Re: [libreoffice-users] Re: How to handle regressions

On 10/2/2014 10:37 AM, Charles-H. Schulz
<charles.schulz@documentfoundation.org> wrote:
> > That said, maybe you didn't mean it as it sounded, so I'll give you the
> > benefit of the doubt...

> Forgive me if I'm intrusive, but there is something I actually do not 
> understand in your situation. If you are not comfortable disclosing this 
> by all means mail me off list: I would like to better understand the 
> situation because I feel there is some deep misunderstanding on how free 
> software projects work and of what your business problem is.

I'm happy to answer any honest and relevant questions.

> You write that you're a one person company and that your customer has 60 
> seats you are -presumably- administering in some ways. Am I correct 
> here?

Yes - I'm an independent contractor, and this is my main client.

> You claim not to have the time to test the build on 60 machines - sure, 
> I get this, but then test it on two machines.

I didn't say I didn't have time to test it on one machine until only
yesterday and this morning.

What I had been trying to make clear was that there was nothing to test
until fairly recently, and even then I don't think there any
downloadable builds (if there are/were, it wasn't obvious in the bug
comments).

> You call the bug in 
> question a major regression, and forget that the people providing 
> quality assurance are indeed volunteers. They either catch a regression, 
> or they don't.

Yes, but ...

a) surely you aren't denying the fact that many - most? - of the
Libreoffice *developers* - especially ones working on core functionality
- are actually *paid* coders, are you?

and

b) surely you aren't suggesting that the developer(s) doing the actual
coding, whether they are volunteers or not, don't have *any*
responsibility for doing *basic* testing of new features - especially
for features that will also be *removing* the old, long established
method the new way is intended to replace (in fact I would suggest that
they should have a *lot* of responsibility in this regard, and that it
should be a major part of their developer agreement they sign when being
given commit access)?

Cut/Copy/Paste into fields is certainly what I would call very,
*extremely* basic functionality that users would expect to actually
work, and apparently - as evidenced by the very existence of this bug -
they *didn't do that*.

> If they don't, automated tests can catch it, or won't. If it were
> major, I am sure it would have already been patched.

<sigh> Inability to cut/copy/paste from/into fields is a *major*
regression - for anyone who uses them.

> Now back to your
> situation. You do not have the time to test the build, no time to do
> quality assurance,

Never said that - but as you have so aptly pointed out above, it is
impossible to test everything... BUT...

It is *much* easier for a developer who is coding a very specific
feature to test *that specific new feature*, than it is for a user to
try to test *every* *single* *feature* of a software like Libreoffice
before each and every release.

Do you not see this? Is this not so obvious as to almost knock your head
off?

> but if I'm correct you are selling something to your customer that
> involves LibreOffice, aren't you?

Nope. I'm an I.T. guy. I manage this clients network, servers,
workstations, phone system, etc.

Installing, managing the installs, and helping users with using
Libreoffice effectively (started with Openoffice many years ago) is only
one, very tiny part of what I do here.

> The point I'm getting at is this one: if you are a professional 
> distributing LibreOffice, that is great, but you must be something else 
> than a user on a user list. You must at least have some expertise, and 
> at the moment, I don't see you being anything else than a user who has a 
> bug but will not test a patch because of whatever reason I will not 
> judge.

The bug was reported on March 24th, confirmed on the 28th (and again by
Sophie on April 15th).

I ran headlong into the bug in early July, found the open bug and
commented on the severity, and asked for consideration of my enhancement
request to bring back the old behavior (even if only as an option).

Lots of follow up comments confirming the bug in July.

Joel asked if it was a regression, making it clear he didn't read the
comments (the comments make it very clear that it works fine in 4.1.x).

A patch was provided almost a month later - so, only a month and a half
ago - saying it was pushed to 'master' (whatever that means - we users
are NOT developers), with no instructions on where to download test
builds, and no requests to test it.

A month later the OP asked if the 4.3 series would be patched or if we
would have to wait until 4.4, and he was rudely (imnsho) told by Joel to
'feel free to submit a patch', which happens far too often.

Who was it on here complaining about me/users not asking for patches top
be applied? "This is what happens more often than not when we do.

> By doing that,

<snipped> everything that follows since it is not applicable (since that
is notr what I am doing)...

> You have complained that this is a regression and not a bug, but
> regressions are not intentional,

No one said they were (or this one was) - but it is also plainly evident
that the developer who pushed this code into production didn't even do
minimal testing.

> I am afraid I don't understand why you're even complaining 
> :-)

Yeah, well, maybe that's part of your problem? You refuse to even try to
understand this from a users perspective.

-- 
To unsubscribe e-mail to: users+unsubscribe@global.libreoffice.org
Problems? http://www.libreoffice.org/get-help/mailing-lists/how-to-unsubscribe/
Posting guidelines + more: http://wiki.documentfoundation.org/Netiquette
List archive: http://listarchives.libreoffice.org/global/users/
All messages sent to this list will be publicly archived and cannot be deleted