Some thoughts about PB in its current state

Everything else that doesn't fall into one of the other PB categories.
User avatar
deseven
Enthusiast
Enthusiast
Posts: 362
Joined: Wed Jan 12, 2011 3:48 pm
Location: Serbia
Contact:

Some thoughts about PB in its current state

Post by deseven »

Hi guys.

A week ago i decided to drop PureBasic completely.
Don't get me wrong, i'm not here to whine about it, but still it was a very hard decision for me, since i started to use it in 2009 and at least half of my public and private projects has been written in PB.

Over those years i slowly moved from windows-only tools to cross platform ones just to finally arrive to pure MacOS applications. And now i can surely say that the only platform on which PB works as it should is Windows. But even there it has many critical flaws.
One particular thing that bothers me a lot is that PB uses many powerful libraries, such as libcurl, ogre3d or scintilla, but their implementation is so limited, that, for example, we can't even define the user agent for our HTTP queries. Yeah, it's 2017, we live in a web-driven world, but we still can't do anything with it besides very basic things. And moreover, the included version of libcurl is lacking zlib support so we have to include external one. What is this, a joke?

And i can go on and on about PB for MacOS.
In two months it will be a year since the release of MBP 2016 which PB doesn't support. It was reported in april.
Many basic things are broken. Some of them are here for years.

That basically means just one thing - PureBasic is unreliable. If you encounter some game-breaking stuff you're basically doomed. There are no possible ways for you to know whether this issue will be fixed at all and when it will happen. Whatever you do there is no way to get the answer. Here is a nice example.

I do understand that it's all coming from limited resources, but it just doesn't look right for me, because everything can be fixed. I see at least two possible ways for PB to be well.

First one is pretty straightforward - opensource the compiler. Just the compiler, leave the IDE as it is and continue to sell the full bundle. Let people fix small bugs and put bounties on huge ones, let them add things, let them argue, let them see that something is actually going on. Don't leave them in the dark alone, it's frustrating.

Second one is to drop all unimportant stuff such as x86 compiler, LTS version (let's put it straight, PB has no "S"-stuff from LTS), crossplatform support, 3d engine and its routines (anyone ever saw any 3d games written in PB?) and focus on the one thing that it already does good - making an easy, small and fast desktop software for Windows. Fred actually tried to step on this path by removing ASCII support, but that's definitely not enough. I also believe that some sort of paid subscription can lead to adequate support and bugfixing.

Here are my thoughts. What are yours?
User avatar
Lunasole
Addict
Addict
Posts: 1091
Joined: Mon Oct 26, 2015 2:55 am
Location: UA
Contact:

Re: Some thoughts about PB in its current state

Post by Lunasole »

It's stupid as for me to think that you can make some commercial game for example, using only tools/libraries/frameworks made by someone else. You surely need to fix/customize them and typically a lot of that is needed if you doing something nice and unique. So as I said it few times already, that's only your problem if you blindly used PB in some huge project, instead of planning ahead and using directly those open-source libraries when it is needed (critical parts where full control required to ensure all will be OK tomorrow).

Any stuff better fits some things while others not, and though PB is almost perfect to code many different things, surely it's not language to code everything... and so on.
Anyway I don't know what is the point for you to drop it completely, instead of use it more wisely and get lot of advantages from that.
First one is pretty straightforward - opensource the compiler. Just the compiler, leave the IDE as it is and continue to sell the full bundle. Let people fix small bugs and put bounties on huge ones, let them add things, let them argue, let them see that something is actually going on. Don't leave them in the dark alone, it's frustrating.
PB will not be even close to current if it will be open-sourced.. That would be just rather a mess like that FreeBasic, not saying about other cons.
Second one is to drop all unimportant stuff such as x86 compiler, LTS version (let's put it straight, PB has no "S"-stuff from LTS), crossplatform support, 3d engine and its routines (anyone ever saw any 3d games written in PB?) and focus on the one thing that it already does good - making an easy, small and fast desktop software for Windows. Fred actually tried to step on this path by removing ASCII support, but that's definitely not enough. I also believe that some sort of paid subscription can lead to adequate support and bugfixing.
x86 is important (and not a problem to support it in modern C++), x-platform is important (well not for me, I was already proposing to drop that mac os^^), 3D-engine is important (it goes fine in many thing, including visualisation or just learning of 3D... wtf it must be for games only).
Idea of some paid support with faster response/fixing looks fine... maybe. That's often used
"W̷i̷s̷h̷i̷n̷g o̷n a s̷t̷a̷r"
IdeasVacuum
Always Here
Always Here
Posts: 6425
Joined: Fri Oct 23, 2009 2:33 am
Location: Wales, UK
Contact:

Re: Some thoughts about PB in its current state

Post by IdeasVacuum »

...there are still more x86 desktop PCs than x64, which is why Microsoft still supports it.

Perhaps a survey of PB Users could be defined, to determine % of each OS and % of Users producing cross-platform apps.

Since each OS is a large market, I'm inclined to think that not many developers are producing cross-platform apps. So, if you are creating apps for one platform, it boils-down to the type of app. Probably most are desktop utility/office type, but the 3D is not just for games, it is useful for Engineering/Architectural/Domestic goods/Medical apps too and generally speaking they do not require the 'whistles and bells' that games use.

If you want to talk about reliability, the first port of call is yourself. Most issues are caused by coding errors by us, not the programming language. A lot of my work requires coding in C, but with PB I can produce better, more reliable apps - faster!

Opensourcing the PB compiler would inevitably be the death of it.
IdeasVacuum
If it sounds simple, you have not grasped the complexity.
User avatar
kenmo
Addict
Addict
Posts: 1967
Joined: Tue Dec 23, 2003 3:54 am

Re: Some thoughts about PB in its current state

Post by kenmo »

I agree that bugfixing should be a bigger priority, and it would be nice to hear updates about bugs.
The team has chosen not to use a bug tracker, for years.

Open-sourcing a language only works if its community is large enough, I don't think PB is there.
Selling just an IDE bundle would be much less sales, less feasible for the team to work on PB.

I totally disagree on what you consider "unimportant stuff" :)
Cross-platform and 32/64-bit are two of the most important features.
Personally: I use PureBasic almost daily, across ~4 computers. If you drop 32-bit and Mac support, you've lost 75% of my usage.
Dude
Addict
Addict
Posts: 1907
Joined: Mon Feb 16, 2015 2:49 pm

Re: Some thoughts about PB in its current state

Post by Dude »

deseven wrote:A week ago i decided to drop PureBasic completely. [...] PureBasic is unreliable. [...] Here are my thoughts. What are yours?
I thought you quit? So why do our thoughts matter? What is the point of your post? Why are you bashing PureBasic as you walk out the door and stirring us up? :evil:
User avatar
Shield
Addict
Addict
Posts: 1021
Joined: Fri Jan 21, 2011 8:25 am
Location: 'stralia!
Contact:

Re: Some thoughts about PB in its current state

Post by Shield »

Dude wrote: I thought you quit? So why do our thoughts matter? What is the point of your post? Why are you bashing PureBasic as you walk out the door and stirring us up? :evil:
It is important to address issues users have with the product in order to improve it; blatant fanboyism doesn't help anyone.
The points deseven mentioned are mostly valid, however I do think that this is exactly the niche PB is filling.

The way I experienced and continue to experience PB is that it fills the very specific niche of being an easy-to-use
(mostly beginner's) language with a comfortably sized standard library. For lack of better terms, it is basically in the sweet spot
between "noob languages" which you can't use for anything larger and large and complex "professional" languages.

This is the only reason I still occasionally use PureBasic for quick prototyping. It is quick to set up and has a little bit of everything.
With a comparatively low number of functions, it is easy to get familiar with all of them and there is normally no need
to constantly refer to the manual. This is an advantage.

Having said that, it is obvious that PB has quite a few limiting factors (both in the language and the library) and there is no point denying that.
I started using PB some 12 years ago and I feel it has never been more quiet on the forums. Deseven moving on is just yet another senior member
leaving the ship, as many of us have done. I check the forums regularly and see fewer and fewer familiar names.

Addressing the reasons why people are leaving might make them stick around just for a little bit longer. :)

Edit: changed "Dude" to "deseven"
Last edited by Shield on Thu Jul 27, 2017 12:24 am, edited 1 time in total.
Image
Blog: Why Does It Suck? (http://whydoesitsuck.com/)
"You can disagree with me as much as you want, but during this talk, by definition, anybody who disagrees is stupid and ugly."
- Linus Torvalds
User avatar
deseven
Enthusiast
Enthusiast
Posts: 362
Joined: Wed Jan 12, 2011 3:48 pm
Location: Serbia
Contact:

Re: Some thoughts about PB in its current state

Post by deseven »

IdeasVacuum wrote:...there are still more x86 desktop PCs than x64, which is why Microsoft still supports it.
Can you please provide a link to the source? I've searched for this statistics, but wasn't able to find any reliable and fresh source except steam hw survey (which says that 95% of PCs are already 64-bit, but obviously gamers are not our typical userbase).
IdeasVacuum wrote:If you want to talk about reliability, the first port of call is yourself. Most issues are caused by coding errors by us, not the programming language. A lot of my work requires coding in C, but with PB I can produce better, more reliable apps - faster!
I can't argue about that because if the links i provided in my first post are not a reliable argument for you then i don't know what will be.
IdeasVacuum wrote:Opensourcing the PB compiler would inevitably be the death of it.
Lunasole wrote:PB will not be even close to current if it will be open-sourced..
Can't say i understand why are you people so against open source. Can you explain your opinion in details?
Lunasole wrote:x86 is important
Maybe it is, but i see that it's just another burden for Fred. There are some bugs which persist in 32-bit version but not in 64-bit and vice versa. I'd say with such limited resources you have to stick with something which will not die completely in the nearest future. Obviously that'll be 64-bit OSes.
Dude wrote:I thought you quit? So why do our thoughts matter? What is the point of your post? Why are you bashing PureBasic as you walk out the door and stirring us up? :evil:
The point is that i want to create a dialog which may lead to some changes. And those changes can be really helpful for PB community which i was a part of for so long. No need to be so aggressive, if you think that everything is good then there's nothing to talk about, why bother?
Shield wrote:I still occasionally use PureBasic for quick prototyping
Yes, it may be a perfect niche for PB, but is that what most people want?

UPD:
Shield wrote:I started using PB some 12 years ago and I feel it has never been more quiet on the forums.
Exactly! I didn't mention it, but i do think that the community is dying and this is a very bad sign. Some changes has to be made.
walbus
Addict
Addict
Posts: 929
Joined: Sat Mar 02, 2013 9:17 am

Re: Some thoughts about PB in its current state

Post by walbus »

On relayse 5.44 i would leave PB
Unfixed old bugs, with a very long bearth, i self can not accept

Now, PB560 was coming and a very nice relayse
A bug free relayse is not what i want
I want a better and faster bug fixing and a better support for the bug handling

I need not a lot new features, i need stability and fixes for known old bugs before

I myself would like to pay for new relayses, everyone should earn money with his work, otherwise it makes no sense, i think..
User avatar
Kukulkan
Addict
Addict
Posts: 1352
Joined: Mon Jun 06, 2005 2:35 pm
Location: germany
Contact:

Re: Some thoughts about PB in its current state

Post by Kukulkan »

Listen all,

deseven told us his opinion and I think we all agree on the one or other issue. But, as he claimed to drop PB, why to discuss? It was his opinion and that's it. We have to accept.

I agree on some points and I disagree on many others. We develop commercial tools using PB for all three platforms. Both in 32 and 64 bit and not only executables but also DLL's (.so, .dylib) for our power users. Yes, we use our own curl and also external cryptography. But hey, try to use C++ or .NET and you will have to include other libs as well. This is why there are QT, wxWidgets and others, too.

We personally do not use the game stuff, but it attracts other developers. The bigger the user basis the bigger the chance that a product is maintained and alive. I have no problem with that.

Even if it is sad that deseven leaves, looking forward to the future and many new users who love PB. Everybody uses the tools that suit him/her best.

Peace :)

Kukulkan
walbus
Addict
Addict
Posts: 929
Joined: Sat Mar 02, 2013 9:17 am

Re: Some thoughts about PB in its current state

Post by walbus »

Nobody (and nothing) is perfect, and this is ok

To time PB is for my the best solution

But, a old martial master has say to my, before many years : "We can all make better" :wink:
uwekel
Enthusiast
Enthusiast
Posts: 740
Joined: Sat Dec 03, 2011 5:54 pm
Location: Oldenburg (Germany)

Re: Some thoughts about PB in its current state

Post by uwekel »

I dropped Purebasic as well, because of too bad Linux support. I wrote a lot of bug reports, but none of them have been fixed. See here: http://www.purebasic.fr/english/viewtop ... 15&t=68135 topic. I will keep in eye on the development, and maybe i return if Linux support has been improved.
PB 5.70 LTS (x64) - Debian Testing, Gnome 3.30.2
walbus
Addict
Addict
Posts: 929
Joined: Sat Mar 02, 2013 9:17 am

Re: Some thoughts about PB in its current state

Post by walbus »

gtk3 is a karma
I have migrated to gtk3 my tools; and after a half year, other gtk3 relayses, a many looking other, catastrophe
So i have changed all back to gtk2
Problems with little buttons, with color buttons, problems with different button sizes on x86/x64, a lot work, and all now for the trash can
For compatipility with all OS, i use now only gtk2 on PB for Linux

To time on 560 i can primary make compatibility to all OS my tools, and PB for Linux runs stable now

Making compatibility to mac OS, is also not ever simple

All over, for my "simple" tools, i can now good work with 560 on all OS

But i will say it again, i need not a lot new functions with new bugs and old bugs

Who wants to build a house needs a stable foundation
Last edited by walbus on Thu Jul 27, 2017 7:04 pm, edited 1 time in total.
User avatar
Tenaja
Addict
Addict
Posts: 1948
Joined: Tue Nov 09, 2010 10:15 pm

Re: Some thoughts about PB in its current state

Post by Tenaja »

I went with PB because I thought it was "cross platform". I love PB for small simple programs. But for anything with any measure of GUI complexity or customization, it is hardly any more cross platform than C with GTK or Qt. I have so many compilerif statements to cover the OS version that I may have well just gone with C, and will probably do that with my next major rewrite of my program that was the reason for coming here.

I do NOT think PB should Open Source. That will dry up the funds for development. What I would like to see would be a better way to add libraries so they behave as if native, and can be easily shared (even if written in C). Maybe this would be a PB wiki, or a PB repository, but the forum is so scattered it is difficult to find stuff. Currently, when people are sharing their libs on Tips & Tricks, nobody seems to add keywords that someone else might be searching for, so a lib might be there, but due to lack of term standardization it can be difficult or impossible to find. A wiki or some other db should be able to resolve this, with tags and sorting.

Sure, a few other guys have made attempts at library depository sites, and some have been helpful, but either they are small, poorly sorted, or poorly documented. And, then they move on and the site shuts down like tailbite. None of this helps the PB community as much as it could.
User avatar
luis
Addict
Addict
Posts: 3876
Joined: Wed Aug 31, 2005 11:09 pm
Location: Italy

Re: Some thoughts about PB in its current state

Post by luis »

deseven wrote: Here are my thoughts. What are yours?
Mine (about its current state) are listed in the page linked below.

I certainly hope x86 will not be removed, but hoping and asking for ascii builds to not be removed didn't help me much so... (btw: I didn't get any tangible compensation in exchange for the loss).

Also I stopped discussing in the forum about how PB could be enhanced and I now use that time for something useful which may produce results.

I still use PB for small toy projects, and when I'm not pulling my hairs off I still have fun with it thanks to its immediateness and simplicity, probably its strongest points.

And yes, in my view the forum deteriorated.
"Have you tried turning it off and on again ?"
A little PureBasic review
User avatar
kenmo
Addict
Addict
Posts: 1967
Joined: Tue Dec 23, 2003 3:54 am

Re: Some thoughts about PB in its current state

Post by kenmo »

I think the forums are fine...

1. Since it's used for coding questions, bug reports, & feature requests, it seems less active when things are mostly-stable and users are happily programming away.

2. It's more active around releases, when people are playing with new features and helping test betas. Maybe we need beta releases more often :)

3. ALL the traditional BB-style forums I know have been less active over years, I think it's a move to repos for code, bug trackers for issues, Google for finding quick solutions, social media for off-topic...
Locked