You can subscribe to this list here.
2001 |
Jan
|
Feb
|
Mar
|
Apr
(57) |
May
(287) |
Jun
(166) |
Jul
(286) |
Aug
(273) |
Sep
(254) |
Oct
(144) |
Nov
|
Dec
|
---|---|---|---|---|---|---|---|---|---|---|---|---|
2002 |
Jan
|
Feb
|
Mar
|
Apr
|
May
(1) |
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
From: Andreas A. <a.a...@th...> - 2001-07-15 12:55:59
|
Hi Alex, >> Ah cool, he's currently working on a PHP Application Server. Very >> interesting but nothing public right now :( > oooh! Jepp, it sounds groovy (once it works ;-)). This article is a bit old, but I did not find more information on this topic. http://www.geocrawler.com/lists/3/Web/195/125/4988613/ Andi |
From: Andreas A. <a.a...@th...> - 2001-07-15 12:55:59
|
Hi Alex, > an entity is a hierarchical definition of a set of data, with names and > rules associated with each 'field' in that hierarchy. > They are byu their nature 'central' in the system, i.e. accessible to all. > Entities can be stored anywhere in the sourc tree, but they will end up in Hmm. Like that? class TCategory { var $categoryId; var $name; var $parentId; function selfCheck() { .. } } class EtyCategory { function Add() { MetabaseStuff; } ... } Andi |
From: Justin F. <je...@ey...> - 2001-07-15 02:40:17
|
Aha. So it is for managing, er, what I might call "persistant" entities, and that means the database. This also looks good, having all that regex stuff/type_checking/ is_required stuff done for ya. Wonderful. Well, I was on a parallel universe. I was _assuming_ some kind of facility that was a framework that "facilitated" declarations of, filling of, walking of, array-based data structures. I have to let this sink in now. What I am finding is a "need" for a kind of data-structure- manipulator-class (DSM Class). I had some idea that the EntityManager was going to give me such a facility. But what I need is probably too specific for a "Manager". I find that I am slowly writing/adding to this DSM Class to do what I want, and shamefully, it ain't turning out too general. The explanation of this need is that I am writing a dynamic report generator, defined by a web form, and thereby letting the user define, not only the data sources, but the "look-and-feel" that the report will use, which will be determined by such a data structure. Then I can pass this around serialized, say, for the session. Or, if the user wishes to "keep" it, it can be written to some user preferences/profile table, maybe mapped to one or a set of reports, or all reports. And so the EntityManager will make that part easy. Man, this BC is like being given a harem for your birthday. You just don't know where to start (or end). _jef ------------------------------- Alex Black wrote: > > > Thank you John, for this answer. It clears up a lot of things, > > and I feel that it is going to be useful. > > > > What seems satisfying is that, if I understand correctly, is > > that if I need a data structure such as: > > > > $setup = array( > > 'table' => array( > > 'border' => "0", > > 'align' => "left", > > 'bgcolor' => "#CCCCFF", > > 'cellspacing' => "2", > > 'cellpadding' => "5", > > 'width' => "400", > > ), > > 'header' => array( > > 'bgcolor' => "#FFCCCC", > > ), > > 'row' => array( > > 'bgcolor' => "#CCFFCC", > > ), > > 'column' => array( > > 'bgcolor' => "#FCFFCC", > > ), > > 'other' => array( > > 'freshmeat' => true, > > ), > > ); > > > > that I, somehow, write this out in XML, and somehow, give > > it an identity/tag/handle/key I can somehow, declare it easily, > > (and it gets dragged out of ./ent ??), can fill it easily, > > use it arguments to other module method calls. It is > > appreciated that a lot of this effort will be offloaded > > at compile time and not run time. > > hi Justin, > > the above example isn't correct. > > here's a (simplified) example: > > entity: user > username > type => text > size => '20' > regex => 'regex_pattern' > field_label => 'User Name' > required => true > email_address > type => email > size => '40' > field_label => 'Email Address' > required => true > > so, you'll have a users table in your database - and it's good to be able to > do: > > EntityManager->Get(); with a primary key and entity, so you can get back a > set of data, or, if you are posting data, Entitymanager will enforce the > rules you defined: > > if you're Adding a user, the username and email_address fields must not be > empty - username has to match the regex provided, and the email address has > to match the standard email regex that's implicitly attached to the type. > > Entitymanager is _only_ for data access. > > > In any case, it sounds good. I will just have to wait, > > now, for the Stradivarius. > > best, > > _alex > > _______________________________________________ > binarycloud-dev mailing list > bin...@li... > http://lists.sourceforge.net/lists/listinfo/binarycloud-dev -- Justin Farnsworth Eye Integrated Communications 321 South Evans - Suite 203 Greenville, NC 27858 | Tel: (252) 353-0722 |
From: Alex B. <en...@tu...> - 2001-07-15 01:57:15
|
> Thank you John, for this answer. It clears up a lot of things, > and I feel that it is going to be useful. > > What seems satisfying is that, if I understand correctly, is > that if I need a data structure such as: > > $setup = array( > 'table' => array( > 'border' => "0", > 'align' => "left", > 'bgcolor' => "#CCCCFF", > 'cellspacing' => "2", > 'cellpadding' => "5", > 'width' => "400", > ), > 'header' => array( > 'bgcolor' => "#FFCCCC", > ), > 'row' => array( > 'bgcolor' => "#CCFFCC", > ), > 'column' => array( > 'bgcolor' => "#FCFFCC", > ), > 'other' => array( > 'freshmeat' => true, > ), > ); > > that I, somehow, write this out in XML, and somehow, give > it an identity/tag/handle/key I can somehow, declare it easily, > (and it gets dragged out of ./ent ??), can fill it easily, > use it arguments to other module method calls. It is > appreciated that a lot of this effort will be offloaded > at compile time and not run time. hi Justin, the above example isn't correct. here's a (simplified) example: entity: user username type => text size => '20' regex => 'regex_pattern' field_label => 'User Name' required => true email_address type => email size => '40' field_label => 'Email Address' required => true so, you'll have a users table in your database - and it's good to be able to do: EntityManager->Get(); with a primary key and entity, so you can get back a set of data, or, if you are posting data, Entitymanager will enforce the rules you defined: if you're Adding a user, the username and email_address fields must not be empty - username has to match the regex provided, and the email address has to match the standard email regex that's implicitly attached to the type. Entitymanager is _only_ for data access. > In any case, it sounds good. I will just have to wait, > now, for the Stradivarius. best, _alex |
From: Justin F. <je...@ey...> - 2001-07-15 01:41:34
|
Thank you John, for this answer. It clears up a lot of things, and I feel that it is going to be useful. What seems satisfying is that, if I understand correctly, is that if I need a data structure such as: $setup = array( 'table' => array( 'border' => "0", 'align' => "left", 'bgcolor' => "#CCCCFF", 'cellspacing' => "2", 'cellpadding' => "5", 'width' => "400", ), 'header' => array( 'bgcolor' => "#FFCCCC", ), 'row' => array( 'bgcolor' => "#CCFFCC", ), 'column' => array( 'bgcolor' => "#FCFFCC", ), 'other' => array( 'freshmeat' => true, ), ); that I, somehow, write this out in XML, and somehow, give it an identity/tag/handle/key I can somehow, declare it easily, (and it gets dragged out of ./ent ??), can fill it easily, use it arguments to other module method calls. It is appreciated that a lot of this effort will be offloaded at compile time and not run time. In any case, it sounds good. I will just have to wait, now, for the Stradivarius. _jef ----------------------- John Donagher wrote: > > On Sat, 14 Jul 2001, Justin Farnsworth wrote: > > > > > It is not clear, understandably, just what the Entity > > Manager is going to do. The words in the documentation > > make sense, but specifically, > > > > 1. Will we be able to define a data structure with it > > like, externally in XML, and then refer to this > > as what we want to pass around in certain places? > > The entities are defined in XML, yes. The 'make' process "compiles" those XML > structures into PHP data structures, so the run-time cost is not incurred. > > What do you mean "pass around"? The EntityManager manages operations having to > do with entities, namely Add(), Set(), Get(), GetList(), and Delete(). If you > need a populated entity object, then the Get() method will return it based on > the primary key. > > > 2. Or, will data structures have to be defined in each > > program "by hand"? > > I don't follow. What data structures are you talking about? > > > 3. Are "easy" declarations even going to be a part of > > the Entity Manager. > > Don't follow this either. The EntityManager is a framework. Your application > will probably extend it based on your needs. binarycloud will provide a base > set of functionality based on the EntityManager, namely a lister, an editor, a > picker. We don't currently have the notion of a pager, that I'm aware of; > That's part of the lister. > > > 4. Many other questions... > > > > I just don't know what this animal is going to offer. > > For certain, what I feel a need for is a facility that > > "fixes" somehow, certain data structures, makes them > > easy to declare, and fill, and allow us to have a "standard" > > set of them, so I can tell, say, Jason, to write this > > or that module and make sure you use the FOOBAT > > data structure as the formal argument to > > > > It's pretty broad in scope. In our application, the EntityManager is _the_ > connection between application logic and data management, storage and retrieval. > > > > > PS: And when is a first cut of Entity Manager likely???? > > This has yet to be determined. There is a significant amount of effort involved > here and it is a really bad time for the three of us who wrote it to work on > the binarycloud "port". Also, the EntityManager depends on the QueryManager, > and some other significant components. > > John > > -- > > John Donagher > Application Engineer > Intacct Corp. - Powerful Accounting on the Web > 408-395-0989 > 720 University Ave. > Los Gatos CA 95032 > www.intacct.com > > Public key available off http://www.keyserver.net > Key fingerprint = 4024 DF50 56EE 19A3 258A D628 22DE AD56 EEBE 8DDD > > _______________________________________________ > binarycloud-dev mailing list > bin...@li... > http://lists.sourceforge.net/lists/listinfo/binarycloud-dev -- Justin Farnsworth Eye Integrated Communications 321 South Evans - Suite 203 Greenville, NC 27858 | Tel: (252) 353-0722 |
From: Alex B. <en...@tu...> - 2001-07-15 00:49:15
|
hi all, I'm thinking about doing multi-site makes this way. This is very, very early thinking so please bear that in mind :) ----- The current contents of binarycloud/user/ is moved into binarycloud/user/default/ (default would be the "default" distro site name) With the idea being that you end up with: binarycloud/ user/ default/ ** current contents of binarycloud/user/* ** site_name/ other_site/ smallsite/ bigsite/ blah/ blah/ blah/ Also, I would like to try and structure the makefile so there is only one, top level user/Makefile - I don't want to have to maintain (x) Makefiles, or a directory of directories ala: SITE_DIRS = \ site_name \ other_site \ blah \ etc. Right now, there is explicit declaration of the _internal_ structure of the user/ tree: USERDIRS= \ conf \ db \ lang \ lib \ mod \ roles \ tmpl \ $(EMPTY) So, what I was thinking... We enforce this structure within binarycloud/ user/ site_name/ _but_ - if there is a missing directory from binarycloud/user/site_name/, we'd grab user/default/missing_directory_name/ and include it in the build version of binarycloud/user/site_name/ So, for example, you could leave out the user/site_name/conf/ directory, and know that it would be grabbed from user/default/conf/ at make-time and put in user/site_name/conf/. That would be great for maintaining a 'base' config. I need to do a little more looking + putzing, but I think this may be possible at the file level for important files that reside in: conf/ and lang/, also. best, _alex -- alex black, ceo en...@tu... the turing studio, inc. http://www.turingstudio.com vox+510.666.0074 fax+510.666.0093 |
From: Alex B. <en...@tu...> - 2001-07-15 00:27:16
|
>> It is not clear, understandably, just what the Entity >> Manager is going to do. The words in the documentation >> make sense, but specifically, >> >> 1. Will we be able to define a data structure with it >> like, externally in XML, and then refer to this >> as what we want to pass around in certain places? > > The entities are defined in XML, yes. The 'make' process "compiles" those XML > structures into PHP data structures, so the run-time cost is not incurred. > > What do you mean "pass around"? The EntityManager manages operations having to > do with entities, namely Add(), Set(), Get(), GetList(), and Delete(). If you > need a populated entity object, then the Get() method will return it based on > the primary key. > >> 2. Or, will data structures have to be defined in each >> program "by hand"? > > I don't follow. What data structures are you talking about? To clarify: an entity is a hierarchical definition of a set of data, with names and rules associated with each 'field' in that hierarchy. They are byu their nature 'central' in the system, i.e. accessible to all. Entities can be stored anywhere in the sourc tree, but they will end up in binarycloud/build/{lang}/user/ent/ >> 3. Are "easy" declarations even going to be a part of >> the Entity Manager. > > Don't follow this either. The EntityManager is a framework. Your application > will probably extend it based on your needs. binarycloud will provide a base > set of functionality based on the EntityManager, namely a lister, an editor, a > picker. We don't currently have the notion of a pager, that I'm aware of; > That's part of the lister. I think xml declarations are as "easy" as we can expect given the capabilities offered by EntityManager. >> 4. Many other questions... >> >> I just don't know what this animal is going to offer. >> For certain, what I feel a need for is a facility that >> "fixes" somehow, certain data structures, makes them >> easy to declare, and fill, and allow us to have a "standard" >> set of them, so I can tell, say, Jason, to write this >> or that module and make sure you use the FOOBAT >> data structure as the formal argument to That is exactly correct. > It's pretty broad in scope. In our application, the EntityManager is _the_ > connection between application logic and data management, storage and > retrieval. yep :) >> PS: And when is a first cut of Entity Manager likely???? > > This has yet to be determined. There is a significant amount of effort > involved > here and it is a really bad time for the three of us who wrote it to work on > the binarycloud "port". Also, the EntityManager depends on the QueryManager, > and some other significant components. So, I'm working on other things, namely external package integration, and a couple minor enhancements to page. This will happen, all, but it's a buch of really complex code that we need to sort through and integrate properly - and that will take time :) _alex -- alex black, ceo en...@tu... the turing studio, inc. http://www.turingstudio.com vox+510.666.0074 fax+510.666.0093 |
From: John D. <jo...@we...> - 2001-07-15 00:15:54
|
On Sat, 14 Jul 2001, Justin Farnsworth wrote: > > It is not clear, understandably, just what the Entity > Manager is going to do. The words in the documentation > make sense, but specifically, > > 1. Will we be able to define a data structure with it > like, externally in XML, and then refer to this > as what we want to pass around in certain places? The entities are defined in XML, yes. The 'make' process "compiles" those XML structures into PHP data structures, so the run-time cost is not incurred. What do you mean "pass around"? The EntityManager manages operations having to do with entities, namely Add(), Set(), Get(), GetList(), and Delete(). If you need a populated entity object, then the Get() method will return it based on the primary key. > 2. Or, will data structures have to be defined in each > program "by hand"? I don't follow. What data structures are you talking about? > 3. Are "easy" declarations even going to be a part of > the Entity Manager. Don't follow this either. The EntityManager is a framework. Your application will probably extend it based on your needs. binarycloud will provide a base set of functionality based on the EntityManager, namely a lister, an editor, a picker. We don't currently have the notion of a pager, that I'm aware of; That's part of the lister. > 4. Many other questions... > > I just don't know what this animal is going to offer. > For certain, what I feel a need for is a facility that > "fixes" somehow, certain data structures, makes them > easy to declare, and fill, and allow us to have a "standard" > set of them, so I can tell, say, Jason, to write this > or that module and make sure you use the FOOBAT > data structure as the formal argument to > It's pretty broad in scope. In our application, the EntityManager is _the_ connection between application logic and data management, storage and retrieval. > > PS: And when is a first cut of Entity Manager likely???? This has yet to be determined. There is a significant amount of effort involved here and it is a really bad time for the three of us who wrote it to work on the binarycloud "port". Also, the EntityManager depends on the QueryManager, and some other significant components. John -- John Donagher Application Engineer Intacct Corp. - Powerful Accounting on the Web 408-395-0989 720 University Ave. Los Gatos CA 95032 www.intacct.com Public key available off http://www.keyserver.net Key fingerprint = 4024 DF50 56EE 19A3 258A D628 22DE AD56 EEBE 8DDD |
From: Justin F. <je...@ey...> - 2001-07-14 19:54:39
|
Alex: First of all, please do not consider this something like pressure. But, this question(s) which I will pose may be relevant to others on this list that are also in a module writing frenzy. Experience that we have had and question that have arisen may be useful to expose. It seems to me, firstly, that one of the "arts" that will be acquired to exploint the huge potential of BC is the knack for isolating the functionality of a module to the just-right-level, nothing really different that the difficulty in OOP of designing a class effectively. Now we are writing/testing/learning BC by writing modules for an Intranet. There will be dozens or hundreds of different reports, and so we break down these reports to kind of three things: 0. A data fetcher // unique, db's etc ------------------ 1. A selector // general 2. A lister // general 3. A pager // general with which one can bolt together any report on any set of data quickly, as well as serve as pieces for a dynamic report generator. The goal is that these are so generally written that these will be the only three modules needed for any and all reports. Immediately, and I mean immediately, inter-module communication is necessary, as the Selector has to pass in a data structure of data to the Lister, and the Selector also has to make some kind of method call with a data structure argument that sets up the list in certain kinds of formats/look/ feel. Then the Pager has to control how many rows to display, and a kind of control data structure has to be passed around. Whew, all of this "works". But these damn data structures are "hand crafted" at this moment and easily "forgotten" and so there are a lot of nuisance errors. Jason and I keep thinking, "Waaall, this Entity Manager is going to solve our problem." It is not clear, understandably, just what the Entity Manager is going to do. The words in the documentation make sense, but specifically, 1. Will we be able to define a data structure with it like, externally in XML, and then refer to this as what we want to pass around in certain places? 2. Or, will data structures have to be defined in each program "by hand"? 3. Are "easy" declarations even going to be a part of the Entity Manager. 4. Many other questions... I just don't know what this animal is going to offer. For certain, what I feel a need for is a facility that "fixes" somehow, certain data structures, makes them easy to declare, and fill, and allow us to have a "standard" set of them, so I can tell, say, Jason, to write this or that module and make sure you use the FOOBAT data structure as the formal argument to $NewModule->some_setup_method($FOOBAT_TYPE); Lordy, I hope this is what will be possible. If not, then I might be forced in to add this kind of convenience. At the present state of things, one gets tired easily at this level of complexity. _jef PS: And when is a first cut of Entity Manager likely???? PPS: If we want anything "next", it is Entity Manager. -- Justin Farnsworth Eye Integrated Communications 321 South Evans - Suite 203 Greenville, NC 27858 | Tel: (252) 353-0722 |
From: alex b. <en...@tu...> - 2001-07-14 19:16:28
|
> Ah cool, he's currently working on a PHP Application Server. Very > interesting but nothing public right now :( oooh! _a |
From: Andreas A. <a.a...@th...> - 2001-07-14 09:52:11
|
Hi Nathan, >> Last news I had was PHPlib was in the process of being integrated? Maybe I >> missed something... > You did, it ain't. Someone wanted to port a few of the "HTML-Wizard > Classes" to PEAR (like OOH-Forms, SQL_Query, etc..) but it still hasn't > happened, the SourceForge site set up to do this hasn't been touched in > since the intial import several months ago. Hmm, I spoke with Christian Koehntopp at the Linuxtag and asked him about PHPLib->Pear integration. He told me it's currently in progress. Ah cool, he's currently working on a PHP Application Server. Very interesting but nothing public right now :( Andi |
From: TAO R. <ron...@ho...> - 2001-07-14 04:14:51
|
Bruce, Could you please try the following steps, and give me the output...??? 1. make clena, then make again 2. cd $BCHOME/base/utils 3. edit processprepend.php remove '-q' in the first line 4. type this command ./processprepend.php $BCHOME/build/en/htdocs/prepend.php $BCHOME $BCHOME/build/en see what is out, and check $BCHOME/builde/en/htdocs/prepedn.php see if the BC_PATH is correct or not. :) ronald >Alex/Ronald > >BCHOME is being set correctly >$ echo $BCHOME >c:/r2/binarycloud > >I tried this as well >c:/r2/binarycloud/ > >Didn't make any difference > > >Permissions is not a problem in Windows. > >The problem is that the arguments are not coming accross from > c:/r2/binarycloud/base/init/Makefile > > ># {{{ Header ># -File $Id: Makefile,v 1.4 2001/06/25 16:57:08 odysseas Exp $ ># -License LGPL (http://www.gnu.org/copyleft/lesser.html) ># -Copyright 2001, Intacct Corp. ># -Author odysseas tsatalos, ody...@ya... ># }}} > >include ../../Makefile.in > >PACKAGEDFILES= \ > Init.php \ > $(EMPTY) > >all: installcode installprepend > >installcode: > @$(INSTALLCODE) $(BUILD_DIR) $(PACKAGEDFILES) > >installprepend: > @$(INSTALLCODE) $(SITEBUILD_DIR) prepend.php > @$(PROCESSPREPEND) $(SITEBUILD_DIR)/prepend.php $(BC_DIR) $(BUILD_DIR) > >"Makefile" 22L, 525C 1,1 All > >---------------------------------------------------------------------------------------------------------------------------- >If I take a sledgehammer to $BCHOME/base/utils/processprepend.php > >/* >AutoGenerate($argv[1], " ># define('BC_PATH', '$argv[2]'); >define('BC_PATH', '$argv[3]'); >"); >*/ >AutoGenerate("c:/r2/binarycloud/build/en/htdocs/prepend.php", " ># define('BC_PATH', '$argv[2]'); >define('BC_PATH', 'c:/r2/binarycloud'); >"); > > > >It writes out the file OK, but I am afraid that something else may be broken > >------------------------------------------------------------------------------------------------------------------------------ > >If this actually worked then how do I deploy it. > _________________________________________________________________________ Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com. |
From: TAO R. <ron...@ho...> - 2001-07-14 03:43:11
|
> >Alex/Ronald > >BCHOME is being set correctly >$ echo $BCHOME >c:/r2/binarycloud this one is ok, not the following one > >I tried this as well >c:/r2/binarycloud/ > >Didn't make any difference > > >Permissions is not a problem in Windows. yes. it is no need to bother the permissions in cygwin... > >The problem is that the arguments are not coming accross from > c:/r2/binarycloud/base/init/Makefile > > >If this actually worked then how do I deploy it. > >In a previous e-mail I saw > I would say, it is not the problem of that Makefile, it is no difference between the new CVS one, and my testing tarball package one, but I will check it out today. > > once you have untarred that file, you'll have a directory: r2 > > if you copy: > > r2/binarycloud.php to your php include path > > and > > r2/binarycloud to one level above htdocs > > and > > r2/htdocs/* to somewhere (anywhere) in your server's document root, > >I don't see an binarycloud.php file anywhere. > >I did notice a win32 patch send out by Ronald for binarycloud.php but my stupid e-mail reader would not read the file. >Can I download it from somewhere? > binarycloud.php in no longer exist and useful, fogget it..... :) > >Bruce > _________________________________________________________________________ Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com. |
From: Bruce D. <br...@ap...> - 2001-07-14 02:54:43
|
Alex/Ronald BCHOME is being set correctly $ echo $BCHOME c:/r2/binarycloud I tried this as well c:/r2/binarycloud/ Didn't make any difference Permissions is not a problem in Windows. The problem is that the arguments are not coming accross from c:/r2/binarycloud/base/init/Makefile # {{{ Header # -File $Id: Makefile,v 1.4 2001/06/25 16:57:08 odysseas Exp $ # -License LGPL (http://www.gnu.org/copyleft/lesser.html) # -Copyright 2001, Intacct Corp. # -Author odysseas tsatalos, ody...@ya... # }}} include ../../Makefile.in PACKAGEDFILES= \ Init.php \ $(EMPTY) all: installcode installprepend installcode: @$(INSTALLCODE) $(BUILD_DIR) $(PACKAGEDFILES) installprepend: @$(INSTALLCODE) $(SITEBUILD_DIR) prepend.php @$(PROCESSPREPEND) $(SITEBUILD_DIR)/prepend.php $(BC_DIR) $(BUILD_DIR) "Makefile" 22L, 525C 1,1 All ---------------------------------------------------------------------------------------------------------------------------- If I take a sledgehammer to $BCHOME/base/utils/processprepend.php /* AutoGenerate($argv[1], " # define('BC_PATH', '$argv[2]'); define('BC_PATH', '$argv[3]'); "); */ AutoGenerate("c:/r2/binarycloud/build/en/htdocs/prepend.php", " # define('BC_PATH', '$argv[2]'); define('BC_PATH', 'c:/r2/binarycloud'); "); It writes out the file OK, but I am afraid that something else may be broken ------------------------------------------------------------------------------------------------------------------------------ If this actually worked then how do I deploy it. In a previous e-mail I saw > once you have untarred that file, you'll have a directory: r2 > if you copy: > r2/binarycloud.php to your php include path > and > r2/binarycloud to one level above htdocs > and > r2/htdocs/* to somewhere (anywhere) in your server's document root, I don't see an binarycloud.php file anywhere. I did notice a win32 patch send out by Ronald for binarycloud.php but my stupid e-mail reader would not read the file. Can I download it from somewhere? Bruce >a little, > >you need to make sure the permissions are correct, i.e. that the dir is >writable by you. > >I don't know how sygwin handles that on windows. > |
From: Alex B. <en...@tu...> - 2001-07-14 01:59:01
|
hi all, I'm actually going to start "enforcing" the separation of the lists soon (yeah, ala judge dredd) :P - please subscribe to the ones that interest you: http://lists.sourceforge.net/lists/listinfo/binarycloud-cvs-commit This list receives a message every time there is a commit, and includes a diff. http://lists.sourceforge.net/lists/listinfo/binarycloud-general This list is for installation questions, and _usage_ questions. http://lists.sourceforge.net/lists/listinfo/binarycloud-announce This is an extremely low-traffic list that I'll send announcements to. http://lists.sourceforge.net/lists/listinfo/binarycloud-dev This is the development list. Use this if you are contributing or intend to contribute. best, _alex -- alex black, ceo en...@tu... the turing studio, inc. http://www.turingstudio.com vox+510.666.0074 fax+510.666.0093 |
From: Alex B. <en...@tu...> - 2001-07-14 01:47:02
|
> You did, it ain't. Someone wanted to port a few of the "HTML-Wizard > Classes" to PEAR (like OOH-Forms, SQL_Query, etc..) but it still hasn't > happened, the SourceForge site set up to do this hasn't been touched in > since the intial import several months ago. heh >> I haven't see anyony on the list say "no, don't add it" so I'm not sure what >> this is in reply to (did I miss something?) > > Yes, people did. yep, and I read 'em. >> >>> However, make room for application frameworks for binary cloud and the like, >>> and keep those separate somehow from the core of PEAR, but at least allow >>> them in somwhow, and now you have something that many, I mean MANY people >>> would be very interested in. >> >> That I agree with. >> > > As do I, but it won't happen. Get over it. Neither phplib nor > binarycloud will make it into PEAR. Just move along and pretend this > never happened. At least, that's what I'm doing. I don't particularly much care, to be honest. I did this because I was interested in seeing the reaction, not because I'm dying to have binarycloud in pear. I do use the pear libs in bc, and I do find them useful, and I do think that there is room for frameworks in the pear model. But I'm not interested in raising a kerfluffle over it. I did that once, I think it was necessary then, but I have no interest in doing it again. >> Inclusion should not necessarily mean "endorsement" it should mean "base set >> of quality requirements met" > > Ahh.. But I get the feeling it does. yes, that ivory tower thing, :) ... it will bit them in the ass one day 8) anyway, I'm sad to see that phplib wasn't integrated, there are some great tools that I think pear could benefit from. natch! have a good one, _alex -- alex black, ceo en...@tu... the turing studio, inc. http://www.turingstudio.com vox+510.666.0074 fax+510.666.0093 |
From: nathan r. h. <na...@ds...> - 2001-07-14 01:29:53
|
On Fri, 13 Jul 2001, Alex Black wrote: > > Last news I had was PHPlib was in the process of being integrated? Maybe I > missed something... > You did, it ain't. Someone wanted to port a few of the "HTML-Wizard Classes" to PEAR (like OOH-Forms, SQL_Query, etc..) but it still hasn't happened, the SourceForge site set up to do this hasn't been touched in since the intial import several months ago. > I haven't see anyony on the list say "no, don't add it" so I'm not sure what > this is in reply to (did I miss something?) Yes, people did. > > > However, make room for application frameworks for binary cloud and the like, > > and keep those separate somehow from the core of PEAR, but at least allow > > them in somwhow, and now you have something that many, I mean MANY people > > would be very interested in. > > That I agree with. > As do I, but it won't happen. Get over it. Neither phplib nor binarycloud will make it into PEAR. Just move along and pretend this never happened. At least, that's what I'm doing. > Inclusion should not necessarily mean "endorsement" it should mean "base set > of quality requirements met" > Ahh.. But I get the feeling it does. -n -- -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- nathan hruby / digital statement na...@ds... http://www.dstatement.com/ Public GPG key can be found at: http://www.dstatement.com/nathan-gpg-key.txt ED54 9A5E 132D BD01 9103 EEF3 E1B9 4738 EC90 801B -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- |
From: Alex B. <en...@tu...> - 2001-07-14 00:59:08
|
>> Stig Bakken wrote: >>> I don't really have any strong opinions on this, so I'll go with >>> the majority. >> >> Okay, since we decided not to add application frameworks like >> PHPLIB to PEAR I'm strongly against adding binarycloud. Huh? Last news I had was PHPlib was in the process of being integrated? Maybe I missed something... >> Sebastian Bergmann I agree with Stig on this. If the pear community doesn't want that, I agree. but and to continue... (if you want the three second version, I think applications frameworks are fine in PEAR so long as they do not dictate to pear, and meet the coding requirements) > I don't understand what it would hurt to add binary cloud, I only see how, > in the end it would help, by making PEAR seem to many as something more > worthwhile. It seems that ultimately what stands in the way of widespread > use of PEAR in the first place is a lack of ease to use the system and set > it up. (I am speaking as someone who is not a seasoned programmer but a web > designer who found himself in a position where he had to learn web > programming). PHP itself has met with great success, I would argue primarily > due to the support (an online manual that is extremely useful), as well as > its flexibility and ease of use. I agree to a certain extent, though binarycloud is not your perfect example of ease-of-use. it requires a good deal of setup and config, because it's a beefy system. I haven't see anyony on the list say "no, don't add it" so I'm not sure what this is in reply to (did I miss something?) > PEAR on the other hand, while valuable, and I see a lot of potential for it, > I don't see it gaining widespread use until people who are not necesarily > coming from a programming background (such as myself) see it as easy to use, I disagree 100%. PEAR is serious s*it, which is exactly what php needs. This kind of development (pear and binarycloud and I'm sure others) requires discipline, and it's complex stuff. While I agree that some of the tools should be "easy" to use because they are clearly defined and well documented, I am not interested in making things "easy" for their own sake. I'm speaking for bc here, though I would venture to guess that the pear people feel sort of the same way. > impliment, and see a real use for it. As it stands there are other database > abstraction layers out there, there are other classes available that give > the average developer all they would need, so their it no real need for PEAR > for most people unless they are a die hard coder, or programmer, or are just > sympathetic to the cause. That is ostensibly true. But: when I went looking for serious, pluggable chunks of code before I thought of binarycloud, I didn't find much. So I went about creating a means for that to happen. As it happens, PEAR is doing stuff I'm not: pear doesn't do system design, it does high quality libraries, which fits perfectly with binarycloud. binarycloud will actually rely on a number of PEAR libs (but that's another thread) :) > However, make room for application frameworks for binary cloud and the like, > and keep those separate somehow from the core of PEAR, but at least allow > them in somwhow, and now you have something that many, I mean MANY people > would be very interested in. That I agree with. Inclusion should not necessarily mean "endorsement" it should mean "base set of quality requirements met" > I don't see why we could not somehow make room for other application > frameworks like binary cloud. After all, if they are based on top of PEAR it > only makes sense to me, especially if someone is willing to spend the time > to port them over. At the least, I argue it could not hurt, only help in the > long run, both those interested in developing binary cloud and those > interested in developing PEAR. As it stands we have all this effor going in > two different directions, seems such a waste. Personally I would love to see > it. Actually, not. I danced around pear for a bit to make sure we weren't doing duplicative things - we are a _little_ with some of the binarycloud core classes, but otherwise I intend to rely on PEAR quite extensively for caching, error handling, and many, many other miscellaneous tasks. There is a small amount of "conflict" but I view it as so small as to be negligible. The code in pear is written with discipline - which is something the php community badly needs. -alex |
From: Alex B. <en...@tu...> - 2001-07-13 19:22:59
|
hi all, I have done another CVS synch to binarycloud.com, the new variable naming conventions are in cvs.. _alex -- alex black, ceo en...@tu... the turing studio, inc. http://www.turingstudio.com vox+510.666.0074 fax+510.666.0093 |
From: Alex B. <en...@tu...> - 2001-07-13 18:59:06
|
a little, you need to make sure the permissions are correct, i.e. that the dir is writable by you. I don't know how sygwin handles that on windows. _a > Ronald, > This is really neat stuff. I did all of that but this is what I got back. Have > I stuffed up somewhere? > > $ make > Building da site > in user > in user/htdocs > in base > in core > in bldr > in init > Could not open file for writing : > in lib > in mgr > Building en site > in user > in user/htdocs > in base > in core > in bldr > in init > Could not open file for writing : > in lib > in mgr > > > >> binarycloud make/r2 Win32 quick tutorial >> ----------------------------------------------------------------------------- >> -- >> 1. go to http://sources.redhat.com/cygwin/ install the latest cygwin >> package, >> up till now, I still not sure which package is a must to install, so try >> hard to install all packages, make sure you set the default TEXT FILE >> TYPE >> to UNIX from the installation tools, sorry :) >> >> 2. get the latest win32 version of PHP for http://www.php.net/, don't ask me >> how to install PHP, ok... >> >> 3. get the latest binarycloud r2 package, there are two ways to do so, first >> you can download the tarball here, http://www.binarycloud.com/download/, >> and unpack it in the cygwin environment, DO NOT UNPACK IT IN NATIVE WIN32 >> ENVIRONMENT, or checkout from the latest CVS sourse tree. (I forgot the >> a/c >> and password for anonymous access, you ask alex, ok...) >> >> 4. set the environment variable BCHOME to c:/path/to/binaryclound (yes, is >> the >> Windows sytle path, not the UNIX one, check your own), try to type these >> in >> cygwin shell. >> export BCHOME='c:/path/to/binarycloud' >> Then try the following to see if it is ok or not. >> echo $BCHOME >> >> You need a text editor you familiar with for the following steps, no metter >> it >> is vi in cygwin or UltraEdit in Windows, but make sure it can work with UNIX >> type EOL, otherwise you will find all the files are very messy. >> >> 5. use a editor to open $BCHOME/base/utils/installcode.sh, add the following >> line to the very begining of the file (the first line, ok...). >> #!/bin/sh (yes, is UNIX style path this time) >> >> 6. use a editor to open $BCHOME/base/utils/processprepend.php and >> xml2php.php >> you can see these in the first line, #!/usr/local/php4/bin/php -q, of >> both >> files, simply replace it by the following line. >> #!c:/path/to/php/php.exe -q (yes, is Windows sytle again) >> >> 7. things done, go back to $BCHOME, and type make, everythins should be >> fine. > > > > _______________________________________________ > binarycloud-dev mailing list > bin...@li... > http://lists.sourceforge.net/lists/listinfo/binarycloud-dev > -- alex black, ceo en...@tu... the turing studio, inc. http://www.turingstudio.com vox+510.666.0074 fax+510.666.0093 |
From: Alex B. <en...@tu...> - 2001-07-13 18:59:03
|
> Metabase is much better than PEARdb I think that as well, but if PEAR DB is better in future, I'll use it. > i can=B4t say much to pear in general right now (besides that there are > some inviduals contributing code which seems to have the only > reason in establishing them as future "php consultants" or something > similar. > "Mr. XY is a former php community member since x years now and > the author of several important PEAR components. > Mr. XY can consult your company on all PHP issues for just $ XXX > per hour".) >=20 > and what can i say to phpclasses ? I like phpclasses. yeah, you have a dig a little. but it isn't an ivory tower, and manuel (it's proprietor) is pretty cool (and the author of metabase and the groovy xml parser) _a > "There are no rules, fear is unknown and sleep is out of question !" >=20 > Anyway ... i had a bad day today, >=20 > Holger >=20 >=20 >=20 > _______________________________________________ > binarycloud-dev mailing list > bin...@li... > http://lists.sourceforge.net/lists/listinfo/binarycloud-dev >=20 -- alex black, ceo en...@tu... the turing studio, inc. http://www.turingstudio.com vox+510.666.0074 fax+510.666.0093 |
From: TAO R. <ron...@ho...> - 2001-07-13 15:51:38
|
did you set the environment variable BCHOME using the windows style path?? actually I stuck on the same things at the very begining... :) ronald > >Ronald, >This is really neat stuff. I did all of that but this is what I got back. Have I stuffed up somewhere? > >$ make >Building da site >in user >in user/htdocs >in base >in core >in bldr >in init >Could not open file for writing : >in lib >in mgr >Building en site >in user >in user/htdocs >in base >in core >in bldr >in init >Could not open file for writing : >in lib >in mgr > _________________________________________________________________________ Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com. |
From: Holger B. <hb...@ts...> - 2001-07-13 15:49:38
|
Metabase is much better than PEARdb i can´t say much to pear in general right now (besides that there are some inviduals contributing code which seems to have the only reason in establishing them as future "php consultants" or something similar. "Mr. XY is a former php community member since x years now and the author of several important PEAR components. Mr. XY can consult your company on all PHP issues for just $ XXX per hour".) and what can i say to phpclasses ? "There are no rules, fear is unknown and sleep is out of question !" Anyway ... i had a bad day today, Holger |
From: Bruce D. <br...@ap...> - 2001-07-13 13:29:44
|
Ronald, This is really neat stuff. I did all of that but this is what I got back. Have I stuffed up somewhere? $ make Building da site in user in user/htdocs in base in core in bldr in init Could not open file for writing : in lib in mgr Building en site in user in user/htdocs in base in core in bldr in init Could not open file for writing : in lib in mgr >binarycloud make/r2 Win32 quick tutorial >------------------------------------------------------------------------------- >1. go to http://sources.redhat.com/cygwin/ install the latest cygwin >package, > up till now, I still not sure which package is a must to install, so try > hard to install all packages, make sure you set the default TEXT FILE >TYPE > to UNIX from the installation tools, sorry :) > >2. get the latest win32 version of PHP for http://www.php.net/, don't ask me > how to install PHP, ok... > >3. get the latest binarycloud r2 package, there are two ways to do so, first > you can download the tarball here, http://www.binarycloud.com/download/, > and unpack it in the cygwin environment, DO NOT UNPACK IT IN NATIVE WIN32 > ENVIRONMENT, or checkout from the latest CVS sourse tree. (I forgot the >a/c > and password for anonymous access, you ask alex, ok...) > >4. set the environment variable BCHOME to c:/path/to/binaryclound (yes, is >the > Windows sytle path, not the UNIX one, check your own), try to type these >in > cygwin shell. > export BCHOME='c:/path/to/binarycloud' > Then try the following to see if it is ok or not. > echo $BCHOME > >You need a text editor you familiar with for the following steps, no metter >it >is vi in cygwin or UltraEdit in Windows, but make sure it can work with UNIX >type EOL, otherwise you will find all the files are very messy. > >5. use a editor to open $BCHOME/base/utils/installcode.sh, add the following > line to the very begining of the file (the first line, ok...). > #!/bin/sh (yes, is UNIX style path this time) > >6. use a editor to open $BCHOME/base/utils/processprepend.php and >xml2php.php > you can see these in the first line, #!/usr/local/php4/bin/php -q, of >both > files, simply replace it by the following line. > #!c:/path/to/php/php.exe -q (yes, is Windows sytle again) > >7. things done, go back to $BCHOME, and type make, everythins should be >fine. |
From: Justin F. <je...@ey...> - 2001-07-13 03:17:05
|
Alex: The Subject is a rather meaningless statement, but here I am back again with the problem that I explained the other day, and for which I didn't get too much sympathy for. I will have to give you a simple example of the "problem" with BC, using your code, not mine, so to speak. I will use your "Hello World" disto. I will use this to prove that something is "rotten". What I think is that it may have to do with the Zend mechanization. So, this is easy, just drop the following into the HelloWorld.php module, in the context evident outside my comments. [== snip, snip ==] echo "</pre>\n<!-- end HelloWorld test module output -->\n"; #=========== jef experiment ===================== #== set some warnings, two ways... == global $Page; $this->setWarning("Called inside class with \$this"); $Page->modules['content'][2]->setWarning("SET BY INDEX"); if($this->HaveWarning || true) { // make true for experiment printf("<h4>WARNINGS (%d)</h4>\n", count($this->WarningStack)); echo "<pre>\n"; for($i=0; $i < count($this->WarningStack); $i++) { printf("%d: %s\n", $i+1,$this->WarningStack[$i]); } echo "</pre>\n"; } else { echo " "; } } function setWarning($msg="Unknown warning") { $this->HaveWarning = true; array_push($this->WarningStack,$msg); printf("setWarning called...with [%s]. Now %d warnings on stack.<br>\n", $msg,count($this->WarningStack)); } var $WarningStack = array(); #==================================================== // }}} // {{{ Vars [== snip, snip ==] NOW, my friend, render the bloody thing. Do you understand what you see? You will see that, somehow, there are two copies of WarningStack somewhere, or Zend does not do dereferencing correctly with this suggested construct to communicate between modules of $Page->modules['content'][2]->do_or_call_something In Page, you are actually "storing" the instantiated classes inside the Page Class in $Page->modules. Considering what little I know about Zend, I do know that Zend makes copies of the vars, but only references the methods after instantiation. IF I PUT THE STACK GLOBAL, EVERYTHING WORKS. Using a stack inside a class that is instantiated in Page, it does not work. I did not confuse this example by putting my "tag identification" of a module, but believe me, that also "does not work". I think this problem may be more sinister than you think, and if I am not all wet here, it has grave ramifications. I think it would be "fixed" by having the instantiations of the modules in a global array, and not inside the Page class itself. But I will let you mull this over. I have been trying to chase down this "bug" in my code for several days. It is just something else, in my opinion, that Zend is doing, and for which BC is assuming a behaviour that does not correspond to reality. You are going to see what is APPARENTLY TWO COPIES OF VARS (WarningStack), viz: setWarning called...with [Called inside class with $this]. Now 1 warnings on stack. setWarning called...with [SET BY INDEX]. Now 5 warnings on stack. _jef -- Justin Farnsworth Eye Integrated Communications 321 South Evans - Suite 203 Greenville, NC 27858 | Tel: (252) 353-0722 |