Is corba still alive and hot in telco ???
rb
I am an avid user of CORBA and would like to pursue it beyond even
CORBA3. I do not find the omg website useful for anything other than
the specs.Is there any source ( journal/magazine/website ) that keeps
tabs on the progress or new updates in the market that corba is
involved in.
Any input on the prospects of corba in the future as per technology as
well as job market shall be appreciated.
Finding articles on the web concentrated around the 1998-99 era makes
me feel that corba is no longer hot.
regards,
rb.
Douglas C. Schmidt
>> I am an avid user of CORBA and would like to pursue it beyond even
>> CORBA3. I do not find the omg website useful for anything other
>> than the specs.Is there any source ( journal/magazine/website )
>> that keeps tabs on the progress or new updates in the market that
>> corba is involved in.
All the major telecom equipment and service providers use CORBA.
Check out
http://www.cs.wustl.edu/~schmidt/TAO-users.html
for an overview of projects (in telecom and elsewhere) that are using
CORBA. You'll find similar lists of companies at other ORB providers,
e.g., http://www.ois.com/corporate/corp-2-1.asp#proj and
http://www.prismtechnologies.com/English/News/PressReleases/2002_05_Alcatel.html.
Take care,
Doug
--
Dr. Douglas C. Schmidt, Associate Professor TEL: (949) 824-1901
Dept of Electrical & Computer Engineering FAX: (949) 824-2321
616E Engineering Tower WEB: www.ece.uci.edu/~schmidt/
University of California, Irvine, 92697-2625 NET: sch...@uci.edu
rb
much appreciated but would still want to know whether there are any
magazines or websites that actively pursue corba as the apple of their
eye.
I have something like "byte" or some such periodical in mind where i
can get all the gossip as well as technology updates in corba.
Suprisingly enough i wonder why i do not see those many new articles
in corba any longer.Resources such as ss7 using corba , the new 3g
wave leveraging on corba would be interesting to know about.
As per individual companies i do believe that iona is among the best
to do justice to it on their website.Certain research articles from
lucent and ibm are also worth the time.
r.b
sch...@macarena.cs.wustl.edu (Douglas C. Schmidt) wrote in message news:<b2banc$9...@macarena.cs.wustl.edu>...
Douglas C. Schmidt
>> much appreciated but would still want to know whether there are any
>> magazines or websites that actively pursue corba as the apple of their
>> eye.
Steve Vinoski and I write a column for the C/C++ Users Journal. Check out
http://www.cs.wustl.edu/~schmidt/report-doc.html
and you'll find all the articles we've published over the past couple
of years.
Ke Jin
> Thanks,
> much appreciated but would still want to know whether there are any
> magazines or websites that actively pursue corba as the apple of their
> eye.
> I have something like "byte" or some such periodical in mind where i
> can get all the gossip as well as technology updates in corba.
> Suprisingly enough i wonder why i do not see those many new articles
> in corba any longer.Resources such as ss7 using corba , the new 3g
> wave leveraging on corba would be interesting to know about.
In telecom industry, CORBA are mostly used in TMN area to replace or
unify various legacy technologies such as CMIP/TL1/etc.. My opinion
is, CORBA is so far the best and dominated technology in this area and
is extending itself from NE/EMS layers toward NMS and OSS layers by
more tightly integrating with J2EE technology. In IN area, CORBA is
typically used to enhance SCP backend, instead of replacing SS7. A
CORBA/TC was available from OMG 4 or 5 years ago but was merely a
meaningless local TC provider API to construct TC packages locally in
order to transfer them over SS7 (this is similar to JINA). Certainly,
with software switch, IN and TMN can largely be unified and CORBA/IIOP
could be used beyond SCP backend. As for CORBA in 3G, 3GPP has many
related and free downloadable docuements/specification. TMF and ITU
also have similar documents/specifications downloadable, but not for
free.
Regards,
Ke
Tom Welsh
<corba...@hotmail.com> writes
>Thanks,
>much appreciated but would still want to know whether there are any
>magazines or websites that actively pursue corba as the apple of their
>eye.
As someone who regularly writes articles, reports, etc. on
middleware, I have been acutely aware for some years that
CORBA is an unpopular topic in the media.
It's not easy to explain, but I suggest a combination of
the following factors:
1. CORBA is quite technical, and quite complicated even when compared
to many other technical subjects. There tend to be two levels at which
you can talk about it - the simplistic and the dauntingly detailed.
2. At the detailed level, the readers would be programmers and software
designers. Apparently this is not an audience sufficient to support a
magazine - not enough people, not likely enough to read magazines,
more likely to consult forums like comp.object.corba or books.
3. At the simplistic level, everyone "knows" that "CORBA is dead".
Microsoft said so. While some rogue journalists might persist in
challenging even the word of Redmond, IBM has now joined Microsoft
in talking Web services up and CORBA down. For a typical piece
of anti-CORBA propaganda, see this paper (which has been
heavily criticised and mocked, but is still on IBM's website):
http://www-106.ibm.com/developerworks/webservices/library/ws-arc3/
4. As Ries and Trout pointed out many years ago in their fine
books on marketing, the public generally remembers only one or
two (at most three) brands in any sector. One of these is usually
the leader, or flavour of the month, and this is the one that
most people will come up with if asked. Today, the middleware
du jour (as far as the media is concerned) is Web services,
with .Net (whatever that may be) and J2EE trailing behind. Yes,
I know .Net and J2EE are not middleware, but that is the
perception of many journalists.
On the whole, I think it boils down to a combination of
CORBA's irreducible technical complexity and commercial
pressures to talk up alternatives. Many (even most) IT
journalists do not really know much about software - the
same applies, sadly, to many analysts. I get the feeling
that they have never been comfortable with CORBA, and are
very happy to avoid writing about it in case they make
fools of themselves.
If you disagree, try this simple test. Name a book on CORBA
that makes it simple and easy to understand for the typical
non-programmer!
--
Tom Welsh
Michi Henning
> It's not easy to explain, but I suggest a combination of
> the following factors:
>
> 1. CORBA is quite technical, and quite complicated even when compared
> to many other technical subjects. There tend to be two levels at which
> you can talk about it - the simplistic and the dauntingly detailed.
I agree. Have look at existing CORBA literature. Either you get the
10 mile altitude view where everything exists only as a concept, or you
get the 10 micron close-up view, where everything is about memory management
and threading strategies -- there is nothing in between.
> 3. At the simplistic level, everyone "knows" that "CORBA is dead".
> Microsoft said so. While some rogue journalists might persist in
> challenging even the word of Redmond, IBM has now joined Microsoft
> in talking Web services up and CORBA down. For a typical piece
> of anti-CORBA propaganda, see this paper (which has been
> heavily criticised and mocked, but is still on IBM's website):
>
> http://www-106.ibm.com/developerworks/webservices/library/ws-arc3/
This is a singularly bad article (and I made this clear in my comments
on that article on the same site). But you are right: many big
vendors have gone cold on CORBA. To me, the reason isn't that web
services are so great, but that CORBA is simply too complex for the
average developer to handle. So, people are looking for the next silver
bullet (i.e., web services) when, in fact, they should be looking at
what is wrong with CORBA and why. The ironic thing is that web services
is right on the path to becoming the same luke-warm thing that CORBA
turned out to be. Web services is simply reinventing the CORBA wheel all
over again. Give them another five or so years, and they will have
"advanced" the web services architecture to the same point that CORBA
reached two years ago. Not a lot of progress there, as far as I can see...
> 4. As Ries and Trout pointed out many years ago in their fine
> books on marketing, the public generally remembers only one or
> two (at most three) brands in any sector. One of these is usually
> the leader, or flavour of the month, and this is the one that
> most people will come up with if asked. Today, the middleware
> du jour (as far as the media is concerned) is Web services,
> with .Net (whatever that may be) and J2EE trailing behind. Yes,
> I know .Net and J2EE are not middleware, but that is the
> perception of many journalists.
Right. And none of these are suitable to solve the world's communication
problems: .NET only runs on Windows and thereby disqualifies itself
immediately as a universally applicable solution; J2EE requires
everything to be written a single programming language, and
therefore also disqualifies itself; web services is mostly wishful
thinking right now and, even if they go and standardize for the next few
years, will never even come close to what CORBA can do in terms of
scalability and performance -- technically, web services is a retrograde
step. And then there is CORBA which, while being a good platform, breaks
the world complexity record. For heaven's sake -- I have to read my own
book every time I write a bit of serious CORBA code. Being generally
regarded as an expert, I *still* can't remember all this stuff and have
to look it up in my own book. There is something seriously wrong with APIs
that I can't remember even after nearly eight years of in-depth experience...
> On the whole, I think it boils down to a combination of
> CORBA's irreducible technical complexity and commercial
> pressures to talk up alternatives.
Right. That last point is the telling one: "commercial pressures to talk
up alternatives". "Talk up" is the operative phrase here. I just wish we
would, for once, concentrate on actually solving the problem instead of
"talking up" something else that, on close examination, doesn't have a
gnat's chance of solving the problem either :-(
The perverse thing is that CORBA needn't be that complex. The complexity
came about largely because inexperience, political infighting, lazyness,
and incompetence -- technically, it is possible to do what CORBA does
(and more) without all the complexity. But, given the kind of
organization the OMG is, that will never happen with CORBA.
> Many (even most) IT
> journalists do not really know much about software - the
> same applies, sadly, to many analysts. I get the feeling
> that they have never been comfortable with CORBA, and are
> very happy to avoid writing about it in case they make
> fools of themselves.
Yes :-) I've been taking great delight in shooting down journalists
and analysts like that over the years ;-)
> If you disagree, try this simple test. Name a book on CORBA
> that makes it simple and easy to understand for the typical
> non-programmer!
There isn't one, to the best of my knowledge. To be fair, you are asking
for a lot though: explaining distributed computing to a
*non*-programmer is a tall order, given that explaining it to
programmers isn't turning out to be all that easy. But your point is
well taken: CORBA is too complex for the majority of programmers to put
up with. That's probably the main reason for why it wasn't all that
successful.
Cheers,
Michi.
Ke Jin
1. Comparing CORBA with WebService within the scope of telco domain
(especially within the backend side) is not comparing apple to apple.
WebService is ideal for web or enterprise application integrations,
where, WebService will obviously be the winner over CORBA. However,
advantages of WebService within these areas become its fatel
disadvantages outside. By emphasizing loose coupling and the highest
flexability, WebService inevitably sacrifices performance, scalability
(measured under given system resources), intuitive object model and
seamless *programming* language coupling (e.g. with the programming
language's own type system).
2. In telco. domain, I would expect WebService to be adopted by
frontend applications or integrations or even to NE configuration
management (as presented by DMTF, or even in place of TL1, for
instance), etc.. but absolately not as a distribute platform for the
backend telco system operation and support. In these areas,
technologies like SS7 (as IN or AIN backbone), which was introduced
almost 20 years ago, CORBA/SNMP/CMIP (in TMN area), etc. will stay.
j2ee, .net (yes, as you said, they are not middlewares), vanilla CORBA
and legacy MOM/TP middlewares or various workflow engines will compete
in AIN SCP, OSS/CRM backend areas.
3. CORBA might be complicated in writting small examples, but is
reasonably simple in developing large distributed
applications/systems. Certainly, to survive, CORBA community should
start to put more effort on showing and defining itself to be a simple
and effective tool, rather than a sophisticated ivor tower.
Regards,
Ke
Tom Welsh <ne...@tom-welsh.co.uk> wrote in message news:<U$xp9gAgy...@nildram.co.uk>...
Gerald Brose
>
> So, people are looking for the next silver
> bullet (i.e., web services) when, in fact, they should be looking at
> what is wrong with CORBA and why. The ironic thing is that web services
> is right on the path to becoming the same luke-warm thing that CORBA
> turned out to be. Web services is simply reinventing the CORBA wheel all
> over again. Give them another five or so years, and they will have
> "advanced" the web services architecture to the same point that CORBA
> reached two years ago. Not a lot of progress there, as far as I can see...
Might be earlier than in five years. Have you had a look at IBM's Web
Service Invocation Framework (WSIF)? In fact, that is a middleware bus
for SOA (service-oriented archicture), just like CORBA was for OO. In
my view, the main differences are:
a) it uses WSDL as its IDL,
b) it's not object-oriented, and does not care for object/service life
cycles
c) services can be bound to different transports, e.g., SOAP, EJB (means
RMI/IIOP), JMS (probably means MQ), etc.
Not at all uninteresting: it seems people did learn that many CORBA
interfaces were chosen in too fine-grained a fashion, and that IIOP
may not always cover all requirements (reliebility, security, etc.)
that you may want in particular cases, so binding to different
transports (e.g., JMS-MQ) is an interesting option.
Regards, Gerald.
--
Dr. Gerald Brose mailto:br...@xtradyne.com
Xtradyne Technologies http://www.xtradyne.com
Schoenhauser Allee 6-7, Phone: +49-30-440 306-27
D-10119 Berlin, Germany Fax : +49-30-440 306-78
Ivor Denny
I heard a comment from an acquaintance of mine recently who said
CORBA's not dead - just buried. This chap is the resident CORBA guru
in a major investment bank in London who is responsible for the
architecture of the banks IT infrastructure. Among my other recent
contacts are numerous telco's. Despite the major recession seen
recently with the saturation of the mobile market, these companies are
still forking out millions in renewable CORBA licences on an annual
basis.
The fact is CORBA works because it's based on proven technology. The
complexity in CORBA that people sometimes speak of, is in-fact usually
a mis-interpretation of it's sophistication. The cockpit of a 747
looks more ‘complex’ than a milk-float’s, but it can
do a whole lot more !. Dumbed-down solutions - J2EE/.net, almost by
definition, are profoundly limited in-terms of the adaptability needed
to meet the integration challenge posed by widely heterogeneous
environments. Remember, ‘legacy’ is often a mis-placed
euphemism for something that is often in-fact very desirable -
‘fit-for-purpose’. So unless implementers are really
prepared to lock themselves into a straight-jacket of proprietary
solutions - until the next fad comes along that is, the standards
approach - the CORBA approach, is the only sensible way.
Maybe the sensible ones are those too modest to make a big noise about
it 
Ivor
CORBA Consultant,
Durham,
UK
Tom Welsh
<ivor_...@yahoo.co.uk> writes
>
>I heard a comment from an acquaintance of mine recently who said
>CORBA's not dead - just buried. This chap is the resident CORBA guru
>in a major investment bank in London who is responsible for the
>architecture of the banks IT infrastructure. Among my other recent
>contacts are numerous telco's. Despite the major recession seen
>recently with the saturation of the mobile market, these companies are
>still forking out millions in renewable CORBA licences on an annual
>basis.
>
>The fact is CORBA works because it's based on proven technology. The
>complexity in CORBA that people sometimes speak of, is in-fact usually
>a mis-interpretation of it's sophistication.
<snip>
>Maybe the sensible ones are those too modest to make a big noise about
>it ;
Thanks for the evidence, Ivor. I'm not surprised, having long
suspected this to be the case. Btw, I love the phrase "not dead -
just buried". That sums up the idea neatly and wittily.
Now, if this is the case - lots of organizations are busily
beavering away using CORBA to good effect, while the media and
analysts are ignorantly telling everyone it is "dead" - the
industry has a problem. Namely, its communication organs are
not functioning properly, with the result that many people
may make decisions based on completely erroneous beliefs.
In the long run, CORBA could be less successful than it should
be, through a process of self-fulfilling prophecy.
I continue to believe that the IT industry's biggest single
problem lies in excessive (even paranoid) secrecy. Vendors
that won't share their precious "intellectual property" are
outdone only by users who carefully hush up all news of failed
projects - and, just as often, of successful ones. ("We must
protect our competitive edge"). Consequently, while we are all
paying the price of learning by trial and error, we are not
reaping the rewards. I am not an economist (thank God!) but
I strongly believe that if everyone cooperated more and
pooled their knowledge and hard-won experience, we would
all be better off. Software is not a zero-sum game!
Nevertheless, there is the occasional ray of sunlight. I heard
recently of a company that was told by a major analyst that
CORBA was dead and they should drop it. (No names, no
pack drill). Having used CORBA to very good effect, the
company responded by... firing the analyst! Nice one.
--
Tom Welsh
Michi Henning
> Now, if this is the case - lots of organizations are busily
> beavering away using CORBA to good effect, while the media and
> analysts are ignorantly telling everyone it is "dead" - the
> industry has a problem. Namely, its communication organs are
> not functioning properly, with the result that many people
> may make decisions based on completely erroneous beliefs.
Now, that is a quotable quote if I've ever heard one: "... many people
may make decisions based on completely erroneous beliefs." That's got to
be the most perfect summation of the entire IT industry I've heard in
many a year :-)
> In the long run, CORBA could be less successful than it should
> be, through a process of self-fulfilling prophecy.
CORBA is less successful than it should have been for quite a number of
reasons. Some of the bigger ones are non-technical ones. In particular,
the outcome of the DCOM-CORBA battle (which has bestowed us with Web
Services) is to blame in no small amount. Vendor attrition is another
one -- it is not a confidence-inspiring thing for a customer to watch
one CORBA vendor after another go out of business. And greed is another
one: the license fees that are being charged by most commercial CORBA vendors
are obscene (to say the least).
> I continue to believe that the IT industry's biggest single
> problem lies in excessive (even paranoid) secrecy. Vendors
> that won't share their precious "intellectual property" are
> outdone only by users who carefully hush up all news of failed
> projects - and, just as often, of successful ones. ("We must
> protect our competitive edge"). Consequently, while we are all
> paying the price of learning by trial and error, we are not
> reaping the rewards. I am not an economist (thank God!) but
> I strongly believe that if everyone cooperated more and
> pooled their knowledge and hard-won experience, we would
> all be better off. Software is not a zero-sum game!
I agree. As long as major industry leaders continue to make decisions
based on a desperate longing to convince the world that each has the
largest set of reproductive organs, this industry will continue to suffer.
> Nevertheless, there is the occasional ray of sunlight. I heard
> recently of a company that was told by a major analyst that
> CORBA was dead and they should drop it. (No names, no
> pack drill). Having used CORBA to very good effect, the
> company responded by... firing the analyst! Nice one.
Good to hear! Always refreshing to find at least some spine in this
industry... :-)
If you want to build a heterogenous distributed system, you are not
going to use DCOM or .NET (obviously), and you are not going to use J2EE
either (obviously). So, what does that leave? Not DCE... The only
remaining candidate is CORBA. That doesn't mean that CORBA is the best
tool for the job, only that, given the available options, CORBA is the
only one that passes the requirements filter. It reminds me of the old days
of troff: people were willing to put up with this horrible and arcane
tool because it was the only one around; if you wanted to typeset
something, you either used troff or you did without typesetting
altogether. CORBA falls into the same camp -- it is at least twice as
complex and big as it needs to be for the available functionality.
Cheers,
Michi.
Ivor Denny
The “dead not burried” quote was all the more poignant and
amusing because it was made to counter the J2EE lobby in the Bank, at
a time when the J2EE bandwagon was at full speed, and made in relation
to the fact that EJB 2.0 (J2EE) mandates RMI-over-IIOP (CORBA) as its
interoperable protocol.
Ivor
Georges Martin
Tom Welsh <ne...@tom-welsh.co.uk> wrote:
> If you disagree, try this simple test. Name a book on CORBA
> that makes it simple and easy to understand for the typical
> non-programmer!
Hmmmm...
"Object Technology: A manager's Guide" by David taylor, followed by
"Instant CORBA" by Robert Orfali
That helped the non-specialist I am to understand a bit of CORBA... :-)
Georges Martin
Tom Welsh
Georges Martin <georges...@deboeck.be> writes
first CORBA book that I ever saw - "The Essential CORBA" by Mowbray and
Zahavi - was reasonably non-technical too.
What I was really trying to say, though, was that books that try to
explain how CORBA actually works tend to get very complicated quite
quickly. (Mind you, this is even more true of much Microsoft software,
such as COM+ and .Net).
In case you had not realised, I am all in favour of CORBA. I think it is
a great project that has been seriously underrated, and when I get the
chance I try to even things up a little.
--
Tom Welsh
Diego Sevilla Ruiz
Tom Welsh <ne...@tom-welsh.co.uk> wrote in message news:<REvYTLAd...@nildram.co.uk>...
[snip]
>
> Nevertheless, there is the occasional ray of sunlight. I heard
> recently of a company that was told by a major analyst that
> CORBA was dead and they should drop it. (No names, no
> pack drill). Having used CORBA to very good effect, the
> company responded by... firing the analyst! Nice one.
Sorry, Tom, but I can't avoid pointing the audience to one of the
talks you gave, and I enjoyed the most:
'"Web Services Will Succeed Where CORBA Failed" True, False Or
Irrelevant?'
http://www.omg.org/attachments/pdf/WebServicesCORBA-Welsh.pdf
It was a nice time in Orlando :)
Best regards!
diego