Hi all - Paul sent a note to zope-coders some time back fishing for some feedback regarding planning for a Zope 2.6 (excepted):
I propose that planning for a 2.6 focus on the following thoughts:
1) This release is more about community-contributed work than ZC-contributed work. ...
Ultimately the things that go in 2.6 will be governed by what people in the community are willing to contribute.
So far the extent of contributions has been minor. Do we need to widen the pool? Or do we instead try to get new contributors on the Zope3 track?
There wasn't much response to this when it was originally posted. Now that 2.5.1 is nearly out the door, it's time to start planning for a Zope 2.6. I've started a 2.6 plan on the dev site at: http://dev.zope.org/Resources/zope_260_plan.html I wholeheartedly agree that 2.6 needs to be significantly a community effort. While I know that many people are engaged in the Zope 3 effort, we also need to get some people engaged on defining and producing 2.6 in the interim. There is not much on the plan right now, so the possibilities are relatively wide open :) Let's get a discussion started to define 2.6. Brian Lloyd brian@zope.com Software Engineer 540.361.1716 Zope Corporation http://www.zope.com
I went to http://dev.zope.org/Resources/zope_260_plan.html and logged in, but couldn't see any way to contribute comments to this page. I could comment on major projects, but none of them fit my objective. -- I've modified HTTPResponse and ZServer/HTTPResponse.py to allow for gzip content encoding on a response-by-response basis. I'm mostly using this with xml-rpc, but it could be generalized and combined with a gzipper- cache manager. I realize that Transfer encoding is really what should be supported, but the fact is that many clients don't handle transfer encoding, but most do handle content-encoding. How can I contribute this? It'll probably need some tweaking to bring it into "standards". -- I've modified medusa/ftp_server to allow specifying the PASV port range in the environment. How do I contribute this? --- Do I just submit both of these to patch manager? Brad Clements, bkc@murkworks.com (315)268-1000 http://www.murkworks.com (315)268-9812 Fax netmeeting: ils://ils.murkworks.com AOL-IM: BKClements
Either you become an official Zope committer with write-access to the CVS or by providing reasonable patches, documentation if needed and updated unit tests. Andreas ----- Original Message ----- From: "Brad Clements" <bkc@murkworks.com> To: <zope-dev@zope.org> Sent: Thursday, February 28, 2002 13:35 Subject: Re: [Zope-dev] Zope 2.6 planning - call for contributors!
I went to http://dev.zope.org/Resources/zope_260_plan.html and logged in, but couldn't see any way to contribute comments to this page. I could comment on major projects, but none of them fit my objective.
--
I've modified HTTPResponse and ZServer/HTTPResponse.py to allow for gzip content encoding on a response-by-response basis. I'm mostly using this with xml-rpc, but it could be generalized and combined with a gzipper- cache manager.
I realize that Transfer encoding is really what should be supported, but the fact is that many clients don't handle transfer encoding, but most do handle content-encoding.
How can I contribute this? It'll probably need some tweaking to bring it into "standards".
--
I've modified medusa/ftp_server to allow specifying the PASV port range in the environment.
How do I contribute this?
---
Do I just submit both of these to patch manager?
Brad Clements, bkc@murkworks.com (315)268-1000 http://www.murkworks.com (315)268-9812 Fax netmeeting: ils://ils.murkworks.com AOL-IM: BKClements
_______________________________________________ Zope-Dev maillist - Zope-Dev@zope.org http://lists.zope.org/mailman/listinfo/zope-dev ** No cross posts or HTML encoding! ** (Related lists - http://lists.zope.org/mailman/listinfo/zope-announce http://lists.zope.org/mailman/listinfo/zope )
I've modified HTTPResponse and ZServer/HTTPResponse.py to allow for gzip content encoding on a response-by-response basis. I'm mostly using this with xml-rpc, but it could be generalized and combined with a gzipper- cache manager.
I'd like this. It would help with the lack of transfer-encoding by Squid.
Brian Lloyd wrote:
Let's get a discussion started to define 2.6.
This isn't exciting by any means unless you're one of the people who package Zope up for distribution, or maybe you're one of the people who manage lots of little Zopes on one system; but I'd like to revive the "grand unified Zope installation and control" proposal that has been floated by many people (including me) in one form or another for some time. Wikiwise, this would wrap up http://dev.zope.org/Wikis/DevSite/Proposals/ZopeStartupProvisions and http://dev.zope.org/Wikis/DevSite/Proposals/InstallationAndConfiguration, at least. To summarize, this would involve - an expanded build program with an installation scheme that would allow multiple versions of Zope to be present on the same system - making that installation 'secure by default' - a registry of Zope installations and one of instances and their configuration settings - a 'zopectl' program or similar that would be able to start and stop instances - a 'zopeinstance' program or similar that would become the _recommended_ way of setting up Zope, by creating an INSTANCE_HOME It would be nice if - the same framework could apply to Zope 3, maybe taking care of that piece ahead of time I'm more than willing to head this up, though I question how long we have before 2.6 to do so.
On Thursday 28 February 2002 04:03 pm, you wrote:
Brian Lloyd wrote:
Let's get a discussion
started to define 2.6.
This isn't exciting by any means unless you're one of the people who package Zope up for distribution, or maybe you're one of the people who manage lots of little Zopes on one system; but I'd like to revive the "grand unified Zope installation and control" proposal that has been floated by many people (including me) in one form or another for some time. Wikiwise, this would wrap up http://dev.zope.org/Wikis/DevSite/Proposals/ZopeStartupProvisions and http://dev.zope.org/Wikis/DevSite/Proposals/InstallationAndConfiguration, at least.
To summarize, this would involve
- an expanded build program with an installation scheme that would allow multiple versions of Zope to be present on the same system
+1 I just did this manually yesterday and it would be *really* nice to have it work like this out of the box. I think it would really improve first impressions of Zope from a site admin perspective. /---------------------------------------------------\ Casey Duncan, Sr. Web Developer National Legal Aid and Defender Association c.duncan@nlada.org \---------------------------------------------------/
Let's get a discussion
started to define 2.6
I'd like to see the ZSyncer Product, or a variant thereof, included in Zope by default. That is, I'd like "Synchronization", to a be a default property of Zope objects, so that objects/content can be pushed and pulled between two Zope installations. I venture a guess that the development/production model is common, and this helps quite a bit in maintaining this model. Ziniti
Let's get a discussion
started to define 2.6
I'd like to see the ZSyncer Product, or a variant thereof, included in Zope by default. That is, I'd like "Synchronization", to a be a default property of Zope objects, so that objects/content can be pushed and pulled between two Zope installations.
I venture a guess that the development/production model is common, and this helps quite a bit in maintaining this model.
Ziniti
Absolutely ... and I would also like to see Richards excellent Call Profiler service become part of the core. It is an (absolutely) essential tool for a number of reasons. My impression from previous threads is that others feel the same way too. I would also like to know what level of effort it would take to ramp-up transparent folders so that it could be considered core ready??? Anyway, my two cents... Eric
On Fri, 1 Mar 2002 15:17, Eric Roby wrote:
Let's get a discussion
started to define 2.6
I'd like to see the ZSyncer Product, or a variant thereof, included in Zope by default. That is, I'd like "Synchronization", to a be a default property of Zope objects, so that objects/content can be pushed and pulled between two Zope installations.
I venture a guess that the development/production model is common, and this helps quite a bit in maintaining this model.
Ziniti
Absolutely ... and I would also like to see Richards excellent Call Profiler service become part of the core.
I'm definitely putting the profiler into 2.6 - there's just an open question of where it gets put. The question was asked on zope-coders, and got no response. I figure if no-one answers within a week of my original posting, I'll just check it in as a product. Richard
Absolutely ... and I would also like to see Richards excellent Call Profiler service become part of the core.
I'm definitely putting the profiler into 2.6 - there's just an open question of where it gets put. The question was asked on zope-coders, and got no response. I figure if no-one answers within a week of my original posting, I'll just check it in as a product.
Richard, FWIW, my own opinion is that it should not take the 'MonkeyPatch' approach. seb
On Fri, 1 Mar 2002 21:25, seb bacon wrote:
Absolutely ... and I would also like to see Richards excellent Call Profiler service become part of the core.
I'm definitely putting the profiler into 2.6 - there's just an open question of where it gets put. The question was asked on zope-coders, and got no response. I figure if no-one answers within a week of my original posting, I'll just check it in as a product.
FWIW, my own opinion is that it should not take the 'MonkeyPatch' approach.
That's my opinion too, but I have had no feedback from zope-coders... Richard
I am now locked out my CMF site, and I can't figure out how to get back in. None of my passwords work any more. Recently I had to reinstall the operating system and change my unix passwords. But now I can't get back into my Zope CMF sites in order to change the Zope passwords. I can log into the top level Zope manager, and look at the Zope tree. I just can't manage_edit my CMF site. When I click on a CMF site in the tree on the left, it does not edit the directory view in the main frame. Instead, it goes to the CMF login page in that frame. I try to log in, and it my password simply does not work. It keeps coming back to the cmf login page. I have had this problem of being forced to log into CMF before (even though I was already logged into the manager), but now my password doesn't work any more. The emergency user password doesn't work either. I tried hacking the python code in AuthEncoding pw_validate to return 1, and that still doesn't let me in. What is going on? Is there any way for me to disable the passwords and get into the cmf folder to see what's wrong? Thanks in advance! -Don
On Mon, 2002-03-04 at 01:36, Don Hopkins wrote:
I am now locked out my CMF site, and I can't figure out how to get back in. [...] I can log into the top level Zope manager, and look at the Zope tree. I just can't manage_edit my CMF site. When I click on a CMF site in the tree on the left, it does not edit the directory view in the main frame. Instead, it goes to the CMF login page in that frame.
Are you using Mozilla, Netscape >= 6.0 or another mozilla based browser? If so, you might be hitting a little odd-but-rfc-allowed mozilla behaviour where, even after asking the user for basic-auth credentials for a protected page, it doesn't send them for other protected pages in the same domain until the web server has responded with 'authorization-required' for every one of them. Because of that, when you're viewing a page in Zope that is protected by cookie based authentication, mozilla will not send basic-auth credentials at first, causing the cookie login form to kick in before zope issues an 'authorization-required' response. I hope this explanation is not too confusing :-) Anyway, try accessing your site with another browser to see if the problem persists. Konqueror, for instance, doesn't have this problem.
Hmm, after looking at the logs, it seems like it's not passing authorization info through the virtual host proxy rewrite rule. When I set Apache up to use a proxy rewrite rule, it asks me to log in instead of showing the folder manager, and the hits come from anonymous: [apache configuration for the name based virtual server DonHopkins.com] RewriteRule ^/(.*) http://www.DonHopkins.com:8080/Don/Hopkins/$1 [p] [zope log of clicking in logged in zope manager tree view on "/Don/Hopkins" cmf folder, which contains a virtual host monster] 216.175.91.16 - dhopkins [03/Mar/2002:23:33:01 -0700] "GET /Don/Hopkins/manage_workspace HTTP/1.1" 302 473 "http://www.donhopkins.com:8080/manage_menu" "Mozilla/4.0 (compatible; MSIE 5.5; W\ indows 98)" 216.218.252.133 - Anonymous [03/Mar/2002:23:33:01 -0700] "GET /Don/Hopkins/manage_main HTTP/1.0" 302 617 "http://www.donhopkins.com:8080/manage_menu" "Mozilla/4.0 (compatible; MSIE 5.5; Win\ dows 98)" 216.218.252.133 - Anonymous [03/Mar/2002:23:33:01 -0700] "GET /Don/Hopkins/login_form?came_from=http://www.DonHopkins.com/Don/Hopkins/mana ge_main&retry= HTTP/1.0" 200 4046 "http://www.donho\ pkins.com:8080/manage_menu" "Mozilla/4.0 (compatible; MSIE 5.5; Windows 98)" 216.218.252.133 - Anonymous [03/Mar/2002:23:33:01 -0700] "GET /Don/Hopkins/default_stylesheet HTTP/1.0" 200 5513 "http://www.donhopkins.com/Don/Hopkins/login_form?came_from=http%3A//www.Don \ Hopkins.com/Don/Hopkins/manage_main&retry=" "Mozilla/4.0 (compatible; MSIE 5.5; Windows 98)" Notice that the proxy requests come from 216.218.252.133 [the server's ip] as Anonymous. Is the auth info not getting passed? But when I change the apache rewrite rule to use external redirect instead of internal proxy, it works and shows me the folder manager, and the hits come from the right place, but the url in the browser is unnecessarily long and nasty. RewriteRule ^/(.*) http://www.DonHopkins.com:8080/Don/Hopkins/$1 [R] 216.175.91.16 - dhopkins [03/Mar/2002:23:35:03 -0700] "GET /Don/Hopkins/manage_workspace HTTP/1.1" 302 473 "http://www.donhopkins.com:8080/manage_menu" "Mozilla/4.0 (compatible; MSIE 5.5; W\ indows 98)" 216.175.91.16 - dhopkins [03/Mar/2002:23:35:04 -0700] "GET /Don/Hopkins/manage_main HTTP/1.1" 200 32200 "http://www.donhopkins.com:8080/manage_menu" "Mozilla/4.0 (compatible; MSIE 5.5; Wind\ ows 98)" I've been able to get it to work with external redirects instead of internal proxy, but the path displayed in the manager doubles it up: Don/Hopkins/Don/Hopkins ... So I put in the first rewrite rule with the ^/Don/Hopkins prefix, and that eliminated the weird double path prefix behavior I was getting. RewriteRule ^/Don/Hopkins/(.*) http://www.DonHopkins.com:8080/Don/Hopkins/$1 [R] Is there a better way for me to rewrite zope virtual host requests, so the authentication headers go through properly? I want to have apache handle https and logging, but let zope handle its own user authentication. And I want the url of the web browser to be short and sweet, not showing :8080 or the virtual host subdirectories. Should I be using fastcgi (or slow cgi) instead of internal proxy, and trying to trick it into passing the authorization in its own way? Thanks a lot for the help! There are so many options it's quite overwhelming... -Don ----- Original Message ----- From: "Leonardo Rochael Almeida" <leo@hiper.com.br> To: "Don Hopkins" <xardox@mindspring.com> Cc: "Zope Developers" <zope-dev@zope.org> Sent: Sunday, March 03, 2002 9:14 PM Subject: Re: [Zope-dev] I'm locked out of managing my CMF folders, becauseit asks me to log in and the password doesn't work.
On Mon, 2002-03-04 at 01:36, Don Hopkins wrote:
I am now locked out my CMF site, and I can't figure out how to get back in. [...] I can log into the top level Zope manager, and look at the Zope tree. I just can't manage_edit my CMF site. When I click on a CMF site in the tree on the left, it does not edit the directory view in the main frame. Instead, it goes to the CMF login page in that frame.
Are you using Mozilla, Netscape >= 6.0 or another mozilla based browser? If so, you might be hitting a little odd-but-rfc-allowed mozilla behaviour where, even after asking the user for basic-auth credentials for a protected page, it doesn't send them for other protected pages in the same domain until the web server has responded with 'authorization-required' for every one of them. Because of that, when you're viewing a page in Zope that is protected by cookie based authentication, mozilla will not send basic-auth credentials at first, causing the cookie login form to kick in before zope issues an 'authorization-required' response.
I hope this explanation is not too confusing :-)
Anyway, try accessing your site with another browser to see if the problem persists. Konqueror, for instance, doesn't have this problem.
Based on your description below, it seems your RewriteRule is not writen correctly. To make proper use of the VirtualHostMonster it should read (all in one line): RewriteRule ^(.*)$ http://localhost:8080/VirtualHostBase/http/www.DonHopkins.com/80/Don/Hopkins... [P,L] The 'L' in '[P,L]' is optional but makes sure this is the last rewrite rule to be applied. The 'localhost:8080' part could be anything as long as it finds the correct Zope. The 'VirtualHostBase/http/www.DonHopkins.com/80/' part (including the '/80' part is really important to make sure Zope knows which server it is pretending to be from and display the proper 'host part' of generated URL's AND AUTHENTICATION MECHANISMS (this is the part that could be biting you). The '/Don/Hopkins/' part before the VirtualHostRoot tells zope which folder it is actually serving and the 'VirtualHostRoot$1' part makes sure Zope knows it is pretending that 'Don/Hopkins' is the root of the URL space. You can think of it as meaning the /Don/Hopkins folder of the localhost:8080 Zope is mounted in the root folder of www.DonHopkins.com:80. Making sure Zope correctly knows where it's pretending to be from is very important in authentication/authorization matters, since browsers promptly discard any authentication request (and cookie requests, for that matter) for a site they are not visiting as a means to prevent cross site exploits. Other than that, Proxying thru apache should pass authentication headers normally, without any other effort on your part. Cheers, Leo On Mon, 2002-03-04 at 03:46, Don Hopkins wrote:
Hmm, after looking at the logs, it seems like it's not passing authorization info through the virtual host proxy rewrite rule. When I set Apache up to use a proxy rewrite rule, it asks me to log in instead of showing the folder manager, and the hits come from anonymous:
[apache configuration for the name based virtual server DonHopkins.com]
RewriteRule ^/(.*) http://www.DonHopkins.com:8080/Don/Hopkins/$1 [p]
[zope log of clicking in logged in zope manager tree view on "/Don/Hopkins" cmf folder, which contains a virtual host monster]
216.175.91.16 - dhopkins [03/Mar/2002:23:33:01 -0700] "GET /Don/Hopkins/manage_workspace HTTP/1.1" 302 473 "http://www.donhopkins.com:8080/manage_menu" "Mozilla/4.0 (compatible; MSIE 5.5; W\ indows 98)" 216.218.252.133 - Anonymous [03/Mar/2002:23:33:01 -0700] "GET /Don/Hopkins/manage_main HTTP/1.0" 302 617 "http://www.donhopkins.com:8080/manage_menu" "Mozilla/4.0 (compatible; MSIE 5.5; Win\ dows 98)" 216.218.252.133 - Anonymous [03/Mar/2002:23:33:01 -0700] "GET /Don/Hopkins/login_form?came_from=http://www.DonHopkins.com/Don/Hopkins/mana ge_main&retry= HTTP/1.0" 200 4046 "http://www.donho\ pkins.com:8080/manage_menu" "Mozilla/4.0 (compatible; MSIE 5.5; Windows 98)" 216.218.252.133 - Anonymous [03/Mar/2002:23:33:01 -0700] "GET /Don/Hopkins/default_stylesheet HTTP/1.0" 200 5513 "http://www.donhopkins.com/Don/Hopkins/login_form?came_from=http%3A//www.Don \ Hopkins.com/Don/Hopkins/manage_main&retry=" "Mozilla/4.0 (compatible; MSIE 5.5; Windows 98)"
Notice that the proxy requests come from 216.218.252.133 [the server's ip] as Anonymous. Is the auth info not getting passed?
But when I change the apache rewrite rule to use external redirect instead of internal proxy, it works and shows me the folder manager, and the hits come from the right place, but the url in the browser is unnecessarily long and nasty.
RewriteRule ^/(.*) http://www.DonHopkins.com:8080/Don/Hopkins/$1 [R]
216.175.91.16 - dhopkins [03/Mar/2002:23:35:03 -0700] "GET /Don/Hopkins/manage_workspace HTTP/1.1" 302 473 "http://www.donhopkins.com:8080/manage_menu" "Mozilla/4.0 (compatible; MSIE 5.5; W\ indows 98)" 216.175.91.16 - dhopkins [03/Mar/2002:23:35:04 -0700] "GET /Don/Hopkins/manage_main HTTP/1.1" 200 32200 "http://www.donhopkins.com:8080/manage_menu" "Mozilla/4.0 (compatible; MSIE 5.5; Wind\ ows 98)"
I've been able to get it to work with external redirects instead of internal proxy, but the path displayed in the manager doubles it up: Don/Hopkins/Don/Hopkins ... So I put in the first rewrite rule with the ^/Don/Hopkins prefix, and that eliminated the weird double path prefix behavior I was getting.
RewriteRule ^/Don/Hopkins/(.*) http://www.DonHopkins.com:8080/Don/Hopkins/$1 [R]
Is there a better way for me to rewrite zope virtual host requests, so the authentication headers go through properly? I want to have apache handle https and logging, but let zope handle its own user authentication. And I want the url of the web browser to be short and sweet, not showing :8080 or the virtual host subdirectories. Should I be using fastcgi (or slow cgi) instead of internal proxy, and trying to trick it into passing the authorization in its own way?
Thanks a lot for the help! There are so many options it's quite overwhelming...
-Don
----- Original Message ----- From: "Leonardo Rochael Almeida" <leo@hiper.com.br> To: "Don Hopkins" <xardox@mindspring.com> Cc: "Zope Developers" <zope-dev@zope.org> Sent: Sunday, March 03, 2002 9:14 PM Subject: Re: [Zope-dev] I'm locked out of managing my CMF folders, becauseit asks me to log in and the password doesn't work.
On Mon, 2002-03-04 at 01:36, Don Hopkins wrote:
I am now locked out my CMF site, and I can't figure out how to get back in. [...] I can log into the top level Zope manager, and look at the Zope tree. I just can't manage_edit my CMF site. When I click on a CMF site in the tree on the left, it does not edit the directory view in the main frame. Instead, it goes to the CMF login page in that frame.
Are you using Mozilla, Netscape >= 6.0 or another mozilla based browser? If so, you might be hitting a little odd-but-rfc-allowed mozilla behaviour where, even after asking the user for basic-auth credentials for a protected page, it doesn't send them for other protected pages in the same domain until the web server has responded with 'authorization-required' for every one of them. Because of that, when you're viewing a page in Zope that is protected by cookie based authentication, mozilla will not send basic-auth credentials at first, causing the cookie login form to kick in before zope issues an 'authorization-required' response.
I hope this explanation is not too confusing :-)
Anyway, try accessing your site with another browser to see if the problem persists. Konqueror, for instance, doesn't have this problem.
_______________________________________________ Zope-Dev maillist - Zope-Dev@zope.org http://lists.zope.org/mailman/listinfo/zope-dev ** No cross posts or HTML encoding! ** (Related lists - http://lists.zope.org/mailman/listinfo/zope-announce http://lists.zope.org/mailman/listinfo/zope )
-- Ideas don't stay in some minds very long because they don't like solitary confinement.
Don Hopkins writes:
... I have seen this often when there was an exception in the action box. In this case, I was able to analyse the problem in one of the following ways:
1. disable cookie authentication You will then get a browser login request. Abort it. And it will show what object could not be accessed. 2. Print the exception in "ZPublisher.Publish" near line 106 after the "tansactions_manager.abort()": import traceback; traceback.print_exc() But maybe, you password is indeed incorrect. Then become an emergency user and fix the problem (learn more about the emergency user on Zope.org). Dieter
Hi, John Ziniti schrieb:
Let's get a discussion
started to define 2.6
I'd like to see the ZSyncer Product, or a variant thereof, included in Zope by default. That is, I'd like "Synchronization", to a be a default property of Zope objects, so that objects/content can be pushed and pulled between two Zope installations.
I would use ZCVSFolder or somewhat likely that included instead of ZSyncer. The problem i see is that we have a development server and more than one production instances in our organisation. If a product comes out of its development state it is moved to the production server. If you use ZSyncer then all Objects "zsynced" over all instances have the same version, but if you use ZCVSFolder instead then you have a version-control mechanism of all objects on the different instances.
I venture a guess that the development/production model is common, and this helps quite a bit in maintaining this model.
Ziniti
_______________________________________________ Zope-Dev maillist - Zope-Dev@zope.org http://lists.zope.org/mailman/listinfo/zope-dev ** No cross posts or HTML encoding! ** (Related lists - http://lists.zope.org/mailman/listinfo/zope-announce http://lists.zope.org/mailman/listinfo/zope )
-- Andre Schubert EMail: Andre.Schubert@km3.de Tel: 03774 6625-78 km3 teledienst GmbH Fax: 03774 6625-79
I'd like to see the ZSyncer Product, or a variant thereof, included in Zope by default. That is, I'd like "Synchronization", to a be a default property of Zope objects, so that objects/content can be pushed and pulled between two Zope installations.
I would use ZCVSFolder or somewhat likely that included instead of ZSyncer.
I think the use cases driving synchronisation requirements are not yet sufficiently understood. Agreed, it's a feature that many people need, but I get the impression it would be premature to include ZSyncer as is in Zope without more detailed planning. On the other hand, there is an argument which says, ZSyncer is a good product used by a lot of people, let's put it in Zope and deal with any issues after the event. seb
seb bacon wrote:
I think the use cases driving synchronisation requirements are not yet sufficiently understood. Agreed, it's a feature that many people need, but I get the impression it would be premature to include ZSyncer as is in Zope without more detailed planning. On the other hand, there is an argument which says, ZSyncer is a good product used by a lot of people, let's put it in Zope and deal with any issues after the event.
Yeah... maybe we need a standard "Zope Expansion Kit" that is a cohesive package of products that aren't in the Zope core, but are officially sanctioned and versioned and released alongside versions of Zope. -- Steve Alexander
Steve Alexander wrote:
Yeah... maybe we need a standard "Zope Expansion Kit" that is a cohesive package of products that aren't in the Zope core, but are officially sanctioned and versioned and released alongside versions of Zope.
I thought that's what the /Products directory in the standard Zoep distributino was? ;-) cheers, Chris
I'd like to see the ZSyncer Product, or a variant thereof, included in Zope by default. That is, I'd like "Synchronization", to a be a default property of Zope objects, so that objects/content can be pushed and pulled between two Zope installations.
Thanks for the enthusiasm but its still too hacky to go into the core. Ideally I'd rather put in the time so this is more transparent in Zope 3. -- Andy McKay
+1, and I say that knowing that it means I have to help. I'm willing to write the docs for whoever works on the code. A gentle reminder on some of the posts in this thread. Please don't respond with "I'd really like <some good idea>." Respond with "I'm willing to do the work for <some good idea>." That's part of the point with Brian's note. You don't _have_ to do the code work. If you write up the docs for <some good idea>, you've likely done most of the work. --Paul Behrens Matt - Grand Rapids wrote:
Brian Lloyd wrote:
Let's get a discussion
started to define 2.6.
This isn't exciting by any means unless you're one of the people who package Zope up for distribution, or maybe you're one of the people who manage lots of little Zopes on one system; but I'd like to revive the "grand unified Zope installation and control" proposal that has been floated by many people (including me) in one form or another for some time. Wikiwise, this would wrap up http://dev.zope.org/Wikis/DevSite/Proposals/ZopeStartupProvisions and http://dev.zope.org/Wikis/DevSite/Proposals/InstallationAndConfiguration, at least.
To summarize, this would involve
- an expanded build program with an installation scheme that would allow multiple versions of Zope to be present on the same system
- making that installation 'secure by default'
- a registry of Zope installations and one of instances and their configuration settings
- a 'zopectl' program or similar that would be able to start and stop instances
- a 'zopeinstance' program or similar that would become the _recommended_ way of setting up Zope, by creating an INSTANCE_HOME
It would be nice if
- the same framework could apply to Zope 3, maybe taking care of that piece ahead of time
I'm more than willing to head this up, though I question how long we have before 2.6 to do so.
_______________________________________________ Zope-Dev maillist - Zope-Dev@zope.org http://lists.zope.org/mailman/listinfo/zope-dev ** No cross posts or HTML encoding! ** (Related lists - http://lists.zope.org/mailman/listinfo/zope-announce http://lists.zope.org/mailman/listinfo/zope )
I'd suggest starting with the combination of Evan's zopemake and zctl scripts. One thing that zopemake could be extended with is an autoconf-style "configure" that figures out where the appropriate version of Python is, which C compiler to use, etc. - C ----- Original Message ----- From: "Paul Everitt" <paul@zope.com> To: "Behrens Matt - Grand Rapids" <Matt.Behrens@Kohler.Com> Cc: <zope-dev@zope.org>; <zope-coders@zope.org> Sent: Friday, March 01, 2002 7:53 AM Subject: Re: [Zope-dev] Zope 2.6 planning - call for contributors!
+1, and I say that knowing that it means I have to help. I'm willing to write the docs for whoever works on the code.
A gentle reminder on some of the posts in this thread. Please don't respond with "I'd really like <some good idea>." Respond with "I'm willing to do the work for <some good idea>." That's part of the point with Brian's note.
You don't _have_ to do the code work. If you write up the docs for <some good idea>, you've likely done most of the work.
--Paul
Behrens Matt - Grand Rapids wrote:
Brian Lloyd wrote:
Let's get a discussion
started to define 2.6.
This isn't exciting by any means unless you're one of the people who package Zope up for distribution, or maybe you're one of the people who manage lots of little Zopes on one system; but I'd like to revive the "grand unified Zope installation and control" proposal that has been floated by many people (including me) in one form or another for some time. Wikiwise, this would wrap up http://dev.zope.org/Wikis/DevSite/Proposals/ZopeStartupProvisions and
http://dev.zope.org/Wikis/DevSite/Proposals/InstallationAndConfiguration,
at least.
To summarize, this would involve
- an expanded build program with an installation scheme that would allow multiple versions of Zope to be present on the same system
- making that installation 'secure by default'
- a registry of Zope installations and one of instances and their configuration settings
- a 'zopectl' program or similar that would be able to start and stop instances
- a 'zopeinstance' program or similar that would become the _recommended_ way of setting up Zope, by creating an INSTANCE_HOME
It would be nice if
- the same framework could apply to Zope 3, maybe taking care of that piece ahead of time
I'm more than willing to head this up, though I question how long we have before 2.6 to do so.
_______________________________________________ Zope-Dev maillist - Zope-Dev@zope.org http://lists.zope.org/mailman/listinfo/zope-dev ** No cross posts or HTML encoding! ** (Related lists - http://lists.zope.org/mailman/listinfo/zope-announce http://lists.zope.org/mailman/listinfo/zope )
_______________________________________________ Zope-Dev maillist - Zope-Dev@zope.org http://lists.zope.org/mailman/listinfo/zope-dev ** No cross posts or HTML encoding! ** (Related lists - http://lists.zope.org/mailman/listinfo/zope-announce http://lists.zope.org/mailman/listinfo/zope )
I would say, make SSL part of the standard z2.py, so you can turn on/off, specify address, etc. of https ports just as you do with http ports (and of course integrated with siteaccess2, etc.)
On Fri, 01 Mar 2002 09:48:08 -0500, marc lindahl <marc@bowery.com> wrote:
I would say, make SSL part of the standard z2.py, so you can turn on/off, specify address, etc. of https ports just as you do with http ports (and of course integrated with siteaccess2, etc.)
Ive never really understood the motivation for wanting https support direct in Zope.... ZServer isnt robust enough to be exposed to the raw internet without risk. Today (and perhaps for the forseeable future, because its not clear that Zope want to take on the responsibility of ensuring it is that robust) if you care about security Zope really needs to be run behind a front-end-proxy, and the two popular choices for proxying, Squid and Apache, are already well endowed for https support. Are there any common scenarios which need the protection given by https, but do not need the protection given by a front-end-proxy? Toby Dickenson tdickenson@geminidataloggers.com
On Fri, Mar 01, 2002 at 03:00:10PM +0000, Toby Dickenson wrote:
Are there any common scenarios which need the protection given by https, but do not need the protection given by a front-end-proxy?
Yes, running zope in intranet environments where the connection to a localhost proxy is not possible (like running it on NT for an intranet service, when you *can't* install an apache just for doing the ssl things) and you don't want to expose the uncrypted data stream to the environment. is that a valid requirement? greetings. christian -- Christian Theune - ct@gocept.com gocept gmbh & co.kg - schalaunische strasse 6 - 06366 koethen/anhalt tel.+49 3496 3099112 - fax.+49 3496 3099118 mob. - 0178 48 33 981 reduce(lambda x,y:x+y,[chr(ord(x)^42) for x in 'zS^BED\nX_FOY\x0b'])
Ive never really understood the motivation for wanting https support direct in Zope.... ZServer isnt robust enough to be exposed to the raw internet without risk. Today (and perhaps for the forseeable future, because its not clear that Zope want to take on the responsibility of
ZServer may not be as robust as Apache or Squid but many, many people expose raw to the internet without problem (and compare that to other products from Redmond). The main motivation is that many people use Zope as a single solution, by installing Zope they can get everything they need to get a web site. By having the ZMI only available through HTTPS by default for example will definitely increase security and make Zope a better all in package. It obviously won't help the enterprise customer. Is there actually a huge amount of risk in this? The patches are there and seem to work ok most of the work seems to have been done already... -- Andy McKay
On 3/1/02 7:30 AM, "Chris McDonough" <chrism@zope.com> wrote:
I'd suggest starting with the combination of Evan's zopemake and zctl scripts. One thing that zopemake could be extended with is an autoconf-style "configure" that figures out where the appropriate version of Python is, which C compiler to use, etc.
Can't the distutils framework be used? It's got all of the utilities to deal with various platform issues already. -- Jeffrey P Shell www.cuemedia.com
On Fri, Mar 01, 2002 at 07:53:35AM -0500, Paul Everitt wrote:
A gentle reminder on some of the posts in this thread. Please don't respond with "I'd really like <some good idea>." Respond with "I'm willing to do the work for <some good idea>." That's part of the point with Brian's note.
You don't _have_ to do the code work. If you write up the docs for <some good idea>, you've likely done most of the work.
--Paul
Hmm ... there was a post on the zope users list concerning logouts from ZMI, and i think that is a thing that could be *really* better (cookie login with a nice form or so) and i would like to contribute but am not sure if i am able to do the implementation work. -- Christian Theune - ct@gocept.com gocept gmbh & co.kg - schalaunische strasse 6 - 06366 koethen/anhalt tel.+49 3496 3099112 - fax.+49 3496 3099118 mob. - 0178 48 33 981 reduce(lambda x,y:x+y,[chr(ord(x)^42) for x in 'zS^BED\nX_FOY\x0b'])
Christian Theune wrote:
On Fri, Mar 01, 2002 at 07:53:35AM -0500, Paul Everitt wrote:
A gentle reminder on some of the posts in this thread. Please don't respond with "I'd really like <some good idea>." Respond with "I'm willing to do the work for <some good idea>." That's part of the point with Brian's note.
You don't _have_ to do the code work. If you write up the docs for <some good idea>, you've likely done most of the work.
--Paul
Hmm ... there was a post on the zope users list concerning logouts from ZMI, and i think that is a thing that could be *really* better (cookie login with a nice form or so) and i would like to contribute but am not sure if i am able to do the implementation work.
The implementation is done--use the cookie crumbler product. The only question is whether the cookie crumbler (under a different name ;-) ) out to be part of the main Zope 2 distribution. Shane
+1 for cookie crumbler Ah right, i didn't look at that before, thats what i thought of. And with the mention of the "Zope Expansion Kit" i think this really should go into core (or somewhere very next to id), including an option to be created with a standard user folder automatically. On Sat, Mar 02, 2002 at 05:18:38PM -0500, Shane Hathaway wrote:
Christian Theune wrote:
The implementation is done--use the cookie crumbler product. The only question is whether the cookie crumbler (under a different name ;-) ) out to be part of the main Zope 2 distribution.
Shane
-- Christian Theune - ct@gocept.com gocept gmbh & co.kg - schalaunische strasse 6 - 06366 koethen/anhalt tel.+49 3496 3099112 - fax.+49 3496 3099118 mob. - 0178 48 33 981 reduce(lambda x,y:x+y,[chr(ord(x)^42) for x in 'zS^BED\nX_FOY\x0b'])
<See below>
This isn't exciting by any means unless you're one of the people who package Zope up for distribution, or maybe you're one of the people who manage lots of little Zopes on one system; but I'd like to revive the "grand unified Zope installation and control" proposal that has been floated by many people (including me) in one form or another for some time. Wikiwise, this would wrap up http://dev.zope.org/Wikis/DevSite/Proposals/ZopeStartupProvisions and http://dev.zope.org/Wikis/DevSite/Proposals/InstallationAndConfiguration, at least.
To summarize, this would involve
- an expanded build program with an installation scheme that would allow multiple versions of Zope to be present on the same system
- making that installation 'secure by default'
- a registry of Zope installations and one of instances and their configuration settings
- a 'zopectl' program or similar that would be able to start and stop instances
- a 'zopeinstance' program or similar that would become the _recommended_ way of setting up Zope, by creating an INSTANCE_HOME
It would be nice if
- the same framework could apply to Zope 3, maybe taking care of that piece ahead of time
I'm more than willing to head this up, though I question how long we have before 2.6 to do so.
Matt, I read 'InstallationAndConfiguration' page, and it made me wonder exactly how that aspect would work. Assuming some sort of distutils based setup [1], would something like the following be the desired approach? [2] #Installs zope in it's current simplest configuration. Equivalent of "python wo_pcgi.py" $ python setup.py #Equivalent of w_pcgi.py $ python setup.py pcgi #Creates SOFTWARE_HOME or ZOPE_HOME (I'm not sure if there's a difference, or which one I would mean) #I guess it would just do this in the current directory $ python setup.py zope_home #Creates an INSTANCE_HOME by prompting the user for a path (if not already given) to install to #Add's instance to some sort of registry [3] $ python setup.py instance [/path/to/instance] I don't really know what else to think about. I haven't ever used ZEO, so I don't have the first clue about that side of things. So, as you may be able to tell, I'm a little hazy on many (most?) aspects of this install/configure idea, but I'm quite keen on it. I've tried to post something that is at least mildly useful more in the hope that it may keep the proposal rolling rather than as a 'me too' or '+1' (as Paul E already told us off for ;-) ). I guess, at the end of the day, I can't really get around the fact it's a 'me too' without any coding commitment [4]. Do you have any concrete ideas for this? cheers tim [1] I've never used distutils before [2] Feel more than free to knock down and tear up anything I've said here :-) [3] Would this 'registry' be some sort of flat text file, or perhaps a zodb store, or...? [4] Apart from the fact that I've never written anything other than web-code (so my python probably isn't up to it), I can't commit time at the moment due to my degree having to take priority :-(. If there were any simply odd-jobs, testing, something I couldn't ballsup, I'd like to help if I can.
Behrens Matt - Grand Rapids wrote:
This isn't exciting by any means unless you're one of the people who package Zope up for distribution, or maybe you're one of the people who manage lots of little Zopes on one system; but I'd like to revive the "grand unified Zope installation and control" proposal that has been floated by many people (including me) in one form or another for some time....
FYI, everyone who's following this: I have hijacked http://dev.zope.org/Wikis/DevSite/Proposals/InstallationAndConfiguration for this purpose. :-)
FYI, everyone who's following this: I have hijacked http://dev.zope.org/Wikis/DevSite/Proposals/InstallationAndConfiguration for this purpose. :-)
Awesome. Exactly how I'd like the default zope install to be structured. :-) Adam
I wholeheartedly agree that 2.6 needs to be significantly a community effort. While I know that many people are engaged in the Zope 3 effort, we also need to get some people engaged on defining and producing 2.6 in the interim. There is not much on the plan right now, so the possibilities are relatively wide open :) Let's get a discussion started to define 2.6.
OK. The things I'd like to see included is the Enhanced Virtual Host Monster that iMeme did, or maybe even the Virtual Host Folder that is being developed now (I haven't tried that one out, but it looks good on paper). I'd also like the functionality provided by the local roles "blacklist" enhancement we did ( http://www.zope.org/Members/regebro/LRBlacklist ). And of course, the bugs that we have reported fixed. :-) Can't think of anything more right now.
I'd like to see ZSQL methods altered so that bind variables could be used in SQL queries. This would improve SQL operations for at least Oracle, which is the one db I know of that uses bind variables to speed it's querie-management. This is filed as a bugreport in the collector previously but has been turned down/put on hold until a newer version of zope. What is the time schedule of Zope 2.6? I ask because I need to start using 2.5 before I suggest more improvements to zope :-) /dario - -------------------------------------------------------------------- Dario Lopez-Kästen Systems Developer Chalmers Univ. of Technology dario@ita.chalmers.se ICQ will yield no hits IT Systems & Services
On Fri, 2002-03-01 at 04:16, Dario Lopez-Kästen wrote:
I'd like to see ZSQL methods altered so that bind variables could be used in SQL queries.
+1, even though MySQL doesn't bind variables (yet, apparently in the works for 4.0). Another thing I would like (and could probably write) is an additional skip_rows parameter (default 0) to Z SQL Methods to complement max_rows. Since this needs to be passed to the DA's query(), and no DA's currently accept this parameter, one of these things need to happen: 1) Trap the TypeError that results from calling a query() that doesn't recognize skip_rows (used as a keyword parameter). In that case, fall back to query(q, max_rows+skip_rows) and return result[skip_rows:] in the Z SQL Method. 2) Add a new DA method query_range(q, skip_rows, max_rows); if AttributeError, fall back like #1. 3) Simply require all DA's to implement it in query() as a keyword parameter to maintain backwards compatibility with older Zopes. Adding this would make Z SQL Methods work a bit more like <dtml-in ... start=skip_rows size=max_rows>. There may be issues with result caching, but I don't think it'll be too much of a problem. -- Andy Dustman PGP: 0x930B8AB6 @ .net http://dustman.net/andy You can have my keys when you pry them from my dead, cold neurons.
If there is any interest in spiffing the Virtual Host Folder up for inclusion in Zope 2.6, I'll do the work. It requires Ordered Folder 0.5.1, and needs just a bit more spiffing. Gary -----Original Message----- From: zope-dev-admin@zope.org [mailto:zope-dev-admin@zope.org]On Behalf Of Lennart Regebro Sent: Friday, March 01, 2002 2:54 AM To: Brian Lloyd; zope-dev@zope.org; zope-coders@zope.org Subject: Re: [Zope-dev] Zope 2.6 planning - call for contributors!
I wholeheartedly agree that 2.6 needs to be significantly a community effort. While I know that many people are engaged in the Zope 3 effort, we also need to get some people engaged on defining and producing 2.6 in the interim. There is not much on the plan right now, so the possibilities are relatively wide open :) Let's get a discussion started to define 2.6.
OK. The things I'd like to see included is the Enhanced Virtual Host Monster that iMeme did, or maybe even the Virtual Host Folder that is being developed now (I haven't tried that one out, but it looks good on paper). I'd also like the functionality provided by the local roles "blacklist" enhancement we did ( http://www.zope.org/Members/regebro/LRBlacklist ). And of course, the bugs that we have reported fixed. :-) Can't think of anything more right now. _______________________________________________ Zope-Dev maillist - Zope-Dev@zope.org http://lists.zope.org/mailman/listinfo/zope-dev ** No cross posts or HTML encoding! ** (Related lists - http://lists.zope.org/mailman/listinfo/zope-announce http://lists.zope.org/mailman/listinfo/zope )
Gary Poster wrote:
If there is any interest in spiffing the Virtual Host Folder up for inclusion in Zope 2.6, I'll do the work. It requires Ordered Folder 0.5.1, and needs just a bit more spiffing.
Why does it require ordered folder? What does that have to do with virtual hosting? cheers, Chris
From: "Chris Withers" <chrisw@nipltd.com>
Gary Poster wrote:
If there is any interest in spiffing the Virtual Host Folder up for inclusion in Zope 2.6, I'll do the work. It requires Ordered Folder
0.5.1,
and needs just a bit more spiffing.
Why does it require ordered folder? What does that have to do with virtual hosting?
Well--the best answer is to suggest you look at it. The next best answer is to say that the ability to order virtual host rules (as in, for instance, the Apache httpd.conf virtual host section) is very convenient. And the full-blown I-can't-help-but-give-an-example is this. The Virtual Host Folder (hereafter VHF) holds implementations of IHost. Each IHost implementation is responsible for mappings to one virtual Zope root. (There is an exception but I'm not going in to it here). So if, for instance, your site is mysite.com and you wanted to have all subdomains (www.mysite.com, anything.mysite.com, etc.) go to a given zope root then you create one host object that uses a regular expression like ^(.*\.|)mysite.com$ and points to your desired virtual zope root. Then, if you in fact want certain subdomains at mysite.com to have a different Zope root, create a new host pointing to your alternate root and use a simple match like special.mysite.com Move this host higher than the catch-all host in the VHF using the OrderedFolder controls and you are set to go. While I did in fact roll my own ordered folder-like mix-in for this purpose initially (so I would have no dependencies), using OrderedFolder allows developers to create their own personalized implementations of the IHost interface (somebody wanted to hook their DNS and email system to their hosts, for instance, and this is not easily generalized) and just drop the new implementations (or subclasses) into the Products folder without making any changes to the VirtualHostFolder distribution itself. And by the way: TO DO list on the VHF before it would be any kind of core candidate: 1) TESTS 2) Some kind of fish bowl to more rigorously plan an API--not convinced I can get critical mass of interest behind it to make it worthwhile though... maybe I'll get around to it and we'll see. 3) OrderedFolder has some ZBabel-related problems on one of my installations; not sure of the cause/fix yet. Gary
On Thursday 28 February 2002 3:00 pm, Brian Lloyd wrote:
Paul sent a note to zope-coders some time back fishing for some feedback regarding planning for a Zope 2.6 (excepted):
I propose that planning for a 2.6 focus on the following thoughts: I wholeheartedly agree that 2.6 needs to be significantly a community effort.
The 2.6 timeframe should allow me to find some time to integrate my Unicode support (in ZPublisher, ZServer, and DTML), if there is agreement that this is the right thing to do.
Toby Dickenson wrote:
The 2.6 timeframe should allow me to find some time to integrate my Unicode support (in ZPublisher, ZServer, and DTML), if there is agreement that this is the right thing to do.
I'd +1 that in a big way :-) Especially combined with landing all Andreas' cool ZCatalog unicode work... cheers, Chris
If I had to make a wild guess, I would say 2.6 might land in the May-June timeframe. I would anticipate a drive for finalization in May, with the release cycle in June. That's not official, just a good guess based on our historical release patterns. On Friday, March 1, 2002, at 06:56 AM, Chris Withers wrote:
Toby Dickenson wrote:
The 2.6 timeframe should allow me to find some time to integrate my Unicode support (in ZPublisher, ZServer, and DTML), if there is agreement that this is the right thing to do.
I'd +1 that in a big way :-)
Especially combined with landing all Andreas' cool ZCatalog unicode work...
cheers,
Chris
_______________________________________________ Zope-Coders mailing list Zope-Coders@zope.org http://lists.zope.org/mailman/listinfo/zope-coders
----- Original Message ----- From: "Brian Lloyd" <brian@zope.com> To: <zope-dev@zope.org>; <zope-coders@zope.org> Sent: Thursday, February 28, 2002 10:00 Subject: [Zope-Coders] Zope 2.6 planning - call for contributors!
I wholeheartedly agree that 2.6 needs to be significantly a community effort. While I know that many people are engaged in the Zope 3 effort, we also need to get some people engaged on defining and producing 2.6 in the interim. There is not much on the plan right now, so the possibilities are relatively wide open :) Let's get a discussion started to define 2.6.
My TextIndexNG could be a powerful contribution to Zope 2.6 ;-) Andreas
Hi! What I'd expect from Zope 2.6 depends a bit on when Zope 3 will be available. If we are talking about a couple of months, I'd prefer only having bug fixes in 2.5.x (and no 2.6 at all). If we are talking about more than half a year, or even a year from now, things look different. The problem is that all time that is invested into Zope 2.6 will be lost for Zope 3 development, but on the other hand I can't imagine how I would convince a customer to fund Zope 3 development if the results will not be useable soon. So what we actually need is a Zope 2.5.x to 3 migration path and plan that justifies investments in either Zope 2.6 or Zope 3. For that, we'll have to answer a few questions, like - Are there any components in the Zope 3 development cycle that can be backported to the 2 series? - Can we build stuff into 2.6 that makes people start thinking the Zope 3 way? - ... I don't want Zope to end up like ArsDigita's ACS. They had a perfectly working 3 series that had all the features you'd expect, but was butt ugly in terms of the actual implementation. Then they started from scratch (like Zope is doing now) and built ACS 4, which was well-designed, but buggy as hell and had only core functionality. The plugins had not been ported yet. Then they started from scratch again and ported to Java (which Zope will not do I guess). Currently there are 500 or so freely available Zope add-ons on zope.org, which will most probably not work on Zope 3, at least not with the "3X" series. And there are even more non-free Zope products people have built on the 2 platform. I have the feeling that many of the add-ons will not be needed for Zope 3 because Zope 3 will do better out-of-the-box. But for many others there must be a migration path. Let's take the database adapters. If Zope 3 does not support the major databases from the beginning, it might not get the momentum it needs. Slightly off topic, I think what Zope (2 AND 3) need really urgently is another layer on top that delivers what the CMF (IMHO) did promise but not deliver to the extent I had expected: A solid foundation for Content Management Systems. This mainly means that we need an excellent DESIGN (i.e. API) on that level. Implementations may vary, but the interfaces should be as stable as possible. Some suggestions for that kind of design (which seems to be much easier to accomplish on the Zope 3 platform, which seems to have implemented some of that stuff already): - Storage should be completely separate from the data model. It should be possible to design a content class and then store it either in ZODB, the file system, an RDBMS or an LDAP server or whatever. - All the basic API (like "store", "delete", "edit", ...) must be free of HTTP specifics, so that I can modifiy content either over a web frontend or over WebDAV, FTP, ... - and even via a "fat client" application like a wxPython application. Currently that one is a mess. - On the HTTP end, skinning/chroming should work with ALL components. Imagine being able to create a design that works with Squishdot, CMF, and your own stuff out of the box, without having to change the implementation of the tools that are skinned. - The security API should really work. Currently there seem to be some really bad flaws with regard to checking for roles (and local roles). (I can go into details with that if somebody asks me and I find the time). - There should be a more rigid development model and tools to implement it. Currently it is hard for Zope developers to know where the business logic has to go. DTML really invited people to mix stuff, and ZPT seems a bit complicated to handle in terms of the actual syntax used. It isn't too readable either. Joachim
On Wednesday 06 March 2002 04:58 am, Joachim Werner wrote:
Hi!
What I'd expect from Zope 2.6 depends a bit on when Zope 3 will be available.
If we are talking about a couple of months, I'd prefer only having bug fixes in 2.5.x (and no 2.6 at all). If we are talking about more than half a year, or even a year from now, things look different.
The problem is that all time that is invested into Zope 2.6 will be lost for Zope 3 development, but on the other hand I can't imagine how I would convince a customer to fund Zope 3 development if the results will not be useable soon.
So what we actually need is a Zope 2.5.x to 3 migration path and plan that justifies investments in either Zope 2.6 or Zope 3. For that, we'll have to answer a few questions, like
- Are there any components in the Zope 3 development cycle that can be backported to the 2 series?
personally i would be interested in a backport of the component architecture, but i think that focusing development efforts on the zope3 core, is a more useful allocation of resources. zope3 will be ready faster the more people are willing to work on it. i know that i've been guilty of having not worked on it, since i have need to finish developing projects now on zope2 before i get to work on it. that said, i really like some of the proposals on the table for 2.6, but i just don't think that backporting zope3 to zope2 is a good use of people's time.
- Can we build stuff into 2.6 that makes people start thinking the Zope 3 way? - ...
in this regard the component architecture would make the most sense... but again it would be fairly much a developer resource, and without the components, services, and utilities themselves it would just be lookup and structure to applications. most of whats in zope3 currently is architecture.
I don't want Zope to end up like ArsDigita's ACS. They had a perfectly working 3 series that had all the features you'd expect, but was butt ugly in terms of the actual implementation. Then they started from scratch (like Zope is doing now) and built ACS 4, which was well-designed, but buggy as hell and had only core functionality. The plugins had not been ported yet. Then they started from scratch again and ported to Java (which Zope will not do I guess).
regarding the acs4 there were many plugins (dude, packages is the preferred nomenclature ;) ported to the acs4 architecture. in fact there are more of them then there were for the 3x platform (partly in due to improved modularity). that platform still lives on and thrives today in the form of the openacs. and includes some services and functionality in the core that i hope zope3 will bring to zope land (package management, workflow, calendaring/events, etc...) the move to java and the fall of arsdigita came as direct result of tasting too much of that poisoned apple, known as venture capital.
Currently there are 500 or so freely available Zope add-ons on zope.org, which will most probably not work on Zope 3, at least not with the "3X" series. And there are even more non-free Zope products people have built on the 2 platform.
I have the feeling that many of the add-ons will not be needed for Zope 3 because Zope 3 will do better out-of-the-box. But for many others there must be a migration path.
i don't know how much discussion there has been on this, but its something worth discussing in more detail, namely the use of the ZopeLegacy system for zope2 products. when things are a little more settled down for zope3, an excellent piece of documentation would be a product porting guide.
Let's take the database adapters. If Zope 3 does not support the major databases from the beginning, it might not get the momentum it needs.
completely apriori, i think these will be a fairly easy thing to port ;) .
Slightly off topic, I think what Zope (2 AND 3) need really urgently is another layer on top that delivers what the CMF (IMHO) did promise but not deliver to the extent I had expected: A solid foundation for Content Management Systems.
just curious, what do you see as the problems with the cmf? <snip good stuff and removed cross-posting> cheers kapil
On Wed, 6 Mar 2002, Joachim Werner wrote:
- All the basic API (like "store", "delete", "edit", ...) must be free of HTTP specifics, so that I can modifiy content either over a web frontend or over WebDAV, FTP, ... - and even via a "fat client" application like a wxPython application. Currently that one is a mess.
+5 Jerome
At 02:10 PM 3/6/2002 +0100, Jerome Alet wrote:
On Wed, 6 Mar 2002, Joachim Werner wrote:
- All the basic API (like "store", "delete", "edit", ...) must be free of HTTP specifics, so that I can modifiy content either over a web frontend or over WebDAV, FTP, ... - and even via a "fat client" application like a wxPython application. Currently that one is a mess.
+5
Use and work with Zope 3. :-) PS: Please remember not to cross-post. Regards, Stephan -- Stephan Richter CBU - Physics and Chemistry Student Web2k - Web Design/Development & Technical Project Management
Hi: I would like to propose my "Paste Reference"/symlink hack for inclusion into Zope 2.6 C U! -- Mario Valente
I would like to second this idea. Adrian... -- The difficulty of tactical maneuvering consists in turning the devious into the direct, and misfortune into gain. - Sun Tzu ----- Original Message ----- From: "Mario Valente" <mvalente@ruido-visual.pt> To: <zope-dev@zope.org>>; <zope-coders@zope.org> Sent: Wednesday, March 06, 2002 3:47 PM Subject: Re: [Zope-dev] Zope 2.6 planning - call for contributors!
Hi:
I would like to propose my "Paste Reference"/symlink hack for inclusion into Zope 2.6
C U!
-- Mario Valente
_______________________________________________ Zope-Dev maillist - Zope-Dev@zope.org http://lists.zope.org/mailman/listinfo/zope-dev ** No cross posts or HTML encoding! ** (Related lists - http://lists.zope.org/mailman/listinfo/zope-announce http://lists.zope.org/mailman/listinfo/zope )
Joachim Werner writes:
- Storage should be completely separate from the data model. It should be possible to design a content class and then store it either in ZODB, the file system, an RDBMS or an LDAP server or whatever. Isn't that already possible (implement your alternative storage that put pickles into your whatever store)
- All the basic API (like "store", "delete", "edit", ...) must be free of HTTP specifics, so that I can modifiy content either over a web frontend or over WebDAV, FTP, ... - and even via a "fat client" application like a wxPython application. Currently that one is a mess. I think, I do this already (with help of a PUT_factory).
... - The security API should really work. Currently there seem to be some really bad flaws with regard to checking for roles (and local roles). (I can go into details with that if somebody asks me and I find the time). Please do...
- There should be a more rigid development model and tools to implement it. I like freedom... Currently it is hard for Zope developers to know where the business logic has to go. I never felt this as a problem... DTML really invited people to mix stuff, and ZPT seems a bit complicated to handle in terms of the actual syntax used. It isn't too readable either. Hm? I was able to explain the ZPT syntax to a newbie in about 1/4 of an hour...
The essential parts: 5 tal attributes (there are more, but rarely used) 4 metal attributes 5 kinds of expressions Really not difficult, less difficult than DTML. Dieter
----- Original Message ----- From: "Brian Lloyd" <brian@zope.com> To: <zope-dev@zope.org>; <zope-coders@zope.org> Sent: Thursday, February 28, 2002 10:00 Subject: [Zope-Coders] Zope 2.6 planning - call for contributors!
http://dev.zope.org/Resources/zope_260_plan.html
I wholeheartedly agree that 2.6 needs to be significantly a community effort. While I know that many people are engaged in the Zope 3 effort, we also need to get some people engaged on defining and producing 2.6 in the interim. There is not much on the plan right now, so the possibilities are relatively wide open :) Let's get a discussion started to define 2.6.
I suggest to add support to retrieve the document source for dedicated WebDAV clients (identified by their useragent string) on the standard HTTP port. Some work on this issue has been done in the past and could be added for 2.6. Andreas
Are you planning to up the python version to 2.2? Because in that case I'd be happy to put in Authentication support in MailHost. Smtplib.py in 2.1.2 doesn't support authentication.
Lennart Regebro wrote:
Are you planning to up the python version to 2.2? Because in that case I'd be happy to put in Authentication support in MailHost. Smtplib.py in 2.1.2 doesn't support authentication.
I don't know how people are currently using Zope with 2.2. There is probably some really nasty problem being masked by the platform (seems to be Linux). See http://collector.zope.org/Zope/200 -- there's at least problems on Solaris and OpenBSD, probably others.
In my own casual experimentation, Zope worked okay by itself on Mandrake Linux 8.2b4 with Python 2.2, but ZEO refused to work. For what it's worth. Gary ----- Original Message ----- From: "Matt Behrens" <matt.behrens@kohler.com>
Lennart Regebro wrote:
Are you planning to up the python version to 2.2? Because in that case I'd be happy to put in Authentication support in MailHost. Smtplib.py in 2.1.2 doesn't support authentication.
I don't know how people are currently using Zope with 2.2. There is probably some really nasty problem being masked by the platform (seems to be Linux). See http://collector.zope.org/Zope/200 -- there's at least problems on Solaris and OpenBSD, probably others.
participants (40)
-
Adam Manock -
Adrian Hungate -
Andre Schubert -
Andreas Jung -
Andreas Jung -
Andy Dustman -
Andy McKay -
Behrens Matt - Grand Rapids -
Brad Clements -
Brian Lloyd -
Casey Duncan -
Chris McDonough -
Chris Withers -
Christian Theune -
Dario Lopez-Kästen -
Dieter Maurer -
Don Hopkins -
Eric Roby -
Gary Poster -
Jeffrey P Shell -
Jerome Alet -
Joachim Werner -
John Ziniti -
kapil thangavelu -
Lennart Regebro -
Lennart Regebro -
Leonardo Rochael Almeida -
marc lindahl -
Mario Valente -
Matt Behrens -
Matthew T. Kromer -
Paul Everitt -
Richard Jones -
seb bacon -
Shane Hathaway -
Stephan Richter -
Steve Alexander -
Tim Hicks -
Toby Dickenson -
Toby Dickenson