Maybe this is fixed. Please check in latest version before release.

Editing pgmanage-connections.conf, click "pgManage Man Page"

Some systems, like FreeBSD, have standard locations, and @prefix@/etc/$(BINARY)/web_root/$(BINARY) doesn't work for them.

There was no error in console. if you have trouble reproducing the error try the usr.ship_track table. ask me for the customer name if you don't know.

Hi. thank you for an exciting project!

I'm not sure if it's something I'm doing, but I can't resize the results pane more than a couple of hundred pixels.

I'm on Windows 10.

Just upgraded my database from PG 9.6.2 to version 10.4 database. When I connect, the welcome screen is showing: "Connected to PostgreSQL v4.2.1", where previously it gave the correct v9.6.2

I have a SQL query which is supposed to return around 4k rows.

In the previous versions of Postage/pgManage, this is not an issue, but in the current version, some of the latter rows in the query returns as Undefined.

Am I missing a setting that must be tweaked or is this a conscious decision to manage performance when expecting big queries being displayed in the browser?

Thanks and great work as usual!

EDIT: I am using v10.0.2

You miss a requirement on Centos 7 libXss:
cannot open libXss.so.1
Solved by "sudo yum install libXScrnSaver"
Then you get:
pgmanage
copying config
/opt/pgManage/resources /pgmanage/pgmanage
got data:
/opt/pgManage/resources/app/pgmanage/pgmanage: error while loading shared libraries: libssl.so.1.1: cannot open shared object file: No such file or directory

4760 closed with code 127
Then I search for this error but not found

Web files can be installed under webroot or dataroot, but config files should go under /etc, so there has to be a separate argument for this.

Postgres 10 Sequences will open in the sql editor but not Postgres 9. It shows up in the list but not in the sql.

For the most part I like beautify but one thing that I think would improve it a lot would be if you removed the line breaks around CASE statements.

Since the editor already wraps lines, I feel like it hurts the formatting more than it helps when CASE statements are split poorly, and they almost always are with the current beautify code.

consider this line the way I wrote it:
, sum(CASE WHEN ordered like '2018-05%' THEN COALESCE(sales_qty,0) ELSE 0 END) as mo_pcs

And how Beautify mangled it:

, sum(CASE WHEN ordered like '2018-05%' THEN COALESCE(sales_qty,0) 
			ELSE 0 
			END) as mo_pcs

There was an error:

....\src\util_canonical.c:canonical: read_file: C:\Users\Alliance 100\AppData\Roaming\pgmanage\sql\secure_5432_postgres\postgres|closed\2017-10-30%2014%3A47%3A33%3A0519%3A0500%20 is a bad path. Path does not exist.

Failed to get canonical path: >C:\Users\Alliance 100\AppData\Roaming\pgmanage\sql\secure_5432_postgres\postgres|/closed/2017-10-30%2014%3A47%3A33%3A0519%3A0500%20<

When you copy from just one field in Pgmanage, it should not do any quotes nor should it have the popup. 99% of the time when you copy from just one field you just want the data by itself.

To anyone who uses pgManage, please reply either way.

We are considering dropping "Web Server" mode.

It would simplify development and allow us to do more desktop integration if we walked away from the "Web Server" mode. Although this mode is the only one where you can use pgManage from a phone or tablet, we aren't aware of anyone using it this way. Let us know if you do.

In the absence of someone speaking up, we'll likely drop "Web Server" mode altogether.

If anyone feels it's useful, we'll consider our options. We may make it an unsupported configuration but I think it's more likely we'll split it off to a new project.

please fix

This is a big deal because this was the whole reason we started a postgres admin app to start with.

There is no reason why I should have ever lost more than thirty seconds worth of work.

Please cause the interface to freeze if there is a problem saving a tab. This will prompt the user to refresh and prevent lost work.

Say I have SQL Query tab opened with 2 SQL code blocks:

SELECT
    column4
    , column3
    , COUNT(DISTINCT column1)
FROM (
    SELECT DISTINCT
        column1
        , column2
        , column3
        , column4
        , column5
        , column6
        , column7
        , column8
    FROM public.my_table
) AS main_table
WHERE 
    column4 != 'C'
    AND column7<=column8
    AND column5 NOT LIKE '09%'
GROUP BY 1,2
ORDER BY 2, 1
;

SELECT
    column1
    , COUNT(DISTINCT column2)
FROM (
    SELECT DISTINCT
        column1
        , column2
        , column3
        , column4
        , column5
        , column5
        , column7
        , column8
FROM (
    SELECT DISTINCT
        column1
        , column2
        , column3
        , column4
        , column5
        , column6
        , column7
        , column8
     FROM public.my_table
) AS main_table
WHERE 
    column4 != 'C'
    AND column7<=column8
    AND column5 NOT LIKE 'notlikehis%'
GROUP BY 1
;

When pressing the Download as a file button, the output file only has this as an output:

SELECT  

There should be a "Save" button between "Cancel" and "Recopy" that also re-copies.

I think this dialog should also be re-done based on the "Postage Options" dialog:

Received a request about the login screen:

"PgManage can be the most powerful Postgres manager on the planet but if it looks like X-Emacs people are not going to flock to it. I think we should start with:

1. Clean up initial login screen. IMO, you don't edit PGPASS, most users have no idea what that is. Instead, check for it. If it is there, load it and populate a drop down menu to select which database to connect to.

2. Off the ability to add a new connection. Add that to .pgpass.

   • This should be a form not a text file

3. Remove pgmanage-connections.conf, that seems redundant when we have .pgpass

4. If we want to offer the ability to edit .pgpass manually, that's cool but we must also offer a way to save it. For example, I can't save my changes to pgmanage-connections.conf or at least it isn't obvious how.

5. Every time we edit, we reload the .pgpass file. Why are we asking if we should reload it, or explicitly requesting a reload?"

While a long query is running if I switch tabs and the query finishes it the tab hangs. I go back to it and it still says running, and stop execution does nothing. The tab is useless until a reload.

Hey there! I am very interested in this but it doesn't work on Ubuntu 18.04.

I tried both the 10.3.0 and 10.1.1 releases and they both do nothing when I click "Log On" after filling out a connection string.

karl@deathstar:/var/log$ cat syslog | grep pgmanage
Jun  8 13:35:57 deathstar gnome-software[3665]: failed to rescan: Failed to parse /usr/share/applications/pgmanage.desktop.dpkg-new file: cannot process file of type text/plain
Jun  8 13:36:00 deathstar gnome-software[3665]: failed to rescan: Failed to parse /usr/share/applications/pgmanage.desktop file: cannot process file of type application/x-desktop
Jun  8 13:36:10 deathstar pgmanage[31898]: Failed to load module "canberra-gtk-module"
Jun  8 13:36:11 deathstar pgmanage.desktop[31898]: copying config
Jun  8 13:36:11 deathstar pgmanage.desktop[31898]: /opt/pgManage/resources /pgmanage/pgmanage
Jun  8 13:36:11 deathstar pgmanage.desktop[31898]: got data:
Jun  8 13:36:11 deathstar pgmanage.desktop[31898]: Open http://<this computer's ip>:15772/ in your web browser
Jun  8 13:37:28 deathstar pgmanage.desktop[31898]: /opt/pgManage/resources /pgmanage/pgmanage
Jun  8 13:37:28 deathstar pgmanage.desktop[31898]: got data:
Jun  8 13:37:28 deathstar pgmanage.desktop[31898]: PGMANAGE IS SHUTTING DOWN
Jun  8 13:37:28 deathstar pgmanage.desktop[31898]: 32128 closed with code 0
Jun  8 13:37:28 deathstar pgmanage.desktop[31898]: got data:
Jun  8 13:37:28 deathstar pgmanage.desktop[31898]: Open http://<this computer's ip>:15772/ in your web browser
Jun  8 13:38:13 deathstar pgmanage.desktop[31898]: got data:
Jun  8 13:38:13 deathstar pgmanage.desktop[31898]: PGMANAGE IS SHUTTING DOWN
Jun  8 13:38:13 deathstar pgmanage.desktop[31898]: 32128 closed with code 0
Jun  8 13:39:01 deathstar gnome-software[3665]: failed to rescan: Failed to parse /usr/share/applications/pgmanage.desktop.dpkg-new file: cannot process file of type text/plain
Jun  8 13:39:01 deathstar gnome-software[3665]: failed to rescan: Failed to parse /usr/share/applications/pgmanage.desktop.dpkg-tmp file: cannot process file of type text/plain
Jun  8 13:39:05 deathstar gnome-software[3665]: failed to rescan: Failed to parse /usr/share/applications/pgmanage.desktop file: cannot process file of type application/x-desktop
Jun  8 13:39:10 deathstar pgmanage[32467]: Failed to load module "canberra-gtk-module"
Jun  8 13:39:10 deathstar pgmanage.desktop[32467]: /opt/pgManage/resources /pgmanage/pgmanage
Jun  8 13:39:10 deathstar pgmanage.desktop[32467]: got data:
Jun  8 13:39:10 deathstar pgmanage.desktop[32467]: Open http://<this computer's ip>:15772/ in your web browser
Jun  8 13:39:38 deathstar pgmanage.desktop[32467]: /opt/pgManage/resources /pgmanage/pgmanage
Jun  8 13:39:38 deathstar pgmanage.desktop[32467]: got data:
Jun  8 13:39:38 deathstar pgmanage.desktop[32467]: PGMANAGE IS SHUTTING DOWN
Jun  8 13:39:38 deathstar pgmanage.desktop[32467]: 32604 closed with code 0
Jun  8 13:39:38 deathstar pgmanage.desktop[32467]: got data:
Jun  8 13:39:38 deathstar pgmanage.desktop[32467]: Open http://<this computer's ip>:15772/ in your web browser
Jun  8 13:39:42 deathstar pgmanage.desktop[32467]: got data:
Jun  8 13:39:42 deathstar pgmanage.desktop[32467]: PGMANAGE IS SHUTTING DOWN
Jun  8 13:39:42 deathstar pgmanage.desktop[32467]: 32604 closed with code 0

When autocompleting after a FROM, we should also list functions. Preferably only ones that return a set.

For example: here https://github.com/pgManage/pgManage/blob/master/src/Makefile.in#L85

I ran DROP SCHEMA CASCADE and pgmanage hung on the tree list refresh. Perhaps it thinks the schema still exists and tries to refresh it?

I do not think the CASCADE is necessary to replicate.

Having things like \N in Object properties shows an unpolished product. It should be NULL or ''

I'm running pgManage v10.1.1 on windows 10 enterprise version 1703

Autocomplete is showing functions after a tablename. in a select call

The license agreement when installing on windows has not been changed.

In the table script there should be a section that goes like this:

/*
ALTER TABLE ... DISABLE TRIGGER ...;
ALTER TABLE ... DISABLE TRIGGER ...;
etc


ALTER TABLE ... ENABLE TRIGGER ...;
ALTER TABLE ... ENABLE TRIGGER ...;
etc
*/

So you can select it, click "Open in new tab", and then write a query between them.

Received a request:

"Increase the text size of the left hand frame identifiers. At FHD they are very small. I don't even want to know what they would look like on QHD or 4k"

What are the "left hand frame identifiers"?

let me know if you need details, but for example, include header won't save to 'when selected'

I just install pgManage on Ubuntu 18.04 with PostgreSQL version 10.3

The opening screen looks like this:

Welcome To pgManage v10.1.1
Connected to PostgreSQL v7.3.0

Great looking, super fast tool to start with. Thanks.

On Linux /bin/sh is bash, but on BSDs it is not.

I clicked in the object browser and there was a hanging "getting list" popup so I decided to close and re-open pgManage.

What happened was I clicked the red "X" and got a dialog saying that it couldn't save my tab. So I clicked "cancel" because I very much wanted to save my tab manually before I quit, to save my work. Instantly pgManage quit and upon re-opening I found I lost work.

for example,

"Empty Values" won't save "(nothing)"

It makes it really hard to figure out where your tab that you just closed went.

Note in the screen shot the dates go from 5/9 to 5/1 back to 5/9.

I've been getting this error on startup ever since upgrading from 10.0.2 to 10.1.1, but removing and replacing with 10.0.0 does not fix it. After confirming the error, pgManage quits immediately.

macOS High Sierra v10.13.3

Text of message:

A JavaScript error occurred in the main process

Uncaught Exception:
TypeError: Error processing argument at index 0, conversion failure from #
at validateState (/Applications/pgManage.app/Contents/Resources/app.asar/node_modules/electron-window-state/index.js:46:34)
at module.exports (/Applications/pgManage.app/Contents/Resources/app.asar/node_modules/electron-window-state/index.js:154:3)
at App.appStart (/Applications/pgManage.app/Contents/Resources/app.asar/main.js:437:20)
at emitTwo (events.js:130:20)
at App.emit (events.js:213:7)

We want to keep the limit, but we need a popup to alert people that it is there.

Several times now I've gone to a table I know is small and it was missing a few records. Then I go to add the records and I get all kinds of weirdness. Then eventually I figure out that there is a limit.

So, ideally, set the limit to 100 records. Then the popup says, limit is 100 records, and shows two buttons, keep limit 100 or show all.

If you have a view in a schema that is in the search path, then the table design will have the drop statements for it but not the recreate statements.

It would be nice if there's a notification feature (e.g. Chrome notification) that users can enable so that they know if a query already finished even when leaving the tab.

please install under a pgManage folder instead

Strangely, the triggers are all missing the schema name so they error as well.

CREATE TRIGGER ds_trg_rcog_cname
   BEFORE INSERT OR UPDATE
   ON rcog_cname
   FOR EACH ROW
   EXECUTE PROCEDURE default_stamp_fn();

This popup slows things down, "Clipboard Settings" should be behind as few clicks as possible (i.e. next to the fullscreen button).
"Data Settings" is for limiting the data shown, but people can already do that in their query. I think it's extraneous.

The "About pgManage" popup (under "Options") has no obvious way to get out of the popup

the DROP should have come before the indexes

-- Index: usr.idx_rship_track_sales_id
-- DROP INDEX usr.idx_rship_track_sales_id;
CREATE INDEX idx_rship_track_sales_id ON usr.rship_track
   USING btree (sales_id);

-- Index: usr.idx_rship_track_shipped
-- DROP INDEX usr.idx_rship_track_shipped;
CREATE INDEX idx_rship_track_shipped ON usr.rship_track
   USING btree (shipped);

-- Index: usr.idx_rship_track_so_id
-- DROP INDEX usr.idx_rship_track_so_id;
CREATE INDEX idx_rship_track_so_id ON usr.rship_track
   USING btree (so_id);

DROP TABLE usr.rship_track01774107942391434;