Giter Site home page Giter Site logo

gumuxiansheng / andro Goto Github PK

View Code? Open in Web Editor NEW
0.0 0.0 0.0 12.17 MB

Automatically exported from code.google.com/p/andro

ApacheConf 0.01% PHP 40.89% HTML 11.65% JavaScript 39.96% CSS 4.50% ColdFusion 0.75% Perl 0.63% ASP 0.71% Lasso 0.33% Python 0.42% Smarty 0.06% NSIS 0.10%

andro's People

Contributors

dorgan avatar

Watchers

 avatar

andro's Issues

Need Javascript function to go to URL with hold variables

Some users stick "hld_" (or hold) variables onto screens, which are passed
along automatically from page to page to keep context.

When a programmer uses x6.json.init(), these variables are automatically
passed along.

However, when a programmer passes to a simple URL, they are all lost.

The new function "linkHolds(url)" accepts a GET string, adds all hold
variables to it, and then jumps to it.

Anybody using this system would use linkHolds() instead of regular links
whenever they need to preserve context.

Original issue reported on code.google.com by [email protected] on 18 May 2009 at 4:50

Need simple way to make project x6

It is not actually possible to easily make a project x6 by default. Need a
command either in YAML or in applib that throws a system easily into x6 by
default.

Original issue reported on code.google.com by [email protected] on 13 May 2009 at 6:08

Follow up to #14, Disallow TAB while auto-complete in progress

Description of Bug:

Hitting TAB while search is in progress produces error.  No crash, but very
confusing to user.

Please tell us how to reproduce this bug:

Go to any auto-complete against a large table, type in value and hit TAB
very rapidly, like customer:  1-2-3-4-5-TAB.  When the TAB hits and focus
leaves the control, there is not yet a complete value.  It's a mess.

SOLUTION: disable TAB on auto-complete fields while search is in progress.

Original issue reported on code.google.com by [email protected] on 28 May 2009 at 4:51

Last-Minute auto-complete refactor

Description of Bug:

If a user is on an auto-complete field (x6select), the searches do not
quite come back fast enough and are not smart about cancelling.  If a user
quickly types "123" they are likely to get "231" with a search result that
does not correspond to what is in the box.

Please tell us how to reproduce this bug:

Use an auto-complete input where the table is > 100 rows and type quickly.

Original issue reported on code.google.com by [email protected] on 27 May 2009 at 1:56

Detect duplicate ID use in DOM

We've been running into bugs that turn out not to be bugs. A programmer
will have two buttons or other items that have the same ID, and then their
behavior becomes quirky. The complete solution to this would be to detect
and prevent multiple items with the same ID from going out to the browser,
that would save a lot of trouble.

Original issue reported on code.google.com by [email protected] on 20 May 2009 at 12:26

Clumsy handling of edit/no save/new

If a user edits a row, but does not save, and then tries to edit a
different row, x6 will attempt a save and give errors if there are any.
Might be better if the request to edit a new row put up a dialog asking if
they want to save the other row.

Original issue reported on code.google.com by [email protected] on 20 May 2009 at 12:24

Error Reporting: secondary errors

Go to the demo, go to orders, create a new order. Type in an invalid
customer #, like 123, and try to save. You will get two errors in a
pop-up, one complaining about bad customer, the next complaining about bad
tax authority. However, the screen does not change these fields to red.

Original issue reported on code.google.com by [email protected] on 13 May 2009 at 6:11

Pre Build Scripts

Pre build scripts would allow you to execute a rename of a table before db
changes take place this way you could rename the table, and make the
changes in the YAML to reflect this.

Of course the builder may have to modified to not complain about creating a
table that it thinks is new that already exists, not sure if it actually
does currently complain.

Original issue reported on code.google.com by [email protected] on 20 May 2009 at 12:32

Documentation of Foreign Keys

Description of Bug:

On the page http://www.andromeda-project.org/foreignkeys.html, the last line of 
the second code block describing table 'orders' is incorrect.  Instead of 
saying:

table orders:
  ...
  foreign_key orders:

it should say:

table orders:
  ...
  foreign_key customers: 

In addition, the second paragraph after that code block should say "the 
column(s) that make up the primary key of the CUSTOMERS table."  There are also 
some other inconsistencies.  For example, in the end of the fourth code block,

    foreign_key father_persons:
        prefix: _father

The last line should be "prefix: father_".

Finally, in the last code block it's unclear what the foreign key column names 
will be: persons_mother and persons_father, or person_mother and person_father.

Please tell us how to reproduce this bug:
N/A

Original issue reported on code.google.com by [email protected] on 5 Aug 2010 at 3:49

More complex chain expressions

Sometimes I really don't like that our chain expressions allow only one
operator per expression. I have learned to be clever about constructing
them, but I don't want to ask new andro developers to learn that. I would
like to retain our ability to validate the expressions while still being
able to allow more complex expressions. LOWEST priority as this is just a
"thinking about" so far.

Original issue reported on code.google.com by [email protected] on 20 May 2009 at 12:30

UI Must execute Fetches

This feature is very important. Don't know if we'll be able to squeeze it
into Release 1. Don't know anybody who actually needs it for current
projects, but it is a hole that it is not there. I have made it priority 8
to show it is either for Version 1.01 or will be bumped to 9 and put into
1.

Example of the problem: User is editing an order on an admin screen. User
puts in customer id. There are fields on the order that have automation_id
FETCH from the customer, but the user will not see them until after saving.
The interface should make a quick JSON call and fill those fields in right
away w/o making the user wait for a save.

Original issue reported on code.google.com by [email protected] on 20 May 2009 at 12:23

recnum in both parent and child tables

Building with this YAML file (extract), on oddity appears in the web
application when filling out an order line: the system seems to be using as
the key the sale_orders' recnum column rather than the sale_order_lines'
recnum.

==== Extract =====
table sale_orders:
module: sales
description: Orders

column recnum:
primary_key: Y
foreign_key customers:
uisearch: Y

table sale_order_lines:
module: sales
description: Order Lines
# nomenu: Y

column recnum:
primary_key: Y

foreign_key sale_orders:
....

=================

Original issue reported on code.google.com by [email protected] on 13 May 2009 at 5:58

Problem pulling updates from SVN

After clicking several times on the SVN update link:
?gp_page=a_pullsvn&svnpull=1&gp_out=process I keep getting the info that
the local version is out of date. Here's the message:

=======================================

Pulling Software Updates From SVN




Application: andro

Complete URL: http://andro.svn.sourceforge.net/svnroot/andro/releases/

Querying for latest version

Latest version is: 2009.03.07.1

Local version is:

Local version is out of date, pulling latest

Pulling code now, this make take a minute or three...

svn export
http://andro.svn.sourceforge.net/svnroot/andro/releases/2009.03.07.1
/var/www/andro_root/andro/pkg-apps/andro-VER-2009.03.07.1/

Code pulled, finished with this application.

Copying files into application directory

cp -Rf /var/www/andro_root/andro/pkg-apps/andro-VER-2009.03.07.1/*
/var/www/andro_root/andro/

Processing Complete

=======================================

This happens on my Ubuntu install from the deb package.

The same action on my Fedora install shows the right message though:


=======================================

Pulling Software Updates From SVN




Application: andro

Complete URL: http://andro.svn.sourceforge.net/svnroot/andro/releases/

Querying for latest version

Latest version is: 2009.03.07.1

Local version is: 2009.03.07.1

Local version is latest, nothing do to.



Processing Complete

=======================================

But I think it's annoying to still get the same alert in the frontpage:

=======================================
New Version of Andromeda Available

Version 2009.03.07.1 is available at the Andromeda Sourceforge repository.
Click Here to go to the Pull Code From Subversion.
=======================================

Original issue reported on code.google.com by [email protected] on 13 May 2009 at 5:59

Allow php pre-insert/update code to prevent action

In an x6 class you can add methods like <table_name>_Before_insert($row)
that are given a row before it is inserted or updated.

When this enhancement goes in, if the routines returns false then the
action is prevented.  To allow the action to proceed you return the row
with or without modifications.

Original issue reported on code.google.com by [email protected] on 8 Jun 2009 at 4:35

Updating of application with no releases

Description of Bug:
When you run "Update from SVN" if an application has no releases it thinks
there are still releases because it still see ".." and "." and tries to
actually download that as a release, which it actually pulls nothing but it
should be trying to pull anything at that point.

Original issue reported on code.google.com by [email protected] on 2 Jul 2009 at 7:00

andro_fix_perms.sh

chmod & chown should have the arguments of -rf instead of just -r this way
hidden files such as .htaccess will also be included.

Original issue reported on code.google.com by [email protected] on 13 May 2009 at 6:11

Skey apparently not available after save new

Description of Bug:

After saving a new row, the skey value does not appear to be available in
x6.data, this breaks custom buttons and other JS code that depends on it.

Please tell us how to reproduce this bug:

Original issue reported on code.google.com by [email protected] on 8 Jun 2009 at 2:09

Regular Text inputs do not allow ampersand

Description of Bug:

Regular Text inputs do not allow ampersand.

Please tell us how to reproduce this bug:

Go to any screen with a text input and try to type an ampersand.

Original issue reported on code.google.com by [email protected] on 8 Jun 2009 at 1:37

Allow Override conventional detail panes

A conventional profile should recognize methods added to a custom class
that provide complete replacements for any detail pane of the parent or any
child.  This is request from a customer, gets high priority.

Original issue reported on code.google.com by [email protected] on 22 May 2009 at 7:23

Weird IE 7 Session Error

This is strange, found this on only one machine. How to reproduce:

1) Open IE 7, log into the demo
2) Open a new IE 7 through start menu
3) In new window, log into the demo
4) In new window, all menu clicks go back to login

If in the new window you type "javascript:alert(document.cookie)" you will
see two values for PHP Session ID, the cookies are somehow doubled in the
2nd window.

Original issue reported on code.google.com by [email protected] on 13 May 2009 at 6:14

History system writes unnecessary rows

The "history" system is used to record changes in one table to a second
table. There is a small glitch that causes an INSERT to write to a history
table when it is not necessary, resulting in a row full of zeroes. The
conditional logic should be improved to prevent this.

Original issue reported on code.google.com by [email protected] on 13 May 2009 at 6:17

Enhance automation SEQUENCE

In a production environment where databases may do master-master
replication, it a good idea to prevent conflicts on auto-incremented
values. A way to manage this is to reserve a range of sequences per
replicated database (or a starting value + increment value).

I would suggest to enhance the SEQUENCE automation with the possibility of
programming this 3 values:
- Minimum value (the minimum 'Last value' of the sequence)
- Maximum value (the maximum 'Last value' of the sequence)
- Increment

As a matter of fact it would be great if any generated sequences (skey,
recnum, etc) could provision such controls.

Original issue reported on code.google.com by [email protected] on 20 May 2009 at 12:26

primary_key allow_empty

allow_empty should be allowed on a multi primary_key table

for instance in a CMS application pages would reference subsections, and
example of subections  could be:

Computers, Laptops
Computers, Desktops


Well if you want to assign a page to just Computers then you should be able
to point it to 

Computers, 

Original issue reported on code.google.com by [email protected] on 21 Jun 2009 at 6:28

Obscure FK Change problem

I am seeing on one of my proprietary apps a small number of occasional
mistakes in SUM amounts. They are fixable by issuing an UPDATE table set
_agg='C'. I am also seeing cases where FETCHes do not appear to have
executed. I believe these are related to changed values of foreign keys
that do not correctly process the various FETCH, DISTRIBUTE and/or SUMS.
The good news is this is the only outright mistake I have seen in any
automation in 4 years!

Original issue reported on code.google.com by [email protected] on 13 May 2009 at 6:07

Column depending on itself is not a circular dependency

Description of Bug:

Andromeda detects circular dependencies in column automations.  However, if
a column has a chain that returns its own value, this should not be a
circular dependency.


Please tell us how to reproduce this bug:

table example:
    column test:
        chain calc:
            test 00:
                compare: some conditional
                return: @test
            test 01:
                return: @some_other_column

Original issue reported on code.google.com by [email protected] on 26 May 2009 at 5:15

Last Minute Search Refactor??

The search does not quite respond fast enough, and if a user types in 3-4
characters it will show results of only the first 2 they typed. As the
very last thing after all bugs are cleared up, perhaps the search should do
an incremental return, and be smarter about cancelling any existing search
if the user types a new key.

Original issue reported on code.google.com by [email protected] on 13 May 2009 at 6:15

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.