why can't we select the browser we want to use anymore via the window menu?
Hi guys It would be good not to clean without a clear vision. For example we cannot register an old browser to browse code via the menu of a window. Now let us think two minutes to see if you can get my point: - I want to unload nautilus, rb, keymapping, athens, Ecompletion, Gofer, NativeBoost, Zinc, â¦.. and reload them via their configuration so that we can manage Pharo with configurations. - Right now we LOST yes LOST the configurations of most of the part of the systems (I just spendt several afternoon on the one of RB in the past and now guess what) because we do not have a process to use them and we are afraid to have 10 packages and 10 classes more in the system. I do not understand why we do not start to put the configuration inside the image. To me this is totally stupid not to do it. - Now without a browser this is nearly impossible to work. So we will remove the old browser but we should go slowly because else I will you do it with emacs outside the image to see if you succeed. So it would be good to focus on real impacting changes. Stef
This makes think of the tick tock model of Intel improvements. http://www.intel.com/content/www/us/en/silicon-innovations/intel-tick-tock-m... Looks like with 2.0 and 3.0 we are going to have two "tick"s in a row. That may be too much. I am afraid even trying out 3.0 when reading about all the moving parts that are changing in all corners. Phil --- Philippe Back Dramatic Performance Improvements Mob: +32(0) 478 650 140 | Fax: +32 (0) 70 408 027 Mail:phil@highoctane.be | Web: http://philippeback.eu Blog: http://philippeback.be | Twitter: @philippeback Youtube: http://www.youtube.com/user/philippeback/videos High Octane SPRL rue cour Boisacq 101 | 1301 Bierges | Belgium Pharo Consortium Member - http://consortium.pharo.org/ Featured on the Software Process and Measurement Cast - http://spamcast.libsyn.com Sparx Systems Enterprise Architect and Ability Engineering EADocX Value Added Reseller On Sun, Nov 24, 2013 at 9:02 PM, Stéphane Ducasse <stephane.ducasse@inria.fr
wrote:
Hi guys
It would be good not to clean without a clear vision. For example we cannot register an old browser to browse code via the menu of a window. Now let us think two minutes to see if you can get my point:
- I want to unload nautilus, rb, keymapping, athens, Ecompletion, Gofer, NativeBoost, Zinc, â¦.. and reload them via their configuration so that we can manage Pharo with configurations.
- Right now we LOST yes LOST the configurations of most of the part of the systems (I just spendt several afternoon on the one of RB in the past and now guess what) because we do not have a process to use them and we are afraid to have 10 packages and 10 classes more in the system. I do not understand why we do not start to put the configuration inside the image. To me this is totally stupid not to do it.
- Now without a browser this is nearly impossible to work. So we will remove the old browser but we should go slowly because else I will you do it with emacs outside the image to see if you succeed.
So it would be good to focus on real impacting changes.
Stef
On 24 Nov 2013, at 21:08, phil@highoctane.be wrote:
This makes think of the tick tock model of Intel improvements.
http://www.intel.com/content/www/us/en/silicon-innovations/intel-tick-tock-m...
Looks like with 2.0 and 3.0 we are going to have two "tick"s in a row. That may be too much.
I am afraid even trying out 3.0 when reading about all the moving parts that are changing in all corners.
Maybe we should give up, define Pharo as finished and be happy. Marcus
Don't say silly things. Now we should focus on the important parts. Stef
On 24 Nov 2013, at 21:08, phil@highoctane.be wrote:
This makes think of the tick tock model of Intel improvements.
http://www.intel.com/content/www/us/en/silicon-innovations/intel-tick-tock-m...
Looks like with 2.0 and 3.0 we are going to have two "tick"s in a row. That may be too much.
I am afraid even trying out 3.0 when reading about all the moving parts that are changing in all corners.
Maybe we should give up, define Pharo as finished and be happy.
Marcus
Am 24.11.2013 um 21:18 schrieb Marcus Denker <marcus.denker@inria.fr>:
On 24 Nov 2013, at 21:08, phil@highoctane.be wrote:
This makes think of the tick tock model of Intel improvements.
http://www.intel.com/content/www/us/en/silicon-innovations/intel-tick-tock-m...
Looks like with 2.0 and 3.0 we are going to have two "tick"s in a row. That may be too much.
I am afraid even trying out 3.0 when reading about all the moving parts that are changing in all corners.
Maybe we should give up, define Pharo as finished and be happy.
I don't think that every statement that has a notion like the one from phil needs a response out of self-defense. This will happen as long as the "researcher guys" and the "industry guys" share the same table. It is just the way it is and we are free to leave statements uncommented. Norbert
Exactly. This researcher/industry - lab/delivery is always present, no matter where one goes. I was thinking more along the lines of: Investing in the industry Thousands of vendors depend on Intel® processors for product development. To help them forge ahead with new product advancements, Intel invests heavily in research that drives innovations at the silicon level and establishes new, industry-wide standards. *Combined with the predictability *of Intelâs tick-tock model, these efforts promote faster, more efficient innovation throughout the industryâyear-in and year-out. --- Philippe Back Dramatic Performance Improvements Mob: +32(0) 478 650 140 | Fax: +32 (0) 70 408 027 Mail:phil@highoctane.be | Web: http://philippeback.eu Blog: http://philippeback.be | Twitter: @philippeback Youtube: http://www.youtube.com/user/philippeback/videos High Octane SPRL rue cour Boisacq 101 | 1301 Bierges | Belgium Pharo Consortium Member - http://consortium.pharo.org/ Featured on the Software Process and Measurement Cast - http://spamcast.libsyn.com Sparx Systems Enterprise Architect and Ability Engineering EADocX Value Added Reseller On Mon, Nov 25, 2013 at 8:40 AM, Norbert Hartl <norbert@hartl.name> wrote:
Am 24.11.2013 um 21:18 schrieb Marcus Denker <marcus.denker@inria.fr>:
On 24 Nov 2013, at 21:08, phil@highoctane.be wrote:
This makes think of the tick tock model of Intel improvements.
http://www.intel.com/content/www/us/en/silicon-innovations/intel-tick-tock-m...
Looks like with 2.0 and 3.0 we are going to have two "tick"s in a row. That may be too much.
I am afraid even trying out 3.0 when reading about all the moving parts that are changing in all corners.
Maybe we should give up, define Pharo as finished and be happy.
I don't think that every statement that has a notion like the one from phil needs a response out of self-defense. This will happen as long as the "researcher guys" and the "industry guys" share the same table. It is just the way it is and we are free to leave statements uncommented.
Norbert
On 25 Nov 2013, at 09:19, phil@highoctane.be wrote:
Exactly. This researcher/industry - lab/delivery is always present, no matter where one goes.
I was thinking more along the lines of:
Investing in the industry
Thousands of vendors depend on Intel® processors for product development. To help them forge ahead with new product advancements, Intel invests heavily in research that drives innovations at the silicon level and establishes new, industry-wide standards. Combined with the predictability of Intelâs tick-tock model, these efforts promote faster, more efficient innovation throughout the industryâyear-in and year-out.
Yes, and tick-tock as well as research all happens in parallel. Now how can we get the resource to do that? In the end, the âstable/unstableâ is very much like a tick/tock: we release, when we find problems in the release (and we find a fixâ¦) we fix the release (tik), while working in tok (big change). It would be better to have payed teams for both⦠Marcus
On 25 Nov 2013, at 10:04, Marcus Denker <marcus.denker@inria.fr> wrote:
On 25 Nov 2013, at 09:19, phil@highoctane.be wrote:
Exactly. This researcher/industry - lab/delivery is always present, no matter where one goes.
I was thinking more along the lines of:
Investing in the industry
Thousands of vendors depend on Intel® processors for product development. To help them forge ahead with new product advancements, Intel invests heavily in research that drives innovations at the silicon level and establishes new, industry-wide standards. Combined with the predictability of Intelâs tick-tock model, these efforts promote faster, more efficient innovation throughout the industryâyear-in and year-out.
Yes, and tick-tock as well as research all happens in parallel.
Now how can we get the resource to do that?
In the end, the âstable/unstableâ is very much like a tick/tock: we release, when we find problems in the release (and we find a fixâ¦) we fix the release (tik), while working in tok (big change).
It would be better to have payed teams for bothâ¦
With âpayedâ I donât mean the money aspect but the fact that someone will look at the bug tracker every day and spend all hours needed to fix even those bugs that he/she does not care about at all and has never managed to run into. Marcus
No you should not. 30 is really usable and we are in alpha. Now removing the old code browser is a goal but we should not fool ourselves. Once all the tools are based on Spec, breaking/changing spec will let us without back door. Stef
This makes think of the tick tock model of Intel improvements.
http://www.intel.com/content/www/us/en/silicon-innovations/intel-tick-tock-m...
Looks like with 2.0 and 3.0 we are going to have two "tick"s in a row. That may be too much.
I am afraid even trying out 3.0 when reading about all the moving parts that are changing in all corners.
Phil
--- Philippe Back Dramatic Performance Improvements Mob: +32(0) 478 650 140 | Fax: +32 (0) 70 408 027 Mail:phil@highoctane.be | Web: http://philippeback.eu Blog: http://philippeback.be | Twitter: @philippeback Youtube: http://www.youtube.com/user/philippeback/videos
High Octane SPRL rue cour Boisacq 101 | 1301 Bierges | Belgium
Pharo Consortium Member - http://consortium.pharo.org/ Featured on the Software Process and Measurement Cast - http://spamcast.libsyn.com Sparx Systems Enterprise Architect and Ability Engineering EADocX Value Added Reseller
On Sun, Nov 24, 2013 at 9:02 PM, Stéphane Ducasse <stephane.ducasse@inria.fr> wrote: Hi guys
It would be good not to clean without a clear vision. For example we cannot register an old browser to browse code via the menu of a window. Now let us think two minutes to see if you can get my point:
- I want to unload nautilus, rb, keymapping, athens, Ecompletion, Gofer, NativeBoost, Zinc, â¦.. and reload them via their configuration so that we can manage Pharo with configurations.
- Right now we LOST yes LOST the configurations of most of the part of the systems (I just spendt several afternoon on the one of RB in the past and now guess what) because we do not have a process to use them and we are afraid to have 10 packages and 10 classes more in the system. I do not understand why we do not start to put the configuration inside the image. To me this is totally stupid not to do it.
- Now without a browser this is nearly impossible to work. So we will remove the old browser but we should go slowly because else I will you do it with emacs outside the image to see if you succeed.
So it would be good to focus on real impacting changes.
Stef
On 24 Nov 2013, at 21:18, Stéphane Ducasse <stephane.ducasse@inria.fr> wrote:
No you should not. 30 is really usable and we are in alpha. Now removing the old code browser is a goal but we should not fool ourselves. Once all the tools are based on Spec, breaking/changing spec will let us without back door.
I donât consider that a problem, on the contrary: eating your own dog food is good. BTW, in the past (old tools, old frameworks) the problem was the same. Yes, that means increased responsibility, so what ? If I break Zinc, nobody can load or commit code any more - though.
Stef
This makes think of the tick tock model of Intel improvements.
http://www.intel.com/content/www/us/en/silicon-innovations/intel-tick-tock-m...
Looks like with 2.0 and 3.0 we are going to have two "tick"s in a row. That may be too much.
I am afraid even trying out 3.0 when reading about all the moving parts that are changing in all corners.
Phil
--- Philippe Back Dramatic Performance Improvements Mob: +32(0) 478 650 140 | Fax: +32 (0) 70 408 027 Mail:phil@highoctane.be | Web: http://philippeback.eu Blog: http://philippeback.be | Twitter: @philippeback Youtube: http://www.youtube.com/user/philippeback/videos
High Octane SPRL rue cour Boisacq 101 | 1301 Bierges | Belgium
Pharo Consortium Member - http://consortium.pharo.org/ Featured on the Software Process and Measurement Cast - http://spamcast.libsyn.com Sparx Systems Enterprise Architect and Ability Engineering EADocX Value Added Reseller
On Sun, Nov 24, 2013 at 9:02 PM, Stéphane Ducasse <stephane.ducasse@inria.fr> wrote: Hi guys
It would be good not to clean without a clear vision. For example we cannot register an old browser to browse code via the menu of a window. Now let us think two minutes to see if you can get my point:
- I want to unload nautilus, rb, keymapping, athens, Ecompletion, Gofer, NativeBoost, Zinc, â¦.. and reload them via their configuration so that we can manage Pharo with configurations.
- Right now we LOST yes LOST the configurations of most of the part of the systems (I just spendt several afternoon on the one of RB in the past and now guess what) because we do not have a process to use them and we are afraid to have 10 packages and 10 classes more in the system. I do not understand why we do not start to put the configuration inside the image. To me this is totally stupid not to do it.
- Now without a browser this is nearly impossible to work. So we will remove the old browser but we should go slowly because else I will you do it with emacs outside the image to see if you succeed.
So it would be good to focus on real impacting changes.
Stef
On 24 Nov 2013, at 21:41, Sven Van Caekenberghe <sven@stfx.eu> wrote:
On 24 Nov 2013, at 21:18, Stéphane Ducasse <stephane.ducasse@inria.fr> wrote:
No you should not. 30 is really usable and we are in alpha. Now removing the old code browser is a goal but we should not fool ourselves. Once all the tools are based on Spec, breaking/changing spec will let us without back door.
I donât consider that a problem, on the contrary: eating your own dog food is good.
BTW, in the past (old tools, old frameworks) the problem was the same.
When Camille last week tried to convince me to kill the old Compiler *now* I said âbut itâs so nice to be able to switch when changing the newâ. He replied: âBut there was just one old Compiler, wasnât there?â And he is right. What we need to do for real is to support multiple versions of code in the same image⦠or multiple images where one can remote-devlope the other. This would solve the problem for real. Marcus
On 24 nov. 2013, at 21:44, Marcus Denker <marcus.denker@inria.fr> wrote:
On 24 Nov 2013, at 21:41, Sven Van Caekenberghe <sven@stfx.eu> wrote:
On 24 Nov 2013, at 21:18, Stéphane Ducasse <stephane.ducasse@inria.fr> wrote:
No you should not. 30 is really usable and we are in alpha. Now removing the old code browser is a goal but we should not fool ourselves. Once all the tools are based on Spec, breaking/changing spec will let us without back door.
I donât consider that a problem, on the contrary: eating your own dog food is good.
BTW, in the past (old tools, old frameworks) the problem was the same.
When Camille last week tried to convince me to kill the old Compiler *now* I said âbut itâs so nice to be able to switch when changing the newâ. He replied: âBut there was just one old Compiler, wasnât there?â
And he is right.
What we need to do for real is to support multiple versions of code in the same imageâ¦
That's work in progress :)
or multiple images where one can remote-devlope the other. This would solve the problem for real.
Marcus
I donât consider that a problem, on the contrary: eating your own dog food is good.
BTW, in the past (old tools, old frameworks) the problem was the same.
When Camille last week tried to convince me to kill the old Compiler *now* I said âbut itâs so nice to be able to switch when changing the newâ. He replied: âBut there was just one old Compiler, wasnât there?â
And he is right.
No :) He will not be the guy that will have to fight with bugs in Opal but you. He would have been right if we would have a script or a way to create a new opal that can compile another one. I'm in favor to remove the old compiler and the old browser but only when we are in a situation where we can continue to work fast.
What we need to do for real is to support multiple versions of code in the same image⦠or multiple images where one can remote-devlope the other. This would solve the problem for real.
Just start to compile in a different environment is a first step in that direction. without a fully environment based system there are solutions but they will take time. Stef
I donât consider that a problem, on the contrary: eating your own dog food is good.
It is a problem. For example, for opal better have in place a mechanism to change it (for example having a process to duplicate it with a different name or make sure that it can compile in a separate environnement) before removing the old one because else if we get a breaking change then it will force us to rollback. For code browser this is the same. It is always smart to have a really simple working in any circumstances. Alain was thinking about that too. Because else when you cannot edit or debug code you are dead. We worked with igor fixing all the LayoutFrame in the system and when we broke (because you always break soetmghitn) the lyout of the browser it was less fun. This is not the question of not eating your own dog food since I'm not using emacs to edit Pharo right now and I would hate to be forced to do it. but if we do not pay attention to our tools and the idea of reflection withitn them then this is what will happen. Right now I cannot fix the Browser in the image. So I opened another one. Stef
BTW, in the past (old tools, old frameworks) the problem was the same.
Yes, that means increased responsibility, so what ?
If I break Zinc, nobody can load or commit code any more - though.
Stef
This makes think of the tick tock model of Intel improvements.
http://www.intel.com/content/www/us/en/silicon-innovations/intel-tick-tock-m...
Looks like with 2.0 and 3.0 we are going to have two "tick"s in a row. That may be too much.
I am afraid even trying out 3.0 when reading about all the moving parts that are changing in all corners.
Phil
--- Philippe Back Dramatic Performance Improvements Mob: +32(0) 478 650 140 | Fax: +32 (0) 70 408 027 Mail:phil@highoctane.be | Web: http://philippeback.eu Blog: http://philippeback.be | Twitter: @philippeback Youtube: http://www.youtube.com/user/philippeback/videos
High Octane SPRL rue cour Boisacq 101 | 1301 Bierges | Belgium
Pharo Consortium Member - http://consortium.pharo.org/ Featured on the Software Process and Measurement Cast - http://spamcast.libsyn.com Sparx Systems Enterprise Architect and Ability Engineering EADocX Value Added Reseller
On Sun, Nov 24, 2013 at 9:02 PM, Stéphane Ducasse <stephane.ducasse@inria.fr> wrote: Hi guys
It would be good not to clean without a clear vision. For example we cannot register an old browser to browse code via the menu of a window. Now let us think two minutes to see if you can get my point:
- I want to unload nautilus, rb, keymapping, athens, Ecompletion, Gofer, NativeBoost, Zinc, â¦.. and reload them via their configuration so that we can manage Pharo with configurations.
- Right now we LOST yes LOST the configurations of most of the part of the systems (I just spendt several afternoon on the one of RB in the past and now guess what) because we do not have a process to use them and we are afraid to have 10 packages and 10 classes more in the system. I do not understand why we do not start to put the configuration inside the image. To me this is totally stupid not to do it.
- Now without a browser this is nearly impossible to work. So we will remove the old browser but we should go slowly because else I will you do it with emacs outside the image to see if you succeed.
So it would be good to focus on real impacting changes.
Stef
I am using Pharo 3.0 in production and it is highly reliable. Doru On Sun, Nov 24, 2013 at 9:08 PM, phil@highoctane.be <phil@highoctane.be>wrote:
This makes think of the tick tock model of Intel improvements.
http://www.intel.com/content/www/us/en/silicon-innovations/intel-tick-tock-m...
Looks like with 2.0 and 3.0 we are going to have two "tick"s in a row. That may be too much.
I am afraid even trying out 3.0 when reading about all the moving parts that are changing in all corners.
Phil
--- Philippe Back Dramatic Performance Improvements Mob: +32(0) 478 650 140 | Fax: +32 (0) 70 408 027 Mail:phil@highoctane.be | Web: http://philippeback.eu Blog: http://philippeback.be | Twitter: @philippeback Youtube: http://www.youtube.com/user/philippeback/videos
High Octane SPRL rue cour Boisacq 101 | 1301 Bierges | Belgium
Pharo Consortium Member - http://consortium.pharo.org/ Featured on the Software Process and Measurement Cast - http://spamcast.libsyn.com Sparx Systems Enterprise Architect and Ability Engineering EADocX Value Added Reseller
On Sun, Nov 24, 2013 at 9:02 PM, Stéphane Ducasse < stephane.ducasse@inria.fr> wrote:
Hi guys
It would be good not to clean without a clear vision. For example we cannot register an old browser to browse code via the menu of a window. Now let us think two minutes to see if you can get my point:
- I want to unload nautilus, rb, keymapping, athens, Ecompletion, Gofer, NativeBoost, Zinc, â¦.. and reload them via their configuration so that we can manage Pharo with configurations.
- Right now we LOST yes LOST the configurations of most of the part of the systems (I just spendt several afternoon on the one of RB in the past and now guess what) because we do not have a process to use them and we are afraid to have 10 packages and 10 classes more in the system. I do not understand why we do not start to put the configuration inside the image. To me this is totally stupid not to do it.
- Now without a browser this is nearly impossible to work. So we will remove the old browser but we should go slowly because else I will you do it with emacs outside the image to see if you succeed.
So it would be good to focus on real impacting changes.
Stef
-- www.tudorgirba.com "Every thing has its own flow"
On 24 Nov 2013, at 21:31, Tudor Girba <tudor@tudorgirba.com> wrote:
I am using Pharo 3.0 in production and it is highly reliable.
I think, too, that Pharo 3 is *amazingly* stable considering what we changed. I am actually quite proud of it. Marcus
On 24 Nov 2013, at 21:33, Marcus Denker <marcus.denker@inria.fr> wrote:
On 24 Nov 2013, at 21:31, Tudor Girba <tudor@tudorgirba.com> wrote:
I am using Pharo 3.0 in production and it is highly reliable.
I think, too, that Pharo 3 is *amazingly* stable considering what we changed.
I am actually quite proud of it.
Marcus
+1
On Nov 24, 2013, at 9:33 PM, Marcus Denker <marcus.denker@inria.fr> wrote:
On 24 Nov 2013, at 21:31, Tudor Girba <tudor@tudorgirba.com> wrote:
I am using Pharo 3.0 in production and it is highly reliable.
I think, too, that Pharo 3 is *amazingly* stable considering what we changed.
I am actually quite proud of it.
me too. Now I just ask for a bit of care else this will be much more difficult to arrive to bootstrap. If we do not make sure that we can unload and reload package for examples. Stef
Marcus
Am 24.11.2013 um 21:31 schrieb Tudor Girba <tudor@tudorgirba.com>:
I am using Pharo 3.0 in production and it is highly reliable.
So it is not just reliable but highly reliable? What does it mean and what do you call "using in production"? Norbert
Doru
On Sun, Nov 24, 2013 at 9:08 PM, phil@highoctane.be <phil@highoctane.be> wrote: This makes think of the tick tock model of Intel improvements.
http://www.intel.com/content/www/us/en/silicon-innovations/intel-tick-tock-m...
Looks like with 2.0 and 3.0 we are going to have two "tick"s in a row. That may be too much.
I am afraid even trying out 3.0 when reading about all the moving parts that are changing in all corners.
Phil
--- Philippe Back Dramatic Performance Improvements Mob: +32(0) 478 650 140 | Fax: +32 (0) 70 408 027 Mail:phil@highoctane.be | Web: http://philippeback.eu Blog: http://philippeback.be | Twitter: @philippeback Youtube: http://www.youtube.com/user/philippeback/videos
High Octane SPRL rue cour Boisacq 101 | 1301 Bierges | Belgium
Pharo Consortium Member - http://consortium.pharo.org/ Featured on the Software Process and Measurement Cast - http://spamcast.libsyn.com Sparx Systems Enterprise Architect and Ability Engineering EADocX Value Added Reseller
On Sun, Nov 24, 2013 at 9:02 PM, Stéphane Ducasse <stephane.ducasse@inria.fr> wrote: Hi guys
It would be good not to clean without a clear vision. For example we cannot register an old browser to browse code via the menu of a window. Now let us think two minutes to see if you can get my point:
- I want to unload nautilus, rb, keymapping, athens, Ecompletion, Gofer, NativeBoost, Zinc, â¦.. and reload them via their configuration so that we can manage Pharo with configurations.
- Right now we LOST yes LOST the configurations of most of the part of the systems (I just spendt several afternoon on the one of RB in the past and now guess what) because we do not have a process to use them and we are afraid to have 10 packages and 10 classes more in the system. I do not understand why we do not start to put the configuration inside the image. To me this is totally stupid not to do it.
- Now without a browser this is nearly impossible to work. So we will remove the old browser but we should go slowly because else I will you do it with emacs outside the image to see if you succeed.
So it would be good to focus on real impacting changes.
Stef
-- www.tudorgirba.com
"Every thing has its own flow"
Am 24.11.2013 um 21:08 schrieb "phil@highoctane.be" <phil@highoctane.be>:
This makes think of the tick tock model of Intel improvements.
http://www.intel.com/content/www/us/en/silicon-innovations/intel-tick-tock-m...
Looks like with 2.0 and 3.0 we are going to have two "tick"s in a row. That may be too much.
I am afraid even trying out 3.0 when reading about all the moving parts that are changing in all corners.
Yep, me too. But then I think it can't be worse than 2.0 :) We have to live with that. There won't be a matured image version anytime soon so ... Norbert
Phil
--- Philippe Back Dramatic Performance Improvements Mob: +32(0) 478 650 140 | Fax: +32 (0) 70 408 027 Mail:phil@highoctane.be | Web: http://philippeback.eu Blog: http://philippeback.be | Twitter: @philippeback Youtube: http://www.youtube.com/user/philippeback/videos
High Octane SPRL rue cour Boisacq 101 | 1301 Bierges | Belgium
Pharo Consortium Member - http://consortium.pharo.org/ Featured on the Software Process and Measurement Cast - http://spamcast.libsyn.com Sparx Systems Enterprise Architect and Ability Engineering EADocX Value Added Reseller
On Sun, Nov 24, 2013 at 9:02 PM, Stéphane Ducasse <stephane.ducasse@inria.fr> wrote: Hi guys
It would be good not to clean without a clear vision. For example we cannot register an old browser to browse code via the menu of a window. Now let us think two minutes to see if you can get my point:
- I want to unload nautilus, rb, keymapping, athens, Ecompletion, Gofer, NativeBoost, Zinc, â¦.. and reload them via their configuration so that we can manage Pharo with configurations.
- Right now we LOST yes LOST the configurations of most of the part of the systems (I just spendt several afternoon on the one of RB in the past and now guess what) because we do not have a process to use them and we are afraid to have 10 packages and 10 classes more in the system. I do not understand why we do not start to put the configuration inside the image. To me this is totally stupid not to do it.
- Now without a browser this is nearly impossible to work. So we will remove the old browser but we should go slowly because else I will you do it with emacs outside the image to see if you succeed.
So it would be good to focus on real impacting changes.
Stef
On 24 Nov 2013, at 21:02, Stéphane Ducasse <stephane.ducasse@inria.fr> wrote:
Hi guys
It would be good not to clean without a clear vision. For example we cannot register an old browser to browse code via the menu of a window.
There where two systems in parallel for making the browser pluggable. I removed one months ago so we have just one. It was really a mess: some caller uses âSmalltalk tools browserâ, others âSystemBrowser defaultâ (subclass of AppRegistry). This was quite complex and everything was duplicated, and everything mixed together. (Smalltalk tools browser returned SystemBrowser default). In short: a mess. It tool quite some effort to clean up. Now the âSmalltalk tools browserâ (ToolRegistry) is not perfect. -> it should provide a way to register multiple implementations for the same tool -> it should provide a Settings interface to switch between the tools -> it maybe should use pragmas instead of the current scheme Yes, unifying everything to us ToolRegistry instead of AppRegistry meant we lost the menu entry⦠but now we have *one* mechanism, not two incompatible ones. Marcus
There where two systems in parallel for making the browser pluggable. I removed one months ago so we have just one.
It was really a mess: some caller uses âSmalltalk tools browserâ, others âSystemBrowser defaultâ (subclass of AppRegistry).
This was quite complex and everything was duplicated, and everything mixed together. (Smalltalk tools browser returned SystemBrowser default).
In short: a mess. It tool quite some effort to clean up.
Now the âSmalltalk tools browserâ (ToolRegistry) is not perfect.
-> it should provide a way to register multiple implementations for the same tool -> it should provide a Settings interface to switch between the tools -> it maybe should use pragmas instead of the current scheme
Yes, unifying everything to us ToolRegistry instead of AppRegistry meant we lost the menu entry⦠but now we have *one* mechanism, not two incompatible ones.
yes but the menu was handy. Next time ask and I will just call the right behavior from it.
participants (7)
-
Camille Teruel -
Marcus Denker -
Norbert Hartl -
phil＠highoctane.be -
Stéphane Ducasse -
Sven Van Caekenberghe -
Tudor Girba