Re: Health warning:

I know a database whose name begins with O that does this.

the user types:
*psql -U postgres *

response:
*PROBLEM: psql: error: could not connect to server: FATAL: Peer
authentication failed for user "postgres"*

*SOLUTION: change "local all all peer" in file
/etc/postgresql/12/main/pg_hba.conf *
*to "local all all trust".create password then change back.*

Postgres database is better than the Big O. we agree.
Then we all become multi-millionares :)

On Tue, 17 Dec 2019 at 19:47, Zahid Rahman <zahidr1000(at)gmail(dot)com> wrote:

> The original problem is same as
>
>
> https://stackoverflow.com/questions/40354551/pg4admin-401-unauthorized-error
>
>
> https://www.postgresql-archive.org/401-Unauthorized-on-first-run-of-pgadmin4-td6029337.html
>
>
> https://superuser.com/questions/1322086/how-to-fix-the-pgadmin-4-v3-0-port-number
>
> I have done two clean operating systems and postgres pgadmin4.
>
> 16 emails but not one response to the problem I logged for two days.
>
> I guess you all like to argue.
>
>
>
>
>
>
> On Tue, 17 Dec 2019, 18:47 Justin, <zzzzz(dot)graf(at)gmail(dot)com> wrote:
>
>> HI Zahid
>>
>> After 16 emails I think we are getting close to CONSTRUCTIVE criticism
>> and to a valid complaint, that needs to be addressed. The point being made
>> is, the install scripts need some validation code on paths, additional
>> dependency checks, and better documentation on how the install process
>> works for those that do not know APT, etc......
>>
>> Take you points, document why the install went sideways/failed , what was
>> done to fix it, and how these changes will fix the problem going forward..
>> I'm sure this will be taken care of...
>>
>> AFter 16 emails no one knows what is wrong with the Installer, all we
>> know
>> 1: it sucks,
>> 2: see this video for proof,
>> 3: see this link that agrees ,
>> 4: GUI is better
>> 5: No one on pgadmin team can program,
>> 6: now for the stern lecture on good programming ..
>>
>> Do you see the breakdown in communication? There is no substance/action
>> item with the complaint. Its just a complaint.
>>
>>
>>
>>
>>
>> On Tue, Dec 17, 2019 at 12:22 PM Zahid Rahman <zahidr1000(at)gmail(dot)com>
>> wrote:
>>
>>>
>>> > But how does a GUI installer fix an issue with a path, typos, missing
>>> dependencies, or mistakes the install scripts? Is there a Linux GUI
>>> installer that works across all the flavors of Linux, what about BSD, or
>>> MAC,
>>>
>>> The answer is simpler then you think.
>>>
>>> Look at this code . Even the though the variable is declared as
>>> Datasource with an if statement it is double checked for validation and
>>> constraint.
>>>
>>> if (value instanceof DataSource)
>>>
>>> I call it VALIDATION and CONSTRAINT code.
>>> As application developers we provide drop lists where we can so we
>>> don't risk the user doing a typo. That's CONSTRAINT code. You can do same
>>> with command line argument. check for argv[1] content , don't just start
>>> execution based on the assumption the user typed correctly. I know the
>>> validation code will probably be 5 times more than the FUNCTIONAL code.
>>> That is necessary for good software.
>>>
>>> VALIDATION / CONSTRAINT code will be far greater than functional code.
>>>
>>> an example:
>>> Don't just open a file .
>>> Don't assume.
>>> Check if the file is open first. That's validation code.
>>>
>>> Check for path values to see if they exist , check for operating system.
>>> For example when you go to download page. You should go to download page
>>> appropriate for your operating system. That is a result of a check being
>>> made.
>>>
>>> There is no reason why a GUI installer cant work across all *nix . Just
>>> validate the operating system then call the relevant source code.
>>> There is no magic in programming. What will happen is exactly what you
>>> wrote will happen after you validated. If you did not validate, just
>>> started doing then that is like crossing the road without looking left or
>>> right. The user is the one who is getting run over by bus. That is to say
>>> the user has been thrown under a bus because the programmer did not look
>>> left ,right , listen.
>>>
>>> This is my own personal view.
>>> There are three components which make up good code or deliverable
>>> product.
>>> FUNCTIONAL code.
>>> VALIDATION code.
>>> CONSTRAINT code.
>>>
>>> I am always and sometimes for fun checking for functional code and
>>> validation. Once in a while one does come across software as a programming
>>> tool or an API which has all three components.
>>>
>>> In your suggestion (MAC/BSD) check for OS type then act appropriately.
>>> that's VALIDATION code. Think of functional code without validation or
>>> constraint code as inappropriate or out of control behaviour.
>>>
>>> Inappropriate behaviour warrants a health warning :) don't you think
>>> ?
>>>
>>>
>>> On Tue, 17 Dec 2019, 16:34 Justin, <zzzzz(dot)graf(at)gmail(dot)com> wrote:
>>>
>>>> Taking this off this list. way of topic for Postgresql Novice
>>>>
>>>> You are right testing needs to be done on a clean install and done on a
>>>> not so clean systems...
>>>>
>>>> But how does a GUI installer fix an issue with a path, typos, missing
>>>> dependencies, or mistakes the install scripts? Is there a Linux GUI
>>>> installer that works across all the flavors of Linux, what about BSD, or
>>>> MAC,
>>>>
>>>> Pathing, system settings, and dependency hell aka DLL Hell have been
>>>> around for decades, no one has come up with a solution to this problem,
>>>> The latest attempt is to just hide the problem, wrap everything in Virtual
>>>> Environment, Docker, or something. The problem does not go away it was
>>>> just morphed into a new set of problems.
>>>>
>>>> Consider the language GO, it compiles all its dependency into a
>>>> single BIN/EXE file, NOT a new idea, Xbase languages did that years
>>>> ago,
>>>>
>>>>
>>>>
>>>>
>>>> On Tue, Dec 17, 2019 at 11:04 AM Zahid Rahman <zahidr1000(at)gmail(dot)com>
>>>> wrote:
>>>>
>>>>> The simple solution is that who ever is testing the pgadmin4
>>>>> installation , if they did it on a clean machine , it would save alot of
>>>>> pain.
>>>>>
>>>>> When I worked in a software house I followed this practice every time.
>>>>> For every project I did a clean install. I know it is time consuming. But
>>>>> it is a best practice needs to be followed.
>>>>>
>>>>> Because I knew I may have a setting or software installed from
>>>>> previous project , so the user environment would be different from
>>>>> development environment.
>>>>>
>>>>> Linux still has problems. Fedora 31 has wifi driver but ubuntu doesn't
>>>>> so I am using USB wifi dongle with my new laptop.
>>>>>
>>>>>
>>>>> On Tue, 17 Dec 2019, 15:46 Justin, <zzzzz(dot)graf(at)gmail(dot)com> wrote:
>>>>>
>>>>>> Zahid
>>>>>>
>>>>>> Open Source means Free Testing and Free code writing, we all
>>>>>> benefit. You can write GUI pgadmin4 installer for Linux
>>>>>>
>>>>>> We welcome CONSTRUCTIVE criticism throwing up a link, or snarky
>>>>>> youtube video then saying "see I'm right", will not get traction with
>>>>>> people it turns them off. Remember we are doing this for FREE
>>>>>> Specifically how does EDB page support your position???
>>>>>>
>>>>>> Enterprise DB maintains the windows installer as a give back to the
>>>>>> Community and promote Postgresql.. Postgresql installers have come a long
>>>>>> way over the years, I remember a time when everything on Linux was
>>>>>> compile then install what a horrible amount of pain, Just getting the CD to
>>>>>> boot to install Linux could take days because of hardware issues.
>>>>>>
>>>>>> When writing criticism first step, Document the issue, and be
>>>>>> specific, take screen shots, note the issue clearly then post it..
>>>>>>
>>>>>> Postgresql has several ways to note and document issues, Mailing
>>>>>> Lists, Wiki, IRC, create a Blog Post and post it on
>>>>>> Planet.postgresql.org
>>>>>>
>>>>>> 99% of linux installers are done from the command line, and yes it
>>>>>> goes horribly wrong at times.
>>>>>>
>>>>>> On Tue, Dec 17, 2019 at 10:33 AM Adam Lipinski <ALipinski(at)exactal(dot)com>
>>>>>> wrote:
>>>>>>
>>>>>>> Zahid,
>>>>>>>
>>>>>>> Please contribute, your view is important. Quoting a commercial
>>>>>>> enterprise and how they choose to spend their development resource is a
>>>>>>> moot point.
>>>>>>>
>>>>>>> Best Regards,
>>>>>>>
>>>>>>> Adam
>>>>>>>
>>>>>>> ------------------------------
>>>>>>> *From:* Zahid Rahman <zahidr1000(at)gmail(dot)com>
>>>>>>> *Sent:* Tuesday, December 17, 2019 11:29:41 PM
>>>>>>> *To:* Gerald Cheves <gcheves(at)verizon(dot)net>
>>>>>>> *Cc:* pgsql-novice(at)lists(dot)postgresql(dot)org <
>>>>>>> pgsql-novice(at)lists(dot)postgresql(dot)org>
>>>>>>> *Subject:* Re: Health warning:
>>>>>>>
>>>>>>> By support you mean I should do free tests and let you know where
>>>>>>> the problem are.
>>>>>>>
>>>>>>> Anyway alot of talk but fail to address the issue , why a company
>>>>>>> "EnterpriseDB" has provided installer. The fact I have not received to
>>>>>>> my original query means you are all using the free installer or managed
>>>>>>> to install it with sheer luck.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> On Tue, 17 Dec 2019, 15:23 Gerald Cheves, <gcheves(at)verizon(dot)net>
>>>>>>> wrote:
>>>>>>>
>>>>>>> On 12/17/2019 10:17 AM, Zahid Rahman wrote:
>>>>>>> > Open source and community doesn't have to mean *crap* but it is
>>>>>>> free.
>>>>>>>
>>>>>>> You don't have to contribute and, likewise, you shouldn't expect
>>>>>>> support
>>>>>>> ... because, as you have indicated, it is free.
>>>>>>>
>>>>>>> - G
>>>>>>>
>>>>>>> --
>>>>>>> siamo arrivati sani e salvi
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Internet Email Confidentiality Footer: This email and any files
>>>>>>> transmitted with it contain privileged/confidential information intended
>>>>>>> for the addressee. Neither the confidentiality of nor any privilege in the
>>>>>>> email is waived, lost or destroyed by reason that it has been transmitted
>>>>>>> other than to the addressee. If you are not the addressee indicated in this
>>>>>>> message (or responsible for delivery of the message to such person), you
>>>>>>> may not copy or deliver this message to anyone. In such case, you should
>>>>>>> destroy this message, and notify us immediately.
>>>>>>>
>>>>>>