public inbox for development@lists.ipfire.org
 help / color / mirror / Atom feed
From: Michael Tremer <michael.tremer@ipfire.org>
To: development@lists.ipfire.org
Subject: Re: Add "reinstall core update" button to Pakfire?
Date: Thu, 12 May 2022 10:20:49 +0100	[thread overview]
Message-ID: <97290638-3ADA-4036-924B-11A59D84DA31@ipfire.org> (raw)
In-Reply-To: <bdd41c67-957c-c839-0419-748c67dcf219@leo-andres.de>

[-- Attachment #1: Type: text/plain, Size: 3041 bytes --]

Hello Leo,

> On 11 May 2022, at 13:18, Leo Hofmann <hofmann(a)leo-andres.de> wrote:
> 
> Hi all,
> 
> I saw this comment in the Pakfire functions library:
> https://git.ipfire.org/?p=ipfire-2.x.git;a=blob;f=src/pakfire/lib/functions.pl;h=d4e338f23ae8ae97d6f18c6d8890d13463dc5d30;hb=refs/heads/next#l762
> 
> And I remembered that sometimes people ask how to reinstall an update.
> For example because they have changed from the "testing" tree back to "stable" and want to get the final release version of the update.
> 
> Would it be worthwhile to add a "reinstall core update" button to the web interface? Or would that lead to people breaking stuff?
> 
> Is decrementing '/opt/pakfire/db/core/mine' still the recommended way to do that?
> https://wiki.ipfire.org/configuration/ipfire/pakfire/testing

Technically, that is all that needs to be done, and we semi-automatically do this when people change from one branch to another one. We then re-install the latest version of the last update which is just a shot in the dark to keep systems somewhat close to the releases - which is always a little bit difficult in a test environment.

However, I am not in favour of making re-installing this too easy, because you can just see some of the outfall in Adolf’s emails from yesterday. It is generally not a supported operation. However, it can be helpful in very few cases. But only very few.

I would also say that on the web UI, we should be very careful about giving people the option to “press a magic button and everything is alright”. We somehow introduced that with the fsck button which I do not think helps anyone that much really, because a healthy filesystem does not manual checking (it is not Windows 98 where you have to “defragment” things from time to time) and if it does, it is normally smart enough to figure that out by itself. If the user has any reason to believe that their filesystem might be corrupt, the filesystem should have noticed that earlier and try to fix itself, or you are in territory where you will have to replace your storage device and re-install the whole OS.

Right now, I do not mind having that button - it does not do much harm after all. But I thought it was a good example to illustrate that just because we can very easily give people the option, we should ask ourselves why people would need this in the first place.

Regarding re-installing an older Core Update I am thinking:

a) The user is testing something and things didn’t work out. Going back isn’t always possible but it might work. In that case, those people should know how to do this. Maybe a command line option is convenient. I wouldn’t object that.

b) You have reason to believe that your system was compromised. Re-installing the latest Core Update does not fix that. Here is where I consider a button on the web UI dangerous. Just pressing that does not fix anything.

Did I overlook any reasons why this option should be there?

-Michael

> 
> Best regards
> Leo
> 


  reply	other threads:[~2022-05-12  9:20 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-05-11 12:18 Leo Hofmann
2022-05-12  9:20 ` Michael Tremer [this message]
     [not found] <88d84c753abc60a6d8655a479c05bfac@leo-andres.de>
2022-05-13 10:19 ` Michael Tremer

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=97290638-3ADA-4036-924B-11A59D84DA31@ipfire.org \
    --to=michael.tremer@ipfire.org \
    --cc=development@lists.ipfire.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox