding a 
tooltray icon, that seems to be a runtime dependent feature.

By the way, if you have a better ant pun than propellant for these 
features feel free to mention it, propellant seems close but not quite 
close enough to me.

--Danno

 

[Instant-user] splash and progress

[Ted T. <ted...@ya...>]

Danno,
1) Looking at SWT it appears that there isn't a widget
which can render html.
2) It looks like you modeled the InstallShield splash
behavior but if I recall correctly InstallAnywhere
just lets you put up an icon (splash screen) which
goes away when the wizard starts. What I'm getting at
is it doesn't seem necesary for instSplash to be the
parent of all of the other install tasks does it?
3) Is there a task for displaying progress of the 'do'
tasks? Also, how is 'doPrompt' differnet from 'prompt?

Ted



__________________________________________________
Do you Yahoo!?
Yahoo! Mail Plus - Powerful. Affordable. Sign up now.
http://mailplus.yahoo.com

[Instant-user] Design and features

[Ted T. <ted...@ya...>]

Danno,
The current design seems to be a bit too tightly
coupled to swing/awt with references in a number of
the net.sf.inform classes. You know this is the
classic case for the application of the bridge and
factory method patterns. 
Yes, I was asking specifically about what install
related tasks you envisioned. Ant already supplies a
great number of tasks that are useful during an
install. However, some things I was thinking that are
missing (and needed in during an install specifically
in the Windows environment) are ways to interact with
the registry and do things like add a shortcut to the
desktop or a icon to the tooltray.

Ted



__________________________________________________
Do you Yahoo!?
Yahoo! Mail Plus - Powerful. Affordable. Sign up now.
http://mailplus.yahoo.com

[Instant-user] Re: Inst.ant - Re: Mail delivery failed: returning message to sender

[Danno F. <sh...@ya...>]

As far as SWT v.s. Swing, I was attempting to design the interface so 
that either could be used, or AWT or simple console interactions. 
 Whether I have coupled to tightly to swing right now rather than to 
keeing it generic is up for discussion, it's been a bit since I've 
looked at it.  I think brining in swt as well as swing would be a goof 
proof of concept for the design.

As for the new ant functionality, I've just beens subscribed to the 
ant-dev list at apache.  There was some discussion when the commit came 
in.  Unfortionately it's not too well documented yet.

As for a plan that the tasks inst.ant will provice, I am not sure what 
is being asked.  How it will interoperate?  It will be an external lib 
brought in via taskdefs.  What tasks need to be written?  Dunno, we 
should exploit as many of the existing ant tasks as we can, that was 
part of the point of writing Inst.Ant was to re-use the core of an 
existing tool rather than creating new java classes to do intaller 
classes.  That being said there will definately be a need for some as 
far as single jar deployment goes.

I've forwarded this to the mailing list that is set up as well, ideally 
I would like to see this grow into a community, so it has the 
possibility of becoming an apache ant subproject, so all design should 
be done in the open as a general rule of thumb.

thanks
--Danno

Ted Toth wrote:

>Danno,
>I've looking at the source base. So what are your
>thoughts on using SWT instead of Swing? Do you know
>where to get information on new functionality in Ant
>1.6? In particluar dynamic properties? Did you have a
>plan for the tasks inst.ant will provide?
>
>Ted
>
>--- Danno Ferrin <sh...@ya...> wrote:
>
>---------------------------------
>  Thanks for offering to help, it will surely be
>appreciated.  I've gone aheadand added you to the list
>of the developers for the project.
>
>Right now my paricipation will be rather limited given
>my work load at work. I assume that you've read my
>web-log entries so you have a good idea wherethe
>design needs to go.  If I had the time myself I would
>be looking intobringinging in what exists of ant 1.6
>and trying to get the property introspectorsand late
>resolution of tasks so that we can play evil games
>with mutableproeprties.  But as you can see there are
>lots of other rough edges thatneed to be smoothed
>over, so knock yourself out.
>
>--Danno
>
>Ted Toth wrote:
>
>  
>
>     Danno,I'm interested in
>
>helping. I've used Ant a bunch forbuilds and general
>utility work and have also done a lot ofinstaller work
>(not my primary job) with InstallAnywhere.
>I'mcurrently not working and had thought about
>starting a similarproject when I came across inst.ant.
>Ted    
>    
>
>    
>__________________________________________________Do
>you Yahoo!?Yahoo! Mail Plus - Powerful. Affordable.
>Sign up now.http://mailplus.yahoo.com  
>
>
>
>
>__________________________________________________
>Do you Yahoo!?
>Yahoo! Mail Plus - Powerful. Affordable. Sign up now.
>http://mailplus.yahoo.com
>
>  
>