[Pharo-project] What about killing PharoWeb images?
Please, don't say "NOOOOO!" immediately. Read the mail carefully, take a deep breath and think cold. After working a couple of weeks in the Metacello configurations and now with Damien's scripts to build Dev and Web images, I came to the question: Do we really need a PharoWeb image? I will give just my opinions of why having PharoWeb images is not a good idea anymore 1) First of all, I don't like the PharoWeb name. Because it is not Web in general, it is seaside. So, in my opinion, a better name would be PharoSeaside. Maybe even new comers may think that the only alternative is seaside. You know I have nothing against Seaside and that I even use it several times. I would like to see AidaWeb images based in Pharo too. 2) Why to still have PharoWeb images if now Seaside images (at least the one click) are based in seaside ? In addition, when I develop in Seaside, I use Seaside images, not PharoWeb, and I think most people do the same. 3) Seaside images (or Lukas images for seaside) are MUCH more stable than the PharoWeb images. Why? Several reasons: 3.a) THEY know the dependencies of the packages, the versions, how to set up, the order of load, etc, NOT us. THEY are the experts in seaside, not us. 3.b) THEY really test their images before make then public. 3.c) We are "force" to release one release per month which has no sense. How says that Seaside is released once a month? That's bad because it means we are usually taking unstable versions. THEY know when a release should be done and when a release is stable or not. 4) Now I created all the Metacello configurations for the PharoDev image and I took me a lot of time. Because I am not the developer of any of those projects. So, I need to understand the dependencies, the order, the versions, etc. But why ??? If Seaside people have already done that, don't duplicate efforts. 5) The instability of PharoWeb images give to our public a bad impression. We waste all of out efforts of Pharo in that image. 6) The idea of Pharo is to be the vehicle. And that's exactly the Pharo role in the Seaside images. We don't need to build PharoWeb images. Even more, do we build PharoGlass images ? NOOOO. Why not ? Because Dale already prepare them and as HE knows much more about Glass than us, his images are better for Glass. 7) Tracking and managing issues always take time. Even analyzing a bug, say "we won't fix" and close it. With PharoWeb images we are having those kind of issues. 8) Our resources are limited. Why not to use them in more important stuff ? For example, building a better and more robust Dev image? Here is my proposal: 1) Completely remove the idea of PharoWeb image. Remove the links from the website and do not create Web images any more. 2) In http://www.pharo-project.org/pharo-download add a description explaining that now Seaside images are based in Pharo, so you can directly download that image. We can even explain a bit about seaside, and even if you want you can put a href called "PharoWeb" which links to the seaside image. However, I would rather call it "PharoSeaside". 3) Tomorrow we can have AidaWeb images based in pharo (please Janko!!) so we can just do the same and put the link to those images. 4) You may notice that Seaside images are a bit "different" than our Dev images. This is mostly due to the preferences, the theme and maybe some tools (ECompeltion instead of OCompletion, etc..). So, what I suggest is to prepare a little script that can be downloadable and those people who want a Seaside image but more "similar" to our Dev images, just run that script. So...what do you think? I would really like to hear your opinion about this. Cheers Mariano
On Jan 4, 2010, at 9:24 AM, Mariano Martinez Peck wrote:
Please, don't say "NOOOOO!" immediately. Read the mail carefully, take a deep breath and think cold.
After working a couple of weeks in the Metacello configurations and now with Damien's scripts to build Dev and Web images, I came to the question: Do we really need a PharoWeb image?
we dont :) I agree with you.
I will give just my opinions of why having PharoWeb images is not a good idea anymore
1) First of all, I don't like the PharoWeb name. Because it is not Web in general, it is seaside. So, in my opinion, a better name would be PharoSeaside. Maybe even new comers may think that the only alternative is seaside. You know I have nothing against Seaside and that I even use it several times. I would like to see AidaWeb images based in Pharo too.
2) Why to still have PharoWeb images if now Seaside images (at least the one click) are based in seaside ? In addition, when I develop in Seaside, I use Seaside images, not PharoWeb, and I think most people do the same.
3) Seaside images (or Lukas images for seaside) are MUCH more stable than the PharoWeb images. Why? Several reasons: 3.a) THEY know the dependencies of the packages, the versions, how to set up, the order of load, etc, NOT us. THEY are the experts in seaside, not us. 3.b) THEY really test their images before make then public. 3.c) We are "force" to release one release per month which has no sense. How says that Seaside is released once a month? That's bad because it means we are usually taking unstable versions. THEY know when a release should be done and when a release is stable or not.
4) Now I created all the Metacello configurations for the PharoDev image and I took me a lot of time. Because I am not the developer of any of those projects. So, I need to understand the dependencies, the order, the versions, etc. But why ??? If Seaside people have already done that, don't duplicate efforts.
5) The instability of PharoWeb images give to our public a bad impression. We waste all of out efforts of Pharo in that image.
6) The idea of Pharo is to be the vehicle. And that's exactly the Pharo role in the Seaside images. We don't need to build PharoWeb images. Even more, do we build PharoGlass images ? NOOOO. Why not ? Because Dale already prepare them and as HE knows much more about Glass than us, his images are better for Glass.
7) Tracking and managing issues always take time. Even analyzing a bug, say "we won't fix" and close it. With PharoWeb images we are having those kind of issues.
8) Our resources are limited. Why not to use them in more important stuff ? For example, building a better and more robust Dev image?
Here is my proposal:
1) Completely remove the idea of PharoWeb image. Remove the links from the website and do not create Web images any more. 2) In http://www.pharo-project.org/pharo-download add a description explaining that now Seaside images are based in Pharo, so you can directly download that image. We can even explain a bit about seaside, and even if you want you can put a href called "PharoWeb" which links to the seaside image. However, I would rather call it "PharoSeaside". 3) Tomorrow we can have AidaWeb images based in pharo (please Janko!!) so we can just do the same and put the link to those images. 4) You may notice that Seaside images are a bit "different" than our Dev images. This is mostly due to the preferences, the theme and maybe some tools (ECompeltion instead of OCompletion, etc..). So, what I suggest is to prepare a little script that can be downloadable and those people who want a Seaside image but more "similar" to our Dev images, just run that script.
So...what do you think? I would really like to hear your opinion about this.
Cheers
Mariano
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
1) First of all, I don't like the PharoWeb name. Because it is not Web in general, it is seaside. So, in my opinion, a better name would be PharoSeaside. Maybe even new comers may think that the only alternative is seaside. You know I have nothing against Seaside and that I even use it several times. I would like to see AidaWeb images based in Pharo too.
I think at some point the images also contained AIDA.
4) Now I created all the Metacello configurations for the PharoDev image and I took me a lot of time. Because I am not the developer of any of those projects. So, I need to understand the dependencies, the order, the versions, etc. But why ??? If Seaside people have already done that, don't duplicate efforts.
There is a configuration for Seaside 3.0 (and probably also Seaside 2.8). This was probably the first configuration every written. Ask Dale.
4) You may notice that Seaside images are a bit "different" than our Dev images. This is mostly due to the preferences, the theme and maybe some tools (ECompeltion instead of OCompletion, etc..). So, what I suggest is to prepare a little script that can be downloadable and those people who want a Seaside image but more "similar" to our Dev images, just run that script.
They are different because they reflect my preferences. I basically build the Seaside images for me. If other people like and use these images, that's great of course. Lukas -- Lukas Renggli http://www.lukas-renggli.ch
On Mon, Jan 4, 2010 at 9:45 AM, Lukas Renggli <renggli@gmail.com> wrote:
1) First of all, I don't like the PharoWeb name. Because it is not Web in general, it is seaside. So, in my opinion, a better name would be PharoSeaside. Maybe even new comers may think that the only alternative is seaside. You know I have nothing against Seaside and that I even use it several times. I would like to see AidaWeb images based in Pharo too.
I think at some point the images also contained AIDA.
No it doesn't. Not at least in the lasts images I have and in the scripts I have.
4) Now I created all the Metacello configurations for the PharoDev image and I took me a lot of time. Because I am not the developer of any of those projects. So, I need to understand the dependencies, the order, the versions, etc. But why ??? If Seaside people have already done that, don't duplicate efforts.
There is a configuration for Seaside 3.0 (and probably also Seaside 2.8). This was probably the first configuration every written. Ask Dale.
yes, that's true.
4) You may notice that Seaside images are a bit "different" than our Dev images. This is mostly due to the preferences, the theme and maybe some tools (ECompeltion instead of OCompletion, etc..). So, what I suggest is to prepare a little script that can be downloadable and those people who want a Seaside image but more "similar" to our Dev images, just run that script.
They are different because they reflect my preferences.
yes, of course. That's why I propose a little "script" that changes that to something more like our "dev" images.
I basically build the Seaside images for me.
Wait... The Seaside one click images and any other image that it is in the seaside website is not for you. I guess they are for the public.
If other people like and use these images, that's great of course.
Lukas
-- Lukas Renggli http://www.lukas-renggli.ch
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
I basically build the Seaside images for me.
Wait... The Seaside one click images and any other image that it is in the seaside website is not for you. I guess they are for the public.
No, they are not designed for the public, I build the images only for me. The fact that I put them on seaside.st is because people asked us to do so and this is really easy. If people like the images and it helps them, then this is great. If you are really into building images, you should build them for YOU and not for anybody else. Otherwise, I think, they are never going to fly. Lukas -- Lukas Renggli http://www.lukas-renggli.ch
On Tue, Jan 5, 2010 at 12:34 AM, Adrian Kuhn <akuhn@iam.unibe.ch> wrote:
Lukas Renggli <renggli@...> writes:
If you are really into building images, you should build them for YOU and not for anybody else. Otherwise, I think, they are never going to fly.
+1
But the concern about this emails was final users, maybe even beginners and new comers. Most of us (who have some time with Pharo) use our own build images, but they do not.
--AA
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
My big concern is that either testing of Pharo will suffer (things that are made difficult tend not to happen), or that a bunch of us will end up with ad-hoc build processes rather than having it done in one place with a chance of getting it right. Bill ________________________________ From: pharo-project-bounces@lists.gforge.inria.fr [mailto:pharo-project-bounces@lists.gforge.inria.fr] On Behalf Of Mariano Martinez Peck Sent: Tuesday, January 05, 2010 5:02 AM To: Pharo-project@lists.gforge.inria.fr Subject: Re: [Pharo-project] What about killing PharoWeb images? On Tue, Jan 5, 2010 at 12:34 AM, Adrian Kuhn <akuhn@iam.unibe.ch<mailto:akuhn@iam.unibe.ch>> wrote: Lukas Renggli <renggli@...> writes:
If you are really into building images, you should build them for YOU and not for anybody else. Otherwise, I think, they are never going to fly.
+1 But the concern about this emails was final users, maybe even beginners and new comers. Most of us (who have some time with Pharo) use our own build images, but they do not. --AA _______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr<mailto:Pharo-project@lists.gforge.inria.fr> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
2010/1/5 Schwab,Wilhelm K <bschwab@anest.ufl.edu>
My big concern is that either testing of Pharo will suffer (things that are made difficult tend not to happen), or that a bunch of us will end up with ad-hoc build processes rather than having it done in one place with a chance of getting it right.
I don't think so. How many people do you think that are using PharoWeb? I even don't know anyone. And we have enough beta testers for Dev images. And the difference between Dev and Web images is that in Web there are also the seaside packages, but there is NO change in the packages load. So, the only issues that can appear in PharoWeb but not in PharoDev are those related to Seaside, and that's seaside, not us. It would be cool to hear also the opinion of the pople who is using pharoweb images (if there is anyone). Cheers Mariano
Bill
------------------------------ *From:* pharo-project-bounces@lists.gforge.inria.fr [mailto: pharo-project-bounces@lists.gforge.inria.fr] *On Behalf Of *Mariano Martinez Peck *Sent:* Tuesday, January 05, 2010 5:02 AM *To:* Pharo-project@lists.gforge.inria.fr *Subject:* Re: [Pharo-project] What about killing PharoWeb images?
On Tue, Jan 5, 2010 at 12:34 AM, Adrian Kuhn <akuhn@iam.unibe.ch> wrote:
Lukas Renggli <renggli@...> writes:
If you are really into building images, you should build them for YOU and not for anybody else. Otherwise, I think, they are never going to fly.
+1
But the concern about this emails was final users, maybe even beginners and new comers. Most of us (who have some time with Pharo) use our own build images, but they do not.
--AA
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
You know someone (me). If the installation is not a moving target, then I can add it to the list of things I do to create a new image, but having Seaside in place has been appreciated. In general, "that's xyz's problem" wears thin. Taken to extremes, it could turn into "sure, the browser is broken, but that's not Pharo's doing; the debugger trashes the image sometimes, but that's the external debugger package that's to blame; you can't build stable web applications, but that's not Pharo's problem, ..." True as some claims of this type might be, we won't be able to get away with it forever, and we should have things that are known to work or look for ways to replace them. Bill ________________________________ From: pharo-project-bounces@lists.gforge.inria.fr [mailto:pharo-project-bounces@lists.gforge.inria.fr] On Behalf Of Mariano Martinez Peck Sent: Tuesday, January 05, 2010 8:37 AM To: Pharo-project@lists.gforge.inria.fr Subject: Re: [Pharo-project] What about killing PharoWeb images? 2010/1/5 Schwab,Wilhelm K <bschwab@anest.ufl.edu<mailto:bschwab@anest.ufl.edu>> My big concern is that either testing of Pharo will suffer (things that are made difficult tend not to happen), or that a bunch of us will end up with ad-hoc build processes rather than having it done in one place with a chance of getting it right. I don't think so. How many people do you think that are using PharoWeb? I even don't know anyone. And we have enough beta testers for Dev images. And the difference between Dev and Web images is that in Web there are also the seaside packages, but there is NO change in the packages load. So, the only issues that can appear in PharoWeb but not in PharoDev are those related to Seaside, and that's seaside, not us. It would be cool to hear also the opinion of the pople who is using pharoweb images (if there is anyone). Cheers Mariano Bill ________________________________ From: pharo-project-bounces@lists.gforge.inria.fr<mailto:pharo-project-bounces@lists.gforge.inria.fr> [mailto:pharo-project-bounces@lists.gforge.inria.fr<mailto:pharo-project-bounces@lists.gforge.inria.fr>] On Behalf Of Mariano Martinez Peck Sent: Tuesday, January 05, 2010 5:02 AM To: Pharo-project@lists.gforge.inria.fr<mailto:Pharo-project@lists.gforge.inria.fr> Subject: Re: [Pharo-project] What about killing PharoWeb images? On Tue, Jan 5, 2010 at 12:34 AM, Adrian Kuhn <akuhn@iam.unibe.ch<mailto:akuhn@iam.unibe.ch>> wrote: Lukas Renggli <renggli@...> writes:
If you are really into building images, you should build them for YOU and not for anybody else. Otherwise, I think, they are never going to fly.
+1 But the concern about this emails was final users, maybe even beginners and new comers. Most of us (who have some time with Pharo) use our own build images, but they do not. --AA _______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr<mailto:Pharo-project@lists.gforge.inria.fr> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project _______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr<mailto:Pharo-project@lists.gforge.inria.fr> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
2010/1/5 Schwab,Wilhelm K <bschwab@anest.ufl.edu>
You know someone (me). If the installation is not a moving target, then I can add it to the list of things I do to create a new image, but having Seaside in place has been appreciated.
Wait...one thing is not to have a PharoWeb image prepared and public so that people can download. But another thing is not to test. We have created a group of 13 beta testers where some of them will tests web images. They will test that using the metacello configuration. And if this is not working, we will make it work. I haven't time to tested by myself but as far as I know both seaside 2.8.4 and 3.0 metacello configurations are perfectly being load in a PharoDev image. So...seaside WILL be in place (if it not already done).
In general, "that's xyz's problem" wears thin. Taken to extremes, it could turn into "sure, the browser is broken, but that's not Pharo's doing; the debugger trashes the image sometimes, but that's the external debugger package that's to blame; you can't build stable web applications, but that's not Pharo's problem, ..."
Of course. Buy we need help. Do you take care of fixing and integrating OB and O2 ? Cool.
True as some claims of this type might be, we won't be able to get away with it forever, and we should have things that are known to work or look for ways to replace them.
Bill
------------------------------ *From:* pharo-project-bounces@lists.gforge.inria.fr [mailto: pharo-project-bounces@lists.gforge.inria.fr] *On Behalf Of *Mariano Martinez Peck *Sent:* Tuesday, January 05, 2010 8:37 AM
*To:* Pharo-project@lists.gforge.inria.fr *Subject:* Re: [Pharo-project] What about killing PharoWeb images?
2010/1/5 Schwab,Wilhelm K <bschwab@anest.ufl.edu>
My big concern is that either testing of Pharo will suffer (things that are made difficult tend not to happen), or that a bunch of us will end up with ad-hoc build processes rather than having it done in one place with a chance of getting it right.
I don't think so. How many people do you think that are using PharoWeb? I even don't know anyone. And we have enough beta testers for Dev images. And the difference between Dev and Web images is that in Web there are also the seaside packages, but there is NO change in the packages load. So, the only issues that can appear in PharoWeb but not in PharoDev are those related to Seaside, and that's seaside, not us.
It would be cool to hear also the opinion of the pople who is using pharoweb images (if there is anyone).
Cheers
Mariano
Bill
------------------------------ *From:* pharo-project-bounces@lists.gforge.inria.fr [mailto: pharo-project-bounces@lists.gforge.inria.fr] *On Behalf Of *Mariano Martinez Peck *Sent:* Tuesday, January 05, 2010 5:02 AM *To:* Pharo-project@lists.gforge.inria.fr *Subject:* Re: [Pharo-project] What about killing PharoWeb images?
On Tue, Jan 5, 2010 at 12:34 AM, Adrian Kuhn <akuhn@iam.unibe.ch> wrote:
Lukas Renggli <renggli@...> writes:
If you are really into building images, you should build them for YOU and not for anybody else. Otherwise, I think, they are never going to fly.
+1
But the concern about this emails was final users, maybe even beginners and new comers. Most of us (who have some time with Pharo) use our own build images, but they do not.
--AA
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
Ok....so, after reading all answers and comments I see the following possible actions: 1) Do nothing and keep things as now 2) Like 1) but change PharoWeb name to PharoSeaside. 3) Don't create PharoWeb images any more but gives the user the possibilities to use Seaside + Pharo. Now, my final point of view: 1) I don't like it because of what I have already said. 2) I like it more than 1) but anyway I don't like it. 3) For me it is the best solution. We can do, the following things: 3.a) Put in www.pharo-project.org/download a text among these lines: "There are few ways to develop Web application using Pharo. The most important and common framework is Seaside. Seaside is the web framework for building sophisticated applications in Smalltalk. Seaside provides what it is called "one click image" which is a compressed file that you only need to uncompress and double click to have a complete IDE to develop with seaside (even with web browser!!!). The link is... Take into account that this image may have different look and feel, preferences and tools from our one PharoDev image. This image is recommended for beginners and new comers. However, there is another way to use Seaside with Pharo and it is installing it by yourself in a PharoDev image. To do this, the best approach is to use the Metacello configuration. To do that, just evaluate the following code in a PharoDev image: Gofer....ConfiguratioOfSeaside...etc This apprach is recommended for people who already have a little of knoweldege with Pharo and Seaside. " 3.b) Using this way, we encourage people to use the Metacello configuration for Seaside and that's the most important thing because we can give feedback on it, fix issues, etc. And this is also shared for example with Gemstone. They will give feedback on us and us on them. We will probable make even simpler the move for applications from Pharo to Gemstone. 3.c) Those beta testers that said that they would test Web images, can just install Seaside exactly the same way a final user will do (use the metacello configuration). So...we are also testing that. 3.d) Not now, but maybe in a future we can even use the Esteban Lorenzano's Loader for Metacello. That would be even simpler. They will be able to take a PharoDev image and just evaluate: Loader load:'Seaside 2.8.4' and wala! So...I don't know how this kind of big decisions are taken. The borad decides? We do a poll ? I don't know. What do we do ? On Tue, Jan 5, 2010 at 11:01 AM, Mariano Martinez Peck < marianopeck@gmail.com> wrote:
On Tue, Jan 5, 2010 at 12:34 AM, Adrian Kuhn <akuhn@iam.unibe.ch> wrote:
Lukas Renggli <renggli@...> writes:
If you are really into building images, you should build them for YOU and not for anybody else. Otherwise, I think, they are never going to fly.
+1
But the concern about this emails was final users, maybe even beginners and new comers. Most of us (who have some time with Pharo) use our own build images, but they do not.
--AA
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
Mariano Martinez Peck wrote:
Ok....so, after reading all answers and comments I see the following possible actions:
1) Do nothing and keep things as now 2) Like 1) but change PharoWeb name to PharoSeaside. 3) Don't create PharoWeb images any more but gives the user the possibilities to use Seaside + Pharo.
The Pharo-Web image is not just the Pharo Seaside image though! I also remember earlier versions of Pharo-Web to have stuff like AIDAWeb and others, but I recently noticed too that was no longer the case. I would go option 4 :) Until Pharo has a well established user friendly package management system, why not look at the Pharo-Web image as Pharo's showcase web development image? A one-click image people can play around with without having to know too much about how to load stuff into their image. On another note, the Pharo team probalby has enough other stuff on their plate to worry too much about what to do with the Pharo-Web image :) -- View this message in context: http://n2.nabble.com/What-about-killing-PharoWeb-images-tp4249050p4255967.ht... Sent from the Pharo Smalltalk mailing list archive at Nabble.com.
On Tue, Jan 5, 2010 at 5:20 PM, Geert Claes <geert.wl.claes@gmail.com>wrote:
Mariano Martinez Peck wrote:
Ok....so, after reading all answers and comments I see the following possible actions:
1) Do nothing and keep things as now 2) Like 1) but change PharoWeb name to PharoSeaside. 3) Don't create PharoWeb images any more but gives the user the possibilities to use Seaside + Pharo.
The Pharo-Web image is not just the Pharo Seaside image though! I also remember earlier versions of Pharo-Web to have stuff like AIDAWeb and others, but I recently noticed too that was no longer the case.
I haven't see AIDAWeb in a Web image, but that's a very good idea.
I would go option 4 :) Until Pharo has a well established user friendly package management system, why not look at the Pharo-Web image as Pharo's showcase web development image? A one-click image people can play around with without having to know too much about how to load stuff into their image.
Because of all the things I already have said: why to duplicate efforts? In that case, you the Seaside one click. What cab be easier than that ? And it would be FAR more stable than ours. Do we create PharoGlassClient image ? no. Do we create PharoMoos image? no. Because it is obvious that they know better than us what versions to include. Of course, if Seaside ready images doesn't exist we will probably do it. Give me only one reason why spending time and resources in PharoWeb ? (taking into account you have other several alternatives better than this)
On another note, the Pharo team probalby has enough other stuff on their plate to worry too much about what to do with the Pharo-Web image :)
There is no such thing of "Pharo team". Pharo team are us. You. The community. Pharo has NO employee. There is the board but doesn't mean they are working full time in Pharo.
-- View this message in context: http://n2.nabble.com/What-about-killing-PharoWeb-images-tp4249050p4255967.ht... Sent from the Pharo Smalltalk mailing list archive at Nabble.com.
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
Mariano Martinez Peck wrote:
Because of all the things I already have said: why to duplicate efforts? In that case, you the Seaside one click. What cab be easier than that ? And it would be FAR more stable than ours. Do we create PharoGlassClient image ? no. Do we create PharoMoos image? no. Because it is obvious that they know better than us what versions to include. Of course, if Seaside ready images doesn't exist we will probably do it. Give me only one reason why spending time and resources in PharoWeb? (taking into account you have other several alternatives better than this)
I don't see it as duplicating efforts, the Pharo-Web image is (should be) more than Pharo-Seaside and having this sandpit image will probably give a lot more feedback than just having a Pharo-Dev image ... in my view. Mariano Martinez Peck wrote:
There is no such thing of "Pharo team". Pharo team are us. You. The community. Pharo has NO employee. There is the board but doesn't mean they are working full time in Pharo.
Settle down mate ... anyway, the Pharo Team is doing great stuff :) -- View this message in context: http://n2.nabble.com/What-about-killing-PharoWeb-images-tp4249050p4256091.ht... Sent from the Pharo Smalltalk mailing list archive at Nabble.com.
On Tue, Jan 5, 2010 at 5:48 PM, Geert Claes <geert.wl.claes@gmail.com>wrote:
Mariano Martinez Peck wrote:
Because of all the things I already have said: why to duplicate efforts? In that case, you the Seaside one click. What cab be easier than that ? And it would be FAR more stable than ours. Do we create PharoGlassClient image ? no. Do we create PharoMoos image? no. Because it is obvious that they know better than us what versions to include. Of course, if Seaside ready images doesn't exist we will probably do it. Give me only one reason why spending time and resources in PharoWeb? (taking into account you have other several alternatives better than this)
I don't see it as duplicating efforts, the Pharo-Web image is (should be) more than Pharo-Seaside and having this sandpit image will probably give a lot more feedback than just having a Pharo-Dev image ... in my view.
Ok, fortunately, we live in a free world :) Everybody can think whatever. And that's cool.
Mariano Martinez Peck wrote:
There is no such thing of "Pharo team". Pharo team are us. You. The community. Pharo has NO employee. There is the board but doesn't mean
they
are working full time in Pharo.
Settle down mate ... anyway, the Pharo Team is doing great stuff :)
Yes, no problem :) But this is not something small to say. For me this is very important. For example, do you know who use to do the dev images since the last 2 or even more years ? Damien. Unfortunately Damien is very complicated with his studies (in Java) and has very little (or almost nothing) to prepare the images. So, it is easy to pretend. But who will do the images right now ? You will ? Anyway, I am glad to hear all the opinions. So...instead of building bad images, unstable and that gives a bad impression to the final user I proposed a simple solution for this. Maybe it is not the best. But this what we have. I even think I would do what I said having good images (but forget this). Damien also use to integrate stuff on OB and O2 for example before release. Who will do that now ? I don't know. That's why I am since 2 weeks writing Metacello configurations for all packages and that's why I asked Damien if he could send me his scrips and instructions of how to build those images. Kind regards, Mariano --
View this message in context: http://n2.nabble.com/What-about-killing-PharoWeb-images-tp4249050p4256091.ht... Sent from the Pharo Smalltalk mailing list archive at Nabble.com.
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
As an example...today I was in a Stef lesson at the Ecole de Mines. They have to do a TP with Pharo, and it was really sad to see Stef asking me in the middle of the class if the dev image from the website was usable or not. Image the students thoughts. I had to put the last image I had "stable" and put it in a USB key and give it to the students. On Tue, Jan 5, 2010 at 5:58 PM, Mariano Martinez Peck <marianopeck@gmail.com
wrote:
On Tue, Jan 5, 2010 at 5:48 PM, Geert Claes <geert.wl.claes@gmail.com>wrote:
Mariano Martinez Peck wrote:
Because of all the things I already have said: why to duplicate efforts? In that case, you the Seaside one click. What cab be easier than that ? And it would be FAR more stable than ours. Do we create
PharoGlassClient
image ? no. Do we create PharoMoos image? no. Because it is obvious that they know better than us what versions to include. Of course, if Seaside ready images doesn't exist we will probably do it. Give me only one reason why spending time and resources in PharoWeb? (taking into account you have other several alternatives better than this)
I don't see it as duplicating efforts, the Pharo-Web image is (should be) more than Pharo-Seaside and having this sandpit image will probably give a lot more feedback than just having a Pharo-Dev image ... in my view.
Ok, fortunately, we live in a free world :) Everybody can think whatever. And that's cool.
Mariano Martinez Peck wrote:
There is no such thing of "Pharo team". Pharo team are us. You. The community. Pharo has NO employee. There is the board but doesn't mean
they
are working full time in Pharo.
Settle down mate ... anyway, the Pharo Team is doing great stuff :)
Yes, no problem :) But this is not something small to say. For me this is very important. For example, do you know who use to do the dev images since the last 2 or even more years ? Damien. Unfortunately Damien is very complicated with his studies (in Java) and has very little (or almost nothing) to prepare the images. So, it is easy to pretend. But who will do the images right now ? You will ? Anyway, I am glad to hear all the opinions. So...instead of building bad images, unstable and that gives a bad impression to the final user I proposed a simple solution for this. Maybe it is not the best. But this what we have. I even think I would do what I said having good images (but forget this).
Damien also use to integrate stuff on OB and O2 for example before release. Who will do that now ? I don't know.
That's why I am since 2 weeks writing Metacello configurations for all packages and that's why I asked Damien if he could send me his scrips and instructions of how to build those images.
Kind regards,
Mariano
--
View this message in context: http://n2.nabble.com/What-about-killing-PharoWeb-images-tp4249050p4256091.ht... Sent from the Pharo Smalltalk mailing list archive at Nabble.com.
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
My experience is that the images on the site probably are not "stable" as-is, but that means the process needs to be fixed rather than killing off useful products - sending people off on their own will not help anything in itself. We also need to put some obstacles in the way of inclusion in Pharo; the latter will have pros and cons, but I suspect it will eventually be necessary for us to have a quality product. Bill ________________________________ From: pharo-project-bounces@lists.gforge.inria.fr [mailto:pharo-project-bounces@lists.gforge.inria.fr] On Behalf Of Mariano Martinez Peck Sent: Tuesday, January 05, 2010 12:01 PM To: Pharo-project@lists.gforge.inria.fr Subject: Re: [Pharo-project] What about killing PharoWeb images? As an example...today I was in a Stef lesson at the Ecole de Mines. They have to do a TP with Pharo, and it was really sad to see Stef asking me in the middle of the class if the dev image from the website was usable or not. Image the students thoughts. I had to put the last image I had "stable" and put it in a USB key and give it to the students. On Tue, Jan 5, 2010 at 5:58 PM, Mariano Martinez Peck <marianopeck@gmail.com<mailto:marianopeck@gmail.com>> wrote: On Tue, Jan 5, 2010 at 5:48 PM, Geert Claes <geert.wl.claes@gmail.com<mailto:geert.wl.claes@gmail.com>> wrote: Mariano Martinez Peck wrote:
Because of all the things I already have said: why to duplicate efforts? In that case, you the Seaside one click. What cab be easier than that ? And it would be FAR more stable than ours. Do we create PharoGlassClient image ? no. Do we create PharoMoos image? no. Because it is obvious that they know better than us what versions to include. Of course, if Seaside ready images doesn't exist we will probably do it. Give me only one reason why spending time and resources in PharoWeb? (taking into account you have other several alternatives better than this)
I don't see it as duplicating efforts, the Pharo-Web image is (should be) more than Pharo-Seaside and having this sandpit image will probably give a lot more feedback than just having a Pharo-Dev image ... in my view. Ok, fortunately, we live in a free world :) Everybody can think whatever. And that's cool. Mariano Martinez Peck wrote:
There is no such thing of "Pharo team". Pharo team are us. You. The community. Pharo has NO employee. There is the board but doesn't mean they are working full time in Pharo.
Settle down mate ... anyway, the Pharo Team is doing great stuff :) Yes, no problem :) But this is not something small to say. For me this is very important. For example, do you know who use to do the dev images since the last 2 or even more years ? Damien. Unfortunately Damien is very complicated with his studies (in Java) and has very little (or almost nothing) to prepare the images. So, it is easy to pretend. But who will do the images right now ? You will ? Anyway, I am glad to hear all the opinions. So...instead of building bad images, unstable and that gives a bad impression to the final user I proposed a simple solution for this. Maybe it is not the best. But this what we have. I even think I would do what I said having good images (but forget this). Damien also use to integrate stuff on OB and O2 for example before release. Who will do that now ? I don't know. That's why I am since 2 weeks writing Metacello configurations for all packages and that's why I asked Damien if he could send me his scrips and instructions of how to build those images. Kind regards, Mariano -- View this message in context: http://n2.nabble.com/What-about-killing-PharoWeb-images-tp4249050p4256091.ht... Sent from the Pharo Smalltalk mailing list archive at Nabble.com. _______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr<mailto:Pharo-project@lists.gforge.inria.fr> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
Let us be pragmatic and focused. I would say that I would prefer to have one single pharo-dev image than a middle pharo-dev image and a pharo-web. We can focus on pharo-dev And let lukas ship a one-click seaside image. With the time the pharo-dev will support easy configuration to load seaside aida and a lot more so we will just offer one good image and let people decide. so I would focus on our customers first :) Stef
I too believe that is the best choice at this moment. Cheers, Doru On 5 Jan 2010, at 20:15, Stéphane Ducasse wrote:
Let us be pragmatic and focused.
I would say that I would prefer to have one single pharo-dev image than a middle pharo-dev image and a pharo-web. We can focus on pharo-dev And let lukas ship a one-click seaside image.
With the time the pharo-dev will support easy configuration to load seaside aida and a lot more so we will just offer one good image and let people decide. so I would focus on our customers first :)
Stef _______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
-- www.tudorgirba.com "Reasonable is what we are accustomed with."
Hello, Yes... we need to control the amount of work invested. And we need to be clear when we reach a point where we just can't provide something. So I vote for not having an official web image. Marcus On Jan 5, 2010, at 9:11 PM, Tudor Girba wrote:
I too believe that is the best choice at this moment.
Cheers, Doru
On 5 Jan 2010, at 20:15, Stéphane Ducasse wrote:
Let us be pragmatic and focused.
I would say that I would prefer to have one single pharo-dev image than a middle pharo-dev image and a pharo-web. We can focus on pharo-dev And let lukas ship a one-click seaside image.
With the time the pharo-dev will support easy configuration to load seaside aida and a lot more so we will just offer one good image and let people decide. so I would focus on our customers first :)
Stef _______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
-- www.tudorgirba.com
"Reasonable is what we are accustomed with."
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
me too On Jan 6, 2010, at 11:26 AM, Marcus Denker wrote:
Hello,
Yes... we need to control the amount of work invested. And we need to be clear when we reach a point where we just can't provide something.
So I vote for not having an official web image.
Marcus
On Jan 5, 2010, at 9:11 PM, Tudor Girba wrote:
I too believe that is the best choice at this moment.
Cheers, Doru
On 5 Jan 2010, at 20:15, Stéphane Ducasse wrote:
Let us be pragmatic and focused.
I would say that I would prefer to have one single pharo-dev image than a middle pharo-dev image and a pharo-web. We can focus on pharo-dev And let lukas ship a one-click seaside image.
With the time the pharo-dev will support easy configuration to load seaside aida and a lot more so we will just offer one good image and let people decide. so I would focus on our customers first :)
Stef _______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
-- www.tudorgirba.com
"Reasonable is what we are accustomed with."
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
2010/1/5 Mariano Martinez Peck <marianopeck@gmail.com>:
3) Don't create PharoWeb images any more but gives the user the possibilities to use Seaside + Pharo. 3) For me it is the best solution. We can do, the following things:
3.a) Put in www.pharo-project.org/download a text among these lines: 3.b)Â Using this way, we encourage people to use the Metacello configuration for Seaside and that's the most important thing because we can give feedback on it, fix issues, etc. And this is also shared for example with Gemstone. They will give feedback on us and us on them. We will probable make even simpler the move for applications from Pharo to Gemstone. 3.c) Those beta testers that said that they would test Web images, can just install Seaside exactly the same way a final user will do (use the metacello configuration). So...we are also testing that.
3.d) Not now, but maybe in a future we can even use the Esteban Lorenzano's Loader for Metacello. That would be even simpler. They will be able to take a PharoDev image and just evaluate:Â Loader load:'Seaside 2.8.4' and wala!
Lukas is right when he said that PharoWeb used to have Aida in. I stopped installing Aida inside PharoWeb images because it was kind of a pain to make Seaside and Aida work happily together. It doesn't take me much time to build PharoWeb images currently because Seaside 2.8 is quite stable. However, you may be right when you say that it is not Pharo's job to build Seaside images. Download counts can be seen on GForge website: https://gforge.inria.fr/frs/?group_id=1299. I also provide some not-up-to-date graphics: http://damiencassou.dabbledb.com/publish/dev-imagesdownloads. These statistics show that web images are downloaded by nearly as much people as the dev images. -- Damien Cassou http://damiencassou.seasidehosting.st "Lambdas are relegated to relative obscurity until Java makes them popular by not having them." James Iry
this is true that the data are really surprising to me. So I was wrong about the uses scenario. So some people really like to use web on advanced bleeding edge software :)
It doesn't take me much time to build PharoWeb images currently because Seaside 2.8 is quite stable. However, you may be right when you say that it is not Pharo's job to build Seaside images.
Damien I think that I would prefer you get your little time pushing pharo-dev :)
Download counts can be seen on GForge website: https://gforge.inria.fr/frs/?group_id=1299. I also provide some not-up-to-date graphics: http://damiencassou.dabbledb.com/publish/dev-imagesdownloads. These statistics show that web images are downloaded by nearly as much people as the dev images.
stef
----- "Lukas Renggli" <renggli@gmail.com> wrote: | > 4) Now I created all the Metacello configurations for the PharoDev | image and | > I took me a lot of time. Because I am not the developer of any of | those | > projects. So, I need to understand the dependencies, the order, the | > versions, etc. But why ??? If Seaside people have already done that, | don't | > duplicate efforts. | | There is a configuration for Seaside 3.0 (and probably also Seaside | 2.8). This was probably the first configuration every written. Ask | Dale. Yes Seaside3.0 was one of the original motivations behind Metacello:) There are Seaside3.0, Seaside2.8 and Pier (for Seaside2.8) configurations in the Metacello Repository and these were built based on information Lukas, so they should match pretty closely the web image content. I agree with the notion of one-click for ease of getting started with a particular framework/project. In the GLASS effort I've often outstripped the one-click images (and appliances) and run into support issues as a consequence. Once a developer is comfortable with the technology then they are likely to want to build their own image and if the one-click was built with a know configuration, then the developer has a reasonable chance of duplicating the code in his/her custom image. Dale
On Mon, Jan 4, 2010 at 7:14 PM, Dale Henrichs <dale.henrichs@gemstone.com>wrote:
----- "Lukas Renggli" <renggli@gmail.com> wrote:
| > 4) Now I created all the Metacello configurations for the PharoDev | image and | > I took me a lot of time. Because I am not the developer of any of | those | > projects. So, I need to understand the dependencies, the order, the | > versions, etc. But why ??? If Seaside people have already done that, | don't | > duplicate efforts. | | There is a configuration for Seaside 3.0 (and probably also Seaside | 2.8). This was probably the first configuration every written. Ask | Dale.
Yes Seaside3.0 was one of the original motivations behind Metacello:)
There are Seaside3.0, Seaside2.8 and Pier (for Seaside2.8) configurations in the Metacello Repository and these were built based on information Lukas, so they should match pretty closely the web image content.
This is really cool. And just because of this I think that we don't need the PharoWeb image, as you have the following options already: - Use metacello configuration by yourself. It is just one line of code and we can even explain in how to do it in the website. Even more, you can load that conf in a Pharo Dev, at that's all :) - Use Seaside one click - Use Lukas script.
I agree with the notion of one-click for ease of getting started with a particular framework/project. In the GLASS effort I've often outstripped the one-click images (and appliances) and run into support issues as a consequence.
Once a developer is comfortable with the technology then they are likely to want to build their own image and if the one-click was built with a know configuration, then the developer has a reasonable chance of duplicating the code in his/her custom image.
Dale
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
Mariano Martinez Peck wrote:
Please, don't say "NOOOOO!" immediately. Read the mail carefully, take a deep breath and think cold.
After working a couple of weeks in the Metacello configurations and now with Damien's scripts to build Dev and Web images, I came to the question: Do we really need a PharoWeb image?
I will give just my opinions of why having PharoWeb images is not a good idea anymore
1) First of all, I don't like the PharoWeb name. Because it is not Web in general, it is seaside. So, in my opinion, a better name would be PharoSeaside. Maybe even new comers may think that the only alternative is seaside. You know I have nothing against Seaside and that I even use it several times. I would like to see AidaWeb images based in Pharo too.
2) Why to still have PharoWeb images if now Seaside images (at least the one click) are based in seaside ? In addition, when I develop in Seaside, I use Seaside images, not PharoWeb, and I think most people do the same.
3) Seaside images (or Lukas images for seaside) are MUCH more stable than the PharoWeb images. Why? Several reasons: 3.a) THEY know the dependencies of the packages, the versions, how to set up, the order of load, etc, NOT us. THEY are the experts in seaside, not us. 3.b) THEY really test their images before make then public. 3.c) We are "force" to release one release per month which has no sense. How says that Seaside is released once a month? That's bad because it means we are usually taking unstable versions. THEY know when a release should be done and when a release is stable or not.
4) Now I created all the Metacello configurations for the PharoDev image and I took me a lot of time. Because I am not the developer of any of those projects. So, I need to understand the dependencies, the order, the versions, etc. But why ??? If Seaside people have already done that, don't duplicate efforts.
5) The instability of PharoWeb images give to our public a bad impression. We waste all of out efforts of Pharo in that image.
6) The idea of Pharo is to be the vehicle. And that's exactly the Pharo role in the Seaside images. We don't need to build PharoWeb images. Even more, do we build PharoGlass images ? NOOOO. Why not ? Because Dale already prepare them and as HE knows much more about Glass than us, his images are better for Glass.
7) Tracking and managing issues always take time. Even analyzing a bug, say "we won't fix" and close it. With PharoWeb images we are having those kind of issues.
8) Our resources are limited. Why not to use them in more important stuff ? For example, building a better and more robust Dev image?
Here is my proposal:
1) Completely remove the idea of PharoWeb image. Remove the links from the website and do not create Web images any more. 2) In http://www.pharo-project.org/pharo-download add a description explaining that now Seaside images are based in Pharo, so you can directly download that image. We can even explain a bit about seaside, and even if you want you can put a href called "PharoWeb" which links to the seaside image. However, I would rather call it "PharoSeaside". 3) Tomorrow we can have AidaWeb images based in pharo (please Janko!!) so we can just do the same and put the link to those images. 4) You may notice that Seaside images are a bit "different" than our Dev images. This is mostly due to the preferences, the theme and maybe some tools (ECompeltion instead of OCompletion, etc..). So, what I suggest is to prepare a little script that can be downloadable and those people who want a Seaside image but more "similar" to our Dev images, just run that script.
So...what do you think? I would really like to hear your opinion about this.
Cheers
Mariano
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
+1 I would definitely support concentrating on the Seaside one click images. As Mariano says, they're pretty stable, so you can get on with developing in them.
From a marketing point of view, the one click images are much easier for someone who is evaluating Pharo for the first time. Also, the existing web dev is not the advertisement for Pharo that we want.
There is a (very) slight complication. Lukas's recommendation for people starting a new project with JavaScript is to use JQueryUI, hence this implies Seaside 3.0. The Seaside 3.0 image on the Seaside site is not an official release: http://n4.nabble.com/2-1Using-the-One-Click-Image-tt975474.html#a975784 http://n4.nabble.com/2-1Using-the-One-Click-Image-tt975474.html#a975784 So the ideal solution is that that version become an official 3.0 one-click, although this was not in the Seaside plan. And maybe at the same time the Seaside team could apply (or allow to be applied) the few changes that Mariano mentioned to make that image seem more familiar. This would also make this image have the same 'look and feel' as the Seaside book. On the negative side of the ledger, there would be less testers/debuggers of the Pharo side; the problem areas of the OB/O2/Refactoring combination would not be exercised by many end users, until a stable combination could be put in the one click. Mariano, do you use the 2.8.4 or the 3.0 image? ...Stan -- View this message in context: http://n2.nabble.com/What-about-killing-PharoWeb-images-tp4249050p4249415.ht... Sent from the Pharo Smalltalk mailing list archive at Nabble.com.
We certainly don't need the web images, but I would like to see them stay to help with testing. If we are building them with unstable versions of Seaside, that is a problem with the process (loading the wrong stuff) not the concept/service (providing a Seaside-ready version of Pharo for testing). We just finished designing a beta program, at least three of volunteers indicated a preference for the web image. Bill ________________________________ From: pharo-project-bounces@lists.gforge.inria.fr [mailto:pharo-project-bounces@lists.gforge.inria.fr] On Behalf Of Mariano Martinez Peck Sent: Monday, January 04, 2010 3:25 AM To: Pharo Development Subject: [Pharo-project] What about killing PharoWeb images? Please, don't say "NOOOOO!" immediately. Read the mail carefully, take a deep breath and think cold. After working a couple of weeks in the Metacello configurations and now with Damien's scripts to build Dev and Web images, I came to the question: Do we really need a PharoWeb image? I will give just my opinions of why having PharoWeb images is not a good idea anymore 1) First of all, I don't like the PharoWeb name. Because it is not Web in general, it is seaside. So, in my opinion, a better name would be PharoSeaside. Maybe even new comers may think that the only alternative is seaside. You know I have nothing against Seaside and that I even use it several times. I would like to see AidaWeb images based in Pharo too. 2) Why to still have PharoWeb images if now Seaside images (at least the one click) are based in seaside ? In addition, when I develop in Seaside, I use Seaside images, not PharoWeb, and I think most people do the same. 3) Seaside images (or Lukas images for seaside) are MUCH more stable than the PharoWeb images. Why? Several reasons: 3.a) THEY know the dependencies of the packages, the versions, how to set up, the order of load, etc, NOT us. THEY are the experts in seaside, not us. 3.b) THEY really test their images before make then public. 3.c) We are "force" to release one release per month which has no sense. How says that Seaside is released once a month? That's bad because it means we are usually taking unstable versions. THEY know when a release should be done and when a release is stable or not. 4) Now I created all the Metacello configurations for the PharoDev image and I took me a lot of time. Because I am not the developer of any of those projects. So, I need to understand the dependencies, the order, the versions, etc. But why ??? If Seaside people have already done that, don't duplicate efforts. 5) The instability of PharoWeb images give to our public a bad impression. We waste all of out efforts of Pharo in that image. 6) The idea of Pharo is to be the vehicle. And that's exactly the Pharo role in the Seaside images. We don't need to build PharoWeb images. Even more, do we build PharoGlass images ? NOOOO. Why not ? Because Dale already prepare them and as HE knows much more about Glass than us, his images are better for Glass. 7) Tracking and managing issues always take time. Even analyzing a bug, say "we won't fix" and close it. With PharoWeb images we are having those kind of issues. 8) Our resources are limited. Why not to use them in more important stuff ? For example, building a better and more robust Dev image? Here is my proposal: 1) Completely remove the idea of PharoWeb image. Remove the links from the website and do not create Web images any more. 2) In http://www.pharo-project.org/pharo-download add a description explaining that now Seaside images are based in Pharo, so you can directly download that image. We can even explain a bit about seaside, and even if you want you can put a href called "PharoWeb" which links to the seaside image. However, I would rather call it "PharoSeaside". 3) Tomorrow we can have AidaWeb images based in pharo (please Janko!!) so we can just do the same and put the link to those images. 4) You may notice that Seaside images are a bit "different" than our Dev images. This is mostly due to the preferences, the theme and maybe some tools (ECompeltion instead of OCompletion, etc..). So, what I suggest is to prepare a little script that can be downloadable and those people who want a Seaside image but more "similar" to our Dev images, just run that script. So...what do you think? I would really like to hear your opinion about this. Cheers Mariano
2010/1/4 Schwab,Wilhelm K <bschwab@anest.ufl.edu>
We certainly don't need the web images, but I would like to see them stay to help with testing. If we are building them with unstable versions of Seaside, that is a problem with the process (loading the wrong stuff) not the concept/service (providing a Seaside-ready version of Pharo for testing). We just finished designing a beta program, at least three of volunteers indicated a preference for the web image.
Bill: On the contrary: I don't want to have beta testers in PharoWeb and then others in Seaside one click images (which are also based in Pharo). My idea is the contrary: JOIN FORCES. If people who were now using PharoWeb image, start to use Seaside images, they will also help to test and debugs. A lot of issues reported in those images in Seaside were actually Pharo issues. In addition, is VERY uncommon that an issue of the "IDE" appears in PharoWeb but not in PharoDev. So, most of the issues in PharoWeb that are not present in PharoDev may be related to Seaside, and even probably, because we are using worng version of the packages. Cheers Mariano
Bill
------------------------------ *From:* pharo-project-bounces@lists.gforge.inria.fr [mailto: pharo-project-bounces@lists.gforge.inria.fr] *On Behalf Of *Mariano Martinez Peck *Sent:* Monday, January 04, 2010 3:25 AM *To:* Pharo Development *Subject:* [Pharo-project] What about killing PharoWeb images?
Please, don't say "NOOOOO!" immediately. Read the mail carefully, take a deep breath and think cold.
After working a couple of weeks in the Metacello configurations and now with Damien's scripts to build Dev and Web images, I came to the question: Do we really need a PharoWeb image?
I will give just my opinions of why having PharoWeb images is not a good idea anymore
1) First of all, I don't like the PharoWeb name. Because it is not Web in general, it is seaside. So, in my opinion, a better name would be PharoSeaside. Maybe even new comers may think that the only alternative is seaside. You know I have nothing against Seaside and that I even use it several times. I would like to see AidaWeb images based in Pharo too.
2) Why to still have PharoWeb images if now Seaside images (at least the one click) are based in seaside ? In addition, when I develop in Seaside, I use Seaside images, not PharoWeb, and I think most people do the same.
3) Seaside images (or Lukas images for seaside) are MUCH more stable than the PharoWeb images. Why? Several reasons: 3.a) THEY know the dependencies of the packages, the versions, how to set up, the order of load, etc, NOT us. THEY are the experts in seaside, not us. 3.b) THEY really test their images before make then public. 3.c) We are "force" to release one release per month which has no sense. How says that Seaside is released once a month? That's bad because it means we are usually taking unstable versions. THEY know when a release should be done and when a release is stable or not.
4) Now I created all the Metacello configurations for the PharoDev image and I took me a lot of time. Because I am not the developer of any of those projects. So, I need to understand the dependencies, the order, the versions, etc. But why ??? If Seaside people have already done that, don't duplicate efforts.
5) The instability of PharoWeb images give to our public a bad impression. We waste all of out efforts of Pharo in that image.
6) The idea of Pharo is to be the vehicle. And that's exactly the Pharo role in the Seaside images. We don't need to build PharoWeb images. Even more, do we build PharoGlass images ? NOOOO. Why not ? Because Dale already prepare them and as HE knows much more about Glass than us, his images are better for Glass.
7) Tracking and managing issues always take time. Even analyzing a bug, say "we won't fix" and close it. With PharoWeb images we are having those kind of issues.
8) Our resources are limited. Why not to use them in more important stuff ? For example, building a better and more robust Dev image?
Here is my proposal:
1) Completely remove the idea of PharoWeb image. Remove the links from the website and do not create Web images any more. 2) In http://www.pharo-project.org/pharo-download add a description explaining that now Seaside images are based in Pharo, so you can directly download that image. We can even explain a bit about seaside, and even if you want you can put a href called "PharoWeb" which links to the seaside image. However, I would rather call it "PharoSeaside". 3) Tomorrow we can have AidaWeb images based in pharo (please Janko!!) so we can just do the same and put the link to those images. 4) You may notice that Seaside images are a bit "different" than our Dev images. This is mostly due to the preferences, the theme and maybe some tools (ECompeltion instead of OCompletion, etc..). So, what I suggest is to prepare a little script that can be downloadable and those people who want a Seaside image but more "similar" to our Dev images, just run that script.
So...what do you think? I would really like to hear your opinion about this.
Cheers
Mariano
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
2010/1/4 Schwab,Wilhelm K <bschwab@anest.ufl.edu>:
We certainly don't need the web images, but I would like to see them stay to help with testing. If we are building them with unstable versions of Seaside, that is a problem with the process (loading the wrong stuff) not the concept/service (providing a Seaside-ready version of Pharo for testing).  We just finished designing a beta program, at least three of volunteers indicated a preference for the web image.
Bill
1. *We* (who sitting here for a while), certainly don't need web images. 2. Newcomers, or entry-level youngsters, definitely will prefer a ready-to-use solutions, which easy to install, run & play with. So, +1 for keeping web images. -- Best regards, Igor Stasenko AKA sig.
On Mon, Jan 4, 2010 at 2:01 PM, Igor Stasenko <siguctua@gmail.com> wrote:
2010/1/4 Schwab,Wilhelm K <bschwab@anest.ufl.edu>:
We certainly don't need the web images, but I would like to see them stay to help with testing. If we are building them with unstable versions of Seaside, that is a problem with the process (loading the wrong stuff) not the concept/service (providing a Seaside-ready version of Pharo for testing). We just finished designing a beta program, at least three of volunteers indicated a preference for the web image.
Bill
1. *We* (who sitting here for a while), certainly don't need web images. 2. Newcomers, or entry-level youngsters, definitely will prefer a ready-to-use solutions, which easy to install, run & play with.
So, +1 for keeping web images.
But Igor, there is nothing easier than a double click. And seaside one click is just a double click no matter in what OS you are. They are MUCH more easier for new commers than a PharoWeb image where they have to download the VM, the sources and the image.
-- Best regards, Igor Stasenko AKA sig.
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
Hi Mariano, I also agree with you. Cheers, Doru On 4 Jan 2010, at 09:24, Mariano Martinez Peck wrote:
Please, don't say "NOOOOO!" immediately. Read the mail carefully, take a deep breath and think cold.
After working a couple of weeks in the Metacello configurations and now with Damien's scripts to build Dev and Web images, I came to the question: Do we really need a PharoWeb image?
I will give just my opinions of why having PharoWeb images is not a good idea anymore
1) First of all, I don't like the PharoWeb name. Because it is not Web in general, it is seaside. So, in my opinion, a better name would be PharoSeaside. Maybe even new comers may think that the only alternative is seaside. You know I have nothing against Seaside and that I even use it several times. I would like to see AidaWeb images based in Pharo too.
2) Why to still have PharoWeb images if now Seaside images (at least the one click) are based in seaside ? In addition, when I develop in Seaside, I use Seaside images, not PharoWeb, and I think most people do the same.
3) Seaside images (or Lukas images for seaside) are MUCH more stable than the PharoWeb images. Why? Several reasons: 3.a) THEY know the dependencies of the packages, the versions, how to set up, the order of load, etc, NOT us. THEY are the experts in seaside, not us. 3.b) THEY really test their images before make then public. 3.c) We are "force" to release one release per month which has no sense. How says that Seaside is released once a month? That's bad because it means we are usually taking unstable versions. THEY know when a release should be done and when a release is stable or not.
4) Now I created all the Metacello configurations for the PharoDev image and I took me a lot of time. Because I am not the developer of any of those projects. So, I need to understand the dependencies, the order, the versions, etc. But why ??? If Seaside people have already done that, don't duplicate efforts.
5) The instability of PharoWeb images give to our public a bad impression. We waste all of out efforts of Pharo in that image.
6) The idea of Pharo is to be the vehicle. And that's exactly the Pharo role in the Seaside images. We don't need to build PharoWeb images. Even more, do we build PharoGlass images ? NOOOO. Why not ? Because Dale already prepare them and as HE knows much more about Glass than us, his images are better for Glass.
7) Tracking and managing issues always take time. Even analyzing a bug, say "we won't fix" and close it. With PharoWeb images we are having those kind of issues.
8) Our resources are limited. Why not to use them in more important stuff ? For example, building a better and more robust Dev image?
Here is my proposal:
1) Completely remove the idea of PharoWeb image. Remove the links from the website and do not create Web images any more. 2) In http://www.pharo-project.org/pharo-download add a description explaining that now Seaside images are based in Pharo, so you can directly download that image. We can even explain a bit about seaside, and even if you want you can put a href called "PharoWeb" which links to the seaside image. However, I would rather call it "PharoSeaside". 3) Tomorrow we can have AidaWeb images based in pharo (please Janko!!) so we can just do the same and put the link to those images. 4) You may notice that Seaside images are a bit "different" than our Dev images. This is mostly due to the preferences, the theme and maybe some tools (ECompeltion instead of OCompletion, etc..). So, what I suggest is to prepare a little script that can be downloadable and those people who want a Seaside image but more "similar" to our Dev images, just run that script.
So...what do you think? I would really like to hear your opinion about this.
Cheers
Mariano
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
-- www.tudorgirba.com "What is more important: To be happy, or to make happy?"
Em 04/01/2010 06:24, Mariano Martinez Peck < marianopeck@gmail.com > escreveu:
Please, don't say "NOOOOO!" immediately. Read the mail carefully, take a deep breath and think cold. After working a couple of weeks in the Metacello configurations and now with Damien's scripts to build Dev and Web images, I came to the question: Do we really need a PharoWeb image? I will give just my opinions of why having PharoWeb images is not a good idea anymore [snipped]
On the whole I appreciate and agree with the reasoning.
Here is my proposal: 1) Completely remove the idea of PharoWeb image. Remove the links from the website and do not create Web images any more.
+1
2) In http://www.pharo-project.org/pharo-download add a description explaining that now Seaside images are based in Pharo, so you can directly download that image. We can even explain a bit about seaside, and even if you want you can put a href called "PharoWeb" which links to the seaside image. However, I would rather call it "PharoSeaside".
I think here we have to work hard to stress this fact since IIUC the /pièce de résistance/ for Pharo is having all production work done by Seaside.
3) Tomorrow we can have AidaWeb images based in pharo (please Janko!!) so we can just do the same and put the link to those images.
While I'm fond of the idea on having AidaWeb very soon as possible, I also think we should put ourselves in Janko's shoes and understand that he'll need to master Metacelllo to arrive at the packaging. . .
4) You may notice that Seaside images are a bit "different" than our Dev images. This is mostly due to the preferences, the theme and maybe some tools (ECompeltion instead of OCompletion, etc..). So, what I suggest is to prepare a little script that can be downloadable and those people who want a Seaside image but more "similar" to our Dev images, just run that script.
I would postpone the publishing of these kind of scripts for a much more stable Pharo and Metacello (or packaging solution ensuing), otherwise you'll have the same work as testing the scripts means creating and testing the packages!!
So...what do you think? I would really like to hear your opinion about this.
my 0.01999999.... -- Cesar Rabak
On Mon, Jan 4, 2010 at 5:22 PM, <csrabak@bol.com.br> wrote:
Em 04/01/2010 06:24, Mariano Martinez Peck < marianopeck@gmail.com > escreveu:
Please, don't say "NOOOOO!" immediately. Read the mail carefully, take a deep breath and think cold. After working a couple of weeks in the Metacello configurations and now with Damien's scripts to build Dev and Web images, I came to the question: Do we really need a PharoWeb image? I will give just my opinions of why having PharoWeb images is not a good idea anymore [snipped]
On the whole I appreciate and agree with the reasoning.
Here is my proposal: 1) Completely remove the idea of PharoWeb image. Remove the links from the website and do not create Web images any more.
+1
2) In http://www.pharo-project.org/pharo-download add a description explaining that now Seaside images are based in Pharo, so you can directly download that image. We can even explain a bit about seaside, and even if you want you can put a href called "PharoWeb" which links to the seaside image. However, I would rather call it "PharoSeaside".
I think here we have to work hard to stress this fact since IIUC the /pièce de résistance/ for Pharo is having all production work done by Seaside.
I am not sure if I understood you correctly. Of course that supports Seaside is one of the most important task in Pharo, but not the only one. I don't know what do you mean with production, but I know few "production" application running in Pharo without Seaside. As Marcus or Gary for example. I really would like Pharo to be the vehicle for all of them: Seaside, AidaWeb, GemTools, Moose, etc.
3) Tomorrow we can have AidaWeb images based in pharo (please Janko!!) so we can just do the same and put the link to those images.
While I'm fond of the idea on having AidaWeb very soon as possible, I also think we should put ourselves in Janko's shoes and understand that he'll need to master Metacelllo to arrive at the packaging. . .
Yes, maybe. But that's not in our discussion. I mean, suppose I don't care how Janko builds his images, I really would like that to be also "PharoWeb". So, it that case we just add one link more to our website: "Pharo Web with Seaside" , "Pharo Web with AidaWeb" , "Pharo Client for Gemstone", etc...
4) You may notice that Seaside images are a bit "different" than our Dev images. This is mostly due to the preferences, the theme and maybe some tools (ECompeltion instead of OCompletion, etc..). So, what I suggest is to prepare a little script that can be downloadable and those people who want a Seaside image but more "similar" to our Dev images, just run that script.
I would postpone the publishing of these kind of scripts for a much more stable Pharo and Metacello (or packaging solution ensuing), otherwise you'll have the same work as testing the scripts means creating and testing the packages!!
Good point. If you want we can achieve that later. Even for 1.1. But at least, for the moment, for the newcomers, I doesn't cost anything a little script to at least set the preferences and theme we use in PharoDev (this has nothing to do with the Package Managment). Thanks for giving me your opinion Carlos. Cheers Mariano
So...what do you think? I would really like to hear your opinion about this.
my 0.01999999....
-- Cesar Rabak
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
csrabak wrote:
...
3) Tomorrow we can have AidaWeb images based in pharo (please Janko!!) so we can just do the same and put the link to those images.
While I'm fond of the idea on having AidaWeb very soon as possible, I also think we should put ourselves in Janko's shoes and understand that he'll need to master Metacelllo to arrive at the packaging. . .
+1 csrabak wrote:
... I would postpone the publishing of these kind of scripts for a much more stable Pharo and Metacello (or packaging solution ensuing), otherwise you'll have the same work as testing the scripts means creating and testing the packages!!
Once Pharo has arrived at a stage where it is extremely easy to manage components/project packages - both from the project and user's point of view - there will still a place to provide a stable environment for those interrested in a dev/web/seaside/aidaweb/glass etc type of image.
From a testing and support point of view, this discussion actually makes me think in a way about the "Pharo-beginner mailing list" suggestion. Pharo's main focuss is on Pharo-Core. On top of Pharo core there may be a number of projects e.g : Pharo-Seaside, Pharo-AIDAWeb, Pharo-Moose, Pharo-GLASS, which all can be loaded into (or un-loaded from) the image (I guess the ConfigurationOfNnnn type scripts). There should be no loss in testers because issues can still be logged, discussed and escallated to the issue culprit, whether it was a configuration script issue or a Pharo-Core issue, no? -- View this message in context: http://n2.nabble.com/What-about-killing-PharoWeb-images-tp4249050p4255309.ht... Sent from the Pharo Smalltalk mailing list archive at Nabble.com.
participants (13)
-
Adrian Kuhn -
csrabak＠bol.com.br -
Dale Henrichs -
Damien Cassou -
Geert Claes -
Igor Stasenko -
Lukas Renggli -
Marcus Denker -
Mariano Martinez Peck -
Schwab,Wilhelm K -
Stan Shepherd -
Stéphane Ducasse -
Tudor Girba