Pascal eXtended Library (aka Asphyre) released!

[ykot]

1. Afterwarp.Net domain now redirects to Asphyre.Net, which is a shorter and more accurate name for PXL project.

2. There were issues yet again with hosting provider, so PXL for a short period of time was hosted purely on SourceForge. However, now the web site is back online. Nevertheless, portions of "Asphyre.Net" web site will be mirrored on SourceForge to make sure the content stays. Please remember that it costs quite a bit to pay for the hosting, domains and so on - I try my best, but budget has shrunk significantly lately.

3. SourceForge, compared to GitHub, seemed more convenient in terms of features for PXL, which is why the library was moved there. GitHub provides great facilities on its own, but it's more source-code centric, whereas PXL is much more than just plain source code. I know that some people may be fans of one provider or another, sorry for any inconveniences this may have caused.

4. The library has been re-released under GPL, this is true. The library is primarily meant for academic and scientific usage, so it grants all four software freedoms, while being true open-source. If you need to use the library under different terms, you can always contact me to arrange a special permission. In that sense, the library is free as in free software, but it may not necessarily be "gratis" (as in free meal), if you want to start making money out of it.

In my perspective, these are very relaxed terms especially considering what the library is capable of and huge effort it takes to develop and maintain it. If you have a problem with this, I'm sorry but these are the only terms under which I will continue to develop and support this project.

By the way, a new version 1.0.1 has been released yesterday. Also the library has been renamed to "Platform eXtended Library" as "Pascal" may incorrectly associate it with older variants of non-Object Pascal, while still keeping "PXL" namespace.

[taazz]

Law is not retroactive. The choice is respected. I don't use sourceforge GPL version only legal MPL/GPL/LGPL GitHub branch v1 version.

law has nothing to do with respect. Your choice is not unlawful it is disrespectful, especially so if you take in to account that up until now he was kind of undecided about the license and he was trying out different approaches. Now that he made hes choice respect it.

[hnb]

4. The library has been re-released under GPL, this is true. The library is primarily meant for academic and scientific usage, so it grants all four software freedoms, while being true open-source. If you need to use the library under different terms, you can always contact me to arrange a special permission. In that sense, the library is free as in free software, but it may not necessarily be "gratis" (as in free meal), if you want to start making money out of it.

In my perspective, these are very relaxed terms especially considering what the library is capable of and huge effort it takes to develop and maintain it. If you have a problem with this, I'm sorry but these are the only terms under which I will continue to develop and support this project.

What a licence mess. You are kidding me? Maybe someday I should also change licence of my Generics.Collections and Sparta packages to pure GPL?  It was also big amount of my time, many years of my work.

Maybe you dislike this but I will create fork of PXL branch v1 called Sparta eXtended Library on GitHub. Sorry. I can develop this on my own hand.

[ykot]

I have copy of PXL under allowed triple licence model from GitHub (MPL 2, GPL 3 and LGPL 3, the last two having linking exception). Maybe I will upload my copy on GitHub again.

If you need to use the library in a closed-source project, I can always provide a special permission for you to do so. Sure, it may sound like a hassle, but at least I will have a privilege to know what project the library is being used in, while still feeling that my work is being appreciated and respected. Sending e-mail sounds like a small price to pay for a big project that I have been supporting for 17 years now, without asking for a single penny for it.

I would not appreciate if you would take some older version with different licensing terms and start to distribute it, this would be quite uncool. It will not have any updates from new version of the library (which if you cannot simply backport as they are GPL) and will just make the whole matter confusing. This entire "auto fork" feature of GitHub is one of the reasons I switched away from it, by the way.

[ykot]

What a licence mess. You are kidding me? Maybe someday I should also change licence of my Generics.Collections and Sparta packages to pure GPL?  It was also big amount of my time, many years of my work.

If you would, I would respect your decision and would not intentionally try to circumvent it.

Seriously, there is something wrong going on with this community...

[hnb]

If you would, I would respect your decision and would not intentionally try to circumvent it.

Seriously, there is something wrong going on with this community...

Thanks. It is open source nature. With community all is fine. You can not constantly change the opinion about licencing... Original author (you) of PXL and true PXL trunk will be mentioned in my small github fork.

Sending e-mail sounds like a small price to pay for a big project that I have been supporting for 17 years now, without asking for a single penny for it.

It is open source nature. Look at my sparta or generics.collections project or for mORMot (much more advanced than PXL or sparta). All is full and true open source.

[taazz]

If you would, I would respect your decision and would not intentionally try to circumvent it.

Seriously, there is something wrong going on with this community...

Thanks. It is open source nature. With community all is fine. You can not constantly change the opinion about licencing... Original author (you) of PXL and true PXL trunk will be mentioned in my small github fork.

Sending e-mail sounds like a small price to pay for a big project that I have been supporting for 17 years now, without asking for a single penny for it.

It is open source nature. Look at my sparta or generics.collections project or for mORMot (much more advanced than PXL or sparta). All is full and true open source.

No its not, its your nature. Open source has no nature other what the users bring to the table.

[hnb]

No its not, its your nature. Open source has no nature other what the users bring to the table.

Be serious. I got the product under the MPL/LGPL with static linking exception. The author himself made that licence, he was fully informed.

Some of my work starts with usage of that PXL product under THE MPL/LGPL with static linking exception. Author changed his mind. So what now? Should I waste my time and drop that project? My boss never fully agree on something like this. GPL or "extra licence" is not an option.

Look at AGG project. Author also changed his mind, no one is using new GPL based licence, all is forked from more liberal older version.

Following your logic: all users of AGG are evil, they don't feel open source licence spirit, any user of AGG/AGGPAS (or fork AggPasLCL contributor like Mattias Gaertner) is "bad guy" like me.

https://en.wikipedia.org/wiki/Anti-Grain_Geometry

[taazz]

No its not, its your nature. Open source has no nature other what the users bring to the table.

Be serious. I got the product under the MPL/LGPL with static linking exception. The author himself made that licence, he was fully informed.

I have no idea how informed he was. But it was plain as day that he was trying various licenses nothing was final even after the so called dual license share.

Some of my work starts with usage of that PXL product under THE MPL/LGPL with static linking exception. Author changed his mind. So what now? Should I waste my time and drop that project? My boss never fully agree on something like this. GPL or "extra licence" is not an option.

Yeah, that's a pile of crap. 1) you already have the license to use it. 2) no one asked you not to use it.

Look at AGG project. Author also changed his mind, no one is using new GPL based licence, all is forked from more liberal older version.

Following your logic: all users of AGG are evil, they don't feel open source licence spirit, any user of AGG/AGGPAS (or fork AggPasLCL contributor like Mattias Gaertner) is "bad guy" like me.

https://en.wikipedia.org/wiki/Anti-Grain_Geometry

An other pile of crap. No, all users of AGG are not evil, the one that forked it might be disrespectful but I haven't followed those events to have an opinion. I see no reason to fork the library at this point there is little or no investment from the community, no bug reports, no patches not even a feature request. You are among the lucky ones that got the library under dual license and that gives you the right to use it. The only reason that would be acceptable for a fork is if there are patches/changes in the library, in that case you are obliged by the license it self to open source it. But simple forking at this time is disrespectful and for me plain wrong.

[ykot]

Look at AGG project. Author also changed his mind, no one is using new GPL based licence, all is forked from more liberal older version.

Similar to your previous examples, this example seems merely an exception rather than a rule. In case of AGG, the original author has passed away, which is kind of exceptional scenario. It is also a bit different in sense that C/C++ user base is significantly larger than Pascal [1].

In case of PXL, the situation is quite different. As sad as it may sound, Pascal community appears to be quite small these days, so PXL, similarly to many other projects, has bus factor close to one. Although there are parts of the project, where others can contribute to, the reality is that it is difficult to find someone else, capable of truly maintaining and further developing it as a whole.

Following your logic: all users of AGG are evil, they don't feel open source licence spirit, any user of AGG/AGGPAS (or fork AggPasLCL contributor like Mattias Gaertner) is "bad guy" like me.

Well, contrary to AGG, the original author of PXL is, hopefully, still alive and can be reasoned with. I'm open to any different licensing options, including the project as a whole, but so far I haven't found any meaningful evidence why it should be done so. Even yourself, you did mention that you wanted to use PXL for something, but that still remains pretty unclear, so I really don't understand why it is such a problem. In fact, from what you are saying, I actually get more convinced that using multiple licensing choices as you initially asked was probably a mistake to start with.

[marcov]

I actually get more convinced that using multiple licensing choices as you initially asked was probably a mistake to start with.

About ten years ago  I wrote this text as a discussion starter about licensing for a local (F)BUG.

[ykot]

About ten years ago  I wrote this text as a discussion starter about licensing for a local (F)BUG.

Similarly, about ten years ago, New Horizons space probe was launched to study Pluto. But how this is not an off-topic?

[marcov]

The point is that people won't even look at a codebase if the license doesn't match free(*) criteria. Simply because it is often a lot of responsibility and bother to take on when you are not even sure if it is what you seek. Then it is better to look for a better licensed alternative, fork an older better licensed version, or start your own.


(*) as in English "free", not "GNU" free.

[ykot]

The point is that people won't even look at a codebase if the license doesn't match free(*) criteria. Simply because it is often a lot of responsibility and bother to take on when you are not even sure if it is what you seek. Then it is better to look for a better licensed alternative, fork an older better licensed version, or start your own.

First, this sounds like a careless generalization. Second, in my own perspective, for academic and scientific usage the current approach is perfectly fine; even for a commercial/closed-source usage there is an option to it. As it is, I think this is a much better approach than making the project closed source and/or abandoning it.

If you try to imply that at the moment you want to open-source something, you should immediately use the most liberal license choices available - isn't it up to each developer to decide? The part you are probably trying to reply to, which you did not quote is:

...so far I haven't found any meaningful evidence why it should be done so.

To clarify: I'm not talking about evidence or any religious discussions about why certain license is better than another, which is probably what you are trying to discuss, but that's an off-topic discussion. What I meant is evidence regarding why current PXL licensing options are specifically unsuitable for a certain specific use-case Hnb, or anyone else, might be experiencing.

If you want to use PXL in a closed-source project - there is an option for it (contact me for a special permission). If you want to make a commercial video game with it - there is an option for it (same as before). If you want to use it for academic/scientific work - current license doesn't restrict you in any way. The only scenario that I can think of, is actually selling source code which contains parts of PXL - even in this case we might be able to negotiate something, but it may or may not be "gratis" for you.

I think this is as relaxing as it may get, while remaining both professional and business-friendly.

[hnb]

Yeah, that's a pile of crap.
...
An other pile of crap.

Licensing is important for companies, and for many normal users. Changing the licence by owner is the reason for the creation of fork (example already mentioned). Maybe in your reality something is different.

Tazzz, what time is good for forking? Should I write a petition to you? Anyone creates a fork when needs.