Lex Spoon lex@cc.gatech.edu said:

Frankly, saying that SqueakL isn't "open source" seems ridiculous. Can you recall what the complaint was? Everyone can download all the source to Squeak, modify it, and redistribute modified versions. In fact, it takes some effort to distribute anything written in Squeak (including Squeak itself), *without* including the source code. If that's not open source, then what is?

Among others, I can't modify and pass on everything - for example, I'm not allowed to pass on the bitmap fonts without including all of Squeak; furthermore, I'm not free to pass it on to anyone I like, because the license requires me to adhere to some foreign country's export regulations. That may not make a difference for you (because as a US citizen, you are already bound by your government's laws), but it does make a difference for the other 96% of the world's population.

Open Source is *not* just being able to peek at the source. For example, even though for all practical purposes qmail by Dan Bernstein rocks, it is definitely *not* Open Source because it has very strict limitations on redistribution of source code (and as Open Source is about free speech, the amount of redistribution allowed is critically important for determining whether something is open source or not).

From the two examples above, it is clear that the Squeak License doesn't

pass the Open Source Institute's Open Source Definition, not the Free Software Foundation's definition of Free Software, so it seems to be inappropriate at the moment to label Squeak as open source. It's a pity, because it is mostly a technicality and quite likely not what Apple meant, but you want to be careful with language that is about legal things, so I think that Squeak should not be recommended as open source.

Lex said:

Why do these guys get to define "open source"?

Get 'em Lex !!!

Jim

"Joshua 'Schwa' Gargus" schwa@cc.gatech.edu wrote:

Because the term (capitalized or not) wasn't in common use before they coined it to differentiate it from Free Software.

I didn't realize it was the same people. I wonder how many people do?

Anyone from the Open Source or Free Software camp who becomes familiar with the workings of the Squeak community would agree that we operate in a manner consistent with the ideals that inspire them (share code, etc.) However, a line in the sand needs to be drawn so that entities (might as well name them) like Microsoft cannot release code that meets the letter of their definition, but actually restricts the freedom of users. Both OSI and FSF retain lawyers to make sure that such loopholes do not exist.

Good example switching it to Microsoft. Still, I am very confident it would be fine. Is there something I don't know about? The font and indemnification and export clauses are nothing for someone trying to use Squeak in practice. The only trouble they've caused in the years Squeak has been available, are for people trying to fit them under definitions like OSI's or Debian's. Squeak really is open source in every way except the strict OSI definition. What could Apple really do to someone using Squeak?

OSI is being too strict in rejecting Squeak License. Worse, they are doing it in a dirty way. When Debian describes software licenses, it is very careful to distinguish general terms like "open" from specific terms like "compliant with Debian's Guidelines". OSI, on the other hand, seems to be playing a political word game, much like FSF with "free": they take a term and change it's meaning to something different than what people expect. Squeak and its website is just one case in point of this -- you worry that businessmen will equate "open source" and "OSI", but that hasn't even happened (yet?) among us technical people.

More generally, isn't it wrong to tie the open source movement into an institution? An institution that several members of the community think is making bad moves? Why do we put up with this, guys? It's already widely accepted that you can be doing the free/open software kind of thing without buying into FSF. Are we now going to repeat the same saga with OSI?

-Lex

PS -- The specificity of "object-oriented" *does* bother me -- it should mean that there are objects floating around, not that you have classes and inheritance. Much worse is the treatment "type" has received.

PPS [ObSqueak] -- Terms are like classes, and their meaning is up to the programmer. Computers are a place to play pretend, and in Squeak, your terms only have to be consistent within one image. If I say a Tomato is a Vegetable in my image, and yours says it is a Fruit, I don't care -- your image can bite my image.

PPPS -- The increasing communication between Squeak objects over the network will change this.

On Mon, 29 Apr 2002, Göran Hultgren wrote:

[...] But I still think it (the phrase on Squeak.org) should either be written differently - or, if people really insist on keeping the current phrase "Squeak is open source." then IMHO it should be accompanied with an explanation of FACTS. [...]

Actually, the current wording is "Squeak comes under an open source license, meaning that you can download and use it for free."

This is only the intro page. All the detail is in the "About" page, including an explanation of the license terms.

All this may sound silly to a lot of you [...]

Indeed. I'm glad you acknowledge that ;-)

-- Bert

Lex Spoon lex@cc.gatech.edu said:

...

you worry that businessmen will equate "open source" and "OSI", but that hasn't even happened (yet?) among us technical people.

Hi all, I'm Mark Miller, and I run an open source project over at http://www.erights.org .

Lex, my perception is that this has indeed happened among both technical and business folks.

I would also argue that this is a good thing, but that's a separate matter. Briefly, the term "open source" is a line in the sand. If the term weren't made institutionally stable somehow, then interests on one side of the slippery slope would rapidly dilute all meaning out of it. I was at a weird closed (!?!) meeting where Bill Joy was trying to sell various leaders of the open source community on the acceptability of the ancestor of SCSL. If OSD hadn't pinned down a set of clear criteria, Bill Joy would have simply trumpeted this license as "open source" and the word would have lost all meaning. We have seen this over and over again with words that were coined to make a needed distinction, like "nanotechnology".

Note that this logic places much higher value on drawing a clear line and sticking with it than it does on getting the line right. Just like we draw a line on "adult" at 18. It's clearly arbitrary, but an arbitrary simple stable line is better than the mess that would result from trying to be more accurate. This is the thresholding of the world into categories that rules can apply to, just as an A-to-D converter thresholds the world into categories that digital logic can apply to. In both cases, the important thing is to remove ambiguity by making arbitrary choices. These choices don't remove actual ambiguity from our understanding of the world, but they enable disparate parts of a system to coordinate with each other in reacting to that world; since they can react to a shared abstraction of the world. Those who care about open-source need a shared distinction more than they need a good distinction. That said, I think the OSD has done a great job at creating a decent distinction.

In any case, I've read section 6 of SqueakL and don't understand the problem with it. I tried looking at the squeak archive, but it's no better indexed than mine ;), so I wasn't able to find an explanation of the problem. If this has already been hashed out, then please respond to me privately rather than on the list.

At 11:05 PM 4/29/2002 Monday, Cees de Groot wrote:

You seem to be thinking that "free software" as a term existed before the FSF used it, and similarly that "open source" as a term existed before (who?) coined it.

It was coined by Christine Peterson, the head of the Foresight Institute, cc'ed on this email.

[...] there are many organizations out there who would like to ride the open source wave but would also like to avoid paying their dues to the community, and when we ('we' as a bunch of guys who are clearly on the right side of things) start diluting the term ourselves, we cannot blame all sort of sleazy types to do the same.

Indeed! Further, we do their job for them.

---------------------------------------- Text by me above is hereby placed in the public domain

Cheers, --MarkM

Mark S. Miller markm@caplet.com said:

Hi all, I'm Mark Miller, and I run an open source project over at http://www.erights.org .

Hi Mark, glad you're here. Contemplating a port of E to Squeak? ;-) I like E, but it runs on a funny platform at the moment...

Note that this logic places much higher value on drawing a clear line and sticking with it than it does on getting the line right.

Thanks for pointing this out so clearly.

In any case, I've read section 6 of SqueakL and don't understand the problem with it.

From some off-list mail I wrote on the topic:

- Clause 6, export law assurances. This is explicitely mentioned in paragraph 5. of the OSD, together with a proposal of what is acceptable (and actually shows up in most OSD-compliant licenses).

But rather let's discuss E on Squeak here...

On Tue, 30 Apr 2002, Lex Spoon wrote:

"Mark S. Miller" markm@caplet.com wrote:

...

Note that this logic places much higher value on drawing a clear line and sticking with it than it does on getting the line right. Just like we draw a line on "adult" at 18. It's clearly arbitrary, but an arbitrary simple stable line is better than the mess that would result from trying to be more accurate.

This is a great example. In common parlance, "adult" is broader than "at least 18".

[snip]

Another wonderful example is "dead" (and "time of death"). Putting aside revivability issues for a moment, it's pretty clear that dying is a continuous process, thus the "moment" of death is a useful fiction. The key to that fiction is that you are on the *correct* side of the fuzzy line :)

Cheers, Bijan Parsia.