.NET

[Jon]

Has anyone played with .NET much? I see the guys here at work knock out some functional GUI apps in no time at all using it. (I've been contracting at MS for about a year now and I swear none of the devs I've met understand pure c/c++ any more, the ideas of using char[] horrifies them).

I was thinking about having a play for when I want to write some non-autoit apps or utils for XP/Vista - and also I thought it would be good to learn in case I ever want to switch from consultancy into programming (the travel I do is very tiring). You can never learn too many languages after all

So, anyone used it for anything more than "hello world" yet? Got any decent newbie tutorials?

Ta.

[Jon]

Adds "cluster fuck" to the lexicon.

[SumTingWong]

Adds "cluster fuck" to the lexicon.

You should also add it to the profanity filter.

[Valik]

Cluster fuck is one of my favorite expressions. I'd called .NET in general a cluster fuck looking at it from the outside.

[Jon]

Miltary term for a situation caused by too many inept officers, cluster referring to the insignia worn my majors and LT. Colenels, oak leaf clusters.

Nice.

[big_daddy]

The term "cluster fuck" is used very frequently in the army.

[BasicOs]

@Jonh I think, not a goodIdea,

That is my experience about this bussiness:

I Stopped migration from MS Visual Studio as soon as .NET was not any more 100% compatible with Visual Studio 6, e.g. the "PRINT FORM" FUNCTION was not anymore working in .NET and more issues, which solutions did not satisfy me, so I invested this energy into improving my Autoit.

I MIGRATED my small VISUAL BASIC 6 database apps into AUTOIT and they are working 100 times better right now. They load so quick as hell.

I just migrated ONE HUGE Visual Basic 6 app, doing 4.000 lines optimized autoit code, and 8 tables.

After that I am required now to migrate another one: a client of mine who has another similar database/forms app to migrate from ACCESS 2000, to .NET, or ACCESS 2007. I prefer into Autoit, but I do not know if I will have time to do (I estimate about 1 month full time approx. or 2 months half time whole migration plus 1 month beta 1/4 of time) if somebody needs CASH$$ and is interested in coop for coding this app. You can pm or tell me. Because I am in another project for this next 9 months and lack of time. This migration is really good paid$ and it is a opportunity to learn how to.

All you can do with .NET, you can do in Autoit, or even more. Only I miss only couple of issues from migration from MS STUDIO, I could not solve so good as i desired:(I know it is not right forum) I tell:

1. the IDE: great Koda soft. could be improved then it could be the matter: To click in any control/button and go into the code directly, i was missing also a couple of small issues more.

2. Some Windows 98 issues in some pcs, e.g. "AUTOIT ERROR" with "xxxxprocesslist not found xxxx" , can you make a timeout for this error in some secs?, or some error management error, I did whatever is recommended in forum, Svenp com error UDF, Opt(errorsfatal 0), it keeps showing same app only win 98.

3. Some times randomly the keyboard get stuck as pc and app boot at the same time from userinit. I do not know why. I know how solve this issue as it shows: pressing left shift, it recovers and you can use Keyboard, but it is a headache. Only XP, and some slower pcs.

4. I do not know how. As they broke some COM support, they break the app. It stops working (It happens not very often), keeps some COM error showing it doesnt work any more till I reghost the pc. It has to do with the WINDOWS system noT to do with the programm itselft

I promise the hell is vb, vstudio, .net and what ever is going on this kind of Stuff, that makes a lot of problems, incompatibilities, lack of easy and powerful udfs, fat installing, big installshield,all goes into FAT CODING.

Edit: I think the propoused migration from Access 2000>autoit could be about 5.000/8.000 line optimized code or may be more. Most of the work should be testing, optimizing, and fixing, coding some modules. I speak of no nobbie.

Has anyone played with .NET much? I see the guys here at work knock out some functional GUI apps in no time at all using it. (I've been contracting at MS for about a year now and I swear none of the devs I've met understand pure c/c++ any more, the ideas of using char[] horrifies them).

I was thinking about having a play for when I want to write some non-autoit apps or utils for XP/Vista - and also I thought it would be good to learn in case I ever want to switch from consultancy into programming (the travel I do is very tiring). You can never learn too many languages after all

So, anyone used it for anything more than "hello world" yet? Got any decent newbie tutorials?

Ta.

Edited October 20, 2006 by BasicOs

[Richard Robertson]

I highly suggest .Net. If you wait a while longer, I will have all of AutoIt's functions ported to .Net. It is harder than I initially thought, but not too bad. I may not do things the way AutoIt does, but it will pretty much produce the same results as I go.

Edit: spelling error

Edited October 20, 2006 by Icekirby1

[this-is-me]

When speaking of .NET, it makes me wonder if it is possible to automatically determine the odd (seemingly) random numbering and naming conventions of .NET controls and therefore interact with them using AutoIt. I realise that there are workarounds like ControlSend, but it would be simpler if one could reference a .NET control inside AutoIt by a fixed name.

[Valik]

<UnsubstantiatedComments>

I believe .NET controls are created on-demand where-as in Win32, they are created when the creation function is called. In other words, it's probable that a control in Win32 exists even if it's never viewed. In .NET, that control would not be created until it first needs to be viewed. You can observe this behavior with some .NET apps. If you start them the same way every time, the controls should all have the same IDs across sessions. However, as is the case with a tabbed application, if you start it with a different tab initially visible, the controls will be created in a different order and so the IDs will be different.

At least that's what I've noticed. That might not be exactly correct. I don't work with .NET enough to know.

</UnsubstantiatedComments>

[Richard Robertson]

You can create the handles without showing the control or window. I think the function is Control.CreateHandle().

Yes, they have inconsistant IDs. That may be an antibotting mechanism.

[Valik]

I doubt it. It's for performance. Why take the time to create a control that's not going to be visible yet? Why not create it the first time it's needed? The fact that there is a method to force control creation only re-affirms what I said above.

Your reason doesn't really make sense, anyway. This is Microsoft's game. They can stop automation tools like AutoIt if they want to. Windows knows, or has the potential to know, simulated input from real input.

[JSThePatriot]

.NET is a good language to learn as M$ is pushing it hard, and its paying off as there are many jobs out there right now that require .NET experience.

I personally dont like .NET due to its very large footprint. I would hate to have to tell a client (that didnt already have a .NET runtime) hey sorry you have to download and install this monster before being able to use my App.

JS

[Richard Robertson]

24 megabytes isn't a lot.

The handle creation is more for things like direct API calls. I've used it a few times on a Form to allow me to use hotkeys.

[Uten]

24 megabytes isn't a lot.

Thats just the download. Expanded on the disk it's quite a loot. Considering the memory footprint it really starts to get annoying unless you are prepared to continue the hardware, OS upgrade cycle every 18 months.

So I think it is time to ask our self if the OS should be allowed to occupy and abuse most of our computing resources (Yes I do consider NET a part of the OS platform). The only thing I do now that I did not do on my 486 based PC is running virtual machines (vmware).

Anyhow. I have done C# and VB.NET. I do like the NET environment. It makes me (feel )really productive. And I suppose NET or JAVA have a place in the corporate environment. I find it a lot easier to do quality assurance of software production in a NET or java environment than in a C/C++ or AutoIt environment.

But I do question the size of the run times and libraries. And more over the need to download huge bug fixes and entire new versions makes me wonder about the OO paradigm (object reuse) and if it is implemented, used and provide the benefits it does promise.

Having to buy a Ferrari just because the business ecosystem earns more on it than on the 2CV does not get me faster from A to B. At least not where I live

[Lazycat]

Not programmed in the NET myself, so can't say about it from "programming" point of view. But from user side, what interesting, while last few years I'm not saw even single good program written with NET. They are all bloat, damn slow and what worst, quite unstable. Using them is big displeasure. Don't know, maybe I'm just unlucky and not found yet NET programs written by good programmer...

Anyway, this can worth to learn because some companies decide to migrate to NET and you can have no choice...

[Richard Robertson]

I'm a good programmer in .Net. I like the flexibility. It can call things that people make and convert to an assembly, as well as direct platform APIs. I had a lot of fun converting XINPUT to .Net, if anyone would like to see it. XINPUT is part of the new DirectX SDK but only for C++, they had no managed copy which is odd because everything else has a managed copy.

[big_daddy]

Currently we only use one piece of software that is written in .Net. It is a Fingerprint Authentication Software that is only available to Banks and Hospitals, but so far I have nothing good to say about it. It requires three processes that eat up 33MB of RAM a piece and half the time are running at 100% CPU because of a Microsoft glitch.

[big_daddy]

Any of these processes end in 'Q'?

Lar.

Nope more like...

Verinex.Biodentify.InteropServer
Verinex.Biodentify.ServiceProvider (system)
Verinex.Biodentify.ServiceProvider (user)

[big_daddy]

I sure hope the company that you work for provides a better product! When we first got these things it was supposed to reduce the amount of support calls we got for resetting passwords. However I think the call volume has doubled or tripled do to the damn software loosing their credentials all the time. Our logon scripts no longer run before the desktop is loaded. RDP Connections render the finger print devices unusable until after a reboot. Oh, and did I mention they require 100MB of RAM to operate!

As you can see I am very unhappy with what we invested over 30,000 dollars in.