Re: ok, maybe i won't find any devs to back akonadi out of kde-pim after all

[ Comments below, in-line ]


On Wed, 21 Mar 2018 08:21:47 -0700, Paul Vixie wrote:
> 
> 
> Pablo Sanchez wrote:
>> ...
>>
>> The operative word is *constructive* If people cannot provide
>> constructive feedback/truth, then they should not bother. General human
>> behavior will shunt the rest.
>>
>> Your past post to Marin was not constructive.
> 
> i apologize, and i resolve to do better in that regard.

I appreciate it Paul.

> i don't know what to say when folks deny that there is a problem,

I spot-read Martin's post (it's on my TODO quue to grok) and in all
his posts he's not denying there's a problem.

In fact, we had a sidebar conversation early on where he said he would
jump in the thread because he felt my initial response was too
performance-centric.  He wanted to ensure people understood that *we*
all acknowledge there are problems.

In short, rest assured that Dan, Martin, Sandro and I are absolutely
not denying there's a problem.

::: Carp about me :::

I'm going to toot my horn a little because I think it may be important
for people to understand my background and my plans to contribute.  I
really don't like doing this because I like to let my work do the
talking ...

I'm semi-retired.

Before that, I had my wee little company (see .sig) that (still)
provides/ed database engineering services.  Those services were
minimally:

o Scalable Database Architecture:  Oracle, SQL Server, MySQL, PG, etc.
o DBA'ing - although generally I felt this was wasting my client's $'s
   as I charge(d) a premium.
o System Admin'ing:  primarily Linux
o Scalability Testing
   - Using a Load Harness, generate a synthetic load
   - Instrument the server stack:  Linux, UNIX or Windows
   - Monitor the DBMS
   - Provide actionable items to the client
o Performance Tuning
   - Database
   - O/S
   - Storage
o SWAT team engagements
   - Dompanies who were losing millions of dollars at quarter end would
     engage my services to resolve performance issues.
o Software development:
   - Performant SQL
   - Program Units:  Stored Procedrues, Functions
   - 'Se habla' vendor SQL:  PL/SQL, T-SQL, etc.
o // other stuff to pay the bills ... //

My client list ranged from the small mom-and-pop companies to large
companies with revenues near 200 billion US.  They all paid the same
rate.

::: Why contribute? :::

My Host O/S is Linux and has been for many years.  I run KDE as my
Desktop Environment.  I have certain tools which only run on Windows
and they're virtualized.

I like to think I'm a practical person so while I love Linux, Office
simply kicks the snot out of LibreOffice.

As a side note, the ERD I created in one of the tickets was done with
my own ERD tool - I bought this software many years ago.  It only runs
on Windows.

However, the majority of the tools I use are on UNIX/Linux.  Once I
semi-retired, I felt the need to give back to the Open Source
community.  The community whose software I used in order to be able to
become semi-retired.

::: What I bring to the Project :::

For kmail2/akonadi, I am using the same techniques I used to find
performance and stability issues for my clients.  What business
clients would pay a lot of $'s for ...

It is very early days on the project.  The first things I found are
egregious performance issues.  These are good finds because they're
easy to determine what is the fix.  In fact, Dan knows.  I suspect I
just confirmed it.

As we're currently in a performance shadow from these problems, we all
understand that I am on hold until they're fixed.

Once they're fixed, I'll resume.

My plan is not only to do SQL tracing.  I plan to run high-resolution
sar captures during the work.  This means that I'll be able to ensure
a wee bit that the new bits will (become) performant even on older
hardware.

> if there are only a small number of people who understand the code
> well enough to fix things, and those people have jobs and families
> and lives, and so are not able (over a period of some years) to
> invest their volunteer efforts, that's understandable, and we still
> owe those people a giant debt of gratitude for whatever they've been
> willing to do.

This is going to sound really stupid of me.  However, I'm old enough
to no longer care how dumb I sound.  :p

I used to think there was a platoon of people who were working Open
Source X.  For example, kmail2.

In fact, it seems there aren't.  There's Dan and Sandro.  (Martin can
tell me how wrong I am!)

There seems to be a platoon of reviewers so that's good.

> i don't think any of them can afford to interrupt their careers with
> a one-month contract to fix akonadi, and that leaves me powerless to
> affect my own outcomes, other than to invest time in
> workarounds. this is frustrating, since when kmail worked for me, it
> was the best thing ever.

I can't speak for Dan and/or Sandro but for me, as I said, I'm
semi-retired.  This project is near and dear to me.  I have committed
to it.  Once I make a commitment, I'm in for the long-haul.

I have been tapped on the shoulder to do some 'other' work (paid-for)
but the CTO knows that kmail2/akonadi is on my plate.  In fact, he
respects that what I'm doing so I'll have time for kmail2/akonadi and
the other 'work.'

If you read this far and are not bored out of your skull, thanks for
reading.

I hope to never have to 'toot my own horn' again ... I feel ... oily.
:p
--
Pablo Sanchez - Blueoak Database Engineering, Inc
Ph:    819.459.1926        iNum:  883.5100.0990.1054