Allow ordering of WebActivator calls across multiple assemblies

[neilduncan]

Hi David,

I have some web activator calls which live in a number of libraries. At the moment, Web Activator ordering only orders the calls inside a single assembly, not across multiple.

I think what I've done here should solve this. Do you want to take a look?

Thanks,
Neil

[davidebbo]

I'm out and won't be able to look at this in depth till August.

Though I do find it a little odd to order across assemblies in the sense that there is no shared understanding of the Ordering even means when assemblies are authored by different people.

[neilduncan]

Yeah, I can understand that point of view. It's not something that I'd
expect many people to run up against. I've certainly been using
WebActivator for a while without needing it!

Perhaps some more about the use case might be helpful?

We've got an internal product that we maintain, and which gets dropped into
various websites as a Nuget package. The product initialises itself using
WebActivator. We now need the ability to customise that initialisation per
website, which is why I'm concerned about the ordering across assemblies.

What about looking at it the other way? Could you see a downside to
ordering across assemblies?

On Friday, July 12, 2013, David Ebbo wrote:

I'm out and won't be able to look at this in depth till August.

Though I do find it a little odd to order across assemblies in the sense
that there is no shared understanding of the Ordering even means when
assemblies are authored by different people.

—
Reply to this email directly or view it on GitHubhttps://github.com//pull/13#issuecomment-20905679
.

[davidebbo]

See related discussion in #10. We should try to get your change in, though I probably won't be able yo get to it for another week. To answer your question, I don't really see a downside, other than potentially breaking scenarios that were making questionable assumption about the order they currently get.

Hopefully, in scenarios where people don't use Order attributes, your change will have no effect on the effective ordering.

[neilduncan]

Great. The commit looks a bit big, but most of the changes are just adding
a new assembly for testing.
The actual change to the code is tiny.

On Wed, Jul 31, 2013 at 9:17 AM, David Ebbo notifications@github.comwrote:

See related discussion in #10#10.
We should try to get your change in, though I probably won't be able yo get
to it for another week. To answer your question, I don't really see a
downside, other than potentially breaking scenarios that were making
questionable assumption about the order they currently get.

Hopefully, in scenarios where people don't use Order attributes, your
change will have no effect on the effective ordering.

—
Reply to this email directly or view it on GitHubhttps://github.com//pull/13#issuecomment-21846879
.

[davidebbo]

Yes, the change looks safe. Would you mind checking that the sorting algorithm is stable? That would guarantee that the ordering will not be affected in the (common) case where no Order is specified.

[davidebbo]

Ok, I see @jskeet says here that it's stable, so that's good enough for me!

[davidebbo]

Could you yank the ncrunch file and resubmit (maybe add to gitignore)? Please reset so the file is not in the history line. Thanks!

[neilduncan]

Argh. Yes, will do.

On Tuesday, 6 August 2013, David Ebbo wrote:

Could you yank the ncrunch file and resubmit (maybe add to gitignore)?
Please reset so the file is not in the history line. Thanks!

—
Reply to this email directly or view it on GitHubhttps://github.com//pull/13#issuecomment-22203537
.

[davidebbo]

Actually, I just did it myself. The change is in! Please try WebActivatorEx 2.0.3 to make sure everything looks good. Thanks for your contribution!