December 1998 - Zope

- ANNOUNCE: Zope 1.9 Final released
by Amos Latteier 23 Dec '98

23 Dec '98

Fellow Zopistas: Just in time for the holidays, we're happy to announce Zope 1.9 final. http://www.zope.org/Download/ This is mostly a bug fix release. We've also updated the MySQL Database Adapter to use the MySQL module in place of the old mySQL module. Use our Tabula-based bugs and features database to submit bugs. http://www.zope.org/Collector/ And most importantly, have fun! -Amos -- Amos Latteier mailto:amos@digicool.com Digital Creations http://www.digicool.com P.S. Here's the changelog: Zope 1.9.0 New Features * Zope now provides better information about products. If a Product has a README.txt, then it will be readable through the product management interface. * If a product encounters an error during startup, the traceback is available through the product management interface. * Cache statistics are displayed in minutes, rather than seconds. * Products can define the attribute __module_aliases__ in their __init__ modules that specifies aliases for modules used in the product. This is useful if module names have changed and the database contains pickles referring to the old modules. The attribute should be a sequence of tuples. Each tuple has a (dotted) module name and a module. Bugs Fixed * Development on ZopeHTTPServer has been frozen in anticipation of the switch to Medusa. Therefor ZopeHTTPServer still does support PUT. Sorry. * ZPublisher.Client used the obsolete rand module. * Errors occurring during product import caused Zope to fail to come up. * Product version information was not displayed correctly. *ZGadflyDA cached table meta-data which lead to misleading information when browsing databases.

1 0

- Can I get the real URL info into the Zope server?
by skip＠calendar.com 22 Dec '98

22 Dec '98

I'm using Andreas' technique of running ZopeHTTPServer behind an Apache server and using mod_rewrite and mod_proxy to pump requests to it. Unfortunately, within the Zope environment I lose all indication of what the original URL was. For instance, if you access http://concerts.dolphin.calendar.com/z/check_request you'll see a number of URL- and host-related fields: BASE0: 'http://127.0.0.2:8043' HTTP_HOST: '127.0.0.2:8043' PARENT_URL: 'http://127.0.0.2:8043' SERVER_NAME: 'concerts.dolphin.calendar.com' URL: 'http://127.0.0.2:8043/check_request' Only SERVER_NAME indicates the front-end server, and that only because I started the server with SERVER_NAME=concerts.dolphin.calendar.com. Is there a way to pump the real URL information across the boundary between the proxy and Zope servers? In output pages I can't simply generate server-relative URLs, because the PATH_INFO (or SCRIPT_NAME) is different between the two servers. The Zope server doesn't know that /z/... in the proxy server maps to /... in the Zope server. Ideally, this would be something I'd pass in from the rewrite rule, since the mapping between proxy and Zope servers is established on a per rewrite rule basis. Apache's mod_rewrite still has a bit more black magic than I comprehend yet. Skip

2 1

RE: [Zope] - Virtual servers
by Paul Everitt 22 Dec '98

22 Dec '98

Chris wrote: > I would say also working correctly under NSAPI and IIS would be pretty > high priorities. I've been trying to get a customer to use Zope to > build something that would fit perfectly, but they have all kinds of > "company standards" that are used to tell you "no" in so many > ways that > I believe OED would be interested in new definitions. Sorry, instead of Apache I should have said "leading web servers". Of course that opens me up to Mac criticism :^) --Paul

1 0

- installing 1.9b3 with pre-configured Apache for Linux
by Joe Van Andel 22 Dec '98

22 Dec '98

Reply-To: vanandel(a)ucar.edu Cc: Fcc: -------- A minor correction to the preconfigured Apache/README. 6) Logged on as *yourself*, using the same UID that you used to install Zope, start Apache: [paul] % ./start (otherwise Apache starts, but can't start Zope). I'm impressed with how quickly I could install the Linux binary version! Joe VanAndel Internet: vanandel(a)ucar.edu National Center for Web: http://www.atd.ucar.edu/~vanandel/home.html Atmospheric Research

1 0

RE: [Zope] - Virtual servers
by Paul Everitt 22 Dec '98

22 Dec '98

Andreas wrote: > Than there is marketing and public relations. (``So you are using what > Webserver? Zope? Why are you not using Apache?'') This is an important issue to me. I've always viewed sitting behind Apache as the #1 focus of our attention. But... > Additionally, you have NOT given one good argument why to do away with > Apache. Anything you've mentioned is also available with > apache/pcgi :) There doesn't have to be _one_ way. There will be _many_ ways. If you did a poll on this list and asked how many people have done production sites with Medusa, you might be surprised at the answer. Thus, I look at Dave's points as _a_ perfectly valid view. --Paul Paul Everitt Digital Creations paul(a)digicool.com 540.371.6909

2 1

- keys() and items() for Request objects... and maybe a bug w/ PATH_INFO
by skip＠calendar.com 22 Dec '98

22 Dec '98

I offer the following implementations of keys() and items() for Z Publisher Request objects: def keys(self): keys = {} for key in self.environ.keys(): if isCGI_NAME(key) or key[:5] == 'HTTP_': keys[key] = 1 keys.update(self.other) lasturl = "" for n in "0123456789": key = "URL%s"%n try: if lasturl != self[key]: keys[key] = 1 lasturl = self[key] else: break except KeyError: pass for n in "0123456789": key = "BASE%s"%n try: if lasturl != self[key]: keys[key] = 1 lasturl = self[key] else: break except KeyError: pass return keys.keys() def items(self): result = [] for k in self.keys(): result.append((k, self[k])) return result They probably aren't useful in general settings, but are useful for people like myself who aren't familiar with what's in a Request and want to explore a bit. Also, when using this, I discovered what may be a bug. It's at least a feature. I am using ZopeHTTPServer. When I examine the REQUEST argument of a URL, among other things I find PATH_INFO: '/check_request' PATH_TRANSLATED: '/home/dolphin/skip/src/Zope-1.9.0b3-src/check_request' SCRIPT_NAME: '/' URL: 'http://127.0.0.2:8043/check_request' I suppose in a Z Publisher context the distinction between elements of PATH_INFO and elements of SCRIPT_NAME is fuzzy, but it seems somehow more intuitive to me that I should see PATH_INFO: '' PATH_TRANSLATED: '/home/dolphin/skip/src/Zope-1.9.0b3-src' SCRIPT_NAME: '/check_request' URL: 'http://127.0.0.2:8043/check_request' instead. In a CGI context you always have a non-empty SCRIPT_NAME. You might or might not have a non-empty PATH_INFO. Given that SCRIPT_NAME and PATH_INFO are holdovers from CGI, I think it makes sense to pile everything into SCRIPT_NAME and leave PATH_INFO empty. Skip Montanaro | Mojam: "Uniting the World of Music" http://www.mojam.com/ skip(a)calendar.com | Musi-Cal: http://concerts.calendar.com/ 518-372-5583

2 1

- Ack! I screwed up slightly...
by skip＠calendar.com 22 Dec '98

22 Dec '98

In the message I just sent out I complained about SCRIPT_NAME being /. Well, it turns out I was starting ZopeHTTPServer with SCRIPT_NAME=/. In reality, if I run ZopeHTTPServer with no SCRIPT_NAME argument I get PATH_INFO: '/check_request' PATH_TRANSLATED: '/home/dolphin/skip/src/Zope-1.9.0b3-src/check_request' SCRIPT_NAME: '' which is still arguably wrong in my opinion, just different than I originally reported. Skip

1 0

- Confera into Zope?
by Mark Interrante 22 Dec '98

22 Dec '98

The subject says it all. Mark

1 0

RE: [Zope] - PCGI with IIS 4.0
by Jeff Bauer 22 Dec '98

22 Dec '98

> My problem is that lured by the simplicity of getting > pcgi to work on linux/apache I have promised to port > my web app to IIS. > I remember that there used to be several references > for getting PCGI to work on IIS, but I can't find > any of them now ;( > and the included readme points to > http://starship.skyport.net/crew/jbauer/persistcgi/ > which is also not very helpful for IIS newbie like me > (I have no idea how to make IIS execute the info file > using pcgi_wrapper.exe, which I guess is the thing I > should do) Hannu, Amos got pcgi-wrapper.exe successfully working with IIS back in August. I'll let him provide the details, then I'll add them to the pcgi docs. ;-) I think one of the things he did was create a file association, .pcgi extensions to pcgi-wrapper.exe. The pcgifile.py sanity tester should work (works with Win32/Apache) and the pcgitime.py is a good starter example to get pcgi working on NT. Keep in mind that process creation is more expensive on NT vs. Unix. So although it should work for you, it may not sustain heavy loads. Don't tell us you're surprised <wink>. Best regards, Jeff Bauer Rubicon, Inc.

2 1

- Products Control_Panel
by Mike Pelletier 21 Dec '98

21 Dec '98

I may be missing something terribly obvious and quite well documented, but I'm still in the dark... What is the Products portion of the Control_Panel _for_ exactly? Mike. -- --- | Mike Pelletier Work: 519-746-1607 /opeware! | Software Developer Home: 519-725-7710 --- | mike(a)zopeware.com Fax: 519-746-7566 http://www.zopeware.com | Zopeware is not endorsed by Digital Creations

2 1