From what I can tell, most contributors to Perlmonks are
here because they enjoy it. They answer questions for free
because they want to and usually they learn a thing or two
themselves in the process.
Right now on Usenet in comp.lang.perl.modules, there is a
discussion about Graham Barr and his lack of response to
bug reports and patch submissions for his libnet
suite of modules. There are maybe 3 general viewpoints:
- Free software, commercial support contract
- Free software, free support
- Free software, limited free support then commercial
An interesting point brought up by the author of
Tangram is that free is not really free, it is a
loss of money and time by the author. Because for every
hour he does something for free, he could have been
earning an hour's pay.
The situation with Graham is in my eyes a very serious one.
Because he has authored several things which are a part of
the core distribution.
So this issue becomes should Perl remain a freeware model?
Should it have been shareware to begin with? And here is
my biggest question: is it only the software that is free
and is anything beyond that asking too much?
So, even though the source tree for Perl proper is just like
Larry showed at his Atlanta talk on Perl 6, the source tree
the libnet modules, which are a core part of Perl,
may soon see divergence due to inaction by its author.
Re (tilly) 1: Perl Commercial Entities?
by tilly (Archbishop) on Mar 19, 2001 at 22:31 UTC
|
I have been programming for under 4 years now. I sincerely believe that it would have been impossible for me (admittedly a motivated learner with an excellent technical background) to have learned as much about programming as I have without Perl being open source (not shareware) and full of people who are generous with their time and knowledge.
This is one of the reasons that I am willing to be generous back. Because I have benefited and I would like to see the torch passed.
Based on my experiences and discussions with others over that time, I honestly believe that Perl is much better for being open source. Were it shareware a number of issues would be raised, starting with fights over how to divide the loot and an attitude of, What's in it for me? By contrast as it stands you can reasonably have people like me come out of nowhere, learn it, and contribute back bug fixes and patches. There is real value in that.
For those reasons and many, many more, were Perl to become shareware I would stop calling myself a Perl programmer the next day.
However there are things that the open source model does not supply. And yes, if you want a support contract you will have to pay. If you want programmers, you will have to pay. If you want a security audit, you will have to pay. If you want books, you will have to pay. And pay. And pay.
So Perl is far from a free product. And by trying to make people pay up front (rather than down the road) you will limit its utility, discourage contributions, and lessen the other sources of revenue... | [reply] [Watch: Dir/Any] |
|
Shareware was just a stage on the road to open source.
I used to buy cheap shareware sometimes. Trinkets like a CD player and an HTML editor. I would not do so now, since Perl provides most of this kind of stuff for free.
Better yet, Perl gives me the tools to build my own tools.
The great thing about Larry Wall is that he is not at all conceited about the value of his brainchild. Larry talks up the Ruby language in interviews. And I doubt Larry would claim that Perl is soooooo great that one should buy it instead of using its competitors for free.
| [reply] [Watch: Dir/Any] |
|
Shareware? No, I don't think it was a step towards OSS, after all Free Software Foundation was started quite some time before the idea of shareware, which was really just a way to provide closed source (and free as in sample) taste of commercial proprietary game(s) coming up.
Great idea, but not to mix with FS nor OSS.
| [reply] [Watch: Dir/Any] |
|
Your post hit home with me. I am in the same position.
I certainly didn't have money in graduate school to buy
all the Perl modules I used and I sure needed them. So
I want to help out too.
But you are missing my point:
And yes, if you want a support contract you will have to pay. If you want programmers, you will have to pay. If you want a security audit,
you will have to pay. If you want books, you will have to pay. And pay. And pay.
But neither of these is the issue. What these people want
is an email response back regarding some patches that they
have worked on to libnet and they are getting
no feedback. They are not whining and complaning, they are
actually fixing libnet and Graham isn't
getting back to them.
| [reply] [Watch: Dir/Any] |
|
Situations like this have come up previously here at the Monastery, and I think tye gave a very useful answer.
| [reply] [Watch: Dir/Any] |
Re: Perl Commercial Entities?
by deprecated (Priest) on Mar 19, 2001 at 23:23 UTC
|
Monks, I have yet to lose my temper on this site. But this post is perhaps one of the most infuriating things I have ever read on this site.
This kind of thinking is what destroys the quality of software.
Software developers like me hesitate to write and distribute
software because of users like you. Users, for some reason or
other, feel that they are entitled to some kind of support.
Well, for commercial software, thats true, but only under the terms
of the software you (dont) read when you buy it. Microsoft, Apple, and IBM
(the three big vendors my clients use) all sell not only the hardware
and (closed source) software, but they sell support contracts for it.
ugh! This really makes me angry! Your inferrance is that because
somebody has written modules that have become core modules in perl
that it is their responsibility to maintain them? What then,
is my motivation for distributing any software? The responsibility
you are ascribing Graham means that there will always be some user
sucking at the teat of free time and support. Let me show you
some examples of why this reasoning sucks.
When I needed a device driver for a PCMCIA Ethernet Card and CDROM
for my laptop, I WROTE ONE. When I needed a
package manager for OpenBSD so that my girlfriend
could manage software on her laptop, i WROTE ONE.
I wrote these pieces of software so that my life would be easier,
not to increase global harmony, not to make your perl more functional,
not to make anyone but myself and my girlfriend happy. I distribute
software I write because I know other people can use it. I dont charge anyone
for it because that is against my morals and ethics. I'm willing
to support users (read: you) when I have the time and when
I am so inclined. If my software becomes a core piece of somebody's
software distribution (as my package manager may become for OpenBSD),
it is still not my responsibility to support or update it. It is
my option!
That having been said, let me move on to an additional subject which is
deeply relevant here. The above examples (the drivers and package managers)
could only have been written with extensive documentation and opensource.
To make a product proprietary and closed source prevents development.
Prevents updating. To make perl 'shareware' or 'limited support' is to
kill not only its community, but its future development. To completely cripple
any chance it has to evolve and progress as a language. Think how many
people use EMACS. Emacs is on something like version 21
right now. How far do you think it would get if emacs went 'shareware' or
'limited support' tomorrow? I'll tell you exactly what would happen.
A clone would appear in two weeks called "freemacs". Users would support
it as rabidly as they had been supporting emacs, and it would get many new
features from eager users. See the progress of awk through berkeley and gnu systems.
If you watch the evolution of software, you often see that when something becomes
opensource, it rapidly gains features (and sadly bugs too) and changes rapidly
immediately, and then it settles down and becomes more consistent. Then all these
new user-contributed changes become part of the core standard by which that
utility is judged. gmake and gnu-ld are other examples of this.
Your thinking is backward. Its also harmful to perl. Its harmful to
software development on a whole. And I really dont like it!
Software development has finally come to the hands of the people.
No more big brother forcing us to use
their tools and their software and their languages.
Please dont try to get rid of the magic. Learn to code. Learn to understand code.
And if you have problems that you can fix -- fix them and help others.
But dont tell me I am required to be generous. I am generous because of
who i am. Maybe that should be another programmer virtue, generosity.
I'll finish ranting, I think ive said my piece.
fuming,
brother dep.
--
laziness, hubris, impatience, and generosity | [reply] [Watch: Dir/Any] |
|
I am entitled to some support at least up to the point at which you tell me that you will not provide it. At that point I am not entitled to any support whatever, and it would be rude for me to ask again.
Most CPAN modules are maintained by developers who do provide support. It is reasonable to expect —unless a developer says up front that he will not be providing any support— that he will be providing it.
Not answering legitimate e-mail is not really polite. If a developer is not going to support his product then he ought not to give the impression that he will support it by providing his email address.
| [reply] [Watch: Dir/Any] |
|
No, you are not entitled to any support, whether or
not the email address is present.
The purpose of that email address is to document who wrote
the software. But they have no obligation to provide
support, answer questions, or supply fixes for problems.
They also have no obligation to accept patches, treat you
politely, etc.
In point of fact you are likely to get all this and more.
But I assure you that if you tell the authors of the
software that their generousity has obliged them to
do anything in the future, they will generally not be
very happy about it. This is something that they
choose to do. If you treat them rudely you will
find that they are liable to choose to be rather
rude back.
By contrast companies truly are obliged to give support.
And you will note that they generally do not give one drop
more support than they feel they have to. IMHO that
reflects human nature. People don't like being obliged to
do things, and so do things they are obliged to do only
grudgingly. People like doing things they voluntarily do,
and are liable to go an extra mile...
| [reply] [Watch: Dir/Any] |
|
| [reply] [Watch: Dir/Any] |
(tye)Re: Perl Commercial Entities?
by tye (Sage) on Mar 19, 2001 at 22:18 UTC
|
Heh! No, we shouldn't be charging for Perl. If Graham can't support libnet, then someone else needs to take over. This has happened before and it will happen again.
-
tye
(but my friends call me "Tye")
| [reply] [Watch: Dir/Any] |
Re: Perl Commercial Entities?
by footpad (Abbot) on Mar 19, 2001 at 23:23 UTC
|
Adding just a bit to some of the previous responses:
....Because for every hour he does something for free, he could have been earning an hour's pay.
That's one way to look at it. Another is consider it an investment in the company and in future earnings. In business terms, it's called a "loss leader," something you do to demonstrate your skills, give back to the community that helped your success, and to drive business to your organization. There's a certainly a balance to be drawn, of course. However, freebies increase business for many companies.
Look at it this way: there are many individuals and organizations that have given away a lot of information, knowledge, and support to the Perl community. They're able to live off the results. (I'm thinking of Damian Conway, among others.)
The situation with Graham is in my eyes a very serious one.
Not really. If no one steps up, as tye and davorg suggested, it's likely that other modules would appear. Also, modules have been unbundled from the core distribution before.
...should Perl remain a freeware model?
Yes; it makes the language accessible for personal and experimental purposes. This provides grass-roots opportunities to compete with more monied organizations who've been less than forthcoming about the internals or who've had a less than stellar record providing quality tools/support. (Do I really need to name names?)
Should it have been shareware to begin with?
I think Perl's success over the past few years is due, in part, to it's free availability.
I know I wouldn't be using it if it cost $250/year to "subscribe" to a development license and obtain bug fixes.
...is it only the software that is free and is anything beyond that asking too much?
Tilly nailed this one. The strength of Perl lies in the compassion, wisdom, talent, and generosity of Larry and those that've contributed to whole.
These contributions and their quality have let Perl become far larger than the sum of its parts.
There's a risk with any software and a cost to learning, support, maintaining, etc. anything. Even Microsoft may fall, given the performance of its stock price over the past year. Heck, look at what happened to AT&T and IBM.
In any event, it's up to you to weigh the risks, costs, and benefits. If you feel relying on libnet is too risky, don't...however, keep in mind the cost of developing a replacement. Personally, I think libnet is a safe investment because even if Barr gives up on it, someone will pick up the torch and carry it on.
That someone could even be you.
--f
| [reply] [Watch: Dir/Any] |
Re: Perl Commercial Entities?
by davorg (Chancellor) on Mar 19, 2001 at 22:23 UTC
|
Am I missing something here? I really can't see what the
problem is. If Graham is too busy to maintain his modules
then someone with more spare time should take over.
And you can get commercial support for Perl thru
people like
ActiveState and
The Perl Clinic.
--
<http://www.dave.org.uk>
"Perl makes the fun jobs fun
and the boring jobs bearable" - me
| [reply] [Watch: Dir/Any] |
A reply falls below the community's threshold of quality. You may see it by logging in. |
Re: Perl Commercial Entities?
by Petruchio (Vicar) on Mar 19, 2001 at 23:16 UTC
|
There is a much remarked-upon ambiguity concerning the
word "free"; it seems you mean that perhaps Perl should
cost money. It can cost money, and I am more than willing
to sell you a copy. I cannot, however, prevent others
from either selling or giving you a copy as well.
When you ask whether it should have been shareware to
begin with, I take you to be asking whether, from a
practical standpoint, it should have been released
under a more restictive license which allowed some
people to sell it more profitably. It seems highly
unlikely that Perl would have proliferated so far
if this had been the case... and the modules which
presently worry you might well not exist.
In order to now change the license, making it more
restrictive, it would be necessary to obtain the consent
of what I would guess to be quite an enormous number of
authors. A centralized administrative structure would
need to be formed, which would impose a considerable
overhead in terms of both effort and expense.
Assuming all this occured, we would get to the real
crux of the question. Would a more commercial model
actually improve things? Of course we all know that
in companies, despite that people are being paid for
their labor, schedules have been known to slip and
bugs to linger. On the whole, even leaving aside the
monumental task of changing the nature of the
development model, the end result would be a net gain
in the reliability of Perl's development?
I doubt it.
Is asking for free things beyond free software asking
too much? In my opinion, asking for free software is
asking too much. You cannot ask of someone that they
write software for you at no cost. Despite having
no obligation to do so, many people do write software
and give you the right to copy and change it. Some
of them are paid to do so, some are not. In either
case, such people deserve our thanks.
Support is much the same as software; I can sell you
support, but I cannot stop other people from selling
or giving you support. The key difference is that support
is a more finite quantity than copies of a piece of
software. | [reply] [Watch: Dir/Any] |
Re: Perl Commercial Entities?
by Albannach (Monsignor) on Mar 19, 2001 at 23:25 UTC
|
for every hour he does something for free, he could have been earning an hour's pay
This only applies if what he's doing for free has absolutely no value to him
personally, and that is hardly likely. I think the common situation is that
an author has a problem, crafts a solution, makes it public, and
gets incredibly valuable thorough peer (and often better) review, suggestions and bug-fixes for free.
The author wins as his problem is solved and probably more thoroughly and robustly than
if he had kept it private. The rest of the comunity wins as another solid tool has been
made available.
Should it have been shareware to begin with?
While I agree with everyone else that Perl simply wouldn't be
what it is today if it had been shareware (though perhaps if it had
been commercial, but that's another debate), more to your current
issue, if it had been shareware and someone stopped supporting
their shareware offering, then the community would have no recourse.
The source is the key. (Gee that sounds way too obvious - my
appoligies to anyone who finds it a waste of bytes.)
--
I'd like to be able to assign to an luser
| [reply] [Watch: Dir/Any] |
Re: Perl Commercial Entities?
by extremely (Priest) on Mar 20, 2001 at 07:50 UTC
|
The situation with Graham is in my eyes a very serious one. Because he has authored several things which are a part of the core distribution.
So this issue becomes should Perl remain a freeware model? Should it have been shareware to begin with? And here is my biggest question: is it only the software that is free and is anything beyond that asking too much?
You left out a HUGE step between these two paragraphs.
One thing here doesn't lead to another. The questions when
someone isn't maintaining something free are:
- Is it a temporary failing? (profer patches publicly and spread the word where they are and that they will be fixed eventually. Hard on the users for a while.)
- Is it time to fork and maintain two sets? (Incompatible
goals or incapable of working together. Hard on maintainers,
which to fix or use?)
- Will the old maintainer "pass the baton"? (Disrupts error reports and confuses end users but easy on the maintainers.)
- Is it time to wrest control away? (Very disruptive
and divisive but prevents the fork'ed code division problems.)
One problem with a personality doesn't lend itself to questioning the whole philosophy behind the release of the
software in question.
The nice thing about free-source software is that it DOES survive the loss of a maintainer well.
Try to get a bug fixed in a commercial program where the
rightsholder is in bankruptcy and liquidation. You may not
EVER find who got the rights to the code
to even ask if they CAN fix it.
I like ya princepawn but this post looks like it was
assembled randomly from multiple sources. The paragraphs hardly go together at all, let alone ask a coherent question. =)
To answer the three and a half real questions: 1. "Yes, it has to, it has been given to the world in such a way that it can't handily be taken back." 2. "No, shareware doesn't work when you have to distribute the runtime engine to get the scripts to work." 3a. "Yes but I take exception to the word 'only' =)" 3b. "Yes! We are lucky Graham blessed us with his work, if he wants to stop helping, what did it cost you?"
--
$you = new YOU;
honk() if $you->love(perl)
| [reply] [Watch: Dir/Any] |
Re: Perl Commercial Entities?
by jlawrenc (Scribe) on Mar 20, 2001 at 02:16 UTC
|
My $0.02.
- Open source software is the best thing that could happen to the IT industry.
- I hope that someday I can give back to the OSS community like they have so generously given to me
- Like businesses, OSS communities need to evolve and develop to address new challenges that arise every day.
- OSS allows me to give BETTER service and support to my clients.
Off the topic:
A note re: libnet support. I did actually discover a build
problem under RH7 and libnet. Didn't know where to send my report as I thought Perlbug was maybe the wrong place...
As silly as this sounds, is there a knowlege base for this type of information/discovery? I would even put here but we kinda need a: Pitfalls and Caveats section.
ie/
Title: cpp on RedHat 7.0 is broken.
Symptom: LWP gets fail such as lwp-download
Affected Modules: IO::Socket
Fix: Obtain fixed cpp from RedHat.
I dunno, just an idea. It probably exists I just don't know about it yet.
J | [reply] [Watch: Dir/Any] |
|
I thought I was our Pitfalls and Caveats section...
duck
| [reply] [Watch: Dir/Any] |
|
|