PB Reliability

aaron · Post by **aaron** » Sun Aug 21, 2005 12:09 am

Wow, Geoff really started a b*tchfest here with all the guys who are grumpy.

Personally I couldn't be happier with the way that Purebasic is being run. There is a nice mix of new features as well as updated older stuff. Ok, so the documentation suffers from french->english translation, but that doesn't affect me when I coding. I've never had any troubles tracking down information in the manuals or understanding what it is saying.

To say that Purebasic is buggy at this point is just ridiculus. I code in it every day and haven't run into anything that has caused a problem for me.I have no doubt that this is true for 99% of Purebasic users. It just works. Its just that the 1% of the people who have trouble (likely much smaller even) are the most vocal.

Trond: take up Fred on the CVS access to the documentation and start fixing it. Sounds like win/win for everyone. I'm impressed with the attention to detail that you are able to comb thru Purebasic with.

Geoff: as for adding more version numbers to keep track of hotfixes... whatever. If you run into a bug, run smartupdate and try again. If there still is a bug, post it on the forum. You don't have to worry whether you are running the latest version or not... it take under a minute to keep up to date with Purebasic. Who cares with version 3.94.1.245.123 had a bug that was already fixed in the latest 3.94 version available?

Post by **Dare2** » Sun Aug 21, 2005 12:13 am

@Trond:

I retract the "Role Model" comment, I shouldn't have done that. I was sinking to your level.

Just got peeved with you. Normally I can ignore a pompous attitude and general rudeness, but the fact you had published someone's personal information on these boards ....

(which, BTW, could land you in jail in some places).

So there. Retracted. Feel better?

dracflamloc · Post by **dracflamloc** » Sun Aug 21, 2005 12:30 am

I love pb. Honestly. Its got flaws as everything does, but I think fr34k, fred and crew are great and very dedicated.

As for professional apps... There are quite a few, some written by myself, some written by others. Paul makes a few, as did inner, and others.

As for updates. Thank god we now have up to date OGRE support, even if its beta and windows only atm.

Like someone earlier said, when you've been in the business for a while you know not to flock to a fresh new version of software for critical apps. Personally I'd much rather the active support of PB with fairly frequent releases than having to wait a year or two for a new version to come out. At least that way you have a choice of what version to run until the new version is fixed.

DoubleDutch · Post by **DoubleDutch** » Sun Aug 21, 2005 12:41 am

Trond:

We're not allowed to report stuff that's will cause PB to become more solid than barely glued together with tape because then people call you "Role model for Anal Retentives" because they want to get on with adding more features.

What is wrong with you, two stupid posts in two days?

Fred, Fr34k, and the rest: Keep up the great work and thanks for fixing my bugs as soon as I report them! All have been fixed great so far...

-Anthony

Post by **Dare2** » Sun Aug 21, 2005 12:42 am

"Release early, release often". The way to go!

PB · Post by PB » Sun Aug 21, 2005 1:10 am

> We're not allowed to report stuff that's will cause PB to become more solid
> than barely glued together with tape

Trond, please feel free to ditch PureBasic and switch to Visual Basic with its
massive runtime DLL at any time. PureBasic is an evolving language which
is worked on very hard every day, and just because there are some bugs
doesn't mean it's "barely glued together", and you damn well know it.

Brice Manuel · Post by **Brice Manuel** » Sun Aug 21, 2005 2:47 am

use of this compiler on this forum it seems to be used more on games stuff then truly serious stuff.

I have been programming since '79 and primarily made a living at it since the late 80's. Indie developers who want to actually make money, make games, as that is where the money is and has always been. I have many close friends who are indie developers who have been making six figures per year since the mid 90's making games. For indie developers the hottest market right now remains games, specifically 2D desktop games and games for the various mobile devices like Palm, PocketPC, etc. I have three friends who are very close to breaking the seven figure per year barrier just on crossplatform mobile sales alone.

The only apps by indie developers that are usually profitable in the six figure range are game development tools, (modeling, art, texturing, programs. Etc).

Contrary to your statement, the main use of PB seems to be apps, which is sad, as PB is a great choice for 2D and is miles above the competition when it comes to 2D.

I own and have used all the current indie languages out there. PB is the best IMHO. Once it matures a bit more, I think FreeBASIC may be a good rival, but right now, PB is easily leading the pack of indie languages. Unlike other developers, Fred has always kept his promises to his users and has not resorted to questionable business practices to elicit more money out of his userbase.

Xombie · Post by **Xombie** » Sun Aug 21, 2005 5:21 am

PureBasic is a great language. I think it's continuous releases and constant attention by the developer (fred) are it's strong points. I can't think of many other developing languages where the primary coder will chat with users and help them out with whatever issues they have. As fred said, nothing is ever perfect and if you keep waiting for a completely stable version, we'll never get anywhere. He's not Microsoft, with a fleet of programmers (which still put out buggy code), he's a guy with a few codevelopers. I have faith in him, hell, he helped me out with some custom libraries when I needed them - without charging or anything. For ... what? $70 that I paid for PB he has my loyalty.

Thanks, fred. And keep up the work. ^_^

griz · Post by **griz** » Sun Aug 21, 2005 6:21 am

I am not in the 'silent majority'.

Pure Basic is wonderfully fresh and I really appreciate it's constant march towards a more feature filled, reliable environment. It's not a static product. I look forward to new features and frequent updates. I believe the new PBOSL (http://pbosl.purearea.net/) is a tremendous gift to the PB community. There are features that some say Pure Basic lacks -- but will Pure Basic lack them in the future? I doubt it. Pure Basic, today, right now, reminds me of Atari's old slogan : Power without the price!

I applaud the PB team for what they've done, and continue to do.

Anden · Post by **Anden** » Sun Aug 21, 2005 9:06 am

Like it or not, but PB is a crossplatform compiler.

Yes, i like that very much (in fact, it's the reason why i gave PB a chance to prove itself worthy for a production environment).
But as so often found in reality, the theory differs much from practice. With the basic functionality of PB **ALONE**, you will never be able to develop a serious, reliable and user-friendly application (comparable to todays standards). Of course you can somehow, but that involves very much platform dependend API coding. And at that point, one loses all of it's platform cross-compatibility.

The PB project has very limited personal resources, that's a fact. We have to accept that, but show-stoppers like the "Compiler bug with string concatenation and procedures" should be released IMMEDIATELY and not weeks after discovering or fixing it (call it a hot-fix

)

And there are definitely missing PRO functions in PB (doubles, unsigneds, gadgets like a grid, gadget color support, drag & drop, ...). Not much you couldn't do with the help of user-libs, API and window callbacks. But that's not the way software ist developed in firms nowadays.

Let's hope the best for V4 ...

okasvi · Post by **okasvi** » Sun Aug 21, 2005 9:42 am

where is that 'silent majority'? |:
only thing im missing in PB is better networking

, but Fred gave winApi support for us to fill holes in PB while its still in constant developement...
and so far i havent ran even into one single bug

oldBear · Post by **oldBear** » Sun Aug 21, 2005 2:07 pm

For those of you who are into developing industrial strength, cross platform, commercial applications, there is a solution - c++. The company I worked for built a half billion dollar business using it.

Most people would do a requirements analysis and select a tool that satisfied their requirements. They wouldn't just go out and buy a saw and then complain to the manufacturer because they couldn't drive nails with it.

pb meets my requirements nicely and I'm a satisfied customer. But then, I had a pretty good idea what I wanted going in and selected a tool that satisfied those requirements.

A previous post offered some suggestions on how to avoid possible problems associated with new releases.

For those things which you feel are lacking, I believe there's a wishlist section in the forum.

For the price, you're probably not going to find a better bargain in a basic language. Of course there's always python or ruby or c++ or whatever but you're probably not going to be as productive as quickly or have as much "fun" with those.

Thanks Fred and team. It a great tool.

cheers,

Anden · Post by **Anden** » Sun Aug 21, 2005 8:30 pm

For those of you who are into developing industrial strength, cross platform, commercial applications, there is a solution - c++. The company I worked for built a half billion dollar business using it.

That days are almost over (at least here in europe). C++ is still strong in the games business and there is loads of "old" code lurking around. But C++ was never very cross-platform (because of the GUI). Java and/or .net (mono) is state of the art now ...

Most people would do a requirements analysis and select a tool that satisfied their requirements. They wouldn't just go out and buy a saw and then complain to the manufacturer because they couldn't drive nails with it.

Quite funny, but not true. If you see a saw (produced by an unknown manufacturer) and that saw is a bargain: Why not give it a chance and buy it right away?
If it satisfied your requirements, buy it for all of your "developers".
If it miserably fails, don't give the saw or the manufacturer a second chance.
And if it is still promising after testing but falls short in the long range: Why not give feedback to improve the product? Thus the product could evolve and there is still the chance for a win-win situation ...

Psychophanta · Post by **Psychophanta** » Mon Aug 22, 2005 7:53 am

Agree with Xombie, and yes, i also have to congratulate Fred one time after another. Really, his work has no price.

geoff · Post by **geoff** » Mon Aug 22, 2005 12:44 pm

First, let me repeat.
I think PureBasic is excellent, I am impressed by what has been achieved so far, and I don't wish to criticise anyone personally. I am especially grateful to Fred and freak who responded quickly to my posting of the Win98x bug in a friendly and helpful manner.

But I am still concerned that some of my points are falling on deaf ears.

Beta version

after the beta5 was out, all I added was the "SaveAll", "CloseAll" menu
and one preference option. So yes, we do take this point seriously.

Partly seriously, the ideal is only bug correction at beta.

The Win98x bug

I did test said problem on my Win98 testing setup, and it worked.
After you reported it, i tried again, and it happens here only once every 20th or 30th time.
As said above, such things can happen even if you do testing. That's life.

On my system it crashes every time I run it, hence my comment, so maybe I was wrong in claiming it was "untested". However, I am not impressed by this "suck it and see" approach. Do we really know what this fault was? Or do we just assume it has gone because it doesn't crop up very often? If you don't really understand what is going on but just take the view "it seems stable now" then these things tend to hang around and kick you when you least expect it. Personally, I am never happy when something "just goes away". I want to know why it went away, and understand it properly.

PureBasic version numbers
Yes, I accept that my suggestion of a version number for every change may be impractical and it is right to post single file changes quicky to help those with problems. This leads to a situation where only users that are using the latest version with all the file updates can expect help. I accept this is reasonable for a product with a small support team.

Windows version

Now if you want to continue using such an old version of Windows, that is fine, and we will do our best to keep it supported, but you have to get used
to the fact that support for more recent, more popular versions of Windows

Fair comment, but in a similar vein to the "why bother with old fashioned RS232" discussed in a previous thread. I think some people's idea of a "typical user" are wide of the mark. From the comments made in this forum I suspect many people have a single PC kept "up to date" with the latest version of Windows and the latest version of PureBasic. They write games, apps connected with the Net, get to understand Windows better and so on.

But others, like myself, have a number of machines running different OS's, used for different applications and maybe connected to hardware, or maybe provide or sell software to others. In the real World PCs and associated hardware are not kept "up to date" with the latest version of everything. Personally, I have a PC running Windows XP Pro, but I also have one running Windows Me, another running Windows 98, and even one running Windows 3.1. Some of these computers won't run XP, nor would it make sense to pay for a version of XP even if they would run it. For example, Win3 runs in just 4MB memory and you can close down the machine using the on/off switch. This is convenient for a laptop just used for GPS mapping. Windows98 has most of the functionality of XP but can be reloaded from a disk image in 30 seconds compared to 6 minutes for XP. There are good reasons, other than finance or inertia to keep these things working.

The PureBasic Product

There is a simple way to achieve the results you are looking for. A bit of simple arithmetic will put it right in no time. Just take the number of coders listed in the About box, multiply by nine, and poof! You've got it. These guys eat, drink and sleep and breathe this stuff for the pure love of the project and for the benefit of the user base. All for sixty euros a pop. What do you want for crying out loud

As I said, I think the product is great value and I appreciate and am impressed by the work of the coders

But I completely disagree with your "arithmetic".
It is not my experience that more coders means better design or higher reliability. Over the years I have been involved in many software projects. The smaller the project, the smaller the number of coders, the less bugs you get. What is the largest software company in the World? Do they produce bug free software? Take a small piece of software like that in your digital alarm clock, do you keep finding bugs in it?

As many of you probably know, project reliability and fault correction is a science in it's own right. Large software projects employ this science. It enables you to monitor the rate of bug correction and make probabilistic estimates of the number of bugs that remain in the code at any stage. Do you thing Microsoft believed Windows XP was bug-free when they released it? Of course not, they understand this stuff.

Bugs will exist but their effects can be reduced by a disciplined approach to change control and version release. Don't expect that a version that "seems stable" is without underlying problems. Create beta versions, and (only) make corrections until the rate of bug discovery reduces to a trickle. Then you have a product for full release that does justice to your efforts.