Borland and Eclipse..... its getting steamy....
From the open letter:
Our customers will get the best solution for Java development, as we bring
the Eclipse ecosystem to our flagship Java IDE, JBuilder, and further
advance JBuilder's leading enterprise development functionality for both
new and existing customers. We will also continue to invest in our support
for the Microsoft Windows platform, .NET and Visual Studio. Our Eclipse
investment will not diminish, in any way, our investments in Delphi, C++,
C#, and Microsoft Windows.
http://bdn.borland.com/article/0,1410,32987,00.html
"As we bring the Eclipse ecosystem to our flagship Java IDE, JBuilder" -
Does this mean JBuilder will now be an Eclipse plugin? Or that JBuilder
is moving to an SWT base rather than Swing? Probably not, but they do
seem to be heading up a new modelling framework or something, press
release here:
http://www.borland.com/news/press_releases/2005/ \
02_28_05_borland_amplifies_eclipse_support.html
(piece back together, client won't lemme know the long url :( )
Mmmmm
Please sign in to leave a comment.
With all these companies clambering onto the hype-wagon, I don't see how
Eclipse (the IDE) can survive the weight. Does anyone actually think
Borland is going to bring anything innovative to the table, let alone
give it away for free?
I see Eclipse (the plugin platform) growing as companies try to find a
way to make a buck with their proprietary plugins, but you'll still have
to pay for them, and they're not going to be integrated with the other
proprietary plugins.
More and more I see Eclipse as two entirely separate things: The IDE,
and the plugin platform. They are very different, and driven by very
different goals, I think. I get the feeling that the users are buying
into the IDE (because it's free and open), and the companies are buying
into the platform (because they can make plugins that are not free and
not open).
It's funny. The more hype that builds up, the less worried I am about it
(the IDE). I wonder if I'm somehow deceiving myself. Anybody else think
the same (or differently)?
--
Rob Harwood
Software Developer
JetBrains Inc.
http://www.jetbrains.com
"Develop with pleasure!"
I think you are (a litle). IMHO you underestimate trends/fashion :).
The biggest problem is that in most companies, platform decisions are taken
by managers, not developers.
Managers are not impressed by features, but by good presentations, good documentation
and trends.
The fact that Eclipse is free is not the point IMHO(only for OS nerds), cause
all the companies that want the platform must use some sort of "IBM partnership",
so they will buy the exepnsive WSAD solutions.
As long as documentation/presentations are not important(and the javadocs
have no comments in it) for JetBrains, IDEA won't be used as a platform,
cause no manager will decide so. I saw this not just once with my own eyes.
Ahmed.
Ahmed Mohombe wrote:
>> It's funny. The more hype that builds up, the less worried I am about
>> it (the IDE). I wonder if I'm somehow deceiving myself.
I think I may be, as well, but not because I underestimate trends. I
think I probably overestimate them! (If you've seen my recent blog,
you'll know what I mean ;)
The question to me is, what does the trend represent? I'm thinking of it
from an Innovator's Dilemma angle. Eclipse, the free/open IDE, has the
power to eventually erode IDEA, even if it is technically inferior. If
people are over-served by a product, they tend to switch to a cheaper
product that serves them 'good enough'. But what happens when that 'good
enough' product is no longer cheaper?
I'll put it in a different way: If Eclipse cost $499 instead of being
free, how many people would be using it? Put us on a level playing field
and we win. If Eclipse, the IDE, starts costing mucho moolah because
Eclipse, the plugin platform, attracts proprietary companies, then it's
back to 2001 when we were competing in the same situation.
We've already faced this kind of competition with Borland, and we
haven't had any problems with it. In fact, we're slowly getting better
at making inroads into manager-driven purchase decisions. It's the
cheaper open-source competition that could threaten IDEA, and I'm not
sure if that will ride the same wave that the proprietary plugin wave is
inevitably going to ride.
Again, this has always been a challenge (again with Borland as an
example). Nothing has changed in that respect. And again, we are working
on it and getting better slowly.
But seriously, I sense there's some flaw in my 'logic'. And yet, at the
same time, I find it bizarrely amazing that JetBrains seems to be
feeling very little pain from Eclipse.
--
Rob Harwood
Software Developer
JetBrains Inc.
http://www.jetbrains.com
"Develop with pleasure!"
I saw your blog. IMHO you are overestimating short term trends and underestimating
long term trends. Long term trends are harder to follow and involve the entire
industry.
Yes, but good enough depends on the situation. In Eclipse case, some very
valuable plug-ins (free) represent this.
They want to switch back, but it's too late :). The company with the very
good product already died or was eaten.
Eclipse will never cost $499. Even if with some required plug-ins it will,
people will always have a "free beer" :). This is what matters for the OS
comunity, not for managers who decide in fact.
>> As long as documentation/presentations are not important(and the
>> javadocs have no comments in it) for JetBrains, IDEA won't be used as
>> a platform, cause no manager will decide so. I saw this not just once
>> with my own eyes.
>>
Sorry but from the first day of OpenAPI, IMHO there was no change/progress
regarding documentation.
Just use some metrics tools to find out how many info do you have in those
javadocs - nothing, and just look how many articles/books are about how to
make plug-ins for Eclipse. I'm not saying that you should compete with them
in this filed, but at least something greater than nothing.
IMHO you will loose the train, cause all the companies will make plug-ins
only for Eclipse.
IMHO your logic flaw is that you think IBM is Borland (and JBuilder is similar
to Eclipse).
IBM made money even out of the Holocaust, and use totaly other strategies
than Borland.
Ahmed.
Ahmed Mohombe wrote:
Both the long and short term trends should be (by my highschool
economics 'logic') that Eclipse erodes IDEA. It doesn't. I can't explain
that with confidence. OBviously, there's something more going on than my
highschool economics can explain.
I personally think usability and intelligence will win out over
'features'. I don't think Eclipse, the open source IDE, can make up the
gap there (at least not in the mid-term).
>> But what happens
>> when that 'good enough' product is no longer lower cost?
I wasn't clear. I meant that 'in order to get good-enough, we need the
base Eclipse IDE plus X,Y,Z plugins which cost the same or more than
IDEA'. I don't fear the lock-in if the lock-in itself doesn't represent
'good enough'. People were 'locked in' to Borland, and they still
eventually switch because it's just not good enough compared to IDEA.
I don't understand your point here. Eclipse, the open source IDE, will
never cost 499, but Eclipse, the good-enough IDE which requires
proprietary plugins, might possibly.
>>> As long as documentation/presentations are not important(and the
>>> javadocs have no comments in it) for JetBrains, IDEA won't be used as
>>> a platform, cause no manager will decide so. I saw this not just once
>>> with my own eyes.
>>>
>> Again, this has always been a challenge (again with Borland as an
>> example). Nothing has changed in that respect. IMHO you are making And
>> again, we are
>> working on it and getting better slowly.
I misunderstood. I thought you meant documentation/presentations for
users and potential clients of IDEA (as in, 'buy our product, because
X,Y,Z, blah blah blah'). We are improving in this way. That's what I meant.
--
Rob Harwood
Software Developer
JetBrains Inc.
http://www.jetbrains.com
"Develop with pleasure!"
Ahmed Mohombe wrote:
But are plugins really the primary selling feature of an IDE? I think a
lot of people are making this assumption, and it may not be true.
Certainly, plugins are important, but if you had to trade off
getting-the-job-done with a wide array of mostly useless or over-priced
plugins, which would you choose?
It's only when that array of plugins becomes low cost and useful that
things get interesting. The open source IDE represented the low cost and
mostly useless. The danger to IDEA would be if mostly useless turned
into useful. People assume that the hype involving the platform will
automatically translate into more-useful plugins for the IDE, but they
forget that they will no longer be low cost. I think it's only a combo
of low cost and useful that could really affect IDEA. This could happen
with the open source IDE, but I don't think that's where the momentum is
really going. I think it's going into the plugin platform.
--
Rob Harwood
Software Developer
JetBrains Inc.
http://www.jetbrains.com
"Develop with pleasure!"
Ahmed Mohombe wrote:
> IMHO your logic flaw is that you think IBM is Borland (and JBuilder is
> similar to Eclipse).
IBM is not Borland, but we're not talking about just IBM here, we're
talking about the 'ecosystem' that Mike Milinkovich is so fond of
talking about. That's the Eclipse Foundation, which is being joined by
companies... like Borland.
Eclipse, the open source IDE, is not like JBuilder, but Eclipse, the
plugin platform with proprietary plugin vendors, does actually look a
bit like JBuilder.
> IBM made money even out of the Holocaust, and use totaly other
> strategies than Borland.
I think both IBM and Borland will profit off of Eclipse, more power to
them. I just don't think it will put JetBrains and IDEA into as much
danger as I had previously thought. That is, as long as we keep
innovating. That is certainly key. I'm not saying we can get complacent,
I'm saying that the Eclipse bark is bigger than its bite.
Sorry for the three part replay. The spam police wouldn't let me post
it. I started censoring words like 'cheap', and 'free', but that didn't
help. So I tried the old divide-and-conquer, and finally found out the
magic phrase 'm a k e m o n e y' which I changed to 'profit'.
--
Rob Harwood
Software Developer
JetBrains Inc.
http://www.jetbrains.com
"Develop with pleasure!"
Well, I don't have your selling numbers :). So if you say that it does not,
than I'm very glad.
>> Yes, but good enough depends on the situation. In Eclipse case, some
>> very valuable plug-ins (fre_e) represent this.
Agree. The small and nice things make the life easyer :). There are however
plug-ins that are just too productive for some tasks so a switch will be
taken.
>>> But what happens
>>> when that 'good enough' product is no longer lower cost?
>> They want to switch back, but it's too late :). The company with the
>> very good product already died or was eaten.
>>
Yes you were very clear. I understood what you wanted to say. I only went
a litle further with the scenario:
If Eclipse will be "good enough", and people will abandon IDEA, there's the
danger that when they want back, it's no more there. Users don't think on
a long term.
>> Eclipse will never cost 499. Even if with some required plug-ins it
>> will, people will always have a "fre_e beer" :). This is what matters
>> for the OS comunity, not for managers who decide in fact.
>>
Yes exactly. But the users, the articles and the studies will always compare
the free version(when it comes to price), even if it does nothing without
the exepensive plug-ins. It's of course unfair but this is how it works.
>> Sorry but from the first day of OpenAPI, IMHO there was no
>> change/progress regarding documentation.
>>
I meant the OpenAPI and plug-in documentation. Based on them, strategic decisions
are made. They are more important IMHO than the casual user that buys a license
at Christmas. I might be wrong but in our team, this was the case.
Regarding the normal/user documentation, IMHO a book(but in English) would
be a very strategic move. People must see the IDEA book in the bookshops,
not only the entire Eclipse shelf that takes the entire show.
Ahmed.
Not the plug-ins but the possibility to make them for sure(and the guarantee
- hence the documentation) if the situation requires it.
Cause you think as a Jetbrains employee :). Managers think differend - mostly
they take decisions to buy more than one license :), and as I said, they
have other decision criteria :).
As I said it's not about the plug-ins but the chance to make them :).
It's like the magic word "fexibility" in every marketing slogan.
Another important aspect not to forget are the "private plug-ins". They are
mosty developed only for a customer(big one of course :) ) - not for the
large masses. Our last 2 Eclipse plug-ins were in this category.
I wanted to use IDEA as a platform, but ... :(.
Ahmed.
Rob
Just look at the sheer number of IDEA plugins that have been written
over the years by motivated an tenacious authors, with - close to - no
help from JetBrains, despite your silent openAPI refactoring, that would
keep breaking them without any notice, nor advice.
It should ring a bell, in your office:
"we must do more for plugin authors"
"we must do more for plugin authors"
"we must do more for plugin authors"
(repeat 20 times)
I don't get why "you" don't get it. There about books about writing
Eclipse plugins. Even Kent Beck wrote one.
IDEA openAPI doesn't even have decent javadoc.
Alain
You might be right, but in our cause we used WSAD (only IBM), not the "foundation
version".
They have a "borg" like strategy :).
Mad dogs are dangerous, cause they spread ... :).
Ahmed.
There's undoubtedly a valuable marketting document that JetBrains could create comparing the cost of a standard IDEA license with the cost of Eclipse plugins necessary to give you equivalent power. I can gaurantee you that the comparison would be in IDEA's favor. Even in the static analysis arena, you can't match IDEA(+InspectionGadgets) for less than $899 per seat (CodePro Analytix, which admittedly bundles a bunch of other cool stuff as well). The free static analysis plugins for Eclipse (Checkstyle, PMD, FindBugs) just don't stack up (fewer rules, no quick-fixes, no batch fixes, no interactive use, ergonomics that would have to stretch to be called crappy). Presumably similar market comparisons could be made with Eclipse plugins for JSP/XML/HTML editing, structural search, J2EE/J2ME support, etc., etc.
--Dave Griffith
Sad to say, the man's right. Things are starting to turn around, but if I had to choose between DevKit and decent OpenAPI javadoc, the javadoc would win everytime.
--Dave Griffith
Ahmed Mohombe wrote:
Like this http://www.manning.com/catalog/view.php?book=fields3
--
Maxim Mossienko
JetBrains, Inc
http://www.jetbrains.com
"Develop with pleasure!"
Maxim Mossienko wrote:
>> Regarding the normal/user documentation, IMHO a book(but in English)
>> would be a very strategic move. People must see the IDEA book in the
>> bookshops, not only the entire Eclipse shelf that takes the entire show.
OK, 1 book is a start...
MOST part of OpenAPI is clear enough, but some parts need more details.
Anyway I'm doubt then if even they add full JavaDocs number of plugins will increase noticeably.
TIA,
Dmitry
Hello Dmitry,
You mean is clear of "any human written sentence".
Oh yeah? Why do you think so?
JavaDocs should be a must and a start, not the end.
Ahmed.
In article <d04tn3$nlk$1@is.intellij.net>,
"Rob Harwood (JetBrains)" <rob.harwood@jetbrains.com> wrote:
Let me dip my oar in.
I find Eclipse very interesting from a technological point of view. It
has a very different focus and feel from IDEA, and that is often a good
thing. For example, the Eclipse meme of "every damn thing in the
dialog" often means a one stop shop for a complicated task where I am
creating a class, and know exactly what I want the skeleton to look
like. On the other hand, the IDEA meme of "several simple steps that
chain together" works very well for general refactoring.
Frankly, if Eclipse used simple XML files to configure products, I would
use it more. Our build system generates IDEA project files for people
to work with, and all the dependencies are correct. I cannot do that
with Eclipse without writing a plugin, and that is not for the faint of
heart. I have asked various Eclipse people how to do this, or to
provide a simple skeleton, and gotten no help. No reason they should
have, but free/open software often has a very strong "write it yourself"
meme that is not seen as often in a proprietary business. For IDEA, I
get more of a "make a business case" feel.
In other words, the products and the authors have a different feel.
Were Eclipse commercial software, its developers would feel more like
the IDEA developers, I suspect, and would be more driven by customer
requests. As it is now, they can go in the direction that the muse
drives them.
So, I do hope JB keeps a very close eye on what you can do with Eclipse
- it does a lot of stuff. It also has a big plugin community - bigger,
I think, than IDEAs. That said, I find IDEA plugins more effective -
Metrics Reloaded is just plain cool. If the field were level, there
would be more pressure on the Eclipse authors to satisfy the needs of
the paying customers, else they would stop paying.
Scott
>That said, I find IDEA plugins more effective -
Metrics Reloaded is just plain cool.
Aw, shucks. Guess that means I have to get off my ass and do another release.
--Dave Griffith
Hello Ahmed,
We write programs, not poems :) So if you want learn programming using Open API docs it's wrong way. But if you have some java experience and IDEA you may use even obfuscated code. One more time documentation not required for WHOLE OpenAPI but for the parts of it such as PSI, Language API and so on.
Last weekend without any docs I use StructureView API and Productivity API easily. (Thanks to Olesya and Anna for good API).
TIA,
Dmitry
>> But are plugins really the primary selling feature of an IDE?
>Not the plug-ins but the possibility to make them for sure(and the guarantee
- hence the documentation) if the situation requires it.
My experience, and what I've read of the market, is that the majority of IDE users never install a single plugin. They use the application directly as it comes out of the installer, and never even think of looking for plugins. A large number will install a single plugin (an SQL browser, typically) when told about it by a friend, and that's all. This is undoubtedly true of Eclipse as well, although this tendency is somewhat masked by the existence of commercial and homebrew Eclipse distros shipping with pre-packaged plugins.
>It's like the magic word "flexibility" in every marketing slogan.
My old boss always said that "flexibility" and "extensibility" are what software providers talk about when they have no idea what their customers actually want. It's an insight that has saved me from making any number of bad software purchases. JetBrains absolutely does not have that problem.
To be honest, I think JetBrains should be glad of the plethora of people writing plugins for Eclipse. Eclipse plugin download repositories provide huge amounts of cheap and valuable market research as to what people are actually interested in when using IDEs that aren't presently available. JetBrains can then use that information to figure out future product directions for IDEA and Resharper. Figure out the functionality the Eclipse plugins provide, and provide it directly in the core product with JetBrain's peerless ergonomics and high quality (not things that most Eclipse plugins are known for). As near as I can tell from the Irida list, that's exactly what JetBrains is doing. I'd love to see innovative Eclipse plugins like Jupiter or ContinuousTesting or some of the collaborative development stuff adapted by JetBrains for inclusion in IDEA. I don't fear that those plugins will cause IDEA to lose to Eclipse because very few of the Eclipse customer base use those plugins.
--Dave Griffith
Amnon I. Govrin wrote:
>> Like this http://www.manning.com/catalog/view.php?book=fields3
I'd rather one book thats good, than hundreds of books that repeat the
same thing, same examples, and don't offer anything different.
I posted a blog, somewhat rehashing this discussion and the discussion
at Javalobby. See
http://idea-log.blogspot.com/2005/03/eclipse-over-hyped-developers-suffer.html
Note: I strategically picked and chose my arguments, so it's a little
(ok, a lot) one-sided, but I think that's fair considering all the hype
IBM is buying. ;)
--
Rob Harwood
Software Developer
JetBrains Inc.
http://www.jetbrains.com
"Develop with pleasure!"
see http://www.sys-con.com/java/readerschoice2004/liveupdate.cfm?BType=9
"Who uses Hibernate"
http://www.hibernate.org/113.html
IDEA mentioned 5 times
Eclipse 19 times.
"Rob Harwood (JetBrains)" <rob.harwood@jetbrains.com> wrote
in message news:d057ff$oqq$14@is.intellij.net...
we need the
more than
It is much easier to get base IDE for free and then to add
only those plugins that needed, paying $50-100 per each.
Even if in the end you'd pay more, the process itself is
easier. It is like buying DVDs one by one instead of buying
the pack out of three movies, even in the pack is cheaper.
Most people prefer to pay in smaller chunks.
I suggested this model to JetBrains two years ago, no one
listened. Now you have it with Eclipse.
American consumer market is all about financing and "how
much monthly payments can you afford?" BTW, what about
financing programs for software? ;)
"Rob Harwood (JetBrains)" <rob.harwood@jetbrains.com> wrote
in message news:d057ji$oqq$17@is.intellij.net...
IDE? I think a
be true.
off
or over-priced
the whole CAD/CAM/CAE/whatever other design industry uses
this model for the long time. Need new shader? Stress
analysis? Head conduction? Aerodynamics? There are plugins
for that. But is is conceptually much easier for CAD because
there is always the same source data, the plugins simply use
different algorithms to calculate the result.
Professional tools market is not a consumer market. At least we hope it's
different. We're not talking about entertainment industry, are we? It's all
about tools that help people do their work more efficiently.
"Maxim Shafirov" <max@jetbrains.com> wrote in message
news:c8a8a1bf5e578c6f233a8db31aa@news.jetbrains.com...
least we hope it's
are we? It's all
efficiently.
Considering number of Java developers out there, Java IDE
market is a consumer market now. And IDEA is consumer
product. If it were not, it would not have nice color
schemes, cool icons, shading/transparency and other bells
and whistles. Gosh, even the Allen wrenches are now made
with colored rubber handle. Because it looks cooler and
stands out. Because there are so many Allen wrenches that
decision is made purely on looks. When the tool is selected
by its looks, this is a consumer item. Check all these
messages "I looked at IDEA and was immediately turned off by
its outdated Swind interface" or vice versa "Eclipse has the
ugliest UI ever".
And no one cares that your simple and rugged wrench is made
of crome vanadium. No one is going to use it for 20 years.