Artima Developer Spotlight Forum - Where is Software Development Heading in 2007?

Articles |
News |
Weblogs |
Books |
Forums

Artima Forums | Articles | Weblogs | Java Answers | News

Sponsored Link •

Artima Developer Spotlight Forum
Where is Software Development Heading in 2007?

61 replies on 5 pages. Most recent reply: Jan 4, 2007 10:47 AM by John C. Walker

Welcome Guest
Sign In

Back to Topic List

Reply to this Topic

Search Forum

Threaded View


Previous Topic		Next Topic

Flat View: This topic has 61 replies on 5 pages [ « | 1 2 3 4 5 | » ]

James Watson

Posts: 2024
Nickname: watson
Registered: Sep, 2005

Re: Where is Software Development Heading in 2007?

Posted: Jan 3, 2007 11:09 AM

> > Perhaps if
> > you could give an example of what you would consider a
> > non-trivial threading problem the discussion might be
> more
> > illuminating.
>
> Of the top of my head, real-time simulation, like physics
> in games. There are lots of dependencies. You can
> parallelize things, sure, but it's not as simple as
> lauching a thread.

I think we need to use a more specific terms here. There are a couple things here. There's the mechanics of threading which is basically making some code run in parallel without errors and there's the process of taking an algorithm or program and structuring it to run in parallel.

The latter definitely varies in difficultly especially when you start talking about non-naive solutions.

What I believe(partially from experience) is that many programmers are struggling with is the former which can be difficult for non-naive solutions but not more than many other programming problems (although debugging threaded code is much more difficult, especially if you use a debugger.) I guess what I am saying is that a lot of people aren't even trying. They've been just using increasing processor speeds and optimizing single-threaded apps and now that doesn't like it's viable for much longer. Multi-threading has been around for a long time, I don't understand why people are acting like it's a brand new thing.

James Watson

Posts: 2024
Nickname: watson
Registered: Sep, 2005

Re: The "maybe" defense?

Posted: Jan 3, 2007 11:28 AM

> > No. I wouldn't. Honestly. If they used Erlang or
> some
> > other language with really good enhanced threading
> support
> > it would be a quite sensible thing to say.
> >
> Of course, your statement didn't say anything about Erlang
> or any other language, you just implied that it was a C++
> problem, so your example is irrelevant. Nevertheless, I'll
> take you at your word.

Why the sudden change? Before my honesty was in question.

> > No, all the accusations that I am lying and your
> > insistence on psychoanalyzing me is aggressive and
> > insulting.
>
> It was you who started talking about my alleged "emotional
> attachment" to C++. Apparently you have a different
> standard for your own behavior than for others.

I explained my statement after you questioned it and you basically called me a liar. How am I supposed to take that?

> > > I don' have any emotional attachment to C++,
> > > I'm not even using it (by choice) on my current
> > project.
> >
> > Then why do you jump to the conclusion that 'I am
> getting
> > my digs in'? Your interpretation of my statement
> clearly
> > shows that you gave it an emotional context.
>
> We weren't talking about C++ at all, and suddenly you made
> a critical statement about it that you weren't able to
> support.

I didn't make a critical statement about it. I questioned whether the tool I was using gave me a different perception. I wish I hadn't mentioned C++. It's really not important to my point. I thought it was pretty much accepted that threading in C++ is more difficult than in Java. Mea culpa.

> Look back at your many, many posts and you'll see
> a guy who has made a lot of criticism of C++. I'm not
> jumping to conclusions, I'm just drawing the most probable
> conclusion based on the facts.

I don't think C++ is a good language choice for most development projects. I'm not alone in this and I'm not trying to hide that. It's not an article of personal faith. It's based on my personal experiences, the experiences of others and some rational thought. If you want to think I have some sort of personal vendetta against C++, fine, be my guest. It's absurd but I really don't care at this point.

> > > ...Don't make too many assumptions about
> > > those parts you know nothing about.
> >
> > Maybe you should consider your own advice. You don't
> know
> > me.
>
> All I know about you is based on what you say here

Exactly. Basically nothing.

Leandro Oliveira

Posts: 21
Nickname: lao
Registered: Aug, 2003

Re: Where is Software Development Heading in 2007?

Posted: Jan 3, 2007 11:35 AM

> I think we need to use a more specific terms here. There
> are a couple things here. There's the mechanics of
> threading which is basically making some code run in
> parallel without errors and there's the process of taking
> an algorithm or program and structuring it to run in
> parallel.
>
> The latter definitely varies in difficultly especially
> when you start talking about non-naive solutions.
>

Ok.

> What I believe(partially from experience) is that many
> programmers are struggling with is the former which can be
> difficult for non-naive solutions but not more than many
> other programming problems (although debugging threaded
> code is much more difficult, especially if you use a
> debugger.) I guess what I am saying is that a lot of
> people aren't even trying. They've been just using
> increasing processor speeds and optimizing single-threaded
> apps and now that doesn't like it's viable for much
> longer. Multi-threading has been around for a long time,
> I don't understand why people are acting like it's a brand
> new thing.

You say people are using increasing processor speeds and that it's not viable for much longer. So, it's not multi-threading that is new, it's the problem that you can't make your programs faster anymore without using it. I'm just making conclusions based on your statements.

Jeff Ratcliff

Posts: 242
Nickname: jr1
Registered: Feb, 2006

Re: The "maybe" defense?

Posted: Jan 3, 2007 11:57 AM

> I explained my statement after you questioned it and you
> basically called me a liar. How am I supposed to take
> that?

Please provide a quote where I claimed you were a liar. BTW what does that have to do with your "emotional attachement" claim anyway. Are you now saying that you mentioned an "emotional attachement" as some kind of paypack for allegedly calling you a liar? I doubt it, but I can't think of any link between the two. Perhaps your just trying to deflect the critism back on me.

> I didn't make a critical statement about it. I questioned
> whether the tool I was using gave me a different
> perception. I wish I hadn't mentioned C++. It's really
> not important to my point. I thought it was pretty much
> accepted that threading in C++ is more difficult than in
> Java. Mea culpa.

Wow, you say that the claim wasn't important and then you almost repeat the claim (the original claim said nothing about Java). I feel like I'm going in circles. I'm temped to ask why you believe that it is "pretty much
accepted that threading in C++ is more difficult than in
Java.", but I don't really want to go through it all again.

> I don't think C++ is a good language choice for most
> development projects. I'm not alone in this and I'm not
> trying to hide that. It's not an article of personal
> faith. It's based on my personal experiences, the
> experiences of others and some rational thought. If you
> want to think I have some sort of personal vendetta
> against C++, fine, be my guest. It's absurd but I really
> don't care at this point.

I don't care whether it's personal or not, I just evalutate your claims.

> Exactly. Basically nothing.
If you really beleive it's not possible to learn anything about your thoughts by reading your posts, than you are certainly wasting a lot of time here, but it's your time to waste.

James Watson

Posts: 2024
Nickname: watson
Registered: Sep, 2005

Re: Where is Software Development Heading in 2007?

Posted: Jan 3, 2007 12:04 PM

> You say people are using increasing processor speeds and
> that it's not viable for much longer. So, it's not
> multi-threading that is new, it's the problem that you
> can't make your programs faster anymore without using it.

Yes but it was a viable solution in the past and has been used in Java, for example, partly because the options for optimization were less. Clustering is basically multi-threading on a network level.

The point is that multi-threading is being used extensively, right now. I suppose for most desktop applications, it wasn't really viable and that is changing. I guess my mostly server-based experience is coloring my view on this.

> I'm just making conclusions based on your statements.

OK. I'm not really sure why you feel the need to point this out. Clearly I've got some sort of foul odor or something.

James Watson

Posts: 2024
Nickname: watson
Registered: Sep, 2005

Re: The "maybe" defense?

Posted: Jan 3, 2007 12:10 PM

> > I explained my statement after you questioned it and
> you
> > basically called me a liar. How am I supposed to take
> > that?
>
> Please provide a quote where I claimed you were a liar.

"I don't understand how the above can be reconciled with the below..."

"Be honest, James..."

"...you wouldn't let them weasel out of it by invoking the "maybe" defense."

"This is a common ploy..."

All of these imply that I am being dishonest on some level.

James Watson

Posts: 2024
Nickname: watson
Registered: Sep, 2005

Re: The "maybe" defense?

Posted: Jan 3, 2007 12:18 PM

> > I explained my statement after you questioned it and
> you
> > basically called me a liar. How am I supposed to take
> > that?
>
> Please provide a quote where I claimed you were a liar.
> BTW what does that have to do with your "emotional
> attachement" claim anyway. Are you now saying that you
> mentioned an "emotional attachement" as some kind of
> paypack for allegedly calling you a liar?

No, I sincerely believe your attacks are emotional and I'm not sure they have anything to do with my statement about C++.

> Wow, you say that the claim wasn't important and then you
> almost repeat the claim (the original claim said nothing
> about Java). I feel like I'm going in circles. I'm temped
> to ask why you believe that it is "pretty much
> accepted that threading in C++ is more difficult than in
> Java.", but I don't really want to go through it all
> again.

You just don't get it. I don't think you want to. I give up.

> > I don't think C++ is a good language choice for most
> > development projects. I'm not alone in this and I'm
> not
> > trying to hide that. It's not an article of personal
> > faith. It's based on my personal experiences, the
> > experiences of others and some rational thought. If
> you
> > want to think I have some sort of personal vendetta
> > against C++, fine, be my guest. It's absurd but I
> really
> > don't care at this point.
>
> I don't care whether it's personal or not, I just
> evalutate your claims.

You pointed out that I have argued against C++ in the past. What was your point? How does that relate to the current question? Are you making some sort of ad-hominem attack or not?

> > Exactly. Basically nothing.

> If you really beleive it's not possible to learn anything
> about your thoughts by reading your posts, than you are
> certainly wasting a lot of time here, but it's your time
> to waste.

OK. What am I thinking now, Kreskin?

Leandro Oliveira

Posts: 21
Nickname: lao
Registered: Aug, 2003

Re: Where is Software Development Heading in 2007?

Posted: Jan 3, 2007 12:32 PM

> The point is that multi-threading is being used
> extensively, right now. I suppose for most desktop
> applications, it wasn't really viable and that is
> changing. I guess my mostly server-based experience is
> coloring my view on this.
>
> > I'm just making conclusions based on your statements.
>
> OK. I'm not really sure why you feel the need to point
> this out. Clearly I've got some sort of foul odor or
> something.

Calm down, I mean you no offense. I made a post earlier in this thread about parallel programming being difficult and you said it was not. Then, I was just clarifying why I think it can be difficult, since you said it was simple. Without this discussion we had, my post would look dumb.
Now it's clear that our conclusions come from experience in different "worlds" (server, desktop).

Jeff Ratcliff

Posts: 242
Nickname: jr1
Registered: Feb, 2006

Re: The "maybe" defense?

Posted: Jan 3, 2007 12:44 PM

> All of these imply that I am being dishonest on some level.

The words you put in bold weren't complementary, I admit, but they are not equivalent to calling you a liar.

Jeff Ratcliff

Posts: 242
Nickname: jr1
Registered: Feb, 2006

Re: The "maybe" defense?

Posted: Jan 3, 2007 12:57 PM

"No, I sincerely believe your attacks are emotional and I'm not sure they have anything to do with my statement about C++."

Huh? How could my "attacks" as you describe them, not have something to do with your statement about C++? Even if they were emotionally based, they'd still be triggered by your statement, right?

"You pointed out that I have argued against C++ in the past. What was your point? How does that relate to the current question? Are you making some sort of ad-hominem attack or not?"

It's not an ad-homienm attack. Your posts have demonstrated that you're generally anti-C++ (it doesn't have to be based on emotion). So when you bring up C++ negatively in a context where it's not the topic of discussion, and you can't really support the statement with facts or arguments, one can reasonably conclude it's part of your anti-C++ agenda. Having an agenda isn't necessarily a character flaw, so pointing it out isn't attacking you, just your argument.

James Watson

Posts: 2024
Nickname: watson
Registered: Sep, 2005

Re: Where is Software Development Heading in 2007?

Posted: Jan 3, 2007 2:19 PM

> > The point is that multi-threading is being used
> > extensively, right now. I suppose for most desktop
> > applications, it wasn't really viable and that is
> > changing. I guess my mostly server-based experience is
> > coloring my view on this.
> >
> > > I'm just making conclusions based on your statements.
> >
> > OK. I'm not really sure why you feel the need to point
> > this out. Clearly I've got some sort of foul odor or
> > something.
>
> Calm down, I mean you no offense. I made a post earlier in
> this thread about parallel programming being difficult and
> you said it was not. Then, I was just clarifying why I
> think it can be difficult, since you said it was simple.
> Without this discussion we had, my post would look dumb.
> Now it's clear that our conclusions come from experience
> in different "worlds" (server, desktop).

I didn't mean to go after you. I'm just getting paranoid that I am not communicating properly. I feel as if everything I am writing is being interpreted the wrong way.

I don't want to make it sound like not understanding threading right-off-the-bat means you aren't smart or talented. It took me a while to really get it and I expect most people are the same way. But I see these really smart developers talking about it as if it were the most difficult thing ever. I'll wager you and most developers have tackled and mastered much more difficult concepts. And I don't claim to be a master of threading either. I know for a fact there are advanced threading issues that I have not mastered. If I ever have a need to do those things, I'll do the needful and work it out, or I'll find some library that does it for me. The best way I have found to deal with threading issues is to avoid them as much as possible with clean OO designs that minimize coupling in a way that also reduces or eliminates thread contention.

James Watson

Posts: 2024
Nickname: watson
Registered: Sep, 2005

Re: The "maybe" defense?

Posted: Jan 3, 2007 2:41 PM

> > All of these imply that I am being dishonest on some
> level.
>
> The words you put in bold weren't complementary, I admit,
> but they are not equivalent to calling you a liar.

I have said repeatedly that the point of my statement was not to 'get a dig in' on C++. You have badgered me on this claiming that I was. If that's not accusing me of being dishonest (a.k.a lying), what is it?

I totally get why you think I was going after C++ in my statement. I should have completely stated it differently. I'm not saying you are making stuff up. But after I attempted to clarify, you refused to believe me. I find that insulting. I think means you have a pretty low opinion of me.

I do have a vague notion that C++ is harder to multi-thread with based on foggy memories of doing it (albeit very long ago) and based on a very recent conversation with someone about C++ who also claimed threading was hard. I think this person is very smart and talented so it planted (or replanted perhaps) the notion that perhaps it's harder to thread in C++ than in Java. But I've never believed this to be a fact. I don't think this could ever be proven or disproven anyway. So when I made my statement, I was thinking that maybe Java makes it easier and that if many people found it more difficult in C++, it might explain why I keep seeing developers saying it's hard.

Like I said before, it was a hypothesis: A tentative proposal made to explain certain observations or facts that requires further investigation to be verified.

If you can't believe this, that's your prerogative but I'm not going to to tell you something different, no matter how think you know about how my mind works.

James Watson

Posts: 2024
Nickname: watson
Registered: Sep, 2005

Re: The "maybe" defense?

Posted: Jan 3, 2007 3:15 PM

> "No, I sincerely believe your attacks are emotional and
> I'm not sure they have anything to do with my statement
> about C++."
>
> Huh? How could my "attacks" as you describe them, not have
> something to do with your statement about C++? Even if
> they were emotionally based, they'd still be triggered by
> your statement, right?

'triggered by' and 'having to do with' are not the same thing. I'm basically saying I feel like you wanted to go after me and you found an excuse to do so. I could be wrong but that's how it looks from here. I don't know why else you would refuse to accept my explanation that I didn't mean it in the way you took it other than that you think I'm a liar.

> "You pointed out that I have argued against C++ in the
> past. What was your point? How does that relate to the
> current question? Are you making some sort of ad-hominem
> attack or not?"
>
> It's not an ad-homienm attack. Your posts have
> demonstrated that you're generally anti-C++ (it doesn't
> have to be based on emotion).

What exactly do you think ad-hominem means? ad-hominem doesn't isn't defined by emotions. It's an attack based on something other that the merits of the argument. In this case, my supposed history as an anti-C++ zealot. Whether I am anti-C++ or not has nothing to do with whether it's actually harder to thread in C++ or not and if you are using it as evidence that I meant to bash C++, well, then it's not ad-hominem, it's an argument for my lying.

> So when you bring up C++
> negatively in a context where it's not the topic of
> discussion, and you can't really support the statement
> with facts or arguments, one can reasonably conclude it's
> part of your anti-C++ agenda.

I don't have an anti-C++ agenda. That's the most ridiculous thing I've read in a while. C++ is actually on the short list of languages I would choose to use, it's just at the end for most purposes. It's been a very successful language and has served many people well. I just think it's past it's heyday and it's not really a good choice for most projects. If I need to write a system driver, I'll probably use C++ (badly, most likely.)

I don't see why you interpret this stance as having an anti-C++ agenda. I get so tired of language discussions becoming a bunch of rooting for your home team.

If I have an anti-language agenda, it's against COBOL. I think it sucks and I'm having to deal with it at my job. I'd very much prefer to work with C++ over COBOL or RPG. But I'm not going to push that agenda here or on any forum. That's stupid. I respect people here more than that.

> Having an agenda isn't
> necessarily a character flaw, so pointing it out isn't
> attacking you, just your argument.

I disagree. Arguing from an agenda means you are not having a real, honest discussion. And you thinking I have an agenda means you are not really taking me seriously. I really wonder where you have gotten this idea because I have never claimed C++ was 'bad', just that I don't think it's the best for most purposes. In any event, if you are going to continue to believe this, feel free to ignore my posts in the future because I really don't care for your characterization of me. If you are afraid I might poison the minds of the impressionable with my anti-C++ fascism, then you can just post that you think I have an agenda anytime you decide I am posting my propaganda. But why bother arguing with me? I'm just some partisan, right?

Jeff Ratcliff

Posts: 242
Nickname: jr1
Registered: Feb, 2006

Re: The "maybe" defense?

Posted: Jan 3, 2007 4:15 PM

James, I'm obviously annoying you greatly. Rather than carrying this discussion forward, I'm just going to shut up. I'm sorry I've offended you.

James Watson

Posts: 2024
Nickname: watson
Registered: Sep, 2005

Re: The "maybe" defense?

Posted: Jan 3, 2007 7:13 PM

> James, I'm obviously annoying you greatly. Rather than
> carrying this discussion forward, I'm just going to shut
> up. I'm sorry I've offended you.

I'm probably being too sensitive. I may be coming down with the flu and I'm kind of cross in general.

But I do really want to get the point across that I'm not just arguing an agenda. I'm can be convinced that I have been wrong. It's not easy but I wouldn't spend the time discussing these things with you if I wasn't considering your arguments seriously. For example, I was once sure dynamic languages (e.g. Python) were not for 'real' work. I've now started using Python on a regular basis because of discussions on forums like this and am looking into using it for production purposes. I fully believe that "foolish consistency is the hobgoblin of little minds."

I'm not here to 'win' the argument either. I feel that if I argue honestly I win, even if I'm wrong. I take this seriously. I'm human and pride can come into play but I do try hard to admit when I am wrong. I am opinionated but I try not to present my opinions as facts. I guess that's why this incident bothered me so much. I felt like you were questioning my integrity. In any event, I would like to put this behind us if you don't mind.

Flat View: This topic has 61 replies on 5 pages [ « | 1 2 3 4 5 | » ]

Previous Topic

Next Topic