BTW, Jesse and I have already had this discussion off-list, but I wanted
to reflect it in the thread. :-)
I couldn't agree more, WRT new projects. Or where people have the
ability to manage their project structures.
My original example was of two source-generation processes in the same
porject, but in reality this *really* should be split into multiple
projects:
project-model
project-parser
project-*
in the same way Maven itself is split up.
HOWEVER, I think we also need to understand that there are going to be
legacy use cases out there, where people don't have control over the
project structure. If someone can't split up their project like above,
do we want to turn them away at the door? In the past, I think Maven has
tended to focus a little too tightly on new project development. IMO, we
need to be more careful to accommodate these legacy cases in the future.
To be clear, I don't want to encourage bad project design. I'm very much
against shoving everything into one build just because it's possible.
But, it should be possible, in order to accommodate those users who are
dealing with bad project design that is outside of their control.
Cheers,
John
Jesse McConnell wrote:
I really wonder about adding in more complexity into the pom with things
like ordering...
one of the attractions of maven imo is that it facilitates making the build
a simple thing, small easily digestable chunks of build process, leveraging
the dependency mechanism to weave it all together. Adding in more phases,
or ordering within a phase just makes the pom get more and more complex
which (in my mind) defeats one of the goals of maven...
sure it is technically possible to glom as much as you can into a pom, but
understandability goes down quickly and we are left with a process that
isn't a scads better then the others, like maven is right now IMO. :)
I understand people are really used to combining as much as they can into
one building entity, but it that going to be our best practice?
jesse
On 2/17/06, John Casey <[EMAIL PROTECTED]> wrote:
IMO a consolidation goal is another workaround. It's definitely possible
now, but if we had phase-ordering, we wouldn't need it, right?
-j
Eric Redmond wrote:
+0 to the pre/post phase. As it has been mentioned a million times
before,
what's the difference between the post of one phase, and the pre of the
next.
However, I am seeing a need for more than a single execution per stage.
I
like John's suggesting alot. It makes sense. Within a particular phase,
I
have a list of goals that need met. With the pre/post thing, it is
effectively saying "You can have at most three goals met per phase".
Another
option is to have some sort of consolidation goal that would then be
called
on a phase, whose definition is an ordered list of goals, I kind of like
this better, as it will keep the syntax cleaner, and honestly, how often
do
you need to cram multiple goals into a phase? One or two at most?
Eric
On 2/17/06, John Casey <[EMAIL PROTECTED]> wrote:
I understand that this is sort of a slippery slope WRT when we stop
adding new phases. While there are major categories for the phases of a
build, things like the following could occur:
I generate a model using Modello, and would like to use my own custom
Antlr grammar to create instances of that model.
Both should fit in generate-sources, but there's a definite order here.
Maybe the solution is to split the project in two, one -model, and
another -parser or something. Still, it seems like we're putting a
band-aid on the problem by adding more phases. Would it be better to
add
control over ordering within a phase? How would that even look in
syntax?
What do you all think?
-j
John Casey wrote:
Hi,
I'd like to add pre/post phases for all of the major lifecycle phases
that don't already have it. I'm starting to see cases where a
particular
packaging maps multiple mojos to the same lifecycle phase, and this
means we cannot control that phase through the old
suppress-and-augment
approach anymore. I'll give you an example:
Say I have two mojos bound to the package phase, for lack of a better
place. The first takes the tested code, and assembles the directory
structure for the archive. The second creates the archive. If I want
to
replace the first step, I can add a 'skip' flag to it, but I *cannot*
bind a new mojo in its place; any new binding will be added after the
second step. Obviously, it makes no sense to prepare an archive
directory structure *after* the archive is created.
This is not the first time we've discussed this sort of thing. We have
pre/post phases for setup and tear-down of integration tests, and
should
probably have something similar for unit tests...not to mention,
install, deploy...
It doesn't seem like a good idea to continue addressing this problem
an
issue at a time in successive Maven releases. Why not make a
reasonable
concession to these use cases, and add pre/post phases to each major
lifecycle phase (those which are themselves pre/post phases are not
what
I consider major).
I'd like to get this into 2.0.3, since it affects some work I'm doing
for a client.
What do you all think?
-john
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
--
jesse mcconnell
jesseDOTmcconnellATgmailDOTcom
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
