Re: [stellation-res] Checkout creating directory structure

["Jonathan Gossage" <jonathan@xxxxxxxxxxxxxx>]

----- Original Message -----
From: "Mark C. Chu-Carroll" <mcc@xxxxxxxxxxxxxx>
To: "Stellation-res" <stellation-res@xxxxxxxxxxxxxxx>
Sent: Tuesday, July 29, 2003 1:56 PM
Subject: Re: [stellation-res] Checkout creating directory structure


> On Tue, 2003-07-29 at 13:48, Jonathan Gossage wrote:
> > ----- Original Message -----
> > <snip>
> > >
> > > Unfortunately, for the moment, yes, this is correct behaviour.
> > > I've been meaning to write up the checkout process for the server,
> > > but I've been too bogged down with other work. Today I took
> > > a vacation day for my birthday, so aside from answering your
> > > mail, I'm not doing work.
> > >
> > > The deal with the system as it stands is that the CLI stuff assumes
> > > that a given workspace contains exactly one project; and it
> > > puts a special metadata directory named SVC in the root
> > > directory of the workspace.
> > >
> > > So, checking the stuff in, I had two choices: require
> > > users to check out into a pre-created project directory
> > > for each project; or add a second layer of directories.
> > > For Eclipse purposes, the former seemed to be the better choice
> > > for the moment.
> > >
> >
> > Do you want me to fix this? The CLI is going to be with us for a month
or
> > two and this is quite painful as it stands.
>
> It's not that painful; once you've created the directory and
> checked out once, you just work in that directory as normal.
> It's far from ideal, but it is not unbearable.
>
> I'd love to see it fixed; but the CLI code is *so* tangled
> and incomprehensible, and does so many strange things
> with the metadata directory in so many places that
> I rather suspect that fixing it is going to be a very
> difficult and error prone process. Changing smaller
> things than that have lead to nightmares.
>
> If you still want to go ahead with it, I'd suggest posting
> a proposal for how you want to fix it to the wiki, and
> we'll see whether it looks doable without wasting an enourmous
> amount of effort.
>
I had a good look at this last night and came to the same conclusion. There
are just too many places where the single project assumption is built in.

Regards

Jonathan Gossage