Prev: Statistician II (#168)
Next: Does Ruby have any advantage over Python to create semantic applications?
From: phlip on 4 Jul 2008 14:45
S2 wrote:

> Yes. I asked to get some devs and PMs together and get an analysis done on
> the migration, what we lose and what we would gain by dropping RoR for
> Java, but "for now we don't have the time. New development should be done
> in Java".

At my day job we use pure XP with RoR. We use promiscuous pair programming,
meaning we switch pairs every 2 hours. We use pure Test Driven Development, with
refactoring & continuous integration. We have a very high velocity and a
vanishing bug rate.

Since I got there, our team has increased to 6, and we will get 2 more as soon
as we remodel. I really doubt we could talk our executives into switching back
to Java, or switching from XP to something else... like "Agile".

Your worksite has issues that prevent your executives from seeing which systems
are working better than others. There's an old adage, "If I had 6 hours to cut
down a tree, I would spend the first 3 hours sharpening my ax".

Your bosses think the more strokes required to cut down a tree, the more
"progress" you made. That's what a 600 line configuration file is - 600 more
strokes with a dull ax.

>> So, no pair programming, and they want to make knowledge transfer easier
>> by reducing the amount of knowledge required. Gotcha!
>
> Exactly. I'm not sure I understand your point?

Exactly... you are not sure? (-;

If you are not pair programming, then knowledge is indeed hard to transfer at
team swap time. So your execs think they can reduce the total amount of
knowledge required. Yet more strokes to cut down that tree!

--
Phlip
From: Robert Dober on 4 Jul 2008 15:03
On Fri, Jul 4, 2008 at 8:21 PM, S2 <x(a)y.z> wrote:
> Robert Dober wrote:
>
>> Of course there was, you stated it yourself above, "developers were
>> quite happy", sorry for my pessimistic look at this.
>
> I don't think they are *that* evil :)
>
>> Apart of the idea that jruby might come in handy this decision really
>> looks odd to me.
>
> No. jruby will not be an option. If that was the case I would not look for
> some other job. I think we/they are going to use the Seam Framework, but
> it's still not sure. They asked me to join the team who gets to evaluate
> the various available frameworks, but every time I look at a hibernate
> config file I get goose-flesh.
>
>> Maybe somebody wants to sink the ship?
>
> I don't think so. I think it's only ignorance.
>
>> Well at least, if they do, they
>> seem to be good at it sigh.
>
> Have some arguments I could tell management? You never know...

Sure, if you trust that they are honest, first thing to do is to ask
for their rationale.
Maybe, after all, it makes sense.
If they cannot share their agenda with you, I was probably a little
bit right, but if they can and do you might be able to
trigger a discussion.

If e.g. you were guessing correctly, and I guess knowing the ship you
are sailing on there are some chances that you have, that they are
trying to get their development sharing spectrum smaller, you might be
able to give them different options (jruby might come into play
again). You might want to figure out some market trends, development
times etc.
In an ideal world you might get a chance to present Ruby's strengths :).

But in order to reason with somebody at first you have to trigger a
dialog, do you think you are in a position to do that?

Well good luck from the bottom of my heart.

Cheers
Robert

From: James Britt on 4 Jul 2008 15:08
S2 wrote:

> No. jruby will not be an option. If that was the case I would not look for
> some other job. I think we/they are going to use the Seam Framework, but
> it's still not sure. They asked me to join the team who gets to evaluate
> the various available frameworks, but every time I look at a hibernate
> config file I get goose-flesh.

You need to be able to explain to them just why you get "goose-flesh".

No doubt everyone on this list is automatically sympathetic to your
plight, but truth is that much of what we see as obvious about Ruby is
not so for others.

I don't see how anyone can expect people to just be instantly
enlightened about the value of using Ruby; folks need facts.*

If Ruby (with Rails or whatever) is the objectively better choice then
you should be able to demonstrate that. There has to be something
tangible, something other than "I like it."

Get some stats, some code, some use cases, and build a solid case. The
best argument tends to center on money. Something that argues for
higher employee retention rate, or lowered training costs. If you can
show that using Ruby means faster turnaround with fewer developers and
lower maintenance, that in the end Ruby == more profit than Java or
Net, you should be OK.*

If you can't make that case, then why would you expect a business to
choose Ruby?






* Or not. Some folks will have ulterior motives for choosing one tool
over another, but you should start be giving people the benefit of the
doubt.



--
James Britt

www.happycamperstudios.com - Wicked Cool Coding
www.jamesbritt.com - Playing with Better Toys
www.ruby-doc.org - Ruby Help & Documentation
www.rubystuff.com - The Ruby Store for Ruby Stuff

From: Jim Morris on 4 Jul 2008 15:45
S2 wrote:
> My company today decided to ditch ruby development and to develop new web
> applications only with Java or .net.
> Current RoR applications will be migrated to Java.
> Sigh. No more ruby for me (in office hours).
> But anyway: are here some Java devs who may suggest some Java frameworks
> similar to RoR? Maybe something that emulates Activerecord and is as
> flexible, powerful and convenient to use as the ActionPack?
>
>

Rather than berate your management for the decision I'll try to answer your actual question ;)

I switch between RoR and Java regularly for different clients.

For Java I use the Spring framework almost exclusively. I do not use Hibernate as I prefer to use
springs DAO framework, and the various jdbc wrappers they provide, I like to get my hands dirty with
SQL, and you can make things very efficient that way too.

Spring has a fairly steep learning curve but is well documented and supported and has many
components, from JMS, WEBMVC, JDBC etc. you can pretty much do everything with the Spring framework.

I use a standard architecture where you have web servers, talking to middleware servers via JMS,
this separates all database access out of the web layer. Then the middleware layer is composed of a
JMS layer calling POJO's and the POJO's call the DAO layer which talks to the database, so
everything is cleanly separated. All database access is in the DAO layer, all Business logic is in
the POJOs. This makes testing pretty easy, I write JUnit tests (using hamcrest matchers) for the
DAOs, and for the POJOS, so I can test database independent of business logic and vice versa.

It is not as much fun as RoR, but it sure is much easier to scale for huge loads.

I use Eclipse as my IDE for Java development.

I would guess it takes me twice as long to write a given application in Java as it would in RoR, not
including the scaling issues.

I write all my support scripts and database management code in Ruby and use the Sequel GEM, so I
still get my ruby fix even when on a Java project.

Good luck


--
Jim Morris, http://blog.wolfman.com

From: Bill Kelly on 4 Jul 2008 17:55

From: "M. Edward (Ed) Borasky" <znmeb(a)cesmail.net>
>
>> * Or not. Some folks will have ulterior motives for choosing one tool
>> over another, but you should start be giving people the benefit of the
>> doubt.
>
> Or better yet, just say, "They may be right," and proceed with the
> accepted environments and projects.

Wait... have you ever actually endured much Java programming? <grin>



HaHaOnlySerious,

Bill



First  |  Prev  |  Next  |  Last
Pages: 1 2 3 4
Prev: Statistician II (#168)
Next: Does Ruby have any advantage over Python to create semantic applications?