Smalltalk › Squeak › Squeak - Dev

A process proposal for 3.10

_‹ Previous Topic Next Topic _›

Classic

List

Threaded

132 messages Options

1 ... 4567

Bert Freudenberg

Re: Removing Etoys (was Re: A process proposal for 3.10)

Well, there always are people asking for project ideas. Working with
Squeak is fun, and if they can help someone with what they are doing,
it's even better and rewarding. I agree that some people come here
with an attitude like "you must do X, or else ...!". That won't work,
of course. But collecting ideas for improvement is still valuable.

- Bert -

Am 25.10.2006 um 16:43 schrieb [hidden email]:

> Hi Ron.
>
> You seem to believe there are people around who are wishing to
> implement
> stuff just because others want it, not because of their own interest.
>
> Well, I'm pretty sure there's none.
>
> So, if the very people who are contributing specific ideas is not
> willing
> to work on them, who will?
>
> If we don't find an answer to this question, me must accept answers
> like
> "you do it and it's done".
>
> Cheers,
> Juan Vuletich
>
>> I agree with you and have been thinking about the implications of
>> scratch
>> your own itch programming. I firmly believe that this community
>> would
>> benefit from listing to users and integrating user feedback into the
>> process. We need to find a way to work together to meet the
>> common goal
>> of
>> relevance for each participating group. There is a lot going on
>> in the
>> community and I believe we would benefit from integrating the many
>> itches
>> that are currently being scratched. Past that we should be
>> looking for
>> ways
>> to better organize and incorporate user feedback into development
>> teams.
>> (I
>> learned this valuable lesson in the business world: There is
>> nothing worse
>> then spending valuable resources developing terrific features that
>> nobody
>> really wants or needs).
>>
>> I also believe that there are a number of programmers willing to
>> participate
>> but are not able to lead these efforts mostly because of time
>> constraints
>> (not ability). We need a way to incorporate private company's
>> contributions, research projects, volunteer efforts, and user's
>> needs into
>> a
>> cohesive process that advances Squeak towards relevance. I'm not
>> sure how
>> we are going to get there, but it will be a very good place when
>> we do.
>>
>> Ron Teitelbaum
>> President / Principal Software Engineer
>> US Medical Record Specialists
>> [hidden email]
>>
>>> From: Stéphane Rollandin
>>> Sent: Wednesday, October 25, 2006 9:10 AM
>>>
>>> too often on this list we see questions or reports like "this
>>> feature is
>>> broken could it be fixed ?" or "that feature would be nice to
>>> have" that
>>> are only answered by a blunt (and in my view quite rude) "then
>>> are you
>>> ready to work on it ?".
>>>
>>> I am tired of this trick. a more correct answer would be in the
>>> line of
>>> "nobody seems to be engaged in doing this, sorry".
>>>
>>> I don't see the need to pressure people, especially publicly, and
>>> especially people who participate in the debate about the future of
>>> Squeak.
>>>
>>> these people should better be thanked for their interest. they are
>>> contributing ideas and opinions, ok it's not code, but it's not
>>> valueless either.
>>>
>>>
>>> regards,
>>>
>>> Stef
>>
>>
>>
>>
>>
>
>
>

Ron Teitelbaum

RE: Removing Etoys (was Re: A process proposal for 3.10)

In reply to this post by Juan Vuletich (dc)

Hi Juan,

I believe that there are a number of people willing to work with a group to
help meet common goals for a number of varying reasons. I also believe that
there are people that are willing to help if an easier path for their
contribution is forged. I have received enough help and contributions from
people in this community to know that this is true. I'm not saying that
someone can come up and say, "You do this, and you do that!" but I am saying
that there are a number of projects that people will work on given better
organization and feedback. The opportunity to learn and work with very
talented people is sometimes incentive enough.

In some cases you are correct, and if the community in general agrees that
we need something but no one is willing to step up and either organize an
effort or do the work themselves then we should consider other incentives to
make it happen. I don't think one size fits all. The only way we will know
what needs to be done is to incorporate better users feedback.

Ron

> From: [hidden email]
> Sent: Wednesday, October 25, 2006 10:43 AM
>
> Hi Ron.
>
> You seem to believe there are people around who are wishing to implement
> stuff just because others want it, not because of their own interest.
>
> Well, I'm pretty sure there's none.
>
> So, if the very people who are contributing specific ideas is not willing
> to work on them, who will?
>
> If we don't find an answer to this question, me must accept answers like
> "you do it and it's done".
>
> Cheers,
> Juan Vuletich
>
> > I agree with you and have been thinking about the implications of
> scratch
> > your own itch programming. I firmly believe that this community would
> > benefit from listing to users and integrating user feedback into the
> > process. We need to find a way to work together to meet the common goal
> > of
> > relevance for each participating group. There is a lot going on in the
> > community and I believe we would benefit from integrating the many
> itches
> > that are currently being scratched. Past that we should be looking for
> > ways
> > to better organize and incorporate user feedback into development teams.
> > (I
> > learned this valuable lesson in the business world: There is nothing
> worse
> > then spending valuable resources developing terrific features that
> nobody
> > really wants or needs).
> >
> > I also believe that there are a number of programmers willing to
> > participate
> > but are not able to lead these efforts mostly because of time
> constraints
> > (not ability). We need a way to incorporate private company's
> > contributions, research projects, volunteer efforts, and user's needs
> into
> > a
> > cohesive process that advances Squeak towards relevance. I'm not sure
> how
> > we are going to get there, but it will be a very good place when we do.
> >
> > Ron Teitelbaum
> > President / Principal Software Engineer
> > US Medical Record Specialists
> > [hidden email]
> >
> >> From: Stéphane Rollandin
> >> Sent: Wednesday, October 25, 2006 9:10 AM
> >>
> >> too often on this list we see questions or reports like "this feature
> is
> >> broken could it be fixed ?" or "that feature would be nice to have"
> that
> >> are only answered by a blunt (and in my view quite rude) "then are you
> >> ready to work on it ?".
> >>
> >> I am tired of this trick. a more correct answer would be in the line of
> >> "nobody seems to be engaged in doing this, sorry".
> >>
> >> I don't see the need to pressure people, especially publicly, and
> >> especially people who participate in the debate about the future of
> >> Squeak.
> >>
> >> these people should better be thanked for their interest. they are
> >> contributing ideas and opinions, ok it's not code, but it's not
> >> valueless either.
> >>
> >>
> >> regards,
> >>
> >> Stef
> >
> >
> >
> >
> >
>
>

Lex Spoon

Re: A process proposal for 3.10

In reply to this post by Marcus Denker

Marcus Denker <[hidden email]> writes:

> On 17.10.2006, at 16:55, Lex Spoon wrote:
> > Also, keep in mind that all the Squeakland people are vitally
> > interested in EToys. There are ETosy-based text books, for goodness'
> > sakes.
> >
> > I do agree that at least one steward for EToys needs to be identified,
> > if there is not one already. Try probing on the Squeakland mailing
> > list, if nothing else.
>
> The SqueakLand people don't use 3.9, and I am quite sure they never
> will.

That's a fascinating observation. I would love to hear from a
SqueakLand person that, in fact, they do not care much to use future
versions of the open-source community-maintained thread of Squeak.

If it's true, that's a big deal for where Squeak goes next. In
particular, it means EToys in Squeak is only important for things like
making presentations, and thus that a lot of it can be cut out with no
one caring.

-Lex

Lex Spoon

Re: Removing Etoys (was Re: A process proposal for 3.10)

In reply to this post by Göran Krampe

[hidden email] writes:

> Markus Gaelli <[hidden email]> wrote:
> > Count me in.
> > Markus
>
> Ok! So I counted 4-5 people. Two questions:
>
> 1. Would it be a real problem if you were instead forced to use the
> Squeakland image etc?
>
> 2. Are you committed in the extent that you would actually join an eToys
> team making it loadable? Juan has already shown that it can be ripped
> out - next step would be making it loadable I presume.
>
> regards, Göran

I use Squeak for presenting material, and EToys is nice for that. It
is useful for connecting different parts of the screen together so
that if you click in one place, then something happens in another
place. EToys makes it easy.

With raw Smalltalk it is harder, especially because it is a little
harder to reference a morph from within another morph. You have to
use global variables, or you have to name your morphs and do things
like "World morphNamed: 'blah'". You have to make up class names for
a lot of little one-off objects.

For these presentations, I would much rather use the developer's
version of Squeak. Frequently I write Smalltalk code in conjunction
with the presentation, which is precisely what the developer's version
is good at. Further, when I am lucky, I get to *present* a
Smalltalk-based tool, which will surely have been developed in Squeak
proper.

All this said, one ultimately has to face the limits of our resources.
If Squeakland folks do not want to do their own modularization effort,
and they do not even want to track the community-maintained version of
Squeak, then the community-maintained version should split ways I
guess, no matter the loss to both sub-communities.

-Lex

J J-6

RE: Removing Etoys (was Re: A process proposal for 3.10)

In reply to this post by Juan Vuletich (dc)

>From: [hidden email]
>Reply-To: The general-purpose Squeak developers
>list<[hidden email]>
>To: [hidden email], "The general-purpose Squeak developers
>list"<[hidden email]>
>Subject: RE: Removing Etoys (was Re: A process proposal for 3.10)
>Date: Wed, 25 Oct 2006 11:43:09 -0300 (ART)
>
>Hi Ron.
>
>You seem to believe there are people around who are wishing to implement
>stuff just because others want it, not because of their own interest.
>
>Well, I'm pretty sure there's none.
>

That's not true. I am working on RecurranceRule's right now. I had the
part I needed weeks ago, but I ditched it to try and get a more complete
solution (which I will likely never use myself) so others will have it. And
when I get done with this, and the project that spawned it, I will look at
what is on the "hit list" that I could do easily (best to get the low
hanging fruit first).

I'm leaning toward a RoR type (but better I hope) implimentation in Seaside,
because I think that would draw more people.

So the truth is, you are right: I plan on doing this for my own interest.
My own interest in this system getting popular so we have more bodies to
write code. Imagine where we would be with one third the people who are
working on Java right now.

I'm only one person (with very little free time), but I don't think I'm
alone in my thinking.

_________________________________________________________________
Get FREE company branded e-mail accounts and business Web site from
Microsoft Office Live
http://clk.atdmt.com/MRT/go/mcrssaub0050001411mrt/direct/01/

keith1y

Re: Removing Etoys (was Re: A process proposal for 3.10)

J J wrote:

+1

I have ported Message eating null.
I just did SSpec.
Installer was an idea that took on a life of its own.
Pier-Magma i did for a site that could easily be done with basic web
tools , but I did it for my own interest too.

I hope these will benefit the community, my own use will often be minimal.

Keith

___________________________________________________________
All new Yahoo! Mail "The new Interface is stunning in its simplicity and ease of use." - PC Magazine
http://uk.docs.yahoo.com/nowyoucan.html

Lex Spoon

Re: "slices" of packages (was: A process proposal for 3.10)

In reply to this post by Pavel Krivanek

"Pavel Krivanek" <[hidden email]> writes:

> On 17 Oct 2006 10:49:11 -0400, Lex Spoon <[hidden email]> wrote:
> > It is just painful reading Stephane's description of package slices.
> >
> > Why not use Package Universes to organize a collection of package
> > versions? There would be a 3.10 release universe. The release team
> > would know the update password, and they could update that universe
> > with new package versions as they see fit.
> >
> > Just give me the word, and I'd be happy to set up a server for you to
> > experiment with.
>
> Maybe I'm wrong but I think that we need firstly better package
> format. Now we are not able to create versioned packages for
> everything we need.

I believe the package format per se is not what we lack, but a good
system for recording and sharing the available packages and their
dependencies.

Now, maybe we talk past each other a little bit. But, SqueakMap and
Package Universes record a lot of information about packages, even
though those packages might not include the information themselves.
For example, it is no problem to record dependency information in the
package-sharing system, even when the packages themselves lack that
information.

Anyway, I could be wrong, but I believe our current suite of package
formats is just fine. For sure, Package Universes seems to do just
fine using the usual suspects: Monticello, changesets, fileouts,
Morphic projects, etc.

-Lex

Lex Spoon

Re: "slices" of packages (was: A process proposal for 3.10)

In reply to this post by Colin Putney

Colin Putney <[hidden email]> writes:
> Second, slices are finer-grained than the packages we have now; they
> include variables separately from classes, and can thus extend
> classes with variables.

If this is what Stephane meant, then indeed I misunderstood. I
thought he was talking about how to deal with *multiple* packages, not
how to deal with things that are smaller than a package.

-Lex

Lex Spoon

Re: "slices" of packages (was: A process proposal for 3.10)

In reply to this post by stephane ducasse-2

stephane ducasse <[hidden email]> writes:

> Package universes are not the solution to the problem I described.
> At the end of the process of course you can get a number of package
> working in the context of an
> universe.
> Now in the process of getting a set of package working together you
> need to identify
> that package 1.5 is working with packageX 2.1 because you want to be
> able to rollback
> or to reload them together. This is why we had a script loader and
> kept all this information.

I believe package universes will do just what you need. It will
record a set of package versions. If and when you decide a package
version is good, you add it to the universe.

To elaborate, the process would be as follows. Start with *some*
reasonable set of packages that (mostly) work together. For each
package you are thinking of adding, load it along with all other
packages you are considering, and test the new package. If everything
seems fine, then add the new package to the universe. Repeat over
time, and your package universe will evolve.

For sure, I have much less experience installing different versions of
packages than you, so maybe I am being naive. What do you see that I
am missing? If you had a way to record a collection of package
versions, what more would you want?

FWIW, the approach is good enough for Debian and Ubuntu, where there
are a lot more packages than in Squeak.

Keep in mind, by the way, that it is especially helpful for you if
*other* developers use the same universe you as a distribution-builder
are working on. Then, all developers in your circle tend to use
approximately the same package versions, anyway.

-Lex

Stéphane Rollandin

Re: Removing Etoys (was Re: A process proposal for 3.10)

In reply to this post by Göran Krampe

[hidden email] wrote:

> You know, I really didn't ask this to offend you -

I know that

I asked it because we
> need to decide something and that means finding out if there actually
> are people willing to pick up this ball. Do not view this as me trying
> to put some pressure on etc. - just say "Yes" if you are interested in
> this task, or don't answer. Simple as that.
>

but THAT is pressure. so sad you don't see it.

> We are trying here to figure out what to do, right? And that will
> inevitable boil down to "effort" in either way - effort throwing it out,
> effort making it reloadable, whatever.
>
> And even though it may sound harsh - if noone *is* interested (for
> various reasons that does not matter in the end, be it lack of interest
> or lack of skills or whatever) then it probably will not get done. And
> we want to know that, because it affects the decision.
>
> regards, Göran
>

I respect your concern and your action, I just have a different point of
view on all this. No time to talk about this today, though..

regards,

Stef

Juan Vuletich (dc)

RE: Removing Etoys (was Re: A process proposal for 3.10)

In reply to this post by J J-6

This is really great to know!

Cheers,
Juan Vuletich

>>From: [hidden email]
>>Reply-To: The general-purpose Squeak developers
>>list<[hidden email]>
>>To: [hidden email], "The general-purpose Squeak developers
>>list"<[hidden email]>
>>Subject: RE: Removing Etoys (was Re: A process proposal for 3.10)
>>Date: Wed, 25 Oct 2006 11:43:09 -0300 (ART)
>>
>>Hi Ron.
>>
>>You seem to believe there are people around who are wishing to implement
>>stuff just because others want it, not because of their own interest.
>>
>>Well, I'm pretty sure there's none.
>>
>
> That's not true. I am working on RecurranceRule's right now. I had the
> part I needed weeks ago, but I ditched it to try and get a more complete
> solution (which I will likely never use myself) so others will have it.
> And
> when I get done with this, and the project that spawned it, I will look at
> what is on the "hit list" that I could do easily (best to get the low
> hanging fruit first).
>
> I'm leaning toward a RoR type (but better I hope) implimentation in
> Seaside,
> because I think that would draw more people.
>
> So the truth is, you are right: I plan on doing this for my own interest.
> My own interest in this system getting popular so we have more bodies to
> write code. Imagine where we would be with one third the people who are
> working on Java right now.
>
> I'm only one person (with very little free time), but I don't think I'm
> alone in my thinking.
>
> _________________________________________________________________
> Get FREE company branded e-mail accounts and business Web site from
> Microsoft Office Live
> http://clk.atdmt.com/MRT/go/mcrssaub0050001411mrt/direct/01/
>
>
>

Milan Zimmermann-2

Re: Removing Etoys (was Re: A process proposal for 3.10)

In reply to this post by karl-8

On 2006 October 25 01:32, karl wrote:

> Milan Zimmermann skrev:
> > On 2006 October 24 18:14, [hidden email] wrote:
> >> Markus Gaelli <[hidden email]> wrote:
> >>> Count me in.
> >>> Markus
> >>
> >> Ok! So I counted 4-5 people. Two questions:
> >>
> >> 1. Would it be a real problem if you were instead forced to use the
> >> Squeakland image etc?
> >
> > Not a real problem, I am just playing .. Recently (as in the last year of
> > so) , I created eToys projects in 3.8, Squeakland and OLPC, and one even
> > in 3.8 Tweak image (which felt like 3.8 eToys from project building
> > perspective), so I am not personally tied to eToys being in 3.10 etc (not
> > venturing into community building discussion :) )
>
> My main objection to get rid of eToy are that eToys now are about to be
> launched as a big world wide education project. I think keeping eToys in
> the image and in working order will benefit Squeak and Squeak users as
> people will be coming from more places to use and contribute to eToys.
> Dropping eToys will confuse a lot of newcomers.
> Karl

I agree with your points. I will note that I did not advocate for removing
eToys, merely replied what it would mean for me; having said that, I also
agree and like the approach of dividing monoliths. If Morphic/eToys are not
reloadable (hopefully temporarily), perhaps there can be a (long) period when
eToys users, use 3.9 or Squeakland or OLPC image, the rest the (hopefully
gradually shrinking and divided) future versions. A note on current status: I
remember having problems between very early 3.9 and Squeakland image projects
loadability. So I am not at all sure that a sufficiently complex eToys
project created in one of the 3 versions is loadable to the other 2. (if I am
wrong someone please correct it) ..

Milan

>
> >> 2. Are you committed in the extent that you would actually join an eToys
> >> team making it loadable?
> >
> > Having limited time as everyone, I would rather switch completely to
> > Squeakland/OLPC for play eToys projects, and along the way spent time
> > studying Tweak and Tweak based eToys...
> >
> > Milan
> >
> >> Juan has already shown that it can be ripped
> >> out - next step would be making it loadable I presume.
> >>
> >> regards, G�ran

1 ... 4567