Menu
▾
▴
dataxi-sysdev — System development list
You can subscribe to this list here.
| 2001 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
(1) |
Sep
|
Oct
|
Nov
|
Dec
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2002 |
Jan
|
Feb
(5) |
Mar
|
Apr
(10) |
May
(4) |
Jun
(4) |
Jul
(2) |
Aug
(1) |
Sep
(2) |
Oct
|
Nov
|
Dec
|
| 2003 |
Jan
(9) |
Feb
(2) |
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2005 |
Jan
|
Feb
(1) |
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2007 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
(1) |
|
From: Vegas C. <Sco...@ae...> - 2007-12-13 08:53:01
|
Start to play at Vegas Club VIP and you can be sure of an exhilarating experience US Players are welcome! Join right now and feel the action in minutes. 750 USD Free just for you! http://www.balletcasino.com |
|
From: Charles D. <cd...@sp...> - 2005-02-12 00:57:55
|
I stumbled upon Dataxi on Freshmeat, and it looks to be almost exactly what I'm seeking. I'm curious, however, as to the status of the project, given that 1.0 was expected in "a few weeks" about 3 years ago, and the mailing list appears to be inactive -- but the web page has no indication that the project is abandoned or up for grabs by a new maintainer; likewise the mailing list. Is the project indeed abandoned, or merely hibernating pending fresh interest? Is it known whether the code has bit-rot or still works nicely w/ current PHP/PEAR/etc versions? Thanks! |
|
From: Jyry K. <jy...@ne...> - 2003-02-03 13:40:25
|
hello
Both funcs now takes an opt $ForceAbs = 0 parameter, that help to return
absolute path even if the path in question contains "related to server
root" marker"~~".
So, dzGetPath("~~/foo", 1) would return "/abs/server/site/foo" instead of
"/site/foo" (that would happen normally when $ForceAbs=0).
--
--Jyry
:-( C:-/ C========8-O C8-/ C:-(
|
|
From: Jyry K. <jy...@ne...> - 2003-02-03 10:33:36
|
Hello you may now declare in dataxi.ini: [sys] no_jslinks=yes if you experience troubles with current JS linking, that may cause some serious slowdown in when sending a page to the browser. -- --Jyry :-( C:-/ C========8-O C8-/ C:-( |
|
From: Jyry K. <jy...@ne...> - 2003-01-28 16:19:11
|
Well, the subject says it all. Anyone interested merging all these new wonderful changes with Sourceforge tree? -- --Jyry :-( C:-/ C========8-O C8-/ C:-( |
|
From: Jyry K. <jy...@ne...> - 2003-01-28 12:13:59
|
Hello Now all the inis, except kone.ini, are fully (?) explained in dataxi-howto.lyx under dataxi/docs/howto -- --Jyry :-( C:-/ C========8-O C8-/ C:-( |
|
From: Jyry K. <jy...@ne...> - 2003-01-28 12:13:06
|
Right, it seems that the page support is now done and finished. Note, that the form title is no longer displayed in Stonga with mode but as a page title, no matter how many pages there are per form. -- --Jyry :-( C:-/ C========8-O C8-/ C:-( |
|
From: Jyry K. <jy...@ne...> - 2003-01-26 08:03:04
|
Hello I made the renaming metioned in the subject due to trouble updating Dataxi installation. Every time I copied Dataxi files from a to b, my dataxi.ini got overwritten that is not very convenient. So, from now on fresh Dataxi installations require renaming dataxi.example-ini to dataxi.ini in order to get the beast working. Maybe a good idea would be to include the Dataxi installation error note that when dataxi.ini is not found? Henkka? -- --Jyry :-( C:-/ C========8-O C8-/ C:-( |
|
From: Jyry K. <jy...@ne...> - 2003-01-25 09:06:14
|
The problem with groups not building up correctly was.. surprise, surprise, due to PHP mem handling when I used soSet like this: $value = &$set->getItem($key) I ran some tests and it worked fine, but apparently not in all places... However, it is recomended to use PHP 4.3.x as it seems to deal with passing by reference much better than previous PHP's. For some reason groups seems to work only on the main page, not on the additional. I have a look into that next. Oh, one more thing: I fixed getPageHtml() to produce valid html... The changes are in dev-cvs, not in sourceforge cvs. -- --Jyry :-( C:-/ C========8-O C8-/ C:-( |
|
From: Jyry K. <jy...@ne...> - 2003-01-24 18:14:22
|
Hello Page support is now *nearly* fully implemented. The only problem - it seems - is that group handling does not work. At all. I hope I can sort it out over the weekend. Cheers, -- --Jyry :-( C:-/ C========8-O C8-/ C:-( |
|
From: Jyry K. <jy...@ne...> - 2003-01-20 18:34:21
|
Hello I noticed that there is no generic piece of code that handles a func call like: $has_right = has_right($user, $groups) There is, however the required code inside dzmenu, but I would rather have it in dzgeneric or similar. For now, the getPagesHtml in dzouthtml will not check for rights. I look into it shorly... %-) --jq |
|
From: Jyry K. <jy...@ne...> - 2003-01-19 18:02:07
|
Hello Now parses form sources fine and builds nice struct. There should be no stopping doing the page swiching in the UI. -- --Jyry "Elämään kuuluu isommat ja pienemmät haaverit. Niistä onneksi selviää sosiaalitu, siis, vakuutuk, siis, ole varovainen. :-( C:-/ C========8-O C8-/ C:-( |
|
From: Jyry K. <jy...@ne...> - 2003-01-17 11:42:49
|
Hello This is the 1st stage on the way to get the multi page support per form working. It now loops through all the forms included in the main form [sys] pages=xxx section/key=value pair. Does not cache the sub pages, tough. Yet. Still, next two steps: 1) Ensure that other pages display on the screen, too. Now the default page (no. 0) is only displayed. 2) Add links to the Nalle instead of plain page titles and sort out the way to call dzevents etc. to change the page in template->PageNo. -- --Jyry "Pornoteollisuudessa liikkuu sadat miljoonat. Sosiaalipornossa vain kymmenet. Karpolla on vielä asiaa." C:-( C:-/ C========8-O C8-/ C:-( |
|
From: Jyry K. <jy...@ne...> - 2002-09-13 13:27:53
|
solib:
------
1) soDa
soDa now has new section SODA_PARENTHESES (or SODA_BLOCK) that takes
column name for which to set parentheses and a comma delimited list of no.
of opening and closing parentheses. For example
addToSection(SODA_PARENTHESES, "colname", "1,0")
would put "(" *before* the "colname" and
addToSection(SODA_PARENTHESES, "colname", "2,3")
would put "((" before "colname" and ")))" after.
2) soIni
Major changes to better handle strange and empty section/sets etc.
soIni seems to work pretty well at the moment.
dataxi:
-------
1) soIni changes: more than half of the code base now uses soIni instead
of soIniData. soIni is a lot more programmer friednly than soIniData
2) dzprint now prints images, too (only bitmap pics, though)
--Jyry
C:-( C:-/ C========8-O C8-/ C:-(
|
|
From: Jyry K. <jy...@ne...> - 2002-09-08 11:19:21
|
How does one prevent query being emptied before query mode is displayd? Say, if I want to display a pseudo form with prefilled form? Where to put and what? --Jyry C:-( C:-/ C========8-O C8-/ C:-( |
|
From: Jyry K. <jy...@ne...> - 2002-08-30 10:11:16
|
Proud to announce new feature for fetch. Table decl can now be replaced by: eval://file::func::table notation. See Dataxi-forms-HOWTO.lyx for more and check out example_fetch.php for details. --Jyry C:-( C:-/ C========8-O C8-/ C:-( |
|
From: Jyry K. <jy...@ne...> - 2002-07-16 08:42:05
|
Password protected form There is a need to password protect a form, so that when ever the form is opened or restored from the viewstack. Maybe setting password=x somewhere in dfataxi block, form options, perhaps? Even more wonderful would be a behaviour when password can be set individually per form, like password=foobar, when password becomes "foobar" for this form, or password=* when user's own password gets promped. Howabout? 4hrs? --Jyry C:-( C:-/ C========8-O C8-/ C:-( |
|
From: Jyry K. <jy...@ne...> - 2002-07-03 12:12:36
|
Hello When trying to edit a satsi that has already been locked, asema reports "read error" to the user instead of "lock error". --Jyry C:-( C:-/ C========8-O C8-/ C:-( |
|
From: Jyry K. <jy...@ne...> - 2002-06-16 20:07:24
|
Hello I got sick and tired with soseqkeys as the mechanism did not skip existing values in the target table. Yes, I know there should be no existing rows in the target table beyond the last key in soseqkeys, and yes, in such a case it is system design flaw, but still. I has happened too many times recently. So, I reorganised soseqkeys and after adding a new column "serie" to support different serias for the same table, I added another one that defines the key column name. Actually, this is a lot better model for two reasons. 1) Table can have more than one sequential keys 2) The "does it already exist in the target table" check can be performed. The down side of the new model is, that transiton from the old table+lastkey table structure to the new table+colname+serie+lastkey is not easy and practially needs some manual labour to do that. However, it is definetly worth it. So be carefull before upgrading and applying new soseqkeys as the old table should be dropped off. The old setTable() method stilll exists as a wrapper for the new setTarget(), but the sooner any code using setTable() is adjusted to use setTarget() the less problems should occur. Cheers, --Jyry C:-( C:-/ C========8-O C8-/ C:-( |
|
From: Jyry K. <jy...@ne...> - 2002-06-07 12:20:16
|
Hello New approach to protect/readonly control properties: There are threee kinds of targets that should be locked agains modification. 1) form, 2) column (all rows) and 3) field (only the current row in col) Each target is locked based on a value in the source field, the very current col/row, a cell. Now, there are atleast two reasonable way to declare a locking, any of those three. 1) one property for each target (form, col or field) 2) two properties, or commadelimited list or params: one to declare target and other for value In the first case, there could be three different properties called "lockcol", "lockfield" and "lockform". Any of these can have a value that will cause the target to be locked. A special value of "*" may be used for "any value", also "**" for "in any case, null or not, always lock". naturally "**" is not very useful with lockform, but does the trick of old "readonly" for "lockcol" and "lockfield". The second approach, with two properties, or more practical maybe with two comma delimited values, could use "lock=target,value" where target is "form", "column" or "field" and value like in the first approach. I would recommend going with the latter. --Jyry C:-( C:-/ C========8-O C8-/ C:-( |
|
From: Jyry K. <jy...@ne...> - 2002-06-06 17:05:45
|
Hello I fixed dzParseMenuTarget so that one can use "http://" or "https://" -- both are considered as "http". --Jyry C:-( C:-/ C========8-O C8-/ C:-( |
|
From: Jyry K. <jy...@ne...> - 2002-06-06 17:03:32
|
Hello
All subari files need to be created, but separate language file is no
longer required.
The new format of subari file is:
[item_name]
exe=something
someother_setting=value
title={
default=som desirable text
fi=Finnish translation
}
If no title if spoecified at all, then the section name if used as title.
Wonderful :-)
--Jyry
C:-( C:-/ C========8-O C8-/ C:-(
|
|
From: Jyry K. <jy...@ne...> - 2002-05-14 17:21:26
|
Hello dzuser has gone major rework to introduce new structure: id, firstname, surname, password, level and language Due to this, all "dzusers" should be dropped to get the new table created correctly. --Jyry C:-( C:-/ C========8-O C8-/ C:-( M. Schumacher will win "Monaco F1 2002". He will be second, too. And probably third. |
|
From: Jyry K. <jy...@ne...> - 2002-05-14 12:43:31
|
Hello Form has certain states that sometimes require doing rpc. Below is listed the spots most likely to run into in such situation: 1) These have the top priority - pre_save (just before write req - pseudo form too) - post_save (right after succesful write req - pseudo form too) 2) These are not *that* important - post_read (after read req, before display) 3) These are the things of the past - pre_runquery (after query [Ok] but before it is run) - post_runquery (after query is run and browse keys are sent to asema) --Jyry C:-( C:-/ C========8-O C8-/ C:-( M. Schumacher will win "Monaco F1 2002". He will be second, too. And probably third. |
|
From: Jyry K. <jy...@ne...> - 2002-05-06 22:33:46
|
Hello Some nice new toys to play with. 1) sogeneric now includes soGetBoolValue -- use it instead of trying to sort out whether some $Value is 1 or 0 (true/false). 2) soSet->getItem now understands opt param $AsType that defaults to "any", but understands int, dec and bool, too. bool eval uses soGetBoolValue in sogeneric. 3) soIniData has (for now) soGetBool but it might be turned into soGetValue($Section, $Key, $AsType =3D "any"). 4) somath now has soSum and soAvg to get sum and avg from arrays Thats it for now. --Jyry C:-( C:-/ C=3D=3D=3D=3D=3D=3D=3D=3D8-O C8-/ C:-( M$-Outlook: Experience a new virus every day! - Koe uusi virus joka p=E4iv=E4! |
