- Zope - Zope lists

- backing up Data.bbb
by Roman Milner 18 Jan '99

18 Jan '99

Following a corrupted Data.bbb disaster I want to automate the backing up of my Data.bbb. My question is - do I need to shut down zope before copying Data.bbb to ensure that it is closed properly? If so, how can I shutdown zope properly from the command line? Thanks, ^Roman

4 4

- Slashdot on Zope
by Gabe Wachob 18 Jan '99

18 Jan '99

I wanted to ask the list what people's thoughts were on the appropriateness of implementing a Slashdot-like site in Zope. I have looked at the Slashdot source code (in perl) and its pretty messy (to put it mildy). I hacked at it to get it to work with the free Sybase server on linux and it works reasonably well, but I haven't gone much beyond that. If a slashdot-like site were implemented on Zope, would an external SQL database be neccesary, or would the internal object database be enough? For searching? Building the Slashdot-style site on Zope has the attractive feature of being very customizable and modular from the get-go (something which Slashdot source code is trying to get to). Slashdot (the site) gets a LOT of traffic -- would a zope server be able to handle the (what I'm guessing is) hundreds of thousands of hits per day? Do we have to wait until the Medusa integration is complete (not that throwing together a slashdot clone can be done before digicool folks can get that done). One thing I'm wondering about is upgrading. Assuming that the slashdot clone is upgraded regularly, how does this affect product instances already in the object store (this is a general zope question). My understanding is that straight pickling is done -- only data is serialized, so if I change methods, I should be able to use new versions of products with objects created by older versions of a product.. But if I add or delete attributes, then I have to be careful... Finally, does the concept of a slashdot-like Product (or set of products) interest anyone? -Gabe

7 6

- DTML and Role information
by David White 18 Jan '99

18 Jan '99

I'm in the process of seeing how I might go about converting some of our existing web-based apps to Zope products. I would like to move the bulk of the user/profiling information into the Zope framework and have each Product/App have its own set of Permissions/Roles/Users. I believe this would be quite easy with Zope. The setup would be something like this: +-BaseDir | +Product_A (has its own Roles and Permissions) | | | +-(acl_users) | +-Product_B (has its own Roles and Permissions) | | | +-(acl_users) (possibly different from Product_A | +-Product_C (has it own Roles and Permissions) | |...etc >From the BaseDir I'd like to have a DTML page which can 'figure out' which Products the user has access to. I basically don't want to provide any access to a certain area below unless I've confirmed they have the right to go there in the first place. I've seen that the Tree tag has a parameter call 'skip_unauthorized', but I'd like to go through the sub-folders using an 'in' (for custom formatting). Would I have to write an External Method which I'd use in an 'if' condition check inside the 'in' loop? If so, what would the correct syntax be? Is there a more direct/builtin way to get at a particular users Role within the current or sub-folders? Thanks in advance, Dave -- ---------------------------------------------------- David R. White Raytheon Electronic Systems Process Support (508) 440-2087 528 Boston Post Rd. Sudbury, MA 01776 davew(a)ed.ray.com ----------------------------------------------------

1 0

- [Zope] zip zap zope (long, rich ?) [Re: Zap]
by Boris Borcic 18 Jan '99

18 Jan '99

<STANDARD_DISCLAIMER> This started as a short word prompted by the idea-announcement of Zap, and ends up as a very long stream of half-baked and not-quite-all topical ideas. Oh well, I guess I'll just have to apologize to the non-readers, congratulate the readers, and warn that I might repeat some of the imbedded ideas in isolated and better presented form, depending on the feedback this (doesn't) get (in this or other fora..hums). </STANDARD_DISCLAIMER> The more I think about Zope, the more I think it is best placed to become "the" balance-tipping Open Software's Dynamics' "killer" app. (I name Open Software Dynamics the Zeitgeist in which digicool's decision to opensoft itself, inscribes itself naturally) I can dream about it in the small, I can dream about it in the big. In my more enthousiastic and litterary moments, I can see it as a cross between Roger Zelazny's Hellrides and The Triumphal March of God's Feast Feeding Birds, straight out of the Apocalypse, with our good Guido van Rossum an onomatopoetic figurehead ! But in any case I need to find myself a niche as a source of source code, if I want a legitimate claim as a significant member of the Open Software Dynamics. So i'll start dreaming in the "virtual small". ========= First idea : Zope+Gadfly+Python 1.5.2+other_bits is really the core of "Full Python". Second idea : "Full Python" like for instance "Prolog", is an artistic achievement opening the door to new *styles* of programming. Third Idea : "Full Python" should Serve Itself First, both as a way to demonstrate itself, and as a way to govern its development to as large a dominance that it will able to achieve. ========= So, to answer the matter of the Zap announcement : The Zopinion project and the ZipZapZope project =================================== The Zopinion project is to have Zope distributed standard, with a possible setup to be used locally by private web page authors maintaining contentful web page systems using Zope, that later automatically unfold to the directory of that end user's internet provider who's not supporting Zope. Ideally, Zopinion could drive its own storing FTP sessions. I believe I've already read that main idea of Zopinion, stated in Zope's list. The moral of Zopinion's name is that we think of it as "a personal opinions server". What's in Zopinion for Digicool, is that Zopinion will push Zope's logo onto WWW sites faster than Zope itself. Of course, Zopinion is far from exercizing all the features of Zope, but such a notice should not deter any real Open Software man. What's in Zopinion for its intended users, is that Zopinion is designed to help individuals with lots of hairy-but-interconnected ideas and opinions, to structure their output in such a way as to provide lurkers with just the appropriate paragraphs and bits. Zopinion is thus the logic of web advertisement technology given back, in the Open Software Spirit, to the private ideas producer. Zopinion's design criteria expresses the idea that mankind has come to a special time, where the competition for bandwidth and mindshare has become so widespread and consistent that the best interest of an intelligent and collaborative source of ideas, like our Guido or Zope's Jim, has become to *serve* his stream of designs with the best dressings for others to copy, freely to the wide public, rather than *hide* them for fear of seeing them stolen. My personal use for Zopinion, is to maintain Zorro's Own Bazaar of Condensing Vaporware (ZOBCV), and my forthcoming "The Y programming language, a short course". [[More mundane Zopinion hype : Zopinion could be augmented to take fully take over and overflow Microsoft's Outlook Express, for reasonably text and code oriented users... that is, manage mail, news and mailing lists... I can see it running as a permanent local proxy.]] === The ZipZapZope project is more ambitious and "way over my head". ZipZapZope is (presumably digicool's) solution to modularly managing Zope distributions, and BaZaars. ZipZapZope is to actual software as Zopinion to condensing vaporware. I envision two aspects/interfaces to ZipZapZope of immediate relevance. One would be a Zope-and-python-oriented hybrid of Debian Linux RPM-FTP distribution and MS Office CDROM installer, functioning over the Web. Ideally, the background logic should automatically construct an Up-to-standards Installer for (at least) the selected items. Don't python libraries provide everything needed ? Another would be, and I guess That One requires user interface innovation, would be a facility for turning, with least effort, mailing lists that are in the state zope(a)zope.org is in, into "self-faqing" branching archives and *persistent* mailthreads, smoothly, with a mail-service logic partly serving newbies with archived paragraphs. Next Step : Zope's conquest of CBZN, the Comprehensive Bazaars over Zope Network. The delicate point with That One is that we will start to need a facility for (Zoped) Bazaars to Rate Each Other. This is a sensitive matter because it pertains to Open Software's Dynamics main alternative to common euro/dollar logic. They are relatively tame perspectives and more radically revolutionary ones. While it isn't always the wisest, I have a tendency to contemplate the more radically revolutionary ones front. It's not always the wisest, because in many cases the tamer perspectives are the right stepping stones to the wild ones. This said, designing for CBZN, I would determine my inspiration by picking a design constraint that pertains to a (yet) imaginary situation. A Kung-fu gesture. I am really a Zelazny fan ! Keyword : snapshots or stripshots. CBZN would "drive or participate" a "management" of "its layer of internet" by mutual snapshot and redistribution services that CBZN sites would provide to ZipZapZope and Zopinion sites. The CBZN main layer should be integrated to full ZipZapZope distribution, I guess. This design constraint brings vast amounts of style to the design... it expresses a personal taste, that I guess must be shared by many Python and Open Software Lovers, that there was, in the good old internet of the textual codes days, something superior to the nowadays' "multimediated internet and programming". I mean 1) the mutual network relationship of CBZN sites should be expressible in terms of this "snapshot" business, 2) CBZN would specialize on *lossy* or *condensating* snapshots, imposing constraints of *data*economy* to the snapshots it takes, 3) CBZN would in particular "have in mind" stripping snapshots of non-linguistic data, or at least, of binary, image, or voice-like data that is not provided with a recognizable or certified equivalent in written semantic (natural or artificial) language. 3) Repeat : CBZN would be choosy on the way it mirrors data automatically in just the right way, that makes Open Software Sources Distributions it's notion of what best content is ! This suggest a couple more avenues. One avenue is a new business, "imagating". Imagating can be sold to big operators as a two-step advertisement strategy, whose ultimate purpose it to permit the big operators to assert their mindshare while showing they are inobstrusive. "Imagation" is an *actually*free* service provided to CBZN-like lossy Web Page snapshots producers and users : plug into semantic descriptions of images contained in strip-shot pages, relatively fitting images taken from Image Banks of images *invisibly*watermarked* with some Big O's watermark ! The second step comes after a while, when "Imagation" has taken hold as an institution : Statistics will be made that will show each Big O's Real Place in the World ! That's the advertisement we need to sell the Big O's ! Or we'll go to the Big O's competition ! Of course, "Imagation" logic can extend to other multimedia snippets that just images, and, besides "retrofitting" stripped pages, will of course serve Zopinion authors too busy to waste their time scanning the images they'd need to make their pages more sexy to readers. (There could be a "reverse side" to imagation, the treatment of illustrious images. Illustrious images, such as portraits of illustrious persons, could be treated to a worldwide collaborative network and friendly competition of "best legend design". Yeah, this would fit the Triumphal Apocalyptic March view, but I guess, in that case, it leaves me with a bit too much of the Main Role to fullfill. [Forthcoming : the YZ rules of blasphemy-classifications and the information Ramadan] ) And of course, plotting packages able to turn a condensed linguistic expression into a graphic automatically, will provide basic imagation services. The second mentionned avenue, is to secure Open Software's Dynamic over names and identifiers. Mmh, while it would be too long to detail independent justifications immediately, think of it as a good idea to maintain a secured archive of Open Software Reference identity, identifiers and litterate snippets library Registry, centered on the works of Open Software's Dynamic's appropriate "conjugates set" of Guido and Jims as Figureheads and Implementations. [One of the first starting tasks is of course for the Python community to answer for itself the question : If Jims validly generalizes Guido, who validly generalizes Jims ? We might say, Jims have Guido-Number between 1 and 2, and we want to make a proper census of people with Guido-Number between 1 and 2, "according to us". For litterature on Guido-Number, see Erdos-Number. Guido Himself is Unique with Guido-Number Zero. For the Larger Communities the question will be rather : Who is like Guido ? Well, Larry, as matter of course, but then ? Who is to Larry as Guido is to Guido's Friends ? Well, Tom and Randall at least, but then ? Well, let's not forget that Guido has a brother : Just ! And let's not forget that Guido's design has a preferred bedmate : Tcl/Tk ! So, Just in a sense "competes" with Larry as a Proper Way to Generalize Guido, just as Tcl/Tk "competes" with Perl as a Proper Way to Generalize Python, for the purpose of designing Our Roster Of Best Friends ! Maybe this allows me, the apparently unjustified freedom in this forum, to cite another person, who has little to do with python and even with programming, as an adequate "pseudo-clone" of Guido, for the purpose of Our Roster, and for the purpose of the quality of the Sampling that we are Dreaming Of for Our Roster, and this person is Mr. John Baez, of sci.physics.research and "This Week's Finds in Mathematical Physics Fame". Below comes the paragraph that was written first in these brackets, and of whom all the above 6 paragraphs are an explicit extension. The Difficult Part of That Work is : Mmmh the problem is the above selection has a programmation that is well-centered in the Open Software's Dynamic's Core, but that it appears to drive towards a form of sample that is formally over-biassed as an example of the ideal sample of persons that we would really want to unite as Authority.] To put it another way. Whatever the community you consider, you may contemplate it with the view that you are exploring the sets of the set of the persons forming the community, and doing your democratic citizen's task by comparing and establishing preferences between such imaginary sets, as to whether there are a good representative subset of their whole, if it came to take decisions or to negociate in the name of their whole. Claim 1 : while there unavoidably exists variance between the conclusions various citizens can draw from such an "imaginary experiment", there nevertheless exists, in normal cases, a strong "common conclusions" component, especially if you do not limit yourself to one community's universe, but maintain the view that all communities share in vaster embedded universes of intelligent beings. (Which means you are actually classifying your sampling-reluniverse pairs of choice) Claim 2 : The Guido-Larry connection and the "Open Software's Dynamic" characterization of "Our Final Sample", as of -10 paragraphs, imho points the way to "building a perfect sample", but for a limited while well characterized community, while at the same time appearing as a perversely biassed sample if we want to view it as a sample of the wider community we would like to scan for Our Wider Sampled Roster Of Best Friends and Friendships. "Perversely Biassed", in the sense that the "obvious" manner to "naively extend the sample", by "lowering its filtration standards", will achieve the reverse of the "obvious" expectation : it will make the bias worse instead of better ! This said, the Guido-Larry connection provides well-defined benefits and guidelines to the trust we want to give to the Open Software Dynamic. Or at least a good figure for something I really think needs to be implemented : an Open Software's Semi-Central Authority of Identity, Identifiers, and Snippets, with an eye on protecting Open Software's choices of names from abusive greed, and another on providing basic support to syncretic scripting language programming, by allowing coders to "mean what they want" in mixed syntax and/or mixed language standard library allusions. Still, all the above doesn't do Justice to the Other Great Dream (OGD) I had for Python, that was to offer to Guido the best equivalent of what he brought to us, the Python coders, by gently demonstrating that we would not mind if he changed his projects, and decided to evolve Python 2, not as a redesign "done right and pure" of Python 1.x, but as an integration of Python 1.5, with its special facilities, in the programming language Oz 2, a marvel by itself, but that may be in need of Guido and Python to really make it to the life it deserves. For instance - and this is really from someone who has no idea of the implementation issues - I have the vague impression that Python is sufficiently flexible to mimic good portions of algorithms expressed in Oz 2, so that Python could "espouse" Oz 2 with the bonus of permitting (with some restrictions) Oz 2 code to refer to the Python Modules, and reciprocally Oz 2 could be made to "espouse" Python by devising a Python-mimicking surface syntax for itself. Note that Oz 2 has something quite close to Tkinter, and should be able to provide Python with all that it yet lacks in Sexpyness and related program transformation aptitudes. Oh, and a last word : securing revenues has been a long-lasting problem and will recur soon, and while I hesitate to draw from Zope, more than the resources that I already draw for free, it may just be the case that it is sensible to mention this issue with the above diarrhea in perspective. Best Regards, Boris Borcic --- Paragraph contents copylefted B.Borcic 1999 with special requirement that the following assertion be appropriately presented whenever the issue arises : "Open software, serving weapons of designfully murderous potential, is a contradiction in terms" Forthcoming : a thoughtful criticism of the limitations of the Bazaar/Cathedral duality. More ADN at www.geneva-link.ch/~zorro (pre-alpha packaging and presentation)

1 0

Re: [Zope] - problem with Apache/stop script?
by Michel Pelletier 18 Jan '99

18 Jan '99

On 15 Jan 1999, Roman Milner wrote: > > Apache/stop has the following: > > # If Principia cranked up, kill it and whack its temporary files. > if [ -f $dot/../var/Main.pid ]; then > kill `cat $dot/../var/Main.pid` > rm -f $dot/../var/Main.pid > fi > > if [ -r $dot/../var/Main.soc ]; then > rm -f $dot/../var/Main.soc > fi > > > Shouldn't this be using pcgi.pic and pcgi.soc? I don't have any > Main.pid. > Yes, this will be fixed in the next version of Zap. For now changing the file names as you suggested will fix it. We're working hard on changing all the Principia things to Zope things. (Note: These files are determined by your PCGI file, which you can customize to place them anywhere or name them anything). -Michel > Thanks, > ^Roman > >

1 0

- FSIMPORT and the root folder
by Timothy J. Grant 18 Jan '99

18 Jan '99

Hi all, I'm messing around with using FSIMPORT, and would like to import my website into Zope using it. My problem is that my Root folder in Zope has no name, and it appears that FSIMPORT requires a folder name for it to work. Thanks for your assistance. -- Stand Fast, tjg. =================================== Timothy J. Grant tjg(a)avalongroup.net Avalon Technology Group www.avalongroup.net (503) 246-3630 voice (503) 246-3124 fax This message may be digitally signed with PGP. A PGP signature guarantees that this message really did come from me. For more information regarding digital signatures and encryption, please contact me.

1 0

- How to circumvent url to object mapping ? (correction)
by Guido Sohne 18 Jan '99

18 Jan '99

I want to have a URL called say, http://www.webstar.com.gh/articles/1998/01/15/Ashanti_Records_Record_Revenu… which should map to a python method/object that queries an external database with the arguments 1998+01+15+Ashanti_Records_Record_Revenues Is there a way to do this *without* using Apache rewrite rules ? ^^^^^^^^^^^^^ I meant to say without but said with ...

4 3

- Need installation recommendation
by LD Landis 18 Jan '99

18 Jan '99

Hi, Ok... Before getting into Zope+Gadfly, et al, I played around with installing Gadfly into the /usr/local/lib/python1.5 directory (since I didn't see any harm in having that there, from my perspective)... I see that gadfly is also installed in the Zope lib... I currently have the gadfly code in two places. Now that I'm ready to install the DCOracle stuff, etc... what is the best way to do that? Do you recommend having separate library areas for each thing (like gadfly, DCOracle, etc)? or just jam them all into the same lib? Is there anything about the DCOracle packaged in the Zope adapter that would prevent just using what is placed elsewhere (e.g. /usr/local/lib/python1.5 or subdir)? Now that I'm getting ready to start running a much more fully featured Python... I'm looking for "the voice of experience" in how to lay it out so that it can be managed. -- Cheers, --ldl ----------------------------------------------------------------------------- LD Landis ldl(a)HealthPartners.Com N0YRQ Voice 612/883-5511 Fax 612/883-6363 HealthPartners, 8100 34th Avenue So, PO Box 1309, Minneapolis, MN 55440-1309 Shape your life not from your memories, but from your hopes. (Borrowed) -----------------------------------------------------------------------------

3 2

Re: [Zope] - ANNOUNCE: DCOracle release
by Kevin Littlejohn 18 Jan '99

18 Jan '99

>>> Anthony Baxter wrote > I'd guess that one of the results is larger than can fit in an integer, > maybe? Actually, it looks like it was in dealing with the input parameters, rather than the output ones - I found in ociCurs.py, around line 124, a piece of code that turns a number into an integer if it has no 'scale'. On commenting that out, it appears to work. If someone with a touch more knowledge than I can comment on the ramifications of that, it'd be appreciated ;) > > also, if timestamp's a number, you don't need the ' 's around it. Yeah, I know - I _did_ mention the example given was one of many attempts, right? ;) (quotes/noquotes * fmt/sqlvar * :float/:integer/nothing on input variables = many different possible combinations to assure myself it wasn't a dumb mistake in syntax ;) Oh, one other thing I wanted to double-check: In db.py (back in ZOracleDA), in the 'columns' method, there's the following code: def columns(self, table_name): c=self.cursor try: r=c.execute('select * from %s' % table_name) except: return () desc=c.description That looks like tempting fate to me... Is the 'select * from ' really not going to do bad things? (I'm presuming it doesn't actually trigger the db into _doing_ a 'select * from' unless it calls fetchone()) Is there a less, um, dangerous-looking way of getting the same result? I only ask because select * from most of the tables here would go away for a _very_ long time... KevinL --------------- qnevhf(a)obsu.arg.nh --------------- Kevin Littlejohn, Technical Architect, Connect.com.au Don't anthropomorphise computers - they hate that.

2 1

- ZOracle DA...
by David White 18 Jan '99

18 Jan '99

Hello, I've been quietly following the development of Zope and I'm VERY impressed and happy to see such a wonderful tool available as open source. I've been anxiously awaiting the release of DCOracle and the ZOracle database adapter and was overjoyed to see them arrive last week. Unfortunately, I've run into a little snag which I was hoping someone could help me through. I downloaded both DCOracle and the ZOracle DA (then realizing the ZOracle DA has DCOracle in the src directory...) I extracted the ZOracle DA product to the <Zope_Home>/lib/python/Products location and I then compiled the DCOracle piece in the <Zope_Home>/lib/python.Products/ZOracleDA/src directory. I ran the DCOracle_test.py file and got no errors. Having thought all was well I moved the Buffer.so and oci_.so shared libs to the <Zope_Home>/lib/python/Products/DCOracle directory. When I restarted Zope, the product appeared to be installed correctly as it wasn't 'broken' (no tracebacks), but when I tried to create a connection I got this traceback: ------------------------------------------------------------- Invalid connection string: webtools/webtools Traceback (innermost last): File /local/app/dc/Zope/lib/python/ZPublisher/Publish.py, line 879, in publish_module File /local/app/dc/Zope/lib/python/ZPublisher/Publish.py, line 595, in publish (Info: /Development/manage_addZOracleConnection) File lib/python/Products/ZOracleDA/DA.py, line 121, in manage_addZOracleConnection (Object: Navigation) File /local/app/dc/Zope/lib/python/Shared/DC/ZRDB/Connection.py, line 139, in __init__ (Object: RoleManager) File /local/app/dc/Zope/lib/python/Shared/DC/ZRDB/Connection.py, line 168, in edit (Object: RoleManager) File /local/app/dc/Zope/lib/python/Shared/DC/ZRDB/Connection.py, line 212, in connect (Object: RoleManager) File lib/python/Products/ZOracleDA/db.py, line 127, in __init__ File lib/python/Products/ZOracleDA/DCOracle/ocidb.py, line 65, in __init__ BadRequest: (see above) ------------------------------------------------------------- I have Zope being published through PCGI. I have an apache webserver which is running as a user which 'knows' where ORACLE_HOME so the PCGI process should 'know' this too. Did I just miss something obvioius? Any help is greatly appreciated, and once again thanks for a GREAT tool! Dave -- ---------------------------------------------------- David R. White Raytheon Electronic Systems Process Support (508) 440-2087 528 Boston Post Rd. Sudbury, MA 01776 davew(a)ed.ray.com ----------------------------------------------------

3 2