[V3dot10] Re: Monticello and 3.10
Edgar J. De Cleene
edgardec2001 at yahoo.com.ar
Mon Aug 13 10:50:11 UTC 2007
El 8/12/07 10:50 PM, "Keith Hodges" <keith_hodges at yahoo.co.uk> escribió:
> Ok, so I thought I would just check what I thought was the case. In 3.10
> I open Monticello browser in the MagmaTester repository and... it doesnt
> even see the mcm files.
> The monticello configuration's browser is for writing configurations,
> but I see no ui for loading one in from a repository. Am I missing
> something obvious.
> Monticello1.5 can see mcm's if MonticelloConfigurations-kph.44 is loaded
> and can load them via the Load button.
> The Monticello15 bootstrap is simply the same code via a file in. This
> is the only reliable way of installing one version of MC on top of
> another unless effort has been made such that the version being loaded
> in doesnt break the loading process as run by the version being
> replaced. However it is impossible to do this for the 3.10 version since
> it has particular differences.
> My personal thought is that whatever testing is needed on 1.5 needs to
> be done, and it should be included in the base 3.10 from the beginning.
> Anything less is a recipe for confusion. I think it has to be this way
> because my 1.5 is one of only two viable ways forward for the
> Monticello1 genre. The other being for someone else to repeat all the
> work that I have already done using the 3.10 as a base. Incidentally
> That is what I did, I started with the version in 3.10, but found that
> Ralphs 'atomic loading' additions to md-308 were superfluous.
> MC15 Sales pitch;
> Imagine this, you are using someone-elses package say, seaside, and you
> change a method in seaside because it doesnt quite do what you want it
> to do. Then a few days later a new update for seaside is released so you
> load it. What happens to your version of that method.
> A. Its overwritten by the updated package's (seaside's) version of that
> method, so when you commit your package to its repo it looses a method.
> B. Your version stays, the updated packages' version is lost if you
> happen to commit an update back to the seaside repo.
> C. It loads the seasides's version, then it loads your version again (so
> the seaside version is preserved in the version history as the one
> before yours) Saving the seaside package spots the override and saves
> the previous seaside version of the method. Saving you package saves
> your override. The integrity of both packages is preserved.
> When this works I think it is really worth having
I very sorry you try to fit English unit measure pieces into Metric system
holes (or viceversa).
If Ralph said I must put your Monticello into base, I put, otherwise must
wait Board nominate you for next release build.
More information about the V3dot10