CUJ and Microsoft C++

N

NPC

Hi,
Is anyone else seeing a pattern with the frequency of Microsoft-related
C++ articles (not including advertisements) appearing in the 'C/C++ Users
Journal' over the last 6 months?

It appears that many of the usual columnists are pushing hard to have the
C++ Standard follow several new inventions in the Managed C++ World.
Mostly, these new inventions appear necessary to solve issues .NET is having
with using only standard C++. Interestingly, it appears that Bjarne
Stroustrop is helping Microsoft resolve some of these issues now.

I've never been too convinced of the latest Microsoft WhizBang (there are
so many dying whizbangs from Redmond after all), but to see articles showing
Bjarne's involvement makes me pause somewhat. I personally feel confident
that he has only the best intentions for Standard C++ in mind. I would like
to see anything Bjarne has to say about C++/CLI - has anyone seen any words
of the sort from Bjarne directly?

Some may think that I'm just another anti-Microsoft kind of guy. I am
not. I am a pro-standards kind of guy - which is why I choose to use
Standard C++. My question to anyone interested in answering is: "Are you
concerned at all about the largest software conglomerate, who also now
boasts > 98% C++ Standards Compliance, leading the charge in how Standard
C++ will change in the future?" Also, "Do you believe that one should only
boast once they have achieved 100% Standards Compliance?" Is 98% good
enough so many years after the Standardization was released? Should it
convince us enough to have Microsoft lead the way for us?

Again, I'm only intersted in 100% Standards Compliance. I hope others do
to. I'm curious.


NPC
 
C

Claudio Puviani

NPC said:
Is anyone else seeing a pattern with the frequency
of Microsoft-related C++ articles (not including
advertisements) appearing in the 'C/C++ Users
Journal' over the last 6 months?

I've stopped finding CUJ relevant a fairly long time ago. Strangely, that
coincided almost exactly with the editorial staff's pitiable attempts to
justify coverage of languages other than C or C++.
It appears that many of the usual columnists are
pushing hard to have the C++ Standard follow
several new inventions in the Managed C++ World.
Mostly, these new inventions appear necessary to
solve issues .NET is having with using only standard
C++. Interestingly, it appears that Bjarne Stroustrop
is helping Microsoft resolve some of these issues now.

Fortunately for us, neither CUJ nor Microsoft has more weight than anyone
else on the standard committee. I'm fully confident that the committee as a
whole will accept only features that will improve the language and flatly
refuse to even consider anything that would make things problematic for the
non-Microsoft world.
I've never been too convinced of the latest Microsoft
WhizBang (there are so many dying whizbangs from
Redmond after all), but to see articles showing Bjarne's
involvement makes me pause somewhat.

That Bjarne is involved in helping Microsoft should also be encouraging. We
can expect more recommendations on how Microsoft could better conform to the
standard than abdications that lead to a bastardization of the language.
I personally feel confident that he has only the best
intentions for Standard C++ in mind. I would like
to see anything Bjarne has to say about C++/CLI
- has anyone seen any words of the sort from Bjarne
directly?

Not I.
Some may think that I'm just another anti-Microsoft
kind of guy. I am not. I am a pro-standards kind
of guy - which is why I choose to use Standard C++.
My question to anyone interested in answering is: "Are
you concerned at all about the largest software
conglomerate, who also now boasts > 98% C++
Standards Compliance, leading the charge in how
Standard C++ will change in the future?"

I'm not in the least concerned, and I don't see this as Microsoft leading
anything, much less anyone following Microsoft.
Also, "Do you believe that one should only boast once
they have achieved 100% Standards Compliance?"

I'd rather see a company admit that they're x% compliant that to have to
guess.
Is 98% good enough so many years after the Standardization
was released?

It's good enough if it supports the subset of the language that a person
uses. The day you encounter code that compiles on another platform that you
use, but not there, then you have something to complain about.
Should it convince us enough to have Microsoft lead the
way for us?

Again, you're turning this into an X-File. There's no conspiracy here.
Microsoft does its thing and it hasn't impacted the standard C++ community
where it didn't want to be impacted. Do you see "far" and "near" anywhere in
the standard? Or any other mutilations that Microsoft and Borland and others
have introduced in the past nearly 20 years? Give the folks on the committee
the benefit of the doubt.
Again, I'm only intersted in 100% Standards Compliance.

You must spend a lot of time on the unemployment line if that's what you're
waiting for. In the real world, if you're mandated to write software for
platform X, you use the tools that are available. You don't get the luxury
of going on strike because your compiler isn't 100% compliant. Even when
there's a choice of compilers, issues like correctness, performance, and
compatibility with existing libraries tend to take precedence over
compliance with the standard.
I hope others do too.

Not likely, unless they don't have to deal with reality.

Claudio Puviani
 
N

NPC

You must spend a lot of time on the unemployment line if that's what you're
waiting for. In the real world, if you're mandated to write software for
platform X, you use the tools that are available. You don't get the luxury
of going on strike because your compiler isn't 100% compliant. Even when
there's a choice of compilers, issues like correctness, performance, and
compatibility with existing libraries tend to take precedence over
compliance with the standard.


Not likely, unless they don't have to deal with reality.

Claudio Puviani

Interesting comments. A bit derogatory and unfounded, which I have to
come to expect from some here (unfortunately). Your comments on my
unemployment status are somewhat funny after seeing so many similar comments
by you to others here... some sort of OCD related to your own fears of
unemployment is my guess.

Nope - never spent a day in the unemployment line. Although, there have
been times when I would have enjoyed the break...

If you believe that there could be no influence by the leaders of the C++
committee, then your view of politics are very idealistic and childish. I
really am simply asking for others opinions about my questions - not real
interested in your judgement of me.

BTW, you may find www.comeucomputing.com interesting in terms of standards
compliance.
 
L

Leor Zolman

BTW, you may find www.comeucomputing.com interesting in terms of standards
compliance.

Out of curiosity, in what sense did you mean he might find Comeau
"interesting"? (I use Comeau, so I'm well aware of its level of Standards
conformance. I'm just wondering what you meant.)
-leor
 
C

Claudio Puviani

NPC said:
Interesting comments. A bit derogatory

It's hard to be complimentary towards blind idealism.
and unfounded, which I have to come to expect
from some here (unfortunately).

16 years of having to deal with real issues concerning C++ compilers instead
of wishing the world to be a perfect place is hardly unfounded. The fact
that compliance is of secondary importance in the real world is not only
founded, it's easily verifiable.
Your comments on my unemployment status are
somewhat funny after seeing so many similar
comments by you to others here...

If you bother to look at the context, you'll notice that what that sentence
says is "you'll suffocate if you hold your breath for 100% compliance." Only
paranoia would lead to it being parsed as "I thing you're unemployed."
However, yes, there have been prior discussions regarding real unemployment.
some sort of OCD related to your own fears of
unemployment is my guess.

I sure hope your programming is better than your pop psychology. My concern,
when the topic is actually employment, is that a disturbing proportion of
programmers is unprepared for the job market. It takes months and dozens of
interviews to find one qualified candidate. That's a serious problem for
both employers and prospective employees.
Nope - never spent a day in the unemployment line.
Although, there have been times when I would have
enjoyed the break...

Once more, that wasn't the gist of my comment, so this is irrelevant.
If you believe that there could be no influence by the
leaders of the C++ committee, then your view of
politics are very idealistic and childish.

This from someone who only accepts 100% compliance. Right.
I really am simply asking for others opinions about my questions

We all see how you react to others' opinions. What you want is for someone
to agree with you. Counterarguments are clearly not welcome.
not real interested in your judgement of me.

My judgment was of your interpretation of the situation, which you
explicitly asked for. I have no opinion of you other than that your view of
the standardization process is skewed.
BTW, you may find www.comeucomputing.com interesting
in terms of standards compliance.

Comeau's compiler is well respected for its compliance and often used to
validate code, but it's not commonly used in production for many of the
reasons I listed.

Next time, you might want to add "please respond only if you agree with me."

Claudio Puviani
 
N

NPC

Again, my response is only to your need to be derogatory toward others. As
before, I see this is a common trend with your replies. Disagreement with
me is nonsensical in the sense that I made no claims. I did want an opinion
about my questions - not about me. My entire response to you has only been
to point out that your comments are very derogatory towards others.

As to your know-all attitude about what the markets are/what employers
need/what employers are looking for/etc., I may suggest that you consider
the stock market for a living rather than beating up on newbies in the C++
newsgroups (which I am not).

I find little need to defend Comeau's compiler. Indicating that its
performance is poor is just plain silly.

Chow.
 
G

Gary Labowitz

As to your know-all attitude about what the markets are/what employers
need/what employers are looking for/etc., I may suggest that you consider
the stock market for a living rather than beating up on newbies in the C++
newsgroups (which I am not).

Please do not send him here. He is OT. (signed) The stock market.

But really folks ... I was considering resubscribing to CUJ after
having let my subscription lapse for about a year. Has the magazine
becomes so bad I shouldn't bother? By "bad" I mean irrelevant to
standard C/C++ coding practices.
The last article I remember reading that interested me was about smart
pointers and reference counting.
 
N

NPC

But really folks ... I was considering resubscribing to CUJ after
having let my subscription lapse for about a year. Has the magazine
becomes so bad I shouldn't bother? By "bad" I mean irrelevant to
standard C/C++ coding practices.
The last article I remember reading that interested me was about smart
pointers and reference counting.


To be honest, I think that CUJ is still a very good magazine. I think it
has been a bit biased recently, but I could really be wrong. In the last
issue (Web Services - May 2004), I would approximate > 70% of the articles
were related to .NET/COM/Microsoft in some way. Tough to ignore Microsoft
in the software world though...
 
M

Mike Wahler

Gary Labowitz said:
Please do not send him here. He is OT. (signed) The stock market.

But really folks ... I was considering resubscribing to CUJ after
having let my subscription lapse for about a year. Has the magazine
becomes so bad I shouldn't bother? By "bad" I mean irrelevant to
standard C/C++ coding practices.
The last article I remember reading that interested me was about smart
pointers and reference counting.

I've been a subscriber of CUJ for many years, and I still
find it interesting and useful.

But of course, YMMV.

$.02,
-Mike
 
L

Leor Zolman

Please do not send him here. He is OT. (signed) The stock market.

But really folks ... I was considering resubscribing to CUJ after
having let my subscription lapse for about a year. Has the magazine
becomes so bad I shouldn't bother? By "bad" I mean irrelevant to
standard C/C++ coding practices.
The last article I remember reading that interested me was about smart
pointers and reference counting.

I've been a bit reluctant to chime in here, because I'm sensitive about
rubbing certain folks associated with CUJ (whom I respect and admire) the
wrong way. But there /has/ been a shift lately, and while I'm aware of
some of the factors contributing to this shift, I'm sure there are many
others of which I am not.

Some background: CUJ means a lot to me. A /real/ lot. Robert Ward started
it as "The BDS C Users Group Newsletter" back in 1981, and the very first
issue coincided with the death of Ed Ziemba, the architect and principal
implementor of the MARC operating system. Ed's death was a devastating
event in too many ways to enumerate. I would most certainly have been a
lot more involved with CUJ in its early, formative years had my life and
wits not been knocked into a very eccentric orbit by Ed's death.

When I more-or-less had regained my footing in 1989, during a phone
conversation with Robert we decided that I'd give a try at joining the CUJ
staff. I lived in Boston at the time, and CUJ was in Kansas. Needless to
say, when I informed my bride (we'd been married about 2 months at the
time) of my sudden employment plan, she was not amused.

We (and our marriage) survived three years in Kansas (Lawrence aint that
bad, it is kind of like a slice of Boston culturally, and anything /but/
flat, but watch out for the chiggers...or rather, don't bother watching
out, because you can't /see/ them), and even thrived. I was low-man on the
totem pole as one of the 3-man "Tech Staff" (Robert, myself and Kenji Hino)
and my ego really, really needed that.

To me, Robert always seemed to fit the textbook characteristics of a
conscientious, empathetic and in-the-trenches small business employer/owner
(and he is also one hell of a nice guy). I made peanuts, and still felt
privileged to be there in that place and during that time. The quality of
CUJ over the years was a direct result of Robert (and Donna, his then-wife,
together comprising the R and D in R&D Publications), on the strength of
their character and work ethic. drawing the very best people in the
business to contribute to CUJ in so, so many ways. All the people that
worked there were incredible, from management to data entry to the
remarkable lady who answers the phone (Hi Loletha!). We all knew we were
involved in something /real/. Even the /marketing/ people were all amazing
folks ;-)

I left in 1992 to begin a solo training career back in Massachusetts.
Kenji left a short while later (I had the privilege to assist him in
acquiring his US citizenship a few year later, by writing him a
recommendation letter. Kenji, wherever you are, I wish you all the best.)
Robert and Donna sold CUJ to CMP, but for years after that, under the help
of Bill Plauger, then Chuck Allison, the editorial decision-making and
story quality remained rock solid.

Between about a year ago and today, something happened. I'm not quite sure
what, but I wouldn't be surprised if it had a lot to do with big corporate
decision-making being at odds with the editorial policies that had defined
what CUJ was for all of its twenty-plus years to that point. Chuck, and
some other folks, left. Much of the content that shows up in the magazine
now would have probably been rejected per the editorial guidelines of CUJ
past. I don't read articles about .NET and such, so I can't attest to the
quality of those articles. I can't even attest to whether or not such
material represents a "necessary" shift or not. I'm not qualified to make
that judgment. But I sure personally don't care for that stuff.

There are still plenty of the kinds of articles I do personally love to
see, there's just more chaff to separate out. On-line, there is lots of
good material in the CUJ Experts Forum.

Today's CUJ is not quite the CUJ that Robert founded, nor the one I worked
for (during my only three years of salaried employment since 1979), nor
even the one it was just a few years ago. But it is still worth subscribing
to and reading, IMHO.
-leor
 
N

NPC

I liked your last post.

The reference to Comeau is simply to point out that there do exist compiler
vendors who really care about 100% compliance. I could go on-and-on about
how great Comeau is, but I'd be preaching to many people with more knowledge
about such things than myself.

To be perfectly honest, I believe that many programmers/designers are all
too willing to throw in the towel when it comes to real standards (as
opposed to industry standards) - the self-proclaimed 'realists', if you
will. If you use tools in such a way that go against a known standard, then
it is only a matter of time before that code breaks. It's always a
rat-race - you can't have compilers with 100% standards conformance the same
day that the new standard is released. However, we are many years beyond
that day now. If you "just get it working now", without regards to how it
will work when your compiler vendor finally gets up to speed, then you are
not doing a very good job (in my opinion).

I believe that Microsoft Visual C++ came a heck of a long ways from 6.0 to
its current status because users (designers and programmers alike)
complained and pointed out how bad it was. Why not clamor about the
remaining 2%? Some of the remaining 2% are important items in my everyday
use of C++. I believe that it's important for users of C++ to be critical
of vendors who call it "good enough". It's just a simple supply and demand
type of thing (well, sort of). We demand something, you provide that
something, you could get the sales for providing what we want.

NPC
 
C

Claudio Puviani

Gary Labowitz said:
Please do not send him here. He is OT.
(signed) The stock market.

Sorry to burst your bubble, but I've been working on Wall Street for years.

Claudio Puviani
 
L

Leor Zolman

I liked your last post.
Thanks.

The reference to Comeau is simply to point out that there do exist compiler
vendors who really care about 100% compliance.


There are a lot of folks now involved with C++ at Microsoft who do in fact
care about 100% compliance. The only issue I'm personally aware of that
faces political opposition is export. I'm not sure I'd lament the effective
banishment of any expectation for full export support from production
compilers (but I would feel great sympathy, if such is in export's future,
for the working stiffs at EDG...)
I could go on-and-on about
how great Comeau is, but I'd be preaching to many people with more knowledge
about such things than myself.

To be perfectly honest, I believe that many programmers/designers are all
too willing to throw in the towel when it comes to real standards (as
opposed to industry standards) - the self-proclaimed 'realists', if you
will. If you use tools in such a way that go against a known standard, then
it is only a matter of time before that code breaks. It's always a
rat-race - you can't have compilers with 100% standards conformance the same
day that the new standard is released. However, we are many years beyond
that day now. If you "just get it working now", without regards to how it
will work when your compiler vendor finally gets up to speed, then you are
not doing a very good job (in my opinion).

I believe that Microsoft Visual C++ came a heck of a long ways from 6.0 to
its current status because users (designers and programmers alike)
complained and pointed out how bad it was. Why not clamor about the
remaining 2%? Some of the remaining 2% are important items in my everyday
use of C++. I believe that it's important for users of C++ to be critical
of vendors who call it "good enough". It's just a simple supply and demand
type of thing (well, sort of). We demand something, you provide that
something, you could get the sales for providing what we want.

I get the impression that, within Microsoft, the same kinds of battles (or
parallel ones, at least) get fought as the ones you hear about throughout
the Standardization process. Progress and righteousness vs.
backwards-compatibility, and all that. The difference is just in the
influence of the mighty dollar.

FWIW, I haven't been shy about voicing my opinion to the few high-ups in
the Visual C++ division I've had the pleasure of meeting (and it /has/ been
a pleasure; all the folks I've met so far in the group have been stand-up,
smart, open-to-suggestion individuals.) Most of them are not at all evasive
about the obvious conflict Microsoft faces wrt these issues. IOW, they
really do care.
-leor
 
C

Claudio Puviani

Leor Zolman said:
I've been a bit reluctant to chime in here, because
I'm sensitive about rubbing certain folks associated
with CUJ (whom I respect and admire) the wrong way.
But there /has/ been a shift lately, and while I'm aware
of some of the factors contributing to this shift, I'm sure
there are many others of which I am not.

Some background: CUJ means a lot to me. A /real/ lot.
Robert Ward started it as "The BDS C Users Group
Newsletter" back in 1981, and the very first issue coincided
with the death of Ed Ziemba, the architect and principal
implementor of the MARC operating system. Ed's death
was a devastating event in too many ways to enumerate.
I would most certainly have been a lot more involved with
CUJ in its early, formative years had my life and wits not
been knocked into a very eccentric orbit by Ed's death.

When I more-or-less had regained my footing in 1989,
during a phone conversation with Robert we decided that
I'd give a try at joining the CUJ staff. I lived in Boston at
the time, and CUJ was in Kansas. Needless to say, when
I informed my bride (we'd been married about 2 months
at the time) of my sudden employment plan, she was not
amused.

We (and our marriage) survived three years in Kansas
(Lawrence aint that bad, it is kind of like a slice of Boston
culturally, and anything /but/ flat, but watch out for the
chiggers...or rather, don't bother watching out, because you
can't /see/ them), and even thrived. I was low-man on the
totem pole as one of the 3-man "Tech Staff" (Robert,
myself and Kenji Hino) and my ego really, really needed
that.

To me, Robert always seemed to fit the textbook
characteristics of a conscientious, empathetic and in-the-
trenches small business employer/owner (and he is also
one hell of a nice guy). I made peanuts, and still felt
privileged to be there in that place and during that time.
The quality of CUJ over the years was a direct result of
Robert (and Donna, his then-wife, together comprising
the R and D in R&D Publications), on the strength of
their character and work ethic. drawing the very best
people in the business to contribute to CUJ in so, so
many ways. All the people that worked there were
incredible, from management to data entry to the
remarkable lady who answers the phone (Hi Loletha!).
We all knew we were involved in something /real/.
Even the /marketing/ people were all amazing folks ;-)

In the early days, the dedication and passion showed in every page. You
couldn't have pried an issue from my fingers with a crowbar.
I left in 1992 to begin a solo training career back in
Massachusetts. Kenji left a short while later (I had the
privilege to assist him in acquiring his US citizenship a
few year later, by writing him a recommendation letter.
Kenji, wherever you are, I wish you all the best.)
Robert and Donna sold CUJ to CMP, but for years after
that, under the help of Bill Plauger, then Chuck Allison,
the editorial decision-making and story quality remained
rock solid.

Between about a year ago and today, something happened.
I'm not quite sure what, but I wouldn't be surprised if it had
a lot to do with big corporate decision-making being at odds
with the editorial policies that had defined what CUJ was for
all of its twenty-plus years to that point.

My take on what happened, and keep in mind that this is from the perspective
of an outsider, is that CUJ lost its mission. They decided to try to harvest
Java and C# programmers (I presume in order to increase sales) with the
empty pretext that C and C++ programmers would have an interest because they
were all languages that used braces to demarcate blocks (my exaggeration,
but certainly not far off the mark). This is without growing the magazine,
which means that the C and C++ content was necessarily diluted. For reasons
that I won't venture to guess, the content that was left was further diluted
with buzzword-driven articles replacing substantive ones.
Chuck, and some other folks, left.

And I take off my proverbial hat to their journalistic integrity.
Much of the content that shows up in the magazine now would
have probably been rejected per the editorial guidelines of CUJ
past. I don't read articles about .NET and such, so I can't attest
to the quality of those articles. I can't even attest to whether or
not such material represents a "necessary" shift or not. I'm not
qualified to make that judgment. But I sure personally don't care
for that stuff.

My opinion, and it's possibly an incorrect one, is that the .NET community
is already served by other magazines, such as MSJ. By leaving its
traditional niche for MSJ's, CUJ just succeeded in making itself -- if not
redundant -- definitely run-of-the-mill. Let's not even get into how many
repetitious articles about Yet AnotherWay to represent data in XML they
published since their shift!
There are still plenty of the kinds of articles I do personally love to
see, there's just more chaff to separate out.

I agree. I sometimes buy it off the shelf for a single interesting article.
But isn't it sad that a magazine that once was jam-packed with must-read
articles is now salvaged by the occasional lone article?
On-line, there is lots of good material in the CUJ Experts Forum.

This is one place that still "tastes" like the old CUJ.
Today's CUJ is not quite the CUJ that Robert founded, nor the
one I worked for (during my only three years of salaried
employment since 1979), nor even the one it was just a few
years ago. But it is still worth subscribing to and reading, IMHO.

I can't agree with you there, but I definitely think everyone should own the
CD with the historical CUJ publications. THOSE have great value.

Claudio Puviani
 
C

Christopher Benson-Manica

Leor Zolman said:
FWIW, I haven't been shy about voicing my opinion to the few high-ups in
the Visual C++ division I've had the pleasure of meeting (and it /has/ been
a pleasure; all the folks I've met so far in the group have been stand-up,
smart, open-to-suggestion individuals.) Most of them are not at all evasive
about the obvious conflict Microsoft faces wrt these issues. IOW, they
really do care.

Surely Bill Gates was not one of those individuals? ;)
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Members online

No members online now.

Forum statistics

Threads
473,743
Messages
2,569,478
Members
44,898
Latest member
BlairH7607

Latest Threads

Top