Compelling reasons for IDEA

Although I am an avid fan of IDEA, I am truly struggling to explain why someone should prefer to use IDEA over a free competitor, such as Eclipse. Of course, I can see the justification for myself, but my explanation so far is not much more than "because it's just better".

Of course I could rant about the same things that many rant about on blogs, etc., but this all seems to be subjective material - far from compelling.

Can anyone provide some substantial arguments that would appear compelling to an observer who is not swayed either way?

70 comments

My opinion:

Number one: productivity.

Else:
- ease of use
- better configuration capabilities
- easy project setup
- less clutter
- better Java editor
- better cvs integration
- nice plugin manager
- responsiveness of the JetBrains support and development team
- frequent bugfix and maintenance releases
- code inspections
- code analysis capabilities
- better Ant integration

Regards

Thomas Gülden
Munich, Germany

0


First argument:

The code analysis and inspection capabilities of IDEA are so much stronger that an average developer will save themselves from at least a couple of bugs per week. The cost of fixing a bug can vary from a few tens of dollars (if found in unit test), to tens of thousands of dollars (if found in production). It doesn't take long at those prices to justify IDEA's really quite cheap license cost. I once did a not-entirely-subjective calculation that showed my IDEA license pays for itself in bugs avoided every six-to-eight work days.

Second argument:

The advanced refactoring and intention capabilities of IDEA save developer time, but they also result in lower costs, as bugs that get created during manual refactoring with other IDEs won't be created with IDEA's automation. Moreover, IDEA's refactorings are so usable and compelling that refactoring is more likely to be executed, resulting in stronger and higher quality code.

Third argument:

In an average month, I will save at least an hour's work just because of IDEA's local history functionality, probably more like two. At those rates, IDEA pays for itself in less than three months just from rework avoided.

Extra added argument:

Hector RULEZZZZZ!!!!!

--Dave Griffith



0

I don't think IDEA has the best cvs intergration than eclipse.

My opinion:

Number one: productivity.

Else:
- ease of use
- better configuration capabilities
- easy project setup
- less clutter
- better Java editor
- better cvs integration
- nice plugin manager
- responsiveness of the JetBrains support and development team
- frequent bugfix and maintenance releases
- code inspections
- code analysis capabilities
- better Ant integration
Regards

Thomas Gülden
Munich, Germany



0

Many arguments are certainly subjective and simply a matter of taste, habit, and personal preferences. However, of all the advantages that - in my opinion - IDEA has over Eclipse I would single out the following. I have found that, unlike Eclipse, IDEA does unintrusively and polietly TEACH you to write better code - as you go - day after day. Every step of the way, it identifies and points out flaws, inconsistencies, or undesirable patterns, and - most importantly - immediately suggests the correct alternative, which is usually one click away! While Eclipse provides some of such functionality, I find that it doesn't even come close to the range and convenience of IDEA. Eclipse often simply underlines the problem without providing a one-click option to correct it. That's why Eclipse users - I see it on a daily basis - tend to ignore those warnings. Practically each time I open in IntelliJ someone else's code that has been edited in Eclipse, the code is sprinkled with yellow warning highlights and red Javadoc errors. I am not talking just about unused variables and imports (that Eclipse does identify, but for some reason Eclipse developers seem to not care about.) IDEA instantly identifies the inefficient/bad coding practices that are not highlited in my colleague's Eclipse editor. Eclipse supporters will argue that Eclipse can be set up to do the same things. However the truth is, the code that I see from Eclipse users (including even very experienced programmers) almost never appears clean (green) in IntelliJ. Amazing, but true. That's a very consistent fact that I observe from project to project, client to client. I suspect that one of the reasons is that it is always a bit more extra work to tidy things up in Eclipse - while in IntelliJ it always comes naturally. IDEA constantly encourages you to write better, cleaner and more elegant code. So, I'd say, this IDE not only increases productivity and is pleasure to use, it makes folks better and more disciplined developers - that is something that is hard to overestimate.

That said, I don't think that IDEA is perfect, it is getting more and more complex, and from time to time I run into issues that do frustrate me. Still, I wouldn't change it for any other IDE (unless I am using AspectJ, which IDEA doesn't have a plugin for.... grrrrr... :) )

0

I have to agree. It seem JetBrains has hired the right usability person.
I like IDEAs ease of use.

Many arguments are certainly subjective and simply a matter of taste,
habit
, and personal preferences. However, of all the advantages that - in
my opi
nion - IDEA has over Eclipse I would single out the following. I have
found
that, unlike Eclipse, IDEA does unintrusively and polietly TEACH you
to wr
ite better code - as you go - day after day. Every step of the way, it
iden
tifies and points out flaws, inconsistencies, or undesirable patterns,
and
- most importantly - immediately suggests the correct alternative,
which is
usually one click away! While Eclipse provides some of such
functionality,
I find that it doesn't even come close to the range and convenience
of IDE
A. Eclipse often simply underlines the problem without providing a
one-clic
k option to correct it. That's why Eclipse users - I see it on a daily
basi
s - tend to ignore those warnings. Practically each time I open in
Intelli
J someone else's code that has been edited in Eclipse, the code is
sprinkle
d with yellow warning highlights and red Javadoc errors. I am not
talking j
ust about unused variables and imports (that Eclipse does identify,
but for
some reason Eclipse developers seem to not care about.) IDEA
instantly ide
ntifies the inefficient/bad coding practices that are not highlited in
my c
olleague's Eclipse editor. Eclipse supporters will argue that Eclipse
can
be set up to do the same things. However the truth is, the code that I
see
from Eclipse users (including even very experienced programmers)
almost nev
er appears clean (green) in IntelliJ. Amazing, but true. That's a very
cons
istent fact that I observe from project to project, client to client.
I sus
pect that one of the reasons is that it is always a bit more extra
work to
tidy things up in Eclipse - while in IntelliJ it always comes
naturally. ID
EA constantly encourages you to write better, cleaner and more elegant
code
So, I'd say, this IDE not only increases productivity and is pleasure
to
use, it makes folks better and more disciplined developers - that is
somet
hing that is hard to overestimate.

That said, I don't think that IDEA is perfect, it is getting more and
more
complex, and from time to time I run into issues that do frustrate me.
Stil
l, I wouldn't change it for any other IDE (unless I am using AspectJ,
which
IDEA doesn't have a plugin for.... grrrrr... :) )



0

What U guys say here is very true.... IDEA just feel like a developer's tools. It work and feel much like part of bloodstream for developer.... This feeling used to be the case for JBuilder in earlier version, but now JBuilder seem like JEnder ...Why ?

I have used Together for ecilipse 2006 trial version and u guys know what ?? This thing crawl like tortise ... everything appear slow and unresponsive (and u talking about running in 2.2ghz Pentium4 and 1gb ram and the UI is badly designed )(earlier version is much usable and work out of the box)... Now u talking about Jbuilder to be written for ecilipse... It seem that the intuitiveness of these program is hampered by the complexity and "IBM's professional image" thing... (If u know what i mean in earlier day when image is more important than functionality) ....

Whenever, i use ecillipse, it just remind me of bureacractic... a program that try to do too much which end up doing nothing good and complex.... ppl endorse it because it is the standard and most importantly it is free...what could u ask for ?? But u pay in productivity .... Eclipse is good as environment for integration as a concept but bad in implementation and real-world ....

For example, whenever debug, why go into debug workspace especially it is slow operation ?? Why can't it just debug and work in same window ?? That's make it easier to use ....
yeah, yeah, yeah....ecillipse as a platform needed to handle this for scalability and for future. ...and so developer say.....

but who really care ??? The developer are paid to produce code now and for result, and not for the future ... In fact, thing only get worse not better, since more feature means more complexity and memory performance issue ....

I fear that later Borland will later go down in history like wordperfect because they are mislead by platform ... like os/2 versus windows ... and yes, ppl say that u cannot go wrong with BigBrother IBM, but then os/2 is bigbrother's pet and that pet has now rest in peace ....

At first, i will agree that intellij will be wise to go ecillipse way like jbuilder , but after using ecillipse and together for ecillipse, it seem like a dead way because it just not there (Watch America Idol and u can just feel who the loser and winner) and the direction seem wrong because design seem a bit counterproductive for development ....

and even if jbuilder get it right, the performance of ecillipse will kill it ...

I rather have a developer's tools maintain its own personality rather than one size fit it all mentality .... afterall different application really deserve to be innovate along the line of productivity design not the common lowest denominator for everything ....

Ecillipse has once again show that bigboy rules but once hype passed, Intellij rule because it really built for developer by developer and not some kind of corporate professional businessman .... at the end of day, efficiency and effectiveness is all matter .........

Perhaps if IBM want to be developer's heaven, for god's sake, buy Intellij and utilize its talent to build the IDE ... and not some guys who still live in the corporate image and dream...

thank

0

Perhaps if IBM want to be developer's heaven, for god's sake, buy Intellij and utilize its talent to build the IDE

There are undoubtedly great stories about why IntelliJ hasn't been purchased by IBM, Sun, Microsoft, Google, BEA, or Oracle. That's the first rank of reasonable suitors, and there are undoubtedly others slightly less likely (Computer Associates? Borland? JBoss?). Unfortunately, I'm pretty sure no one on this forum without a JetBrains bug after their names knows those stories, and they don't seem to be talking. Probably just that the peak of IDEA's manifest technological coolness happened to coincide with the bottom of the "let's purchase technology companies" cycle, back in 2002.

--Dave Griffith

0

Jetbrain probably have the burning passion of developing Intellij for the developers and not to sellout to huge company for money.... In fact, if history has it, any innovative product that being acquired die a slow death, since that product no longer have the visionary founder behind it, it lost sight of what it is ..... the entrepreneurial founder or orginal developer probably give up because having to work bureacratic big company that seem to overdevelop it into somekind of superApplication that do everything but done nothing at all useful..... I see it in Silverstream, together (Hey, isn't the founders of Jetbrain from Together originally)...

I think those people in Jetbrain has really the smart brain to realize that their "baby" if left in hand of corporate professional "Big daddy company" will die a political death .... and so we left with ecillipse that return us to few generation behind visual studio .... In fact, i think Intellij is the benchmark to which all IDE compete including studio....

Some of the features of intellij are so innovative and easy that sometime I wonder why those genius in borland and other company doesn't even think of it, let alone implement it....
Then I realized that it is not that these genius can't do it , but rather they are mislead by corporate and arrogant executive who thought they know it all to implement enterprise features instead .... those which we never used day to day but needed to coerce sale and make impressive product marketing ..... and so they lose touch of what customer want ........ Now borland is selling off their IDE business, because they thought the survival is in some kind of lifecycle support.... I believe that there still lives in IDE business, and it is the sign that borland businessman would rather enter into area where bigBoys play...... but beware, the borland name seem to diminish over the years, and i hope it doesn't drive to been called "borWater" and it is drive out of competition in land to exile to survive in water....

I see company been successful when they switch industry like Nokia from rubber-making to phone manufacturer but nowhere in software industry, has any company has successfuly make that kind of transition which undermine their core competency ......... If borland make it , it will induce into genius world of record ..... start praying for Borland , the IDE icon .

0

the peak of IDEA's manifest technological coolness .. back in 2002.

>


The simplest/most obvious/only way for Jetbrains to be cool again is to
add Rails (and Ruby) to their languages portfolio.

History could easily repeat itself for JetBrains, and it could even work :
- start with a plugin (not for JBuilder this time, but for IDEA).
and later
- develop a full blown IDE for Ruby and Rails.

They'd have to review the pricing though: 500$ for an IDE won't make it
with the Ruby crowd (TextMate - an advanced editor - costs 39? and is
the de facto "official" Rails development tool on MacOS).

Alain

0


+The simplest/most obvious/only way for Jetbrains to be cool again is to
add Rails (and Ruby) to their languages portfolio.+

The only problem with that idea is that, in spite of all the buzz, the penetration numbers for Ruby just aren't particularly impressive: http://www.tiobe.com/index.htm?tiobe_index . Finishing Maxim's Python plugin and putting it in product would have many times the impact. If JetBrains can execute the Demetra roadmap by summer, I'm willing to score them with all the manifest techological coolness points that anyone could want.

--Dave Griffith


0

Hello Dave,

DG> The only problem with that idea is that, in spite of all the buzz,
DG> the penetration numbers for Ruby just aren't particularly
DG> impressive: http://www.tiobe.com/index.htm?tiobe_index . Finishing
DG> Maxim's Python plugin and putting it in product would have many
DG> times the impact. If JetBrains can execute the Demetra roadmap by
DG> summer, I'm willing to score them with all the manifest techological
DG> coolness points that anyone could want.

It's not Maxim's, it's mine. :) And I'm not aware of any plans to put it
in the product.

You're welcome to contribute to it, by the way. :)

--
Dmitry Jemerov
Software Developer
JetBrains, Inc.
http://www.jetbrains.com
"Develop with pleasure!"


0

> They'd have to review the pricing though: 500$ for an
> IDE won't make it with the Ruby crowd

So are you saying that the Ruby crowd won't pay for a professional IDE (not being funny but TextEdit is not even close to that)? If that is the case then either:

1> Ruby doesn't have enough importance for companies to buy it for their developers
2> Ruby developers would not appreciate (or pay for) the benefits of a real IDE

If either one of those things is true then it's hardly going to be profitable for JetBrains to develop. JetBrains are a company first and foremost and they must drive a profit, if their business people see a market they might do it, but comments like yours don't exactly make the Ruby market sound like a lucrative proposition...

0

Dave,


> in spite of all the buzz,

(Thanks for saying "buzz", and not "hype".
see:
http://blog.codahale.com/2006/02/17/six-ground-rules-for-rails-sucks-articles/
)


> the penetration numbers for Ruby just aren't particularly impressive:
> http://www.tiobe.com/index.htm?tiobe_index .

Do you think MPS was a safer bet?
Other indices tell another story:
- Book sales:
"python is up 20% vs. the same period last year, but Ruby is up 1552%!
(Perl is down 3%.) "
http://radar.oreilly.com/archives/2005/12/ruby_book_sales_surpass_python.html
- Amazon sales rank (#922),
- Rails newsgroups traffic (300+msg/day),
- Ruby and Rails downloads exploaded,
- web apps: http://www.emilychang.com/go/ehub/ (here you can say "Hype
Hype")
- hundreds or Registered developers :
http://wiki.rubyonrails.com/rails/pages/WorkingProfessionallyWithRails
- dozens of registered development firms :
http://wiki.rubyonrails.com/rails/pages/DevelopmentFirms
- most web hosters now support Rails :
http://wiki.rubyonrails.com/rails/pages/RailsWebHosts
etc, etc..

I'm not trying to sell Rails to java developers, because I know they'll
come to it on their own. After all, who doesn't want to get rid of the
XML spagettis, frameworks salads and inflating APIs?
The simplest web project in java has you jump through half a dozen
misaligned hoops : servlets, Sping, Struts, Hibernate, etc...
With Rails, there is ONE hoop: Rails. If you want to learn it, there is
ONE book. Read it and you're an expert. Almost. Read teh second
book/learn Ruby, and you're a master. Almost.
There is one picture that sums it up nicely and perfectly:
http://www.garbett.org/upload/books.jpg


No, I'm trying to sell the idea to JetBrains. You could be the first to
bring refactoring to Rails and Ruby (and css, javascript, yml, html,
sql, ..). I understand you want to be careful after the Fabrique fiasco,
but the entry fee for rails is low, IF you can do it with a plugin.

For info, the hot tool in the Rails world is TextMate, a 39? MacOS
advanced editor. TextMate ows a big part of its success to its unanimous
adoption by the Rails core developers team. It's written by 1 man, and
runs on 1 platform - MacOS -, and he's already sold ten(s? of) thousands
of it. He's rich because we're happy, and people (I know 1 at least) are
buying Macs because of TextMate (and Locomotive).
The tool is open => the users are doing the "hard" work of extending it:
supporting new languages, adding features, etc..
Just this week-end, basic navigation and code extraction was added this
way, in a few lines of codes.

For example, here is the command - macro- that navigates between a view
and its associated controler :
(Note that the command is written in Ruby !)
if current_file.controller? if result = current_file.find_method_backwards view_file = File.join(current_file.rails_root, 'app', 'views', current_file.modules.join('/'), current_file.controller_name, result[0] + ".rhtml") TextMate.open view_file else TextMate.message("Move the caret inside the action whose view it is you would like to switch to.", :text => "No Action Found") end elsif current_file.view_action? ... end ]]>


Note: TextMate is just an editor, and it can't be extended very much
that way.
That's where IDEA could shine and where JetBrains could make a difference.

Alain


0

- start with a plugin (not for JBuilder this time,
but for IDEA).


I'm all for an intelligent Ruby (on Rails) IDE, and I would sure cash out to buy one produced by JetBrains, but starting as an IDEA plugin isn't the brightest IDEA.

Such a plugin would only target starting ruby developers migrating from java, who used IDEA for their java work. Now, follow me:

  • The total number of ruby developers aren't that impressive (compared to mainstream languages)

  • The number of ruby developers migrating from java is even less impressive

  • The number of ruby developers migrating from java who used IDEA should be... tiny, at best.


I think the best plan would be starting with a standalone product, providing some features that are the hallmark of JetBrains products, and evolve from there.

0

Rob

> So are you saying that the Ruby crowd won't pay for a professional IDE

"professional" and "enterprise" are no justification for an inflated price.
See the competition:
- TextMate : 39?
(used by all the REAL Rails Professional)
- Eclipse/RadRails : free.
- Komodo: 29$/295$.

> 1> Ruby doesn't have enough importance for companies to buy it for
their developers
It's not Ruby, it's Rails.
Rails is not everything for everybody, but in the web-app worlds, it's
the horse on which to bet.

> 2> Ruby developers would not appreciate (or pay for) the benefits of
a real IDE
Thanks to Ruby, writing a web-app in Rails is pain-free, so a real-IDE
is not a MUST for writing good code.
Though, as Rails success will grow, bigger applications will become more
common, and that's where a refactoring tool like IDEA would be handy,
and could make a difference. That where JB can play a role.



> If either one of those things is true then it's hardly going to be
profitable for
> JetBrains to develop. JetBrains are a company first and foremost and
they must drive a profit,

If they don't loose money, it's already a good move: JetBrains will
reinforce it presence in the languages' tool business.
They already cover css, javascript, html, and python through a plugin.
What would it take to stretch it and cover Rails. I don't know, but it's
worth asking.

If IDEA can be extended to work with Rails and Ruby, they could
repackage it as a non-Java IDEA, at a fraction of the price.
That could be profitable. Only JB people can answer that question.


> if their business people see a market they might do it,
like Fabrique? like MPS?
> but comments like yours don't exactly make the Ruby market sound like
a lucrative proposition...

If you're a professional developer, you're probably already sold to the
idea :
"learn 1 new language every year".

If you're in the web-apps business:
1/ buy the book,
2/ spend the week-end on it
3/ do 1 real app in Rails.
I bet you'll be sold to it and won't look back;

Don't mistake people sharing their pain-relief-related enthusiasm with hype.
Rails works, and that's it. If JB can be a player in that game, we could
all benefit.

Alain

0

Hello Alain,

>> if their business people see a market they might do it,
>>
AR> like Fabrique? like MPS?

Just wanted to clarify one important thing: MPS is a research project. The
first product based on MPS technology will not be MPS itself.

--
Dmitry Jemerov
Software Developer
JetBrains, Inc.
http://www.jetbrains.com
"Develop with pleasure!"


0

- Book sales:
"python is up 20% vs. the same period last year, but Ruby is up 1552%!

Ouch.
So 123,653 more Python books have been sold and 789 more copies of Ruby on Rails.

0

Marcus

> starting as an IDEA plugin isn't the brightest IDEA.
> I think the best plan would be starting with a standalone product,
> providing some features that are the hallmark of JetBrains
products, and evolve from there.


If it's a plugin, it can also be repackaged as a full-blown IDE by
combining it to a trimmed-down version of IDEA.
As a plugin, it could be free, or priced like the companion product.
As an IDE, it could be priced to better match the competition and the
Ruby world.

Alain

0

Dmitry

AR> like Fabrique? like MPS?

>

Just wanted to clarify one important thing: MPS is a research project.
The first product based on MPS technology will not be MPS itself.



I was mistaken by MPS presence and presentation in the products page:
http://www.jetbrains.com/products.html

Alain

0


>> - Book sales:
>> "python is up 20% vs. the same period last year, but Ruby is up 1552%!

Ouch.
So 123,653 more Python books have been sold and 789 more copies of
Ruby on Rails.



That's a brain fart, is it?

A.

0

Alain Ravet wrote:

>>> - Book sales:
>>> "python is up 20% vs. the same period last year, but Ruby is up 1552%!
>> Ouch.
>> So 123,653 more Python books have been sold and 789 more copies of
>> Ruby on Rails.


That's a brain fart, is it?

A.

A., don't you think that absolute numbers are much higher for Python than
for Ruby?
So it's at least hm, just a little bit pretentious to throw percentages around.
Maybe Haskell has also seen a decent increase in book sales.

BTW one of the 789 buyers was me. While I think ruby is nice, it has some major
flaws, too - for example internationalization, which really surprised me.

0

Stephen

A., don't you think that absolute numbers are much higher for Python than
for Ruby?
So it's at least hm, just a little bit pretentious to throw
percentages around.
Maybe Haskell has also seen a decent increase in book sales.


The huge relative increase is interesting because it shows that things
are moving fast in the Ruby and Rails areas => it's worth watching.


BTW, I was wrong about it's amazon sales ranking.
#922 was yesterday, today it's #864
http://www.amazon.com/gp/product/097669400X/ref=pd_kar_gw_1/102-8509905-8696131?n=283155

This doesn't cound the pdf version that they sell directly on their site
(I only bought this one).


Alain

0

Thanks for your reply Alain. Don't get me wrong I'm not saying that Ruby/Rails isn't any good, to be honest with you I have barely worked with it (just a few sample apps etc) so I have absolutely no idea. What I do know is that I have never seen any job asking for Ruby/Rails (in my area). You say to "use it on a real project and you'll be sold" but who is actually doing these projects??

Excluding that, my post was basically talking about the business viability of such an IDE, rather than the viability of the Ruby/Rails platform as a whole. The main issue still remains that if the Ruby/Rails community will not pay for a professional-grade IDE then who are JetBrains going to market this product to??

That's not a question I can answer, I'm not saying there is no market, but you seem to be fairly in tune with the Ruby/Rails market and yet even you do not seem confident that Ruby/Rails developers will pay what it would take to make developing this project financially viable. The only blackspot here (and only JetBrains can decide this) is what is the line that decides whether the project could be financially viable for them?

This truly seems more like something for an enthusiastic plugin developer or startup company than JetBrains. But that is just my (uneducated as far as Ruby/Rails is concerned) opinion. :)

0


> You say to "use it on a real project and you'll be sold" but who is
> actually doing these projects??


"These projects" are just plain web-applications. Nothing special here.
If it can be done in PHP/JSP/ASP and doesn't rely on a very special
architecture/library, it can - most probably - be done in Rails, faster,
with less code.

If nothing else, write your own bug tracker, blog, or personal site with
Rails. Basic (shared) hosting with Rails support is also very cheap:
2$/month (Railsplayground.com, f.ex), 30$/year (Dreamhost, with the
coupons), etc..


> The main issue still remains that if the Ruby/Rails community will not
> pay for a professional-grade IDE then who are JetBrains going to market
> this product to??


There is a market with developers if the - value based - price is right.
There is a profit for JB if cost <= value-based price * license.
The cost of this new product is minimal IF it can be build as an IDEA
plugin.


I may be wrong but I suppose that IDEA's language-API could handle Ruby
easily, as it already works with Javascript and Python.
If it's true then all it would take is for JB to write and support this
plugin.
They could also package it with a reduced version of IDEA, and sell it
as a separate product/full-blown IDE.
The total cost would be a fraction of IDEA's development running cost,
and they'd have a new product to sell.

JB has already put a lot of energy and money in developing smart
features for css, html, xml, javascript, CSV, SVN, etc....
It would be a shame not to reuse what can be reused if it's as simple as
writing a plugin.


My 5.X license is gathering dust and without this plugin I won't renew
it. I know a few web-developers around me who are in the same situation.
We are only a small percentage, but we exist.


Alain

0

In article <dtcfe4$khi$1@is.intellij.net>, alain.ravet@biz.tiscali.be
says...

I'm not trying to sell Rails to java developers, because I know they'll
come to it on their own. After all, who doesn't want to get rid of the
XML spagettis, frameworks salads and inflating APIs?
The simplest web project in java has you jump through half a dozen
misaligned hoops : servlets, Sping, Struts, Hibernate, etc...
With Rails, there is ONE hoop: Rails. If you want to learn it, there is
ONE book. Read it and you're an expert. Almost. Read teh second
book/learn Ruby, and you're a master. Almost.
There is one picture that sums it up nicely and perfectly:
http://www.garbett.org/upload/books.jpg


I doubt it. If you write your code correctly, you don't have to
understand all that much about Spring, Struts, or Hibernate. We use
those very things and people have a fast ramp-up time. Also, when we we
need to write non-web apps, their experience is directly transferable.

And the projects, scale to at least the small and medium-sized apps we
pursue.

I'll say Spring support would make FAR more of an impact than Rails.
--
-


David H. McCoy


-


0

David ,

I'll say Spring support would make FAR more of an impact than Rails.


A RailsPlugin and/or IDE is probably more about
- gaining new customers, in a new domain;
- retaining existing ones that are dropping Java for their web-apps,
for their sanity's sake,
than offering a new tool to existing ones who work in legacy java shops.

Additionaly, it's also about:
- extending JetBrains visibility a little, in the Rails/Ruby world.

I doubt it. If you write your code correctly, you don't have to
understand all that much about Spring, Struts, or Hibernate.


I doubt that very much too: using without knowing the ins and outs!!

In each domain - ORM, templating, etc.., many frameworks provide similar
services so you first have to 1/identify the available products,
2/test+evaluate each one, 3/choose 1 framework.

Once you have chose your toolset, you have to:
- subscribe to and read a mailing list
- buy books, learn,
- learn how to test,
- download, install,
- configure, integrate,
- track new version, patches, bugs, incompatibilities

, and pray that everything will work fine once combined.

And each time one of those products is updated you need to
- evaluate, test, upgrade
- make sure there is no imcompatibilities

This picture is not a joke, it's the real thing
http://www.garbett.org/upload/books.jpg
(Actually, you can remove the Ruby book from the Rails pile, as you can
go a long way without it, initially.).

Also, when we we need to write non-web apps, their experience is directly transferable.


Are you looking for a silver-bullet?
FYI, when working in Rails you use plain and standard Ruby in many
different places :
- add code in html views(like JSP)
- configure the project
- make
- deploy the code
- fill the DB with test data
and also
- extend the editor (ex: TextMate)

Alain

0

Thanks for saying "buzz", and not "hype".

I have nothing against Ruby, and can fully believe that it would be a better world if Ruby eventually "won" over Perl, Python, PHP, etc. I just don't think that adding Ruby/Rails support to IDEA would either be all that profitable for JetBrains or all that compelling technologically. If they were able to do JavaScript, Ruby support is clearly well within their comfort zone, technologically. It also made a bit more market sense, since JavaScript probably has 20x more active professional developers than Ruby does.

0

In article <dtctrg$6sn$1@is.intellij.net>, alain.ravet@biz.tiscali.be
says...

David ,

I'll say Spring support would make FAR more of an impact than Rails.


A RailsPlugin and/or IDE is probably more about
- gaining new customers, in a new domain;
- retaining existing ones that are dropping Java for their web-apps,
for their sanity's sake,
than offering a new tool to existing ones who work in legacy java shops.


An numerically insignificant number of of new customers and a fewer
number of switchers.

Additionaly, it's also about:
- extending JetBrains visibility a little, in the Rails/Ruby world.

I doubt it. If you write your code correctly, you don't have to
understand all that much about Spring, Struts, or Hibernate.


I doubt that very much too: using without knowing the ins and outs!!


I don't care what you doubt. You aren't here. We've had very good
success with people ranging from junior to senior developers. Just
because you cannot manage it doesn't mean it cannot be done. Idea has
features that other IDEs cannot match.

In each domain - ORM, templating, etc.., many frameworks provide similar
services so you first have to 1/identify the available products,
2/test+evaluate each one, 3/choose 1 framework.

Once you have chose your toolset, you have to:
- subscribe to and read a mailing list
- buy books, learn,
- learn how to test,
- download, install,
- configure, integrate,
- track new version, patches, bugs, incompatibilities

, and pray that everything will work fine once combined.


Come on. You cherry pick the absolute worst case scenario that doesn't
exist in reality. "Download and install"? "Learn how to test"? I guess
if one was to find those two actions daunting, actually using a tool
would seem like a challenge.

And each time one of those products is updated you need to
- evaluate, test, upgrade
- make sure there is no imcompatibilities


And it all can be managed, easily. We've been using this stuff for
almost two years and across several versions and have encountered
significant problems. And we've managed to share the codebase of your
Struts/Spring/Hibernate framework across seven projects. Just to define
what HAS happened we've had a total of 3 issues, the most serious was an
Xdoclet problem(used to generate Hibernate hbm files) that took 1 day to
resolve.


This picture is not a joke, it's the real thing
http://www.garbett.org/upload/books.jpg
(Actually, you can remove the Ruby book from the Rails pile, as you can
go a long way without it, initially.).

Also, when we we need to write non-web apps, their experience is directly transferable.


Are you looking for a silver-bullet?
FYI, when working in Rails you use plain and standard Ruby in many
different places :
- add code in html views(like JSP)
- configure the project
- make
- deploy the code
- fill the DB with test data
and also
- extend the editor (ex: TextMate)


Siler bullet? Why don't you just say "when you use a hammer..."? So you
managed to learn use Ruby without
- subscribe to and read a mailing list
- buy books, learn,
- learn how to test,
- download, install,
- configure, integrate,
- track new version, patches, bugs, incompatibilities

How is that possible? :)

The fact is that with a small amount of effort, one can use code that
scale up to larger projects and down to smaller ones if one picks the
write code. Spring, for example, has proven to be compatible from 1.0 to
1.2.4 and promises to be a drop-in replacement for 2.0 and so far,
they've delivered the goods. Hibernate 2.x and 3.x can co-exists and
we've had no need to upgrade Struts beyond 1.1.x.

But we have out of the box support for all our projects for
sophisticated transaction support, caching, profiling, security, Quartz
scheduling, etc with many books and forums to answer pretty much any
question that arises.

We've won business based on this code base. So, IMO, Jetbrains can
expend resources targetting a micro-minority or spend time enhancing
their core base all the while attracting more users to that core by
adding the features that people use.

Alain


--
-


David H. McCoy


-


0

In article <MPG.1e63ddc04f1ba5a09896aa@news.intellij.net>, fake@mail.com
says...

I doubt it. If you write your code correctly, you don't have to
understand all that much about Spring, Struts, or Hibernate.


I doubt that very much too: using without knowing the ins and outs!!


I don't care what you doubt. You aren't here. We've had very good
success with people ranging from junior to senior developers. Just
because you cannot manage it doesn't mean it cannot be done. Idea has
features that other IDEs cannot match.


I don't mean for that "I don't care" to come off so harsh, but I find,
WRT forums, nothing more frustrating than people who attempt to contract
one's direct experience with their opinion.

I've got, for example, a junior developer who was able to take modify
one piece of code that handles searching using Hibernates Query by
example, query by criteria, and Spring's convienience classes ALL
without really understanding how it works simply by examining the
existing code.

In a similar vein, I've added caching support to one of our objects
using AOP and the open source OSCache. I wrote an interface that uses
OSCache for this particular implentation and the caching is configured
via Spring.

By using the put or get method on the interface contained in that
inteceptor one can add caching to one of our Business Object interfaces
all without understanding AOP, Spring, or even OSCache.

Also, a added a simple interceptor for profiling, in about 2 hrs, that
prints for every Business object interface call the name of the object
in question, method, execution time, arguments, and memory used before
and after the call. By adding this interceptor to the business object
definition in our business.xml file, one can add this to any call,
again, without fully understanding AOP, Spring, or even the profile
interceptor itself.

It can be and has been done.

--
-


David H. McCoy


-


0

FWIW, here is my motivation for posting those Rails-related suggestions:
nostalgy.

I have found memories of a time where IDEA was the single ray of light
that guided us in the dark seas of tangled code and files. Over the
years it became an extension of my fingers and brains. Competitors
started to appear but IDEA was able to stay ahead of the pack.
Over the years, a lot of good features were added around a lot of other
languages : javascript, css, html, xml, jsp, etc... How could one live
without it? Impossible.
When they started working with dotNet, I didn't think "treason,
treason", I just thought: "they're a business, and languages are their
stuff". What is good for them is good for us.
Then the language API came, and MPS, and the python plugin. I thought:
"languages are really their stuff"


Then Rails came into my life, and it was such a relief! No more pain, no
more XML spaghetti, no more humonguous APIs, no more frameworks
switching, no more of all that crap. My only regret: leaving IDEA
behind. In this world, IDEA is not required. It would be nice though:
syntax errors highlighting, quickdoc, debugging, code inspection,
refactoring, etc, etc... It's not available here, but you know what:
it's OK. But it would be nice though, and I would pay for that. I think.


But things change fast in the Rails worls. After all, it's only 2 years old.
Over the week-end, a team of developers enhanced TextMate in such a way
that it's now offering many of the good things that were once unique to
IDEA: smart navigation and code extraction
This morning new delivery baffled me:

http://blog.inquirylabs.com/2006/02/21/syncpeople-on-rails-09-wow-what-a-weekend/
And all is done through short "macros", written in plain Ruby.


I'm afraid it's already too late for JetBrains to play a serious role in
the Rails and Ruby world: things are moving too fast, and a half-backed
solution wouldn't do it.


Bye bye IDEA, and thanks for the years of developing with pleasure.
I'll stick to 5.X.

Alain

0

Please sign in to leave a comment.