Am 01.09.2005 um 00:12 schrieb Jecel Assumpcao Jr:
In short: the Self world lacked its Andreas and Ian which was the main factor why Self and Squeak followed different paths.
But if you then look at the development of Tweak/Croquet and the latest discussions on the vm-list, it seems they learned their lesson and now are using a much more closed (hostile?) form of development...
So something clearly did not went the right way with Squeak from their perspective.
One thing that the Traits discussion again has shown that it is very hard to innovate in Squeak. There is a huge hostility against doing new things, even if the alternative is complete standstill.
Marcus
Marcus Denker wrote:
Am 01.09.2005 um 00:12 schrieb Jecel Assumpcao Jr:
In short: the Self world lacked its Andreas and Ian which was the main factor why Self and Squeak followed different paths.
But if you then look at the development of Tweak/Croquet and the latest discussions on the vm-list, it seems they learned their lesson and now are using a much more closed (hostile?) form of development...
So something clearly did not went the right way with Squeak from their perspective.
One thing that the Traits discussion again has shown that it is very hard to innovate in Squeak. There is a huge hostility against doing new things, even if the alternative is complete standstill.
Marcus
Jecel- Thanks for the history lesson. I think I agree with your points.
Marcus- But if one wants the status quo, surely one just adopts a specific Squeak. Traits would seem at this point to be the best of both worlds: 'old' Squeak is unaffected, and experimenters can try out the newer technology.
David
Marcus Denker wrote on Thu, 1 Sep 2005 01:05:22 +0200
Am 01.09.2005 um 00:12 schrieb Jecel Assumpcao Jr:
In short: the Self world lacked its Andreas and Ian which was the main factor why Self and Squeak followed different paths.
But if you then look at the development of Tweak/Croquet and the latest discussions on the vm-list, it seems they learned their lesson and now are using a much more closed (hostile?) form of development...
My impression is the opposite. From the very start Squeak has been about rumors, followed by neat demos, with public releases only happening much later. Both Tweak and Croquet started out following this tradition and then became much more open last year.
So something clearly did not went the right way with Squeak from their perspective.
The first major part of Squeak that was developed in a relatively public way was the 3.3 modules and that led to a lot of anger and people leaving our community. While I don't think there was any relation at all, I can understand it if some people hesitate to do things this way again.
My impression is that most of the core Squeakers like the open source way but aren't fanatics about it. Only John Maloney has spoken out against it and feels Scratch is better as a closed product.
One thing that the Traits discussion again has shown that it is very hard to innovate in Squeak. There is a huge hostility against doing new things, even if the alternative is complete standstill.
Things have always been this way. I have told this story several times, but perhaps some people still haven't heard it. Back in 1998 I had to make a serious choice in my project: either go with Squeak or develop my own technology (with the goal of making it more Squeak compatible over the years). My own technology would be based on Self so I would have to buy a Sun machine (without a monitor!) for US$5000 as a development system. If I adopted Squeak instead I could get the most high end PC possible (to compensate for Squeak's lower performance) and would still have money left over for other investments. And to me a strong community is far more important than technology (see Linux vs BSD). So I poked around to see what people's reaction would be if I tried to get the simplifications I felt were needed into the core of Squeak. My impression was that a vocal (and possibly large - it is hard to tell) part of the community would reject my changes and I would end up with a fork, all alone with a few people cheering on from the sidelines. We will never know if I was right, of course. My decision was that if I couldn't have the community anyway, then I might as well not compromise on the technology.
Meanwhile, other people have been following other paths. Some are very conservative while others tried new things. I think there is very good and allows the community to accept things as they demonstrate their value rather than having to bet on them beforehand. Once the demonstrations are available, however, rejecting them without trying doesn't make sense.
-- Jecel
Marcus Denker wrote:
Am 01.09.2005 um 00:12 schrieb Jecel Assumpcao Jr:
In short: the Self world lacked its Andreas and Ian which was the main factor why Self and Squeak followed different paths.
But if you then look at the development of Tweak/Croquet and the latest discussions on the vm-list, it seems they learned their lesson and now are using a much more closed (hostile?) form of development...
Hey, what's your gripe, man? What exactly do you mean by "the latest discussions on the vm-list" (the latest discussions were about socket problems; before that we had discussions about how to use SVN/CVS, before that we were discussion problems with the 64bit changes - I fail to see how that relates in any way or shape with the above) and what exactly do you mean when you talk about a "hostile" form of development?
From my point of view, Tweak and Croquet follow the same form of development that Squeak used to follow: A small, dedicated group of people moving the project forward, making sure it is focused on a goal and making sure that it gets done. There has been no change in our form of development that I can see.
So something clearly did not went the right way with Squeak from their perspective.
Yes, that is true. There were some things that went "wrong" and that we felt could only be addressed by doing our own systems. Most importantly: We had lost control over our own destiny. That may be not too much of a problem if you're a hobbyist but if you are serious about doing something with Squeak on the scale of either Tweak or Croquet (or Scratch or any of the Impara projects for that matter) you *must* be able to control your own future. It is not acceptable to stand helplessly on the sidelines when something very bad is about to happen and all you can do is scream and kick (and if you do, you get flamed etc). It is not acceptable if you need a fix in your system to go through the endless processes to get it approved. It is simply not doable. We have always recommended that people do exactly that even back in the SqC days and most of those groups who were serious took the advice. Such as Interval or Exobox or StableSqueak. And now, Scratch, Tweak and Croquet do the same (which goes to show that we actually take our own advice of the past).
So the lesson that we learned here is that we were right in the past, that you do have to take your future into your own hands and that you have to avoid error 33 like hell (we paid for it, too). The lesson is to stay focused, to keep the ball running. Discussions can be helpful but only up to a point - we all know (and you are just leaning it again) that people will generally resist *any* kind of change. At times, when it's important enough you just have to shove it down people's throats. And some will not like that one bit and you'll loose them. And others will join you *because* of the very changes. That's life.
One thing that the Traits discussion again has shown that it is very hard to innovate in Squeak. There is a huge hostility against doing new things, even if the alternative is complete standstill.
Heh. I've been saying this for years. Wanna know what a good alternative is? Roll your own. Then you've got control over your own destiny and can build the best system of all. Yiiihaaaa! ;-)
Cheers, - Andreas
Hi andreas,
I agree with the rest but not that:
Heh. I've been saying this for years. Wanna know what a good alternative is? Roll your own. Then you've got control over your own destiny and can build the best system of all. Yiiihaaaa! ;-)
Are you suggesting that the two to three guys that spent their nights harvesting the changes that are useful for everybody should stop to do that and that we simply fork, that we do not collect all the good stuff that is been done and that your non-hobbyst projects can benefit on (new compiler, method annotations, fixes....)
Been a lonely player can be good for you but this is not the way we want to create and support our community. I think that with a good amount of respect and communication we all have to win.
Stef
PS: I think that StableSqueak failed for all kinds of reasons and in particular the lack of openness of SqC related group at that time, because Ginsu was what MC is not from the metamodel perspective back at that time.
PSPS: this was fun to see the remark of mike on the VM setup because I was not able to understand anything related to the discussion but its seems that he had a point. BIt gave me the bad impression that the VM maker is a rich club where kids should not play.
stéphane ducasse wrote:
I agree with the rest but not that:
Heh. I've been saying this for years. Wanna know what a good alternative is? Roll your own. Then you've got control over your own destiny and can build the best system of all. Yiiihaaaa! ;-)
Are you suggesting that the two to three guys that spent their nights harvesting the changes that are useful for everybody should stop to do that and that we simply fork, that we do not collect all the good stuff that is been done and that your non-hobbyst projects can benefit on (new compiler, method annotations, fixes....)
Mostly, I am suggesting to have some fun and stop being so goddamn serious and frustrated. How can you get anything done if you aren't having fun doing it?
Cheers, - Andreas
But we have fun.... We are also doing some stuff that are less fun but are necessary to get the community rolling. My view of the world is that there is nothing like a free lunch and that everything is synergetic. I want to have a powerful Squeak/Tweak/... community to protect my own freedom (you know been an idiot not teaching java...)
Stef
I agree with the rest but not that:
Heh. I've been saying this for years. Wanna know what a good alternative is? Roll your own. Then you've got control over your own destiny and can build the best system of all. Yiiihaaaa! ;-)
Are you suggesting that the two to three guys that spent their nights harvesting the changes that are useful for everybody should stop to do that and that we simply fork, that we do not collect all the good stuff that is been done and that your non- hobbyst projects can benefit on (new compiler, method annotations, fixes....)
Mostly, I am suggesting to have some fun and stop being so goddamn serious and frustrated. How can you get anything done if you aren't having fun doing it?
Cheers,
- Andreas
Hi Andreas,
Thanks for these comments. They will be very useful for me hopefully soon.
Cheers, Juan Vuletich
----- Original Message ----- From: "Andreas Raab" andreas.raab@gmx.de To: "The general-purpose Squeak developers list" squeak-dev@lists.squeakfoundation.org Sent: Thursday, September 01, 2005 1:29 AM Subject: Re: Smalltalk and Self
So something clearly did not went the right way with Squeak from their perspective.
Yes, that is true. There were some things that went "wrong" and that we felt could only be addressed by doing our own systems. Most importantly: We had lost control over our own destiny. That may be not too much of a problem if you're a hobbyist but if you are serious about doing something with Squeak on the scale of either Tweak or Croquet (or Scratch or any of the Impara projects for that matter) you *must* be able to control your own future. It is not acceptable to stand helplessly on the sidelines when something very bad is about to happen and all you can do is scream and kick (and if you do, you get flamed etc). It is not acceptable if you need a fix in your system to go through the endless processes to get it approved. It is simply not doable. We have always recommended that people do exactly that even back in the SqC days and most of those groups who were serious took the advice. Such as Interval or Exobox or StableSqueak. And now, Scratch, Tweak and Croquet do the same (which goes to show that we actually take our own advice of the past).
So the lesson that we learned here is that we were right in the past, that you do have to take your future into your own hands and that you have to avoid error 33 like hell (we paid for it, too). The lesson is to stay focused, to keep the ball running. Discussions can be helpful but only up to a point - we all know (and you are just leaning it again) that people will generally resist *any* kind of change. At times, when it's important enough you just have to shove it down people's throats. And some will not like that one bit and you'll loose them. And others will join you *because* of the very changes. That's life.
Cheers,
- Andreas
On Aug 31, 2005, at 6:05 PM, Marcus Denker wrote:
Am 01.09.2005 um 00:12 schrieb Jecel Assumpcao Jr:
In short: the Self world lacked its Andreas and Ian which was the main factor why Self and Squeak followed different paths.
But if you then look at the development of Tweak/Croquet and the latest discussions on the vm-list, it seems they learned their lesson and now are using a much more closed (hostile?) form of development...
Certainly not hostile.
So something clearly did not went the right way with Squeak from their perspective.
Speaking only for myself, I see Croquet development becoming much more of a Bazaar in the future. Also, I don't see Croquet's current Cathedral-ness as being a direct response to perceived difficulties innovating in Squeak. It is the result of many factors, which I won't list because I haven't thought about them enough (since I think the situation is temporary anyway).
Josh
One thing that the Traits discussion again has shown that it is very hard to innovate in Squeak. There is a huge hostility against doing new things, even if the alternative is complete standstill.
Marcus
Marcus Denker wrote:
One thing that the Traits discussion again has shown that it is very hard to innovate in Squeak. There is a huge hostility against doing new things, even if the alternative is complete standstill.
I disagree. We did not get almost anyone saying "you should not do this because I depend on Squeak to be backwards compatible". I was almost sorry to not hear anyone ask even the really critical question: "but does the RB work with Traits?". And quite a few people have cheered Traits on.
There is at the moment nothing stopping Traits, a quite significant change to the Smalltalk language, from entering the Squeak base image. There're lots of good reasons for this - its a careful idea, lots of effort has been put into papers that explain it all, lots of implementation work over a few years... and the result is we're almost ready. That "almost" is all we're waiting for, as far as I'm concerned.
Coming back to the topic, I think the community around Squeak has created lots of opportunities for all of us.
Andreas Raab wrote:
From my point of view, Tweak and Croquet follow the same form of development that Squeak used to follow: A small, dedicated group of people moving the project forward, making sure it is focused on a goal and making sure that it gets done.
SqueakMap allows everyone to innovate in parallel, including at lower commitment levels.
[snip]
It is not acceptable to stand helplessly on the sidelines when something very bad is about to happen and all you can do is scream and kick (and if you do, you get flamed etc). It is not acceptable if you need a fix in your system to go through the endless processes to get it approved. [snip]
MC can allow us (the wider Squeak community, including Squeakland, Tweak and so forth) to evolve our systems without really forking (because stuff can be easily merged around), and without really blocking on one another.
We have always recommended that people do exactly that even back in the SqC days and most of those groups who were serious took the advice. Such as Interval or Exobox or StableSqueak. And now, Scratch, Tweak and Croquet do the same (which goes to show that we actually take our own advice of the past).
I hope that from 3.9 and on people can eat their cakes and leave them whole, too. Create whatever they want, take whatever they want from Squeak, and give back what they see fit, without there being any need to:
At times, when it's important enough you just have to shove it down people's throats.
But we'll see.
Daniel PS - have you all noticed that we actually have a rewrite rule editor in the RB now? today I changed all the occurrences of OBClassNode on: <something> to (<something>) asNode, inside the relevant package, in 2 minutes flat, with no manual fiddling. It's Magic I tell you, Magic!!
squeak-dev@lists.squeakfoundation.org
-
Andreas Raab -
Daniel Vainsencher -
David P Harris -
Jecel Assumpcao Jr -
Josh Gargus -
Juan Vuletich -
Marcus Denker -
stéphane ducasse