Boost logo

Boost :

From: David Abrahams (david.abrahams_at_[hidden])
Date: 2002-01-29 15:10:35


I note that the instructions you gave for WinCVS are a lot more like my
narrative. However, there's no substitute for feedback from a user. If you
really found the other instructions clearer, I guess we should go with them.

-Dave

----- Original Message -----
From: "Beman Dawes" <bdawes_at_[hidden]>
To: <boost_at_[hidden]>; <boost_at_[hidden]>
Sent: Tuesday, January 29, 2002 2:44 PM
Subject: Re: [boost] Release procedures?

> At 02:52 PM 1/28/2002, David Abrahams wrote:
>
> >The command-line instructions are essentially the same as the detailed
> ones
> >I posted in response to your request,
>
> I don't want to hurt your feelings, but I found the other set of
> instructions easier to follow. Perhaps that's just because I had already
> done a number of experiments by the time I read them.
>
> > except that:
> >
> >1. Mine didn't deal with the multiple merge issue (easy modification)
>
> But IMO this is the trickiest issue, so needs to be treated carefully.
>
> >2. These gloss over the "make sure it builds step", which in general
> isn't
> >possible without either checking in /everything/ you've got and tagging
> it
> >so you can get it back, then checking out everything on the release
> >candidate branch, or using a separate copy of the CVS tree like I
> >suggested.
>
> The problem is that the most appropriate approach is very situational.
For
> example, the "separate copy of CVS tree" work well for developers with a
> high speed Internet connection and a simple change that doesn't require
> more that a few tests to validate the change. If either of those isn't
> true, it may not be the preferred approach. Thus we really need a better
> indication that there are several possible approaches.
>
> >3. These are a lot less explicit.
>
> I guess that's a matter of taste. I found the intermixed prose and cvs
> commands to be harder to follow than just the annotated cvs commands.
>
> But if you want to try to improve it, feel free. The important point is
> that we now know how (in terms of branch candidates, and so forth) we want
> the release procedure to work, so we can go ahead with a release.
>
> >It seems to me that it might be better to integrate the multiple merge
> >instructions into what I already wrote.
>
> I considered that, and decided that it wouldn't be useful to try to hide
> the cvs distinction between a first merge and subsequent merges. But if
> you think you can improve the wording, please go ahead and change it.
>
> --Beman
>
>
> Info: http://www.boost.org Send unsubscribe requests to:
<mailto:boost-unsubscribe_at_[hidden]>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>


Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk