[Mono-dev] Bounty for ClickOnce with Mono

Sebastien Pouliot sebastien.pouliot at gmail.com
Mon May 14 10:42:59 EDT 2007


On Mon, 2007-05-14 at 16:02 +0200, "Andrés G. Aragoneses [ knocte ]"
> I would probably be interested in offering a bounty for someone to hack 
> into Mono to allow ClickOnce deployments work under Linux.
> I suppose that I would not require CAS to be completed because the 
> assembly would run on FullTrust mode on the client machine, 

Maybe not, maybe some. It depends on how you want to setup your policies
to decide if code is Fulltrust or Untrusted. 

If you want to use the CAS policies/evidences (e.g. StrongName) then
some work is still needed (mostly testing and probably fixing existing

If you use an OoB mechanism then CAS isn't required (to offer Fulltrust
to the deployed code) but you have to define/implement your own.

> so the 
> project would be divided into two deliverables:
> a) Allowing to deploy a ClickOnce solution via Apache+Mod_mono, server side.
> b) Hacking into the ClickOnce Firefox extension so as to make it 
> cross-platform (in Linux, use Mono instead of .NET).

Not sure how's the FF extension is done on Windows but Mono's embedding
interface is very different from MS. This may ends up into two very
different extensions.

> Then, my questions are:
> 1) How would you rate this effort economically? Opinions?
> 2) How much time would you estimate for it to be developed?

IMO a large part of it depends on how you will decide between Fulltrust
and Untrusted - and this affects how much compatible this will be with
MS implementation.

> 3) Would the first target (a) allow deploying Gtk# applications or 
> otherwise I should create one more deliverable item for this?
> 4) Which is the best mean to publish this bounty?

apart from this mailing-list, Monologue

> 5) Would any of the mailing-list members be interested in this bounty?
> 6) I would probably add to the bounty the requisite of the code being 
> submitted upstream to the Mono repository, so the person in charge of 
> the project would depend on the reviews of the Mono Team. Therefore, I 
> would like to know the opinion of the parties that would be in charge of 
> reviewing this work:
> - Would you review the work in a reasonable time frame or otherwise 
> you're too busy for this task?

It mostly depends on how many area are affected, as it affects the
number of people required to review this. If the code is reviewed
progressively (i.e. not a single code drop at the end), includes unit
tests where appropriate (and other tests with instructions) then I would
assume the review process can be done within a reasonable time frame.

However it's hard to predict what people will be doing in a few months,
still harder when the review dates are unknown ;-)

> - Would it be adequate to give a percentage of the bounty to the 
> reviewer? If yes, which percentage do you recommend?

Sebastien Pouliot  <sebastien at ximian.com>
Blog: http://pages.infinit.net/ctech/

More information about the Mono-devel-list mailing list