- Zope - Zope lists

RE: [Zope] RFC: Multiple databases
by Alexander Staubo 28 Jun '99

28 Jun '99

Unless you invent some special magic, you'd lose transactions, too, wouldn't you? Or is that built into the ODB engine? (I don't know exactly how this thing is built.) Under BFS (BeOS) and NTFS 5 (Windows 2000) you don't even have to store properties externally, as these file systems support file attributes. On NTFS 4 (NT 4.0 and earlier) you could store the metadata as separate streams inside each file; this is a barely documented, largely overlooked feature of NT that is rarely used, and is in fact there for Mac resource-fork compatibility. But it would certainly work. I think it's a good idea, if only because FTP and WebDAV (at least as far OS integration is concerned) suck. -- Alexander Staubo http://www.mop.no/~alex/ "`This must be Thursday,' said Arthur to himself, sinking low over his beer, `I never could get the hang of Thursdays.'" --Douglas Adams, _The Hitchhiker's Guide to the Galaxy_ >-----Original Message----- >From: zope-admin(a)zope.org [mailto:zope-admin@zope.org]On Behalf Of >Anthony Pfrunder >Sent: 28. juni 1999 02:29 >To: zope(a)zope.org >Subject: [Zope] RFC: Multiple databases > > >[The following is also in the Collecter under features. I'm >posting this >as it relates to the above messages.] > >My thoughts on extensions to Zope: > >I'm <will be> working on constructing an additional Data.fs plugin >(dropover FileStorage.py) which may be of use. The idea is to >store stuff >on the native filesystem and put properties etc in prop.x (ie the >properties for file1.txt are in prop.file1.txt). The >advantage of this is >that it allows ppl to have a smb connection to a shared directory hence >allowing them to more easily update the files. Of course, >this means that >you lose the ability to undo changes. > >This database could be enhanced by allowing it to reconize Microsoft >Office documents and store properties directly in the document itself. >This means that the properties now travel with the document. > >The second part of this proposal is to produce a Product which >allows you >to "mount" a database onto a subpath. This is very similiar >to importing >the database except that they are kept separate. These means that >toplevel stuff goes into one Data.fs and the department >Data.fs contains >/Department/... stuff. This way, you can have a consistent >style across >various webservers just by running multiple copies of Zserver. In >addition, you could make toplevel.fs read-only to protect it. This >Product should be able to handle different filesystems (ie the current >one, the above one and others) > > >Cheers, > >Anthony Pfrunder > > >_______________________________________________ >Zope maillist - Zope(a)zope.org >http://www.zope.org/mailman/listinfo/zope > >(For developer-specific issues, use the companion list, >zope-dev(a)zope.org - http://www.zope.org/mailman/listinfo/zope-dev ) >

1 0

RFC: Multiple databases
by Anthony Pfrunder 28 Jun '99

28 Jun '99

[The following is also in the Collecter under features. I'm posting this as it relates to the above messages.] My thoughts on extensions to Zope: I'm <will be> working on constructing an additional Data.fs plugin (dropover FileStorage.py) which may be of use. The idea is to store stuff on the native filesystem and put properties etc in prop.x (ie the properties for file1.txt are in prop.file1.txt). The advantage of this is that it allows ppl to have a smb connection to a shared directory hence allowing them to more easily update the files. Of course, this means that you lose the ability to undo changes. This database could be enhanced by allowing it to reconize Microsoft Office documents and store properties directly in the document itself. This means that the properties now travel with the document. The second part of this proposal is to produce a Product which allows you to "mount" a database onto a subpath. This is very similiar to importing the database except that they are kept separate. These means that toplevel stuff goes into one Data.fs and the department Data.fs contains /Department/... stuff. This way, you can have a consistent style across various webservers just by running multiple copies of Zserver. In addition, you could make toplevel.fs read-only to protect it. This Product should be able to handle different filesystems (ie the current one, the above one and others) Cheers, Anthony Pfrunder

1 0

ZClasses problem
by Alex Rice 28 Jun '99

28 Jun '99

I'm trying to create some ZClasses via an external method, but I'm running into a problem: I can't getattr() the product's methods. For instance, I have a Product called Etexts, containing a zclass etext a dtml method etext_add (etext constructor) a dtml method etext_addForm (etext constructor input form) a factory etext_factory (etext factory) I can go anywhere in my Zope heirarchy, and in an external method, I can getattr() to manage_addProduct/Etext, but when I try to getattr() to manage_addProduct/Etext/etext, etext_add, or etext_addForm, I get an AttributeError. I don't have this problem within Zope-- for instance in my browser I can successfully call and create a ZClass with http://some_long_path/manage_addProduct/Etexts/etext?id=spam What am I doing wrong to not be able to get to the same method via an external method? Is this the correct way to programatically add ZClasses? As usual, I'm kind of emulating what goes on in the Zope user-interface. TIA Alex Rice | alrice(a)swcp.com | http://www.swcp.com/~alrice Current Location: N. Rio Grande Bioregion, Southwestern USA

1 0

RE: [Zope] Tree tag and SQL methods?
by Alexander Staubo 27 Jun '99

27 Jun '99

Here you go:      -- Alexander Staubo http://www.mop.no/~alex/ "`This must be Thursday,' said Arthur to himself, sinking low over his beer, `I never could get the hang of Thursdays.'" --Douglas Adams, _The Hitchhiker's Guide to the Galaxy_ >-----Original Message----- >From: Phillip J. Eby [mailto:pje@telecommunity.com] >Sent: 28. juni 1999 00:37 >To: Alexander Staubo; Zope Mailing List (E-mail) >Subject: RE: [Zope] Tree tag and SQL methods? > > >Can you post the source of your tree tag call as well? Thanks. > >At 07:21 PM 6/27/99 +0200, Alexander Staubo wrote: >>No offense taken -- yes, of course I have tested it. :-) >> >>I have not tested using the #tree tag with the second output >though, but >>that would meaningless (the second output should be rendered as leaf >>nodes). I even tried having the second query output _all_ data in the >>table, to no avail. Here's my query: >> >> >>select distinct publication, count(publication) as pub_cnt >>from articles >>where publication != '' >>group by publication >>order by publication >> >>select title >>from articles >>where publication =  >> >> >>Here, pb is the publication argument. I assure you, the second part of >>the query does execute correctly. >> >>-- >>Alexander Staubo http://www.mop.no/~alex/ >>"`This must be Thursday,' said Arthur to himself, sinking low over >>his beer, `I never could get the hang of Thursdays.'" >>--Douglas Adams, _The Hitchhiker's Guide to the Galaxy_ >> >>>-----Original Message----- >>>From: Phillip J. Eby [mailto:pje@telecommunity.com] >>>Sent: 27. juni 1999 18:07 >>>To: Alexander Staubo; Zope Mailing List (E-mail) >>>Subject: RE: [Zope] Tree tag and SQL methods? >>> >>> >>>Don't be offended by this suggestion, but, have you tried >>>having the tree >>>print out all the variables that would affect the 'second' >>>query, and then >>>plugging them into manage_test of the query? My guess is >>>you'll find that >>>there's something you've overlooked as far as how output of >>>the first query >>>gets fed into the second, or else the second is just plain wrong. >>> >>>At 03:36 PM 6/26/99 +0200, Alexander Staubo wrote: >>>>Doesn't improve anything: While the "second" part of the >query -- the >>>>one that's supposed to collect the leaf articles -- clearly >>>executes, no >>>>data is displayed. Only the first level -- the publications >list -- is >>>>rendered, with "plus" icons. Clicking on the icon runs the >>>query but no >>>>data is there. >>>> >>> >> >> >>_______________________________________________ >>Zope maillist - Zope(a)zope.org >>http://www.zope.org/mailman/listinfo/zope >> >>(For developer-specific issues, use the companion list, >>zope-dev(a)zope.org - http://www.zope.org/mailman/listinfo/zope-dev ) >> >

2 1

RE: [Zope] Startup error Zope 2.0.0a3: Error 139
by Alexander Staubo 27 Jun '99

27 Jun '99

I really, really hate to be a pest, but could somebody at DC please look this way and at least make a small attempt at helping me out? Python 1.5.2 and Zope both compiled with gcc 2.8.1; relevant? -- Alexander Staubo http://www.mop.no/~alex/ "`This must be Thursday,' said Arthur to himself, sinking low over his beer, `I never could get the hang of Thursdays.'" --Douglas Adams, _The Hitchhiker's Guide to the Galaxy_ >-----Original Message----- >From: Alexander Staubo [mailto:alex@mop.no] >Sent: 25. juni 1999 16:07 >To: Zope Mailing List (E-mail) >Subject: RE: [Zope] Startup error Zope 2.0.0a3: Error 139 > > >I hear rumours that error code 139 indicates memory corruption. Also I >just noticed the crash below has Python producing a core dump, anybody >want to look at it? > >-- >Alexander Staubo http://www.mop.no/~alex/ >"What the hell, he thought, you're only young once, and threw >himself out of the window. That would at least keep the element of >surprise on his side." >--Douglas Adams, _The Hitchhiker's Guide to the Galaxy_ > >>-----Original Message----- >>From: Alexander Staubo [mailto:alex@mop.no] >>Sent: 25. juni 1999 15:43 >>To: Zope Mailing List (E-mail) >>Subject: [Zope] Startup error Zope 2.0.0a3: Error 139 >> >> >>I just installed Zope 2.0.0a3 (no CVS patches) on a Slackware >Linux 2.2 >>(libc6, Python 1.5.2) box. Of course, it refuses to start: >> >>firewall:/usr/local/zope-2.0.0a3# python z2.py -d 192.168.1.1 >>-w 8080 -f >>2121 >>zdeamon: Fri Jun 25 15:32:33 1999: Houston, we have forked >>zdeamon: Fri Jun 25 15:32:33 1999: Houston, we have forked >>zdeamon: Fri Jun 25 15:32:33 1999: Hi, I just forked off a kid: 24905 >>zdeamon: Fri Jun 25 15:32:35 1999: Aiieee! 24905 exited with >>error code: >>139 >> >>Curiously, no tracebacks? >> >>-- >>Alexander Staubo http://www.mop.no/~alex/ >>"What the hell, he thought, you're only young once, and threw >>himself out of the window. That would at least keep the element of >>surprise on his side." >>--Douglas Adams, _The Hitchhiker's Guide to the Galaxy_ >> >> >>_______________________________________________ >>Zope maillist - Zope(a)zope.org >>http://www.zope.org/mailman/listinfo/zope >> >>(For developer-specific issues, use the companion list, >>zope-dev(a)zope.org - http://www.zope.org/mailman/listinfo/zope-dev ) >> > > >_______________________________________________ >Zope maillist - Zope(a)zope.org >http://www.zope.org/mailman/listinfo/zope > >(For developer-specific issues, use the companion list, >zope-dev(a)zope.org - http://www.zope.org/mailman/listinfo/zope-dev ) >

1 0

RE: [Zope] Tree tag and SQL methods?
by Alexander Staubo 27 Jun '99

27 Jun '99

No offense taken -- yes, of course I have tested it. :-) I have not tested using the #tree tag with the second output though, but that would meaningless (the second output should be rendered as leaf nodes). I even tried having the second query output _all_ data in the table, to no avail. Here's my query:  select distinct publication, count(publication) as pub_cnt from articles where publication != '' group by publication order by publication  select title from articles where publication =   Here, pb is the publication argument. I assure you, the second part of the query does execute correctly. -- Alexander Staubo http://www.mop.no/~alex/ "`This must be Thursday,' said Arthur to himself, sinking low over his beer, `I never could get the hang of Thursdays.'" --Douglas Adams, _The Hitchhiker's Guide to the Galaxy_ >-----Original Message----- >From: Phillip J. Eby [mailto:pje@telecommunity.com] >Sent: 27. juni 1999 18:07 >To: Alexander Staubo; Zope Mailing List (E-mail) >Subject: RE: [Zope] Tree tag and SQL methods? > > >Don't be offended by this suggestion, but, have you tried >having the tree >print out all the variables that would affect the 'second' >query, and then >plugging them into manage_test of the query? My guess is >you'll find that >there's something you've overlooked as far as how output of >the first query >gets fed into the second, or else the second is just plain wrong. > >At 03:36 PM 6/26/99 +0200, Alexander Staubo wrote: >>Doesn't improve anything: While the "second" part of the query -- the >>one that's supposed to collect the leaf articles -- clearly >executes, no >>data is displayed. Only the first level -- the publications list -- is >>rendered, with "plus" icons. Clicking on the icon runs the >query but no >>data is there. >> >

2 1

RE: beginner's questions
by Jens Vagelpohl 27 Jun '99

27 Jun '99

Michel wrote: >Specify a -p argument to the ZopeHTTPServer process. alternatively >you can hardwire it, see the very end (port=9673) of the >ZopeHTTPServer/ZopeHTTPServer.py file. That did it... i didn't find it when i looked at the file the first time. thanks. >I would not suggest running ZopeHTTPServer for a grade. It is not >production software. It was intended to allow people to quicky try >out Zope back before we had ZServer, but now that there is ZServer, >and it is just as simple to use as as ZopeHTTPServer, I would suggest >trying that. well, looking at the ZOPE site discouraged me a little... they say ZServer is for "ZOPE hackers" and such. since i don't have any python experience yet (just today i bought o'reilly's "learning python") i didn't think i could pull it off in the timeframe i have. what i need to have up and running is a demonstration, that's all. i am sure ZopeHTTPServer can handle it :) jens

1 0

ftp editing methods
by Alex Rice 27 Jun '99

27 Jun '99

Is there a way to prevent ZServer from saving all docs as dtml docs instead of as dtml methods, when editing with xemacs/efs? Alex Rice | alrice(a)swcp.com | http://www.swcp.com/~alrice Current Location: N. Rio Grande Bioregion, Southwestern USA

1 0

Multiple domains from one server process?
by Andreas Kostyrka 27 Jun '99

27 Jun '99

Hi! I've been wondering how to serve more than one site from one Zope installation. (Why? 1. Resources. 2. Sites are sometime linked on an organization level that make this kind of sharing sensible.) At the moment I have done the following: http://www.mtg.co.at/ is the real root of the Zope installation http://www.mtg.co.at/sites/photoderm.at/ is one subsite. http://www.photoderm.at/ points in the Apache RewriteRule to mtg.cgi/sites/photoderm.at/$1. Now this basically works ok, but there are certain problems: When accessing http://www.photoderm.at/, Z adds a BASE tag with href http://www.photoderm.at/sites/photoderm.at/ In this case this works ok, as Z finds the correct stuff via acquisition, but in the more general case (when one of the subsites has a sites object itself), this would break. I've fixed it the quick and dirty way by adding the following to the standard_html_header of the site, but this is not perfect, nor clean. ...   <BASE HREF="">    ... It seems that this should be doable somehow better, as codeit (http://codeit.com/zsetup.html) offers to do this. Andreas -- Win95: n., A huge annoying boot virus that causes random spontaneous system crashes, usually just before saving a massive project. Easily cured by UNIX. See also MS-DOS, IBM-DOS, DR-DOS, Win 3.x, Win98.

2 2

ANN: ZCatalog 1.0.0 alpha
by Michel Pelletier 27 Jun '99

27 Jun '99

Greetings Zopeheads. After much sweat and toil and 4 rewrites, we are pleased to announce the release of ZCatalog, the Zope Cataloging and Searching Product, version 1.0.0 alpha. The Catalog can be found at: http://www.zope.org/Download/ZCatalog We recommend you read the ZCatalog Tutorial, written by Mike Pelletier (Note that 'Mike' Pelletier is a completely different person than 'Michel' Pelletier, which is me) which is available on the web at the above location. ZCatalog requires Zope 2.0.0alpha3. It will not work with any previous versions of Zope. Remember, this is *alpha* software. It has limitations, and there are several optimizations/features we have planed for it which are not included in this release. We would, however, like the tired to be well kicked. Please report any bugs you find related to ZCatalog to the Collector: http://www.zope.org/Collector We hope you find some interesting things to do with the Catalog! -Michel

4 3