"Clark F Morris" <cfmpublic@[EMAIL PROTECTED]
> wrote in message
news:vlsbs35f58padvgl9kqc0borbgp2u5lgn1@[EMAIL PROTECTED]
> On Tue, 26 Feb 2008 21:24:10 GMT, "William M. Klein"
> <wmklein@[EMAIL PROTECTED]
> wrote:
>
>>Probably all of you who MIGHT care already know about it, but there were
>>several
>>significant announcements by IBM today. Therefore, I won't post links
to
>>all of
>>them, but some that might interest some in CLC are:
>> (I think these links will work, but they may be "individualized" for
>> me. Let
>>me know and I can look for other links if these don't work generally.)
>>
>>HARDWARE:
>> IBM System z10 Enterprise Class -- The forward-thinking mainframe for
>> the
>>twenty-first century
>>
>>
http://www.ibm.com/vrm/newsletter_10577_2548_62970_email_DYN_1IN/WKlein12584487
>>
>>OPERATING SYSTEM (preview)
>> Preview: z/OS V1.10 -- Raising the bar and redefining scalability,
>>performance, availability, and economics
>>
>>
http://www.ibm.com/vrm/newsletter_10577_2548_63032_email_DYN_1IN/WKlein12584487
>>
>>ISPF (yes, it is "well and getting enhancements")
>> IBM ISPF Productivity Tool V5.10 enhancements deliver increased
>> efficiency for
>>ISPF users
>>
>>
http://www.ibm.com/vrm/newsletter_10577_2548_63020_email_DYN_1IN/WKlein12584487
>>
>> ***
>>
>>For a summary of today's announcements, go to:
>>
>>
http://www.ibm.com/vrm/newsletter_10577_2548_62925_email_DYN_1IN/WKlein12584487
>
> What is striking is what announcements were not there. There was NO
> announcement of COBOL sup****ting floating decimal. There was NO
> announcement of COBOL being able to interoperate with 64 bit JAVA.
Some people think that Java and COBOL are in the same boat...
http://www.infoworld.com/archives/emailPrint.jsp?R=printThis&A=/article/08/02/21/08NF-java-react_1.html
Personally, I'm not persuaded of that yet, although it is true that Ruby
and
PHP scripting have made large inroads into traditional Java territory.
I still think that the paradigm is im****tant and that means Java will have
a
place, for a lot longer than COBOL. However, it is also undeniable that
people
are expecting to build applications (usually Web Based) much quicker than
has traditionally been the case, and scripting facilitates that.
As one of the "small developers" mentioned in the article, I have to say
that C# and ASP.NET, along with LINQ and SQL Server, meet all of my
development requirements, and the DotNET framework is now my sole target
platform. (OK, Linux too, if you include Mono, but I'm not specifically
targeting Linux...). Most of what I do now is Web based and designed
around
SOA, and I look on desktop applications as "quaint"... :-)
The fact is that the traditional idea of cutting thousands of lines of
code
to sup****t an application is no longer viable, and scripting and component
based software are inheriting the earth.
That's fine as far as new development goes, but we currently have a huge
Legacy of COBOL and other code that is still in production.
I have learned from my own experience that it is NOT difficult to refactor
this code PROVIDED:
1. You have the right tools; Visual Studio is excellent for doing it.
Legacy
code can be "componentized" and wrapped so it can run in modern
environments
(like DotNET), even if it has to do so as "unmanaged code" initially.
2. The code is well written in the first place. Clean structured COBOL is
easy to refactor; spaghetti code isn't. If code can be "isolated" into
functions or components, the job is easy, and you DON'T need to go to a
Java
conversion.
In fact, the main argument for doing the Java conversion is so that new
age
programmers can maintain code (and they need to do this less in an OO
language, than a non-OO one, if they play their cards right...) in a
language they are familiar with. If this language is also perceived as
being
"past its sell-by date", then there may be no point in converting to it.
Instead, re-write or re-script what needs change and do it in whatever
language facilitates that, whether it is C#, PHP, or Ruby...
Ruby is generating fierce interest in Europe, even more than elsewhere.
The days of "one size fits all" or "one language development" are over.
These days it is about leveraging as much value as possible from legacy
code, and that means combining it with other code that can push it into
new
environments.
> There was NO announcement of any sup****t for the features of the 2002
> standard. There was only an announcement of COBOL XML sup****t. From
> what I can see COBOL for IBM is a cash cow that is going to go away.
I agree.
What was of particular interest to me (and not referenced in Bill's post)
is
the recent salary and overtime shake up at IBM.
http://www.pressconnects.com/apps/pbcs.dll/article?AID=/20080124/BUSINESS/801240335
There are some VERY disgruntled employees who are being told to take a 15%
drop in salary at a time when the company is still making a profit. How
many
will vote with their feet, and what effect this will have (if any) remains
to be seen.
Pete.
--
"I used to write COBOL...now I can do anything."
|
