GPL and some open-sorser: I indignant!
I apologize in advance that I will retell long-known things, but I ran into it myself the other day, and I just can't help saying it :)
[summary: I wanted to use the so-called “free software” for my needs, spent a lot of time, and eventually left with nothing].
The story is. There is a customer selling some software product. For this product, it is proposed to write a very useful, but in general, an optional module (M). In my opinion, the module consists of two large algorithms, one of which (M1) in its pure form does not exist in nature (that is, it must be composed), and the second (M2) is essentially a library procedure, although and very nontrivial.
')
Of course, you can write yourself as M1, and M2, but why reinvent the wheel? And I went for the M2 on the Internet. And what did I see there in these open spaces?
At first glance, everything is not bad. Such modules are available, and even in large quantities. But as it often happens, a closer look instantly cuts off more than half of the projects - incomprehensible semi-abandoned undertakings, “experimental versions” that are not recommended by the authors themselves for real use, some grains extracted from large systems, not the fact that they work independently of everything else .
With this, everything is clear, the normal situation. But then I found myself in a position of very little choice: either buy (for big, by the way, money) a module from a commercial supplier, or use the open source. And then I got :) The fact is that the M2 module is a non-trivial thing, as I said, and almost all the implementations somehow left the university laboratories. And there, as they say, let's face the facts:
- Half of all projects are written in Java. We are not suitable (need a code, customer requirement).
- The quarter is somehow unavailable for "left" reasons (the main developer wrote a diploma and got rid of; the project is closed, we all do other things, and we have no time to mess around with the old code; everything is done to write a scientific article, and use in real life - Um, hammer - and so on). It's a shame, but overall it’s also understandable.
- But the remaining fourth is, in fact, the subject of my indignation, and in fact - spitting at those to whom this project can be at least somehow interesting.
Again, let's face the facts: in universities (I myself work at a university) we write software to do research, compose publications, and go to conferences. Such a life forms a kind of parallel value system, with its pros and cons. In this case, the obvious disadvantage is the development of software with the sole purpose of supporting the work of its own narrow group. Those. Software users - well, five to ten people, and the rest are just not interested in them.
In an ideal world, universities that are tax-efficient societies should benefit from this society. And the development of software (possibly free), which does not reach the hands of the industry - a good step in this direction. Here are just a few who take this step.
The favorite story in practice is Linux and the GPL. A typical statement: “we write on Linux, under Linux we compile everything, do not ask about other platforms, we are not interested”. Well, well, you love Linux, your business, you have the right. I have nothing against Linux. But how is it that “you are not interested”? Those. 90% (and more) of potential users of your software are not interested in you? Where, then, is justice, where do taxes go? ;)
And the GPL is generally the invention of the devil. If the software is published under the GPL license, it cannot be used in a package that is not covered by the GPL license. The authors argue as follows: they say it would be wrong if the company took the GPL-software, finalized it, and then competed with us. Logical and reasonable. But how could one reasonably deal with this? Require to distribute the GPL-code for free, even with a commercial program? Of course. Disallow release of modified software under a non-GPL license? Let it be so! But after all, they go further - suppose I have a text editor that archives files in ZIP when writing text to disk. It would seem that I am not a competitor to the archiver from any side. However, you cannot use the GPL ZIP-archiver (you cannot even call the executable file via the command line)! Release, my friend, your entire text editor under the GPL license! (There is, of course, a much less evil LGPL, but the M2 module I need, distributed by the LGPL, did not come across).
That is, in fact, all this software turns out to be unsuitable for serious use by anyone. Do you want "free software"? Yeah. First, use our operating system, and then also sign the contract in blood, that from now on all your software (in which our library is only a small part) will be distributed "freely." Such a model can be called whatever you like, but not "free."
Result: all the resources that dozens of universities spent on different versions of the M2 module did not give me anything. I negotiate with a commercial company.
Findings:
If you are writing free software, think at least a little about your users. You are welcome! Let there be at least some documentation (sometimes it comes to idiocy: source files are megabytes, and all documentation is a readme file for 100 lines. Well, if you have written so much code, can you not find time for a more decent introduction?)
Let your project be at least a bit portable to other platforms besides your favorite.
Don't GPL! Give commercial users any additional conditions: let them pay you for the right to use your software, let them undertake to release any modified versions of your project for free (that is, let them upgrade and debug your code for their money), let them stuff your name in the About box. But do not chop at the root of their opportunities! The ban on calling GPL zip-archiver from commercial software (under any circumstances!) Is nonsense and a great loss for everyone. In the end, the bulk of the popular “free software” for the developer is not the GPL.
Sorry to boil over. A week spent searching. Now, apparently, we will pay big money to capitalists, who will continue to develop their own (quality product), even more moving away from the seemingly attractive open source servers.
[summary: I wanted to use the so-called “free software” for my needs, spent a lot of time, and eventually left with nothing].
The story is. There is a customer selling some software product. For this product, it is proposed to write a very useful, but in general, an optional module (M). In my opinion, the module consists of two large algorithms, one of which (M1) in its pure form does not exist in nature (that is, it must be composed), and the second (M2) is essentially a library procedure, although and very nontrivial.
')
Of course, you can write yourself as M1, and M2, but why reinvent the wheel? And I went for the M2 on the Internet. And what did I see there in these open spaces?
At first glance, everything is not bad. Such modules are available, and even in large quantities. But as it often happens, a closer look instantly cuts off more than half of the projects - incomprehensible semi-abandoned undertakings, “experimental versions” that are not recommended by the authors themselves for real use, some grains extracted from large systems, not the fact that they work independently of everything else .
With this, everything is clear, the normal situation. But then I found myself in a position of very little choice: either buy (for big, by the way, money) a module from a commercial supplier, or use the open source. And then I got :) The fact is that the M2 module is a non-trivial thing, as I said, and almost all the implementations somehow left the university laboratories. And there, as they say, let's face the facts:
- Half of all projects are written in Java. We are not suitable (need a code, customer requirement).
- The quarter is somehow unavailable for "left" reasons (the main developer wrote a diploma and got rid of; the project is closed, we all do other things, and we have no time to mess around with the old code; everything is done to write a scientific article, and use in real life - Um, hammer - and so on). It's a shame, but overall it’s also understandable.
- But the remaining fourth is, in fact, the subject of my indignation, and in fact - spitting at those to whom this project can be at least somehow interesting.
Again, let's face the facts: in universities (I myself work at a university) we write software to do research, compose publications, and go to conferences. Such a life forms a kind of parallel value system, with its pros and cons. In this case, the obvious disadvantage is the development of software with the sole purpose of supporting the work of its own narrow group. Those. Software users - well, five to ten people, and the rest are just not interested in them.
In an ideal world, universities that are tax-efficient societies should benefit from this society. And the development of software (possibly free), which does not reach the hands of the industry - a good step in this direction. Here are just a few who take this step.
The favorite story in practice is Linux and the GPL. A typical statement: “we write on Linux, under Linux we compile everything, do not ask about other platforms, we are not interested”. Well, well, you love Linux, your business, you have the right. I have nothing against Linux. But how is it that “you are not interested”? Those. 90% (and more) of potential users of your software are not interested in you? Where, then, is justice, where do taxes go? ;)
And the GPL is generally the invention of the devil. If the software is published under the GPL license, it cannot be used in a package that is not covered by the GPL license. The authors argue as follows: they say it would be wrong if the company took the GPL-software, finalized it, and then competed with us. Logical and reasonable. But how could one reasonably deal with this? Require to distribute the GPL-code for free, even with a commercial program? Of course. Disallow release of modified software under a non-GPL license? Let it be so! But after all, they go further - suppose I have a text editor that archives files in ZIP when writing text to disk. It would seem that I am not a competitor to the archiver from any side. However, you cannot use the GPL ZIP-archiver (you cannot even call the executable file via the command line)! Release, my friend, your entire text editor under the GPL license! (There is, of course, a much less evil LGPL, but the M2 module I need, distributed by the LGPL, did not come across).
That is, in fact, all this software turns out to be unsuitable for serious use by anyone. Do you want "free software"? Yeah. First, use our operating system, and then also sign the contract in blood, that from now on all your software (in which our library is only a small part) will be distributed "freely." Such a model can be called whatever you like, but not "free."
Result: all the resources that dozens of universities spent on different versions of the M2 module did not give me anything. I negotiate with a commercial company.
Findings:
If you are writing free software, think at least a little about your users. You are welcome! Let there be at least some documentation (sometimes it comes to idiocy: source files are megabytes, and all documentation is a readme file for 100 lines. Well, if you have written so much code, can you not find time for a more decent introduction?)
Let your project be at least a bit portable to other platforms besides your favorite.
Don't GPL! Give commercial users any additional conditions: let them pay you for the right to use your software, let them undertake to release any modified versions of your project for free (that is, let them upgrade and debug your code for their money), let them stuff your name in the About box. But do not chop at the root of their opportunities! The ban on calling GPL zip-archiver from commercial software (under any circumstances!) Is nonsense and a great loss for everyone. In the end, the bulk of the popular “free software” for the developer is not the GPL.
Sorry to boil over. A week spent searching. Now, apparently, we will pay big money to capitalists, who will continue to develop their own (quality product), even more moving away from the seemingly attractive open source servers.
Source: https://habr.com/ru/post/94168/
All Articles