VB6 is DEAD!

[3com]

I've actually been studying Lua a lot lately

u

Apparently I also have to study it.
So far I've been using FPS scripts in the game I'm developing, but I find that LUA is much more flexible than FPS script, and it is quite similiar to VB, in other words more familiar to me.

I'm on the 5th level and I need to animate the lava you can see in the image below.



The problem here is that you have to be an artist in 3d graphics (I'm not), or work with the models included in the program, or buy models (entities, segments, prefabs, etc. to third parties.

Personally I prefer to create my own models, this introduces in the game a little about yourself, say it does more original.

Working with particles (rain, fire, snow, fog etc), do not know yet if LUA can help me on this.

@ Shaggy Hiker
Unfortunatelly you're right. BTW Torque3d is open source, perhaps somebody can continue, if needed.

[szlamany]

I was in a way playing devils advocate with my postings...

I am an ISV. I should add that near the end of my VB6 decade I was commissioned to create a VB.Net/MS SQL in-house CRM. On top of this was an ASP.Net web portal for Furtune 500 companies to submits jobs to this customer of mine. Heavy ASP.Net - membership provider, grids with serious update panel stuff. Libraries for the perfect upload experience - pushed the limits of what was available for use in ASP.Net.

VB.Net in-house product came out really nice - I came out bloodied and battered from the ASP.Net side.

I've always been one to make a general maintenance type of program - one that makes screens and menus and grids from some kind of "control tables". Maximize re-use.

The VB.Net in-house product mentioned above - with table adapters and all that bloat and "specificity" was far from useful in some future venture of mine.

I had dozens of clients running a VB6 app that was common among them all - all using control tables in their respective DB's to show screens for maintenance and display and reporting.

With all that said I decided to pick and choose - the right tool for each job.

Front-end is web based - a perpetually loaded page that uses heavy AJAX with jQuery to create tabs for maintenance.

It's 10,500 lines of JavaScript.

It's 3500 lines of VB.Net web-methods running in an IIS environment.

I basically created my own ASP.Net so I could be totally in charge. I use open source tools like jQPlot to make beautiful graphs - the open source community for JavaScript and JQuery stuff is great.

I use MS for what is good for my needs. The MS SQL database is a requirement by my customers (health insurance, municipalities, labor unions - law firms). Using VB.Net for the backend - perfect for MS SQL and also means I can also effortlessly do things like create Excel spreadsheets - all needs of my customers. VB.Net is by far the only tool I could use for backend - imo...

The front-end - I did some searching - reading books. jQuery seems like a stable and safe path - it's proven great. The web-app is 2 years in production - I add new features all the time. All my clients benefit from it.

I see this as a long term - 10 year - investment.

[Shaggy Hiker]

Well since i started this Let me chime in...

So i have realized at some point I will have to re-write my vb6 apps that are making me a living right now.
The question is when should i do it. I have read many studies saying that a rewrite can be disasterous, here is one:

http://www.joelonsoftware.com/articl...000000069.html
http://onstartups.com/tabid/3339/bid...-Software.aspx

Ok, so here is my train of thought. I think im going to wait till vb6 breaks. In the mean time i will continue to see what develops with vbrichclient. I am also learning lazarus which i think is a viable vb6 replacement/ native/ cross platform.

Logically it seems that when vb6 apps break all others like lazarus/qt/etc that compile native will break also

In the mean time i am Going to clean up my code and remove as much ms dependencies as possible.

That sounds like the right choice. After all, the sunset is somewhere off over the horizon at this point. It will arrive for all of us, but why hasten towards it unless you need to?

So time will tell, those that are capable i encourage you to help the vbrichclient project, i believe its the only way foward for exusting v 6 code. And let me just say the creator of vbrichclient showed me how to do proper multi threading in vb6 and to this day i am amazed by it, because everyone said it was impossible. Its not. The automation software i wrote run 24/7 and never crashes running multiple threads with vb6 created app. You can hear my station at www.sadanceradio.com

I don't know who this 'everyone' happens to be. I knew it was possible when this own sites WokaWidget (a mod now wandered off) showed how to do it well over a decade back (it had to be at least 2003, or I wouldn't have seen it).

[Shaggy Hiker]

As for the point some others have brought up in the discussion, who are:
"... not selfemployed - and work for a company who dictates which tools they have to use" -
not much to comment about that - there's no larger lever to influence decisions, other than
trying to apply a bit of "soft-pressure" in a team-meeting or something, e.g. how great
your "latest experiment with Node.js" worked out - things like that.

Olaf

That happens all the time. One person likes Node.js, the next argues for Dojo, another for JQuery, another for a blended approach. I don't know that we ever have a meeting where there isn't some discussion as to where we are going and why. It doesn't really help all that much. Everybody needs to be aware of the alternatives, but one of the current issues is not that there is NO alternative, it's that there are too many. A robust number of initiatives MUST fail, but which ones will they be? At this point, the number of alternatives guarantees only that there will be a large number that fade away. They won't necessarily be the worst alternatives, either. History shows repeatedly that superior tech is not guaranteed to win out. So....in the current crop of interesting options, which one is going to survive for ten years?

I agree with your statement about MS needing to support tech longer, but underlying that is a desire that the world remain just a bit static. The last ten years has seen the death of the PDA, the rise of the smartphone, the rise of the tablet, and a steep decline in the PC. The last twenty years has seen all of that, plus the invention of the PDA, and the prevalance of the GUI (twenty years back predates Windows 95, before which there was only Apple with it's minority market share). Dilletante doesn't believe the dektop will survive the next decades, let alone twenty years, and I'm not sure I disagree (at least not the 20 year window). We simply don't know what will happen.

I don't know of any other company that has maintained the viability of software it no longer sells than MS. You don't get updates for the VB6 IDE, but it still runs, and will still run for years yet to come. Furthermore, the apps created by the language may run for longer still. Is there any other company that has that kind of longevity for software it no longer sells or supports? You can rule out Apple, as they have abandoned their legacy stuff a couple times. I think you can rule out pretty nearly every Open Source project, as they freely move on, though they do so in fits and starts.

I'd like to see MS support stuff longer, too, but I also don't know any company that does it better.

[NeedSomeAnswers]

As for the point some others have brought up in the discussion, who are:
"... not selfemployed - and work for a company who dictates which tools they have to use" -not much to comment about that

Was that me Olaf?

All that post for so little response :0(

was my post not controversial enough?

ok i will bring back one point here that surely we can disagree on :0)

I believe that it is no bad thing to advise a new dev to learn .Net, as there are tons of jobs as a .Net developer (on the basis that most new dev's if looking at a career in Software Dev will end up working for someone else rather than themselves.

[Shaggy Hiker]

It was a typo but it was an apt one, given the length of the thread, and I have to ask (because your puns so oft fly below the radar of my wit) was this deliberate:-
? Because if it was then you, sir, deserve cake.

I'll take that cake. I started out with comparing it to cheese, and was trying to think of the stinkiest cheese I knew, but while I was thinking of that, I realized that wine was also aged...and had more puntential than cheese.

[Schmidt]

Just because a handful of people can knock out fantastic apps in VB6 does not mean VB6 is a viable development tool.

Too one-sided again for my taste - because you could write-up the same thing for any language
or platform, as e.g.:
"Just because a handful of people can knock out fantastic apps in C# or VB.NET does not mean .NET
is a viable development tool."

Doesn't really make sense either way though...

There's always those who are less advanced on a certain learning-curve - and there's others
who came very far with their tool of choice in a certain domain or two - then able to produce
"stunning stuff".

No matter if C, C++, Python, FreeBasic, VB or whatever language.

The purpose of languages and dev-tools didn't really change - one has to "transform an Input" -
delivering an "Output" - and for any language there's great libraries which can help with the
more complex things in this regard, shortening the needed code you have to write on
"ground-level".

So, aside from a given language you will have to learn surrounding (class- or flat-)libraries as well,
to become more productive - and if your problem lies e.g. in "bio-informatics", then it will also
help a great deal when you know a bit about the specific domain you write your "transformational
solution" for.

And as for the "modern features" of "modern languages" - they are over-hyped IMO, since for any
given "transformation-problem" in computer-science, there exist dozens of different approaches
(all well-working and near the optimum) to skin a particular cat with a given language (be it
"pure procedural ones" - or "pure OOPish ones" or "pure functional languages" or the new
"chimera-like languages" which try to offer a broader choice of possibly to apply solution-patterns.

Some languages and "patterns" a bit more suitable for certain problems - but then there's
other problems which can be solved a bit better with again different tools.

So, don't conclude from the *tool* someone is using, about "effectiveness".

There's often entire teams which center around "the one guy, who solves the really difficult things" -
when this guy is gone, then productivity can drop by a magnitude sometimes (no matter what
tool the team was using).

Olaf

[FunkyDexter]

The wrath of FunkyDexter has come

Not wrath so much as frustrated dissapointment. I'm no good at wrath.

Ok, so here is my train of thought. I think im going to wait till vb6 breaks. In the mean time i will continue to see what develops with vbrichclient. I am also learning lazarus which i think is a viable vb6 replacement/ native/ cross platform.

Logically it seems that when vb6 apps break all others like lazarus/qt/etc that compile native will break also

That sounds right on the money to me. I've always maintained that a premature rewrite can lead to disaster. Just make sure you've got a plan in place for when the time comes... and it sounds like you have.

one needs to be a masochist

Well, hey... it takes all sorts. Actually I'd revise my metaphor having seen some of the comments in this thread. I probably should have been talking about flash and ajax which are much more analogous to Silverlight's niche. I just didn't think of them at the time.

But the MS-announcement already broke him (financially), because the potential customer-base
was breaking away along with it...

That's the thing though, I don't think it would have broken him because I don't think potential clients would be walking away... yet. They will eventually as their fears that browsers will stop supporting Silverlight grow but I think that's a very gradual process. It's certainly taken a long time for clients to jump ship on VB6. Knowing an application was developed in Classic certainly can act as a negative factor when a client makes a purchasing decision but I'm not convinced it's an overriding one. They'll buy based on price, functionality and the perceived reliability of ongoing support (which will encompass the fact it was developed in VB6 but will also encompass factors like the longevity of the software vendor). I get the impression from the likes of you and Axisdj that you're concerned about the potential for lost custom (and fully understand that concern) but I don't get the impression you've actually experienced it yet on a significant scale - you'd know though so correct me if I'm wrong on that. So given that MS stopped pushing VB6 well over a decade ago and the world didn't end for you yet, I suspect Silverlight Guy's app has a similar life span ahead of it.

So, what did the guy do wrong exactly?
Was it his fault?

Nothing and no. He was just in the wrong place at the wrong time. Sadly that's life.

If that wasn't a "fool me once"-experience, then what is?

That's the thing. I don't think it was a "Fool me once" experience. At worst it was a "Got unlucky once" experience. I've worked in a variety of languages over the years and seen more than a few come and go. I thought Java had had a pretty good continuum since I used it back in the J2 days until I dug out an old project about a year ago and fired it up. The download sites for half the libraries I'd used have magically disapeared so getting that running again would have been a nightmare (though I must admit I gave up with really trying). I can't blame Sun because the libraries were third party... but that's the ecosystem when working in Java. My point is that deciding not to go near MS again for fear of getting burned is the wrong reason to make that choice because you stand a fair chance of getting burned wherever you go.

Switching over to .NET is just *one-of-many* alternatives for VB6-developers

Now there's something we can definitely agree on. I do think it's a fairly natural path to follow due to a reasonably familiar syntax and the existence of the old VB libraries in the .Net framework but I'd no more advocate the automatic selection of .Net any more than I'd advocate the automatic elimination of it. If there's one thing I'm dogmatic about it's that dogma's always a bad thing (there are probably exceptions to that though).

As for the point some others have brought up in the discussion, who are:
"... not selfemployed - and work for a company who dictates which tools they have to use" -
not much to comment about that

Yeah, I've thought a few times when I've seen that argument expressed that, while it may provide a good justification for why they used to .Net it doesn't provide much of an argument as to why you should.

[KGComputers]

Attachment 115517

Tired from work, did some searching on Hosting VB 6.0 forms in VB.NET. And here's one interesting result..

[NeedSomeAnswers]

As for the point some others have brought up in the discussion, who are:
"... not selfemployed - and work for a company who dictates which tools they have to use" -
not much to comment about that

Yeah, I've thought a few times when I've seen that argument expressed that, while it may provide a good justification for why they used to .Net it doesn't provide much of an argument as to why you should.

My point was not that my company dictates my tools but that it insulates me against the vagaries of the market and languages dying out. My company takes the hit not me.

I was trying to explain to Olaf why there will be many people who view .net as a good tool ,and i wanted to show some of the arguments against that he puts forward are just are not things which affect me or many others.

As for why people should use .Net, well not everyone should i do agree you should be looking for the tools for the required problem, but i like using .net because;

- I am more productive in it, as in i can code the same app faster in .Net than i can in VB6, and in part that is due to some of the elements of the .Net language that VB6 doesn't have, e.g. Lists and Lambda expressions e.t.c. (yes i know Olaf this is Subjective and you would probably say the other way round but i find it true for me.)

- The IDE is far superior

- You can make nicer looking apps easier.

- Also i can make Desktop apps and Web apps and Mobile apps in the same language

[SJWhiteley]

Too one-sided again for my taste - because you could write-up the same thing for any language
or platform, as e.g.:
"Just because a handful of people can knock out fantastic apps in C# or VB.NET does not mean .NET
is a viable development tool."

Doesn't really make sense either way though...

...

Exactly my point. But that's the argument FOR keeping VB6 by VB6ers.

But the reality is that huge swaths of people are creating great apps using .NET, rapidly, consistently and of high quality. The corollary being that there are a handful of people knocking out really sh*ty apps in .NET; doesn't make .NET any less viable.

As you state:

The purpose of languages and dev-tools didn't really change - one has to "transform an Input" -
delivering an "Output" - and for any language there's great libraries which can help with the
more complex things in this regard, shortening the needed code you have to write on
"ground-level".

.NET has made it easier to achieve this. I know you don't believe it, but it is true. In addition, .NET makes it viable for 'toy' developers, functional programmers, OOP, MVC, etc. to play in one playground.

Regardless, some tools are more effective than others, and lend them to greater efficiency and effectiveness. As I've said before, programmers are tools, who use tools to create tools for others to use.

In all, you have posted an argument for using .NET and abandoning VB6 to support of legacy code. Horses have been put out to pasture; blacksmiths need not apply to the DOT.

[SJWhiteley]

...

Sorry to say, but many of you .NETers are doing exactly as the members of
that third group (bashing and ridiculing the innocent).

No, apart from a few instances, not at all. What is being bashed is this steadfast and bullheaded approach to development. As a developer with more than a few years under their belt, they should realize that the end will come to their comfort zone, either sooner or later.

This was the reason for the great swath of unemployed tech workers (the tech bubble); for example, highly paid network engineers finding themselves unemployed with 2.5 kids and a $300k mortgage. Sad, but when grandma can go to Best Buy to purchase a router and WAP, your job is pretty much redundant.

VB6 is and was a great tool. Just like the Model T (I work on the Henry Ford plant, so have to get a Ford reference in). It's a marvelous, simple, tool. But not today.

Keep your skills sharp by developing in both mature and cutting edge technology.

Personally, at this point, I don't list any development tools (programming or PLCs) on my resume because it is meaningless (young developers, of course, must pad as much as they can).

[Shaggy Hiker]

.NET has made it easier to achieve this. I know you don't believe it, but it is true. In addition, .NET makes it viable for 'toy' developers, functional programmers, OOP, MVC, etc. to play in one playground.

I rather like that statement, though possibly not for the reason you meant. I enjoy coding, so it really is a playground. Or, perhaps it could be likened to some sandbox where there are all kinds of shovels, buckets, toy trucks, bricks, pipes, and what not. You don't have to use all the toys to create a castle, but there are plenty to play with if you want. From that perspective, .NET is a sandbox that has lots of toys. That doesn't mean that you can create more interesting designs, since that's up to the mind of the user, but if your primary interest is in playing with the toys....Net has a LOT of them.

As I've said before, programmers are tools,

Hey...what you talkin' bout!?!?!

[Shaggy Hiker]

VB6 is and was a great tool. Just like the Model T (I work on the Henry Ford plant, so have to get a Ford reference in).

There are plenty of people who still maintain Model T's to this day. They generally are pretty passionate about them. It's just human nature.

[Schmidt]

Was that me Olaf?

All that post for so little response :0(

Sorry - had to pick a bit, a little "stressed" to keep up with all the replies.

was my post not controversial enough?

A bit - e.g. your story about the company who was abandoning a promising
project due to MS cancelling Silverlight, pretty much proved my point I thought.

That the developers "learned something" whilst developing against Silverlight,
is not really surprising - you always learn something that's useful for your next
project, which deepens your understanding for roughly similar problems.

What I was trying to point out was simply, that the ones who decide to invest money
(and in your case this were not the developers who were hired for the Silverlight-project),
were not able to get a decent ROI, because of the too short "deprecation-period" for
the MS-tool they were choosing.

IMO that answers also the thing you brought in a later reply - you were protected
"by the company" who took the risk... nonetheless the company lost money, whilst not
really being responsible for the cancelling of the project (the writing-off of the monetary
efforts which were going into the project till that point in time).
MS in this case acting irresponsible due to much too short deprecation-periods.

ok i will bring back one point here that surely we can disagree on :0)

I believe that it is no bad thing to advise a new dev to learn .Net, as there are tons of jobs as a .Net developer
(on the basis that most new dev's if looking at a career in Software Dev will end up working for someone else
rather than themselves.

Sure - but still there's a whole lot of smaller companies (offering jobs for only 3-5 developers),
who are not *that* well off with regards to their cash-reserves, to be able to buffer more than
one "mistake" (due to wrongly believing in the stability of a highly advertised new tech).

Olaf

[techgnome]

So i have realized at some point I will have to re-write my vb6 apps that are making me a living right now.
The question is when should i do it. I have read many studies saying that a rewrite can be disasterous, here is one:

http://www.joelonsoftware.com/articl...000000069.html
http://onstartups.com/tabid/3339/bid...-Software.aspx

Ok, so here is my train of thought. I think im going to wait till vb6 breaks. In the mean time i will continue to see what develops with vbrichclient. I am also learning lazarus which i think is a viable vb6 replacement/ native/ cross platform.

That sounds like the right choice. After all, the sunset is somewhere off over the horizon at this point. It will arrive for all of us, but why hasten towards it unless you need to?

That sounds right on the money to me. I've always maintained that a premature rewrite can lead to disaster. Just make sure you've got a plan in place for when the time comes... and it sounds like you have.

Right... you shouldn't re-write your software when updating it... you should be re-designing it... clear up all those old bugs in the system and introduce all new ones. (I say that somewhat tongue in cheek).

But waiting until VB6 breaks to start re-development is just as bad of an idea as prematurely rewriting. Because if it's broken for you, how much longer will it be before it starts breaking for your clients? That's is huge risk. At best, the month that VB6 breaks, you should have something else ready, or close to ready, to roll out. VB6 has another 10 years left on it... assuming for argument sake, you determine it will take 2 years to re-design and develop your app in what ever new technology/language you use. That gives you 8 years to keep rolling. Meanwhile, sometime before that, you'll need to make that decision what road you do want to go down. Today it may be Lazarus... two years from now, that could change again... That's part of the fun and joy of this line of work, things can change just like that. It's almost as fickle as the stock market. Who knows, maybe in 5 years we'll all be writing Lua code.

-tg

[Schmidt]

[the "Silverlight-guy"]

That's the thing though, I don't think it would have broken him because I don't think potential clients would be walking away... yet.

In my example it did happen - have got "anxious questions" in this regard from
my own customers (luckily being able to appease them) - also heard from other
(often selfemployed) VB6-developers, that they experienced the same thing.

So, already the announcement (even a "rumor" would suffice for some guys) has the potential,
to cut deeply into your ROI.

Aside from that, NSA already wrote:
"A team at my last work did exactly what you describe, they developed a whole application in SilverLight, then just
as they were due to complete, the company pulled the project for fears about the future direction of Silverlight."

That's exactly the thing I was trying to point out - you risk loosing a lot of invested money,
when you don't take good care from which source you choose what tech, to saddle your solution on.

And MS is one of the worst in this regard (since the VB6-era).

It's certainly taken a long time for clients to jump ship on VB6.

No, as pointed out above just now, that's not true - I've experienced these customer-concerns
myself already around 2002/2003 or so, when still developing LOB-Apps.

So given that MS stopped pushing VB6 well over a decade ago and the world didn't end for you yet, I suspect Silverlight Guy's app has a similar life span ahead of it.

As said, I encountered the problem early - and it could have become a problem for "us" -
but the customer who expressed these concerns then decided to switch over to a policy of
"Java-generally for all InHouse-Apps" - along with hiring their own Java-developers who
reimplemented the App (with some help from us, so we got at least some last "consulting-
money" from their "Java-throughout"-decision).

Our small company at this point already switching over to more hardware-driven products,
where this problem didn't occur again - but I heard from other VB6-devs that they had similar
"anxious customers"... and that surely cut into their ROI (especially winning new customers
get's a lot tougher, when this kind of anxiety due to allegedly "old tech" is making the rounds).

I don't think it was a "Fool me once" experience.

Well, for him it surely was - not sure if you ever was selfemployed - but believe me,
you'll see things entirely different, when it was *your* "half a million" which went "down the loo"...

I've worked in a variety of languages over the years and seen more than a few come and go.

Oh really?
May I ask which were the ones with your "and go"-attribute?

I thought Java had had a pretty good continuum since I used it back in the J2 days until I dug out an old project about a year ago and fired it up. The download sites for half the libraries I'd used have magically disapeared ...

Sorry to interrupt - but Java is well-known for binary and source backward-compatibility
of newer versions, which are able to run older Bytecode, but also to compile sources from
older versions without larger problems (there was only [Enum] which became a reserved
Keyword - so you had to rename identifiers with that name appropriately in the older sources).

Bad example, really bad - if you dig out an age-old project - not having saved the dependencies
locally along with it, then it's your fault, not Javas.

My point is that deciding not to go near MS again for fear of getting burned is the wrong reason to make that choice because you stand a fair chance of getting burned wherever you go.

No, that's not true - if you choose C++ and QT you will not get burned.
If you choose C and GTK+ you will not get burned.
If you choose Java you will not get burned.
If you choose Python you will not get burned.
If you choose Javascript you will not get burned.

There's many languages which are standardized, careful changes made "by committee"
(or by broad agreement in the community, if it's an OpenSource-driven project).

Olaf

[Schmidt]

No, apart from a few instances, not at all. What is being bashed is this steadfast and bullheaded approach to development.

Sorry can't see where my approach to choose VB6 as my current tool for Win-Desktop-Apps is
"bullheaded".

It does this Job just fine currently (and also in the next decade).

As a developer with more than a few years under their belt, they should realize that the end will come to their comfort zone, either sooner or later.

That's true for you .NETers as well - perhaps at the same time when Windows-Forms-based Apps
go out of scope, VB6-developers will have to react as well.

VB6 is and was a great tool. ... It's a marvelous, simple, tool. But not today.

It's a bit self-contradictory the above sentence - but perhaps you meant:
A great simple tool, following KISS - and able to do "marvelous things" to this day.

Keep your skills sharp by developing in both mature and cutting edge technology.

Oh, I do - I do ... there's not only MS which provides developers with "mature tech" believe me.

Olaf

[Schmidt]

You seemed to have missed one of my post. I'll redirect you to it:
"...Your Silverlight 'guy' just simply made the wrong decision..."

Nah, you're out of the loop - FunkyDexter already agreed, that "the guy did nothing wrong".

However you should translate that to: maybe I should make more educated decisions in the future and wait for a language/platform/etc. to gain a little more stability first.

Well, exactly that's what I'm doing the whole time with .NET - it just doesn't seem to
gain any points, when I compare its "stability" with other alternatives.

Olaf

[Schmidt]

But the reality is that huge swaths of people are creating great apps using .NET, rapidly, consistently and of high quality.

Same holds true for those who are creating great apps using VB6, rapidly, consistently and of high quality.
Huge swaths of people who know what they're doing?
Same percentage in both camps.

The corollary being that there are a handful of people knocking out really sh*ty apps in .NET; doesn't make .NET any less viable.

Same sentence holds true for VB6:
"...that there are a handful of people knocking out really sh*ty apps in VB6; doesn't make VB6 any less viable.

... .NET has made it easier to achieve this. I know you don't believe it, but it is true.

It's not true of course, I know you don't believe me, but - well, believe me...

Writing things like that, is not really sufficient in a good argumentation, is what I'm trying to say I think.

As I've said before, programmers are tools, who use tools to create tools for others to use.
In all, you have posted an argument for using .NET and abandoning VB6 to support of legacy code.

Erm, whut?

Really, the above doesn't make much sense to me.

Olaf

[Schmidt]

I'm not against you not going to .Net if you have valid points because that is your decision, but how does .Net not gain any points whenever it comes to stability? Considering this:

1) It has been out since February 13th, 2002
2) It has been updated 9 times since it's release, 4 of those are major updates
3) It's future is not in jeopardy

What constitutes it not being stable?

dilettante already addressed (with a link to an article) your point 3)

A framework needs stability (as in "no changes, no additions") for some years,
to become a helpful tool - and the constant additions due to overboarding "featuritis"
don't really help - it's got overly-large, overly-complex and over-weight - it's instable
"by definition" (in engineering there's a reason why you develop parts well-isolated,
building separate "blocks" (libs) - after pre-defining stable interfaces).

We already had that with the Win32-API (with *.h files for interfaces) or COM-libs
(with *.tlb files which defined the interfaces) - parts then easily updateable in a more or
less *isolated* fashion - not that whole mega-blob of interdependent Mini-Classes,
the .NET-framework evolved into. From the viewpoint of an engineer, such a blob
is horrible design. They already did better - and recognized that now with more
focus on native compilation and back-paddling to the stable COM-ABI.

You can get a small taste of what I mean in the link below - especially in the comments which follow.
http://devproconnections.com/net-fra...problems-ahead

MS is currently acting *entirely* confused, nobody there seeming to know,
"what needs to be done next" - but everybody of course working full-steam "on it".

That alone is making me wait a bit longer, until they get their feet under them again.

If you are more sure, what I should currently "move on to" - please feel free to
make recommendations aside from moving to .NET.

Is it C++ and Metro/WinRT perhaps ... hmm - also here, nobody at the side of MS
delivers anything but nebulous "everything will be good" statements.

Is it .NET-native (vNext) perhaps?
Much too early to make good guesses, what these new activities finally will churn out.

Of course VB6-developers will just sit there, because they don't know anything else.
Thank god we have you, who constantly remind us, that the only tool *you* know
is the way to go.


Olaf

[dilettante]

I think a good case can be made to move to Java.

Despite attempts here that claim it fails to live up to "write once run anywhere" this is largely true, far more so than for .Net or the dead half-baked Mono. This directly gives you access to Windows, Mac, Linux, and even many Unix platforms as well as mainframes, and it is the native language (with some changes) of Android as well.

While straight Java carries many of the same issues .Net does (no surprise, since .Net is a distant relative of Java's infrastructure), at least it isn't a single-platform dead end.

There are also tools to help ease the transition from Basic to Java. Basic4Android is growing in popularity, along with its free cousin B4J if not the old Jabaco development system.

And for cases where you can accept the penalty of interpreted, sometimes JITted here and there p-code you can always use an obfuscator. When you can't you can use AOT compilers in many environments, or a Jar-to-Exe compiler though this has to statically link a ton of code to implement things the VM normally does. Sort of like the bloat you get using the Xamarin tools to target supported mobile OSs as opposed to the wildly unpopular WinPhone OSs, each of which has enormous compatibility breaks with the others.

[FunkyDexter]

Sorry - had to pick a bit, a little "stressed" to keep up with all the replies

You do seem to have been left holding the fort single handed. Keep it up, though, this is an enjoyable debate now that some of the heat's gone out of it.

No, as pointed out above just now, that's not true - I've experienced these customer-concerns
myself already around 2002/2003 or so, when still developing LOB-Apps.

Then here's my question: why would you continue to advocate it for new development? I know you like it as a language to work in and I can understand why but if it's going to lose you customers then using it for new development is crazy.

I can understand some of the frustration of those who wrote an app in VB6 in 2000. That could, potentially, create real problems for them over the next few years. I could understand if they went out of business in the next few years. I can't understand how someone can reach a point 14 years later where they're still in business and still haven't dealt with those problems. And I really can't understand how someone could spend the next 14 years continuing to make the same mistake by continuing to write apps in VB6.

I honestly make no judgement about someone continuing to use VB6. If it's the right choice for you then it's the right choice for you. But I don't understand the mentality that says "I'm going to make a choice. It's the wrong choice because it will cost me money. I'm going to make it anyway. It's Microsoft's fault that it's costing me money."

FunkyDexter already agreed, that "the guy did nothing wrong".

I think you're poking fun at me here (in which case, carry on) but just in case you miss-understood, I meant he didn't do anything unreasonable or stupid. Nothing that should merit other developerson a forums several years later pointing at him and saying "you're an idiot, mate". On the other hand, I would say that he made what seemed like a reasonable choice at the time that later turned out to be a bad one through no fault of his own.

[NeedSomeAnswers]

Sorry - had to pick a bit, a little "stressed" to keep up with all the replies.

A bit - e.g. your story about the company who was abandoning a promising
project due to MS cancelling Silverlight, pretty much proved my point I thought.

That the developers "learned something" whilst developing against Silverlight,
is not really surprising - you always learn something that's useful for your next
project, which deepens your understanding for roughly similar problems.

What I was trying to point out was simply, that the ones who decide to invest money
(and in your case this were not the developers who were hired for the Silverlight-project),
were not able to get a decent ROI, because of the too short "deprecation-period" for
the MS-tool they were choosing.

IMO that answers also the thing you brought in a later reply - you were protected
"by the company" who took the risk... nonetheless the company lost money, whilst not
really being responsible for the cancelling of the project (the writing-off of the monetary
efforts which were going into the project till that point in time).
MS in this case acting irresponsible due to much too short deprecation-periods.

Personally i think its a bit of both, had our company done more analysis of the product i believe they would never have chosen Silverlight, I certainly wouldn't have is it was my choice, I used Adobe Flex for a while around 6-7 years ago (basically the adobe version of Silverlight, which Silverlight was in direct competition with which compiled into Flash) and it was clear to me even then that these languages (both Flex & Silverlight) would struggle for adoption.

The pattern in the web development world was moving away from propriety run times in the web, and with the release of HTML5 it the likes of Node and Less it completely killed any need for these languages.

My company i feel jumped the gun a little using a language that had gained very little traction in the market but i can see why some would feel burned if they did develop in it. (i was working for a pretty large company and your right they could afford to take the hit where others couldn't)

.Net is not a tool for everyone, and while i can see your problems with it, what i don't understand is yours and others hostility towards .Net. MS Yes i can see why some would be upset with there decision to end vb6 development, but .Net did not kill VB6 MS did.

Sure - but still there's a whole lot of smaller companies (offering jobs for only 3-5 developers),
who are not *that* well off with regards to their cash-reserves, to be able to buffer more than
one "mistake" (due to wrongly believing in the stability of a highly advertised new tech).

Olaf

As for .Net being a risk as a language choice .Net has plenty of Traction to the point where often it is the expected tool by our customers.

I have just moved jobs 3 months ago, to a smaller company of around 30-40 employees (from one of around 2000 employees) and 5 developers, and we extensively use .Net. Part of this is dictated by our Market, we sell software to Lawyers and Barristers and Local Authorities, and they are fairly MS Biased.

.NET has been a money earner for both my last company and my new one and for us has been a fairly safe bet. It is still seen by our customers as a safe platform

[NeedSomeAnswers]

I think a good case can be made to move to Java.

No thanks!

[NeedSomeAnswers]

[QUOTE]I think a good case can be made to move to Java.[/QUOTE

double post

[Schmidt]

You do seem to have been left holding the fort single handed. Keep it up, though, this is an enjoyable debate now that some of the heat's gone out of it.

Wasn't me who brought the heat into the debate - and I know for example Carlos well enough,
that I can ensure you, he's not somebody who insults others deliberately and unprovoced -
I've never seen him troll a forum or newsgroup - and we know each other for years now.

The person who almost always enflames things, is now gone for a while - and we enjoy the
immediate effects of that decision apparently. No "He-mad?" picture-storm, no other cheap
and easy to accomplish "no-real-arguments"-replies, cobbled together in a minute - and only
thought to ridicule, enrage and flood the opponent - not exactly fair if you ask me, since when
you sit at the receiving end of such nonsense, you cannot really do anything about it without
getting insulting in your attempts to point out this idiotic behaviour explicitely to those, who
apparently "have that much fun with it" (thereby forcing you to offer even more of an attack-
surface, because then the next replies, of the "stop-whining"-kind will kick in - all with much
glee on the side of the kids of course, who apparently think they just won an argument... sigh.

Would like a more immediate reaction by the moderators, when such discussion-disrupting
picture-nonsense starts the next time (when it is solely thought for ridiculing the differing
opinion of other members).

I know we're in Chit/Chat here and the "come-on, where's the fun"-thing is kind of expected -
but if that fun deteriorates into the "pointing fingers, muhaha, look at these idiots"-kind of fun,
as it did in most of the other threads about this topic, then the "other side" can only come up
with extremely frustrated replies, which then makes it look as if the ones who started the mess,
"were right all the time".


[Customer-hesitation, in choosing products, based on "labeled-deprecated" technologies]

Then here's my question: why would you continue to advocate it for new development? I know you like it as a language to work in and I can understand why but if it's going to lose you customers then using it for new development is crazy.

Because we are all in the same boat apparently?
My Silverlight-guy example was trying to point out, that this can happen even when
you choose so called: "bleeding edge tech".

Several others were pointing out to me, that "the risks lurk everywhere in our profession" -
and that there is no real "safe bet" to make anymore.

My point being, with a different behaviour on the side of MS, with more prolonged release-
cycles, longer periods until tool-deprecation, the "bets on MS-tools" would get a lot more
safer until some ROI comes in.

There's some languages and platforms which offer (much) more long-term-stability.
I've pointed them out already (those which are standardized or OpenSource-driven).

But i consider .NET as not being among them - I truly don't.

When the next VB6-Newbie comes along in our forum, asking if VB6 is a good idea to
start with - I can recommend VB6 without hesitation, because it just *is* a great KISS-
tool with a very low entry-point (learning-curve-wise) - not much struggling with
understanding the environment, or pitfalls of the language first - easy to get to
the meat of the current learning-task very fast and undistracted by other things.

When I'm being asked directly, what tool to use for professional development,
and the one who asks already has VB6-experience (and in most cases it *is* for
the Desktop), then my answer will also be: continue with VB6 - no reason to switch,
because you can get biten equally easy when you choose .NET.

When a *common* question is in the room - "what's the best tool to use today" -
then I only argue *against* those who recommend .NET - because that's not the
recommendation I would make - I would recommend C++,HTML/JS/jQuery instead -
no real mistake possible with those choices - many of you misundertand my counter-
arguments against .NET-recommendations as "recommendation of VB6 instead".

Please read older replies of mine more carefully in this regard.

[still the Customer-hesitation-topic]

I can't understand how someone can reach a point 14 years later where they're still in
business and still haven't dealt with those problems. And I really can't understand how
someone could spend the next 14 years continuing to make the same mistake by
continuing to write apps in VB6.

That's because there is just no alternative currently where "to move on to" with a
large VB6-based Desktop-App.

Even *if* you rewrite everything you already have running (well-tested, with hundreds
of small bugs eliminated over the last 10 years), with e.g. .NET-Winforms (which most
of you currently work with) the VB6-developer will accomplish exactly nothing in terms
of greater "long-term-stability" for the next 20 years, because WinForms will go out of
scope (with a high probability) at the same time as VB6.

Porting a large "I made a living from it" VB6-LOB-App with hundreds of Forms, hundreds
of Modules and hundreds of Classes to .NET-Winforms just doesn't worth the efforts.

Now, what other alternative techs does the VB6-guy have, to port his App over to,
when it's not .NET-WinForms?

Shall he use .NET+WPF? - seems dead to me - and the posting count in the WPF-forum
here tells the same story...

Shall he re-write for Metro/WinRT/XAML - not sure if that's a good idea at this point
in time either.

That's what I mean with "stability" when we talk about .NET or other "bleeding edge"
tech from MS - there is no clear roadmap - no horse to place a relative safe bet on
(aside from moving to C++ which at least as a language is more stable - but there
you currently also have to wait, what GUI-toolkit of MS will make it in the next years,
not really sure if XAML will be adopted by "everybody" wholeheartedly - maybe they
paddle-back again also on that front, with a renewed incarnation of MFC-based GUIs).

So the only thing remaining for the VB6-guy, in case he wants to start *today*
(no need to yet, just saying) - is to rewrite "for the Web" - but now in this field
(albeit ASP.NET being a very good tool) - there's very hard competition and many
other "nice sisters" to choose from. And even good old classic *.asp is not that bad
either on the serverside (compatible with VB6-source, when you leave out the Type-
specifiers - and serverside code tends to be small compared to the clientside JS-
framework-code you have to write for a modern looking Web-App which offers
"Desktop-feeling").

I honestly make no judgement about someone continuing to use VB6.
If it's the right choice for you then it's the right choice for you.
But I don't understand the mentality that says "I'm going to make a choice.
It's the wrong choice because it will cost me money.
I'm going to make it anyway. It's Microsoft's fault that it's costing me money."

Although being long-winded already, let me explain exactly that again (because
you still have it all backwards):

You have a large VB6-App today (hundreds of Forms and Classes), are selfemployed and still around.

- You invested perhaps 0.5Mio into it until 2003 - but already were in the phase of an "OK-ROI" (looking good for the future)
- Your ROI was cut by an MS-deprecation-announcment - now you encounter difficulties to win new (more "anxious") customers
- Still it is high enough "to survive" - but your cash-resources to "move over to something else" simply aren't there (due to the announcement)
- You keep going, satisfying at least your existing customers, weeding out bugs - polishing things...
- You buy a 350$ Unicode-Controls-Suite, which helps a great deal with a much nicer look - and it won you two new customers "overseas"
- We have 2007 - your codebase and Form-Count is now twice as large - your accumulated time-investment into your software is >1Mio (15 man-years)
- Still the ROI is not as high as it could be - but getting better slowly, because you *do* win new customers due to "mouth propagation" and "being still there"-effects
- Since 2008 you encounter MS behaving more and more "erratically", so you start ruling out "moving to .NET" from your "to-potentially-port-to" list
- till now, watching in amazement - what MS apparently can get away with - potentially porting to a Web-App goes high up in your list

So that's basically the story of a small VB6-shop, still around (company not crashing).

MS hurt you with your ROI (which with a true upward-compatible native VB7 would have looked much better) -
but as it was, it didn't cause your insolvency - though it reduced the income of the small company to a level,
which didn't allow a complete re-write (you had to say good bye to 3 of your 5 developers you started with,
but the remaining two were busy enough already with maintenance and implementing new feature-requests,
keeping things going as they were).

[Silverlight-guy]

On the other hand, I would say that he made what seemed like a reasonable choice at the time that later turned out to be a bad one through no fault of his own.

Thanks, for the recognition again - and that we've established at least that - so my efforts were not entirely wasted...

Olaf

[SJWhiteley]

There are plenty of people who still maintain Model T's to this day. They generally are pretty passionate about them. It's just human nature.

The parallels are uncanny

[Schmidt]

Personally i think its a bit of both, had our company done more analysis of the product
i believe they would never have chosen Silverlight,

Nah, "our silverlight company" was highly motivated after consuming all the advertising -
Mr. Silverlight-guy also hanging around here in the forum, feeling much encouraged
by all the positive things he read here, any "MS-bleeding-edge-tech" recommended as
"always a good idea to move on to". He perhaps got the impression, that any choice was a
good one, as long as it was *not VB6* (check) - but out of the .NET-basket (double check).

I certainly wouldn't have is it was my choice, I used Adobe Flex for a while around 6-7 years ago

Hah, traitor - how dare you... <g>

The pattern in the web development world was moving away from propriety run times in the web,
and with the release of HTML5 it the likes of Node and Less it completely killed any need for these languages.

Well, at the point in time, this trend was not yet obvious to our Silverlight-guy -
HTML5 still in the "cooking"-phase - and JS still that slow, that Silverlight-demos
were able to run circles around it performance-wise, at appropriate MS-Dev-Conferences.

.Net is not a tool for everyone, and while i can see your problems with it,
what i don't understand is yours and others hostility towards .Net.

At least from me, you will see no real hostility - IMO you just get "unwelcome information"
into the wrong throat - arguments which you feel "tainting your own perception of things"
(guy didn't ever use .NET - how dare he post such things about it).

Believe me, I *did* experiment with .NET - I even followed the implementation-progress
of Mono very tightly (in my Linux-phase), experimented with each new release of MonoDevelop
for a while, mostly playing with their GTK# GUI-implementations (their .NET-Windows-Forms-
implementation still not ready and compatible enough at that time).
Also tried out ASP.NET - but also Java-Server-Faces (JSF) which came out as a pendant to
the MS-efforts at roughly that time).
Also experimented with Vala (an already native compiling C# like language):
https://wiki.gnome.org/Projects/Vala
https://wiki.gnome.org/Projects/Vala...arpProgrammers

Just to point out a few things to back-up the: "guy knows what he's talking about"-thing.

MS Yes i can see why some would be upset with there decision to end vb6 development,
but .Net did not kill VB6 MS did.

That's my entire point here - to sharpen the recognition of what MS *continues* doing.
The *functionality* .NET provides is great - I ruled it out for other reasons.

On *only* the tech-side these are:
- the necessary "blob"-deployment (instead of only choosing the COMponents your App needs).
(the little "Mandelbrot-competition" we had, showed quite clearly what I mean... the .NET
implementation was developed against version 2.0 of the framework, to offer roughly the
same "reach" as the VB6-counterpart - many customers still use XP).

- the VM-inherent, larger startup-times

- the Garbage-Collector (it's really not that easy, to reliably wrap flat C-libs behind
.NET-classes, when the *order* of potentially necessary C-Handle-freeing becomes important)

- the potential of the GC, to "kick in" at unpleasant times

- the MSIL not being "true native code" (the "getting hot"-time for the Jitter I've already mentioned -
but in addition, there *is* a better protection of "Intellectual Property" with native compiled code)

Note, that I left out "general performance" - because that's roughly equal (as soon as the MSIL
was jitted) to the VC6-compiler which works underneath the VB6-native-compile-option...

But the other points above *are* facts, which I've always considered a big show-stopper
for me (and the reason why I took a good look at Vala - sadly their support for the Win-
platform is lacking - it's kind of difficult to "get it to work properly over here" (basically the same
difficulties you will encounter with Mono, when you want to integrate it as an alternative compile-
option into your VS-IDEs).

As for .Net being a risk as a language choice .Net has plenty of Traction to the point where often it is the expected tool by our customers.

A valid point - but not that much for the small VB6-shop, who made it "through the trouble" so far.
Please take a look at my reply to FunkyDexter, where I expand a bit more on this point.

Olaf

[Schmidt]

The parallels are uncanny

What's wrong with "polishing my Model T, keeping it in shape" I might ask?
There's a universe where I can order tuning-parts from - and my "lap-times"
still hold up to the competition - so, no reason to change it yet...

Olaf

[homer13j]

No thanks!

Reminds me of the old Mendhak joke:

Knock, knock.
Who's there?
(long pause)
Java.

[Shaggy Hiker]

Wasn't me who brought the heat into the debate - and I know for example Carlos well enough,
that I can ensure you, he's not somebody who insults others deliberately and unprovoced -
I've never seen him troll a forum or newsgroup - and we know each other for years now.

You know Carlos AND (you think probably) know Fatina?

Small world!

I was thinking about some of the stuff NSA said and realized that in my group (if you can call it that), there is almost a gleeful desire to use a tool that nobody else is using. So, if one person says Node.JS is the shizzle, and somebody else decides to try it, then the first person seems almost obligated to switch to Dojo. If that doesn't sound like a good idea....well, can't argue with that, but there are several other oddities about the situation, as well. Still, I think I envy you associating with a few other people who work in the same language....I think...except that I have no idea what that is like, so I really don't know.

Would like a more immediate reaction by the moderators, when such discussion-disrupting
picture-nonsense starts the next time (when it is solely thought for ridiculing the differing
opinion of other members).

Not going to happen in Chit-Chat. This has been a discussion of some substance, but there have been some in the past, equally contentious, but regarding more fundamental and controversial subjects (if anybody remembers the immigration thread...). When people feel strongly enough about a subject, you are going to get some odd responses. To some extent, that's got to ride in Chit-Chat or else the community will end up censoring too many views.

My point being, with a different behaviour on the side of MS, with more prolonged release-
cycles, longer periods until tool-deprecation, the "bets on MS-tools" would get a lot more
safer until some ROI comes in.

Silverlight is kind of the outlier here. When that first came out, I thought MS was nuts to do it, so I wasn't surprised that it failed. Other than that, the time-frame during which they support their technology and keep it viable is in the 10-20 year range (closer to 10 than 20, as noted earlier) that you were asking for. After all, VB6 apps still run fine on all Windows OS (except mobile, where they never ran) now some 12 years after the replacement came out. They kind of suggest that the apps will keep running for another decade, too (though there are other interpretations of that statement).

When I'm being asked directly, what tool to use for professional development,
and the one who asks already has VB6-experience (and in most cases it *is* for
the Desktop), then my answer will also be: continue with VB6 - no reason to switch,
because you can get biten equally easy when you choose .NET.

I would answer that you need to look at the job market. You will get higher pay, on average, for C# than for VB.NET, but there are lots of jobs in either. Are there more jobs there than in VB6? I would expect that to be the case (and I would expect that most announcements for VB are looking for .NET not VB6). After all, you can't buy VB6 from MS anymore, so the only people working in it are getting it from MSDN subscriptions, e-bay, other legacy sources, or piracy. I would be surprised if any company was willing to go with any of those routes.

When a *common* question is in the room - "what's the best tool to use today" -
then I only argue *against* those who recommend .NET - because that's not the
recommendation I would make - I would recommend C++,HTML/JS/jQuery instead -
no real mistake possible with those choices - many of you misundertand my counter-
arguments against .NET-recommendations as "recommendation of VB6 instead".

There are mistakes to be made with ALL of those choices. Three of those are web, while the fourth is C++. Is there a good RAD tool that works with C++? Even where I work, where, as I noted, people delight in using different things from each other, NOBODY works in C++, though several of us know it. It's a great language, and was one of the first I learned, but it's not the choice for creating LOB apps.

As for JS, it may well be around for the duration, but I don't believe it's an evolutionary end, as I've stated before. It's a really poor language in many regards. We can do much better, and I have no doubt that we will. Typescript is headed in that way, and I've heard of other languages that will compile to JS. Eventually, I expect JS to end up as the ASM of the web world...at best (or else I expect it to die entirely). JQuery will live and die on the same timeline, and HTML....isn't a language.

On the other hand, there are lots of jobs in web development, and JS, JQuery and HTML are all things that you should be pretty familiar with if that is the career you are seeking. They won't do a thing for you in desktop, though. So, it all depends on what career you are looking for, because a job that requires those items will not be advertised the same as a job for a desktop developer.

Even *if* you rewrite everything you already have running (well-tested, with hundreds
of small bugs eliminated over the last 10 years), with e.g. .NET-Winforms (which most
of you currently work with) the VB6-developer will accomplish exactly nothing in terms
of greater "long-term-stability" for the next 20 years, because WinForms will go out of
scope (with a high probability) at the same time as VB6.

That seems like a reasonable bet. There are some suggestions that it won't be the case, but I think it's still more likely than not.

Porting a large "I made a living from it" VB6-LOB-App with hundreds of Forms, hundreds
of Modules and hundreds of Classes to .NET-Winforms just doesn't worth the efforts.

That's situational. I'm actually in that boat. I have a large VB6 LOB app which HAS to go somewhere (the source code was lost in a robbery, of all things). However, that provides an opportunity to re-write and fix many of the design mistakes made. There are ALWAYS design improvements that could be made. I have never taken a VB6 app and ported it to .NET without greatly improving usability in one area or another, not because of differences in the language, but because I better understood the problem domain and had things that should be added.

So, porting a LOB app as a static port, where the features don't change, is a bad idea. Putting out a new version with several improvements...that's not so bad. I doubt there is a program that can't be improved. Last night I realized that a certain form that I wrote a year ago is a far-from-ideal means to handle certain data entry (it has only two numericupdown controls on it, and even that can get significant improvement). There's always improvements to be made on a large app. That's when you port, not when there are no changes worth making.

Shall he re-write for Metro/WinRT/XAML - not sure if that's a good idea at this point
in time either.

Unfortunately, it may be, but it's a non-starter if we are talking LOB.

So the only thing remaining for the VB6-guy, in case he wants to start *today*
(no need to yet, just saying) - is to rewrite "for the Web" - but now in this field
(albeit ASP.NET being a very good tool) - there's very hard competition and many
other "nice sisters" to choose from. And even good old classic *.asp is not that bad
either on the serverside (co
mpatible with VB6-source, when you leave out the Type-
specifiers - and serverside code tends to be small compared to the clientside JS-
framework-code you have to write for a modern looking Web-App which offers
"Desktop-feeling").

That web thing is just a fad. It'll never last.

More seriously, there are apps that the web doesn't handle at all, yet. I would assume that there are lots of VB6 LOB apps that are in that camp.

- Your ROI was cut by an MS-deprecation-announcment - now you encounter difficulties to win new (more "anxious") customers
- Still it is high enough "to survive" - but your cash-resources to "move over to something else" simply aren't there (due to the announcement)

That's not a programming issue, that's a marketing issue. Your customers are going to be concerned whether or not THEIR investment in your software is a good one, not whether you are using a certain technology. If they see you as a drunken meth-head, then it doesn't matter what your app does or what technology you are using. If you convince them that YOU are going to be around, and that this is something you are focused on, then they won't be asking what technology you use. In other words, if you are foolish enough to present your app in the most negative light possible, you get what you deserve. If you come in saying, "well, it does what you need now, but MS could change something and I'll just turn my back on you." then you are dead no matter what you are selling.

[Shaggy Hiker]

That's my entire point here - to sharpen the recognition of what MS *continues* doing.
The *functionality* .NET provides is great - I ruled it out for other reasons.

I'm not convinced that you did. On the other hand, I'm also not convinced that humans are intelligent, so I will just live with the doubt.

On *only* the tech-side these are:
- the necessary "blob"-deployment (instead of only choosing the COMponents your App needs).
(the little "Mandelbrot-competition" we had, showed quite clearly what I mean... the .NET
implementation was developed against version 2.0 of the framework, to offer roughly the
same "reach" as the VB6-counterpart - many customers still use XP).

That wasn't a good reason to target 2.0 of the framework. XP got 3.5 as an update, just as it got 2.0, except that 2.0 was a mandatory update while 3.5 was not. A better reason to target 2.0 was....why not? 3.5 offered many significant changes, but not ones that woul win speed races. The biggest additions were LINQ and lambdas, which make for VASTLY shorter code, but are slower than the old-fashioned methods. However, had you targeted 4.0 or 4.5, then you'd have the Task parallel library, which is lighter than any threading available to VB6. The Task parallel library arose from the recognition that multi-threading is going to become increasingly important as CPU advances are in core numbers and multi-thread support rather than in clock speed. The 4.0 Task parallel library is pretty complete, while 4.5 adds a couple convenient features.

- the Garbage-Collector (it's really not that easy, to reliably wrap flat C-libs behind
.NET-classes, when the *order* of potentially necessary C-Handle-freeing becomes important)

- the potential of the GC, to "kick in" at unpleasant times

You see this? In what way? In all these years, I've never seen any impact of the GC on anything.

- the MSIL not being "true native code" (the "getting hot"-time for the Jitter I've already mentioned -
but in addition, there *is* a better protection of "Intellectual Property" with native compiled code)

But that's not true. There isn't ANY protection for intellectual property in either type of code for the desktop other than obfuscation. The thing that protects people is that it's almost always easier to write the program yourself than to crack some other code. When that's not the case, the code is cracked and cracked quickly, it doesn't matter whether it is proprietary or not. The economics of software is that it can be sold because it is cheaper in time and money to pay somebody else for their work than it is to do it yourself. Moti Barski covered this in the final chapter of his book on Battle Programming (though in an inverted fashion). You aren't protected by native code. You are protected by the cost of writing vs the cost of cracking vs the cost of purchase. The rest is all a mirage.

Note, that I left out "general performance" - because that's roughly equal (as soon as the MSIL
was jitted) to the VC6-compiler which works underneath the VB6-native-compile-option...

So, since the .NET native group has stated repeatedly that they will be doing .NET native for the desktop, which removes both the JIT cost, the startup cost, and the (invalid) concern over native code, what will that do to your position?

[FunkyDexter]

Wasn't me who brought the heat into the debate

No, I didn't think you were and hope I haven't given that impression. I won't comment on the other protagonists but I agree we probably should have handled it quicker. I think you can blame me for that. As a new mod I sought the opinions and advice of the more senior mods instead of just dealing with it and that introduced some delays. I also hoped that asking people to stop in the thread would be enough but it wasn't. As I grow in confidence I hope I'll do a better job for you.

The only other thing I would say on the matter is that the handling wasn't influenced at all by which side of the debate the protagonists were on - only by their behaviour.

Anyway, moving on...


I think I probably agree with you that VB6 is a good starter language to learn on. Personally I was taught Ada as my first language. That's not just dead, it's fossilised. Actually, you challenged me earlier to name some language I'd worked in that had fallen away and Ada's a good one to cite. You can still compile a program in it but just you try writing a gui or a web page

I think I might agree with you about VB6 for new development where folks have got VB6 experience too. That means placing a bet on two things: 1. the desktop's going to be around for a while yet and 2: while MS don't want to commit to it, when push comes to shove they'll probably continue to support 6's "it just works" state for as long as they support the desktop. I think I'd probably take both those bets. The only real proviso I'd put on it is that the developer should understand that Microsoft aren't committing to that and that they may lose customers as a result of that choice. They own the choice and they don't get to blame MS for it.

For the more general "what is the best tool for LOB development" question we differ in that I probably would recommend .Net. It's worked for me and looks to have long legs yet. If the desktop does disappear I can be confident it'll support me through other eco-systems as well.

That's because there is just no alternative currently where "to move on to" with a large VB6-based Desktop-App

Here I do disagree with you. There are plenty of other choices and many of them have been identified in this thread. Every day that you don't take one of them that's a choice you've made. For a few years after MS moved away from VB6 I'd have had some sympathy but each passing day diminishes the "It's MS's fault" argument. Every customer you lose because you decided not to take that path is being lost more and more because you chose not to switch and less and less because MS stopped developing VB6. You've already said that you choose to continue to develop in VB6 because it suits you. Well, you make that choice knowing that it might cost you some customers. I'm not saying it's a bad choice. If you're so expert in VB6 that you can churn out high value applications in half the time that it would take you to do so in another language then, demonstrably, it's a good choice. But whichever it is, it's your choice. Own it.

Other than that I'd only argue that you keep referring to .Net as "Bleeding Edge". It really isn't bleeding edge any more. I'm not sure it's even "edge" anymore. It's "handle".


edit>

In all these years, I've never seen any impact of the GC on anything.

Oh I have. These days I work supporting data scientists and the nature of their work Is about doing massive numbers of small calculations. The problem I see more frequently isn't the GC jumping in at unwelcome times, though, it's that it doesn't jump in soon enough. That's the thing with any Garbage Collector, though, it's an attempt to abstract away an extremely difficult but extremely critical function. You try writing any significant program in C without accidentally overloading the stack by forgetting dispose of some resource or other. Hell, that was such a common problem they named a web site after it. But .Net does fail in this regard if you're seriously pushing the envelope.

[dilettante]

Ada is still alive, at least as much as it ever was. The latest standard came out in 2012.

GC has nothing whatsoever to do with the stack. GC is all about the heap.


Desktop .Net programming is also in "it just works" legacy status:

.NET Goes Open Source

Microsoft has formed the .NET Foundation with the idea that the best future for the technology is open source. Is this a positive or a negative in the life of .NET?

It has been said that when a company is bored with a product the best way out is to open source it. It gets rid of the problem and, who knows, if enough people are taken in it might even get the applause for "doing the right thing".

Microsoft is divesting itself of responsibility for .Net as it moves onward.

[Shaggy Hiker]

That would be about the way that Sun divested of Java: It didn't really happen, but they made it more open.

.NET is actively being advanced and developed. I wouldn't mind if they would fix on a set of features and leave it at that, but it hasn't happened, yet. It's not even cllose to being legacy.

[Schmidt]

I'm not convinced that you did.

Well, I'm conviced that I did.

[musings about humans intelligence]

On the other hand, I'm also not convinced that humans are intelligent, so I will just live with the doubt.

Then live with your doubts, whatever you prefer...

[runtimes and deployment]

That wasn't a good reason to target 2.0 of the framework. XP got 3.5 as an update, ...

Then recompile the .NET-Mandelbrot-App which targetted 2.0 with 3.5, if you prefer.
Doesn't change a thing - .NET Apps still are less easy to deploy than VB6-Apps.

A preinstalled VB6-runtime in the right version is a given on all current Win-Versions.
You cannot seriously claim that the same thing holds true for the different framework-versions.

[GC]

You see this? In what way? In all these years, I've never seen any impact of the GC on anything.

I recommend google, before posting stuff you are not really sure about -
I am sure - and if it helps - a colleague of you also already disagreed.

[DeCompiling, DeObfuscation]

But that's not true.

It is - same as with the point above - if in doubt, ask Google before posting things
you're not sure about.

There's a lot of de-obfuscators for .NET.
Well, did the googling for you, choosing the second link:
http://stackoverflow.com/questions/3...scator-for-net

You aren't protected by native code.

I wrote, that native code offers *better* protection than MSIL (no matter if the IL-code was obfuscated or not).
Please read my posts more carefully in the future.

As soon as you come up with the readable sourcecode of the VB6-native-compiled mandelbrot-app
(using a "NativeCode-to-VB6-code" converter of your choice, wherever you may find it, good luck),
I promise, to invest about five minutes of my time too, posting the decompiled sources of the .NET-
version of the same App.

So, since the .NET native group has stated repeatedly that they will be doing
.NET native for the desktop, which removes both the JIT cost, the startup cost,
and the (invalid) concern over native code, what will that do to your position?

Wouldn't that be speculation Shaggy?

So, let me revise my opinion about this renewed .NET (which then deserves
perhaps another name - something with 'next' in it would be fine IMO),
at the time this native-feature will be broadly available and well-working.

If the GC is still included (as currently planned for .NET-native), then I will perhaps
not change much about my opinion ... when you have a list, then you have a list,
you know.

Olaf

[Schmidt]

You know Carlos

Sure - we exchanged E-Mails for years now - doesn't mean, that we plan
a marriage anytime soon, but he's a nice guy - but I already wrote that.

AND (you think probably) know Fatina?

No, that's not what I wrote - what about citing me correctly?

Small world!

Not sure about the "deeper meaning" of that particular snippet - just in case you mean
it as an allegation of potential "double-or-triple-accounts" of mine, you mark me a liar -
therefore a clarification would be nice.

[plea about more decency]

Not going to happen in Chit-Chat.

Well, if FunkyDexter didn't mean his reply to me in this regard as a joke,
then I have a little hope - still only a plea of mine, in case it doesn't fly,
then I have at least tried.

When I'm being asked directly, what tool to use for professional development,
and the one who asks already has VB6-experience ...

I would answer that you need to look at the job market.

Again, read more carefully what I wrote...
When somebody is asking, if he shall continue with professional VB6-development,
then this one is not an employee - employees who work for somebody else, usually
do not have free coice about the tools they are using, hence these would not ask.
So, "answering that with the Job-market" is pointless.

[C++,HTML/JS/jQuery as safer choices, compared to .NET]

There are mistakes to be made with ALL of those choices.

No, there are not, they are throughout much safer investments than choosing .NET.
I gave reasons for that already, and will not repeat myself.

I have a large VB6 LOB app which HAS to go somewhere
(the source code was lost in a robbery, of all things).
However, that provides an opportunity to re-write and fix many of the design mistakes made.

At this point Shaggy, I will end the discussion with you, because I don't see
how you can conclude from a VB6-App (which you have no Source-Code for),
what potential "design-mistakes" it contains - I had the impression already earlier,
but the above clearly indicates (to me at least), that you're telling "stories" - and
I usually avoid discussions with people who try to make a fool out of me.

Have a nice day.

Olaf

[NeedSomeAnswers]

Nah, "our silverlight company" was highly motivated after consuming all the advertising -
Mr. Silverlight-guy also hanging around here in the forum, feeling much encouraged
by all the positive things he read here, any "MS-bleeding-edge-tech" recommended as
"always a good idea to move on to".

That's a really bad reason to move to a new tech, essentially being swayed by the marketing.

The company i was working for around 6-7 years ago was doing a lot of web development, and the guys i was working with were much more experienced then me, and we looked at Silverlight (and Flex the adobe alternative which we built a few extra-net sites in), and despite the benefits of both at the time, we decided that we wouldn't build any major apps in either of them.

The take up of the SilverLight plugin was an issue, and even though Flex targeted the Flash player, we saw the trend even then to move away from run-times in the web for business apps. With SilverLight in particular as well there was the problem with it extensively using the GDU and if you had slow graphics cards in your PC it didn't perform nearly as well.

So my point is the experienced Web Developers i worked with saw SilverLight as a dead end pretty early on, so using a tech like that (when it is still very new and has low adoption) is a risk that could have been avoided if you have the right people making the decisions.

Thats not a pop at your guy btw, i just feel SilverLight could have been and was avoided by many who saw it as to risky)

- .NET Apps still are less easy to deploy than VB6-Apps.

Now i will have to flat out disagree with that statement. I have yet to have a .Net app where i had any install issues with at all really. Deployment has never as far as i am concerned been a problem with .Net apps, i have never seen a difference.

Hah, traitor - how dare you... <g>

I am far from being an MS only guy, despite knowing vb6 and .Net, you need to evaluate which tool is best for the job.

At least from me, you will see no real hostility - IMO you just get "unwelcome information"
into the wrong throat - arguments which you feel "tainting your own perception of things"
(guy didn't ever use .NET - how dare he post such things about it).

But your tone does comes across as pretty Anti .Net at times and you sometime seem to bundle all users of .net together as one entity, although this may have been just after one of Niya's posts :0)

.Net doesn't work for you, but it obviously does for so many others, which is kind of the point i have been trying to make. You certainly don't taint my perceptions, but you put up good arguments which make the debate interesting.

.NET is a good tool, it might not be a good tool for everyone but it does a good job for many others. For me it works well as a RAD development tool, the downsides which you point out, don't affect the business i work for (or my previous ones).

In my example it did happen - have got "anxious questions" (paraprase - about the use of certain tech)

I agree this can happen, it just depends on the customer and the market. Some customers do care about what technology is being used and make purchasing decisions based upon it, others don't care at all as long as it works.

There's some languages and platforms which offer (much) more long-term-stability.
I've pointed them out already (those which are standardized or OpenSource-driven).

But i consider .NET as not being among them - I truly don't.

We will have to disagree on this, i make my living developing and managing a team writing apps primarily (although not exclusively in .Net) and our customers see it as an established platform and a very comfortable with it. The Market we sell into like many many others, is behind the bleeding edge curve, and they will be buying desktop apps for quite some time yet. Yes they want some web and Mobile mixed in but they are stuck primarily on the desktop.

.Net is established and our products sell, and sell well in part because of the technology we use. My last company (which was a lot larger) was making Millions out of selling .Net products, and my current company is very profitable for a small company also primarily selling .Net apps.

Just to point out a few things to back-up the: "guy knows what he's talking about"-thing.

I don't think any of us has the opinion that you aren't knowledgeable Olaf, we may disagree with you but you are pretty much the only person on the VB6 camp that actually puts forward good arguments. Dilettante does too sometime but he is just as likely to enflame the debate, it depends what mood he is in :0).

The person who almost always enflames things, is now gone for a while

While i agree with you there , Carlos was not completely innocent in this thread and his arguments are not nearly as well constructed . He has tended to just say .Net is crap (which just as some of the comments you have found insulting, some .Net developers also found these insulting)

When somebody is asking, if he shall continue with professional VB6-development,
then this one is not an employee - employees who work for somebody else, usually
do not have free coice about the tools they are using, hence these would not ask.
So, "answering that with the Job-market" is pointless.

What about those starting out there career, or still studying? the Job Market is very important i would say to many of them, i know it was to me, and they do have a choice.

Certainly were i live it is now very very difficult to find job asking for VB6, where as it is relatively easy to find jobs asking for .Net, and i don't see why this isn't a good argument as to why new developers shouldn't learn and use .Net.

Now i am not saying it is the only choice or even the best choice as all circumstance are different but for me and many other it has proven a good choice, i have been developing in / managing .Net teams for around 6-7 years now and i would bet i will still be using it for another 10 years+ easy and maybe longer as the .Net tools for Web & Mobile development have improved rapidly and are really quite nice to develop in and produce good quality apps. With something like the telerik control suite you can even build .Net apps and have it compile them to Native mobile apps if you want.

I am in no way Anti VB6, i still support apps written in VB6, and i can see why in some cases it will be the right tool to do new development in, but in general i think that when advising people on what tools to learn .Net is a better bet in particular if you want to make a living out of software development right now imho.

I had the impression already earlier,
but the above clearly indicates (to me at least), that you're telling "stories" - and
I usually avoid discussions with people who try to make a fool out of me.

I think you are being unfair there Olaf, Shaggy has been on this forum for along time now and i have read a fair amount of his post across all sorts of topics. Shaggy Style of writing is often like that i.e. story like and full of analogies and puns :0)) and in no way did i read into it that he was trying to make a fool out of you.

He may disagree with you but he was engaging in honest debate.

and finally ...

A valid point - .. olaf

I will take that :0) i might even put it in my Signature!

[ColinE66]

Not sure why anybody should attach any 'suspicion' to Olaf's claim that he knows Carlos. I can confirm that he had a life before VBF as I myself have been corresponding with him since 2008. Pretty sure that I'm not alone in that regard as he is well-known to many VB6-ers and very helpful when approached outside of the context of social fora.

Just for the record...