[EuroPython] Python conference software
lac at strakt.com
Thu Jan 26 20:29:41 CET 2006
In a message of Thu, 26 Jan 2006 15:56:51 +0100, Dario Lopez-Kästen writes:
>uh, this discussion, IMNSHO, is rapidly becoming non-interesting. And I
>am a bit upset at seeing the same discussion popup again and again.
>So, can we please just decide on what software to use? I am not
>interested in, at least as I perceive it, this covert blame-game that is
>currently going on.
This is not blame. This is a complete and unreconciable difference of
opinion as to what is desirable in the way to do work. There is an
important difference. There is a reason why Python has dozens and
dozens of webframeworks and workflow systems. People who start off
using any one often end up really unhappy because the design of the
system does not match the way that they want to structure the work.
And when they get really unhappy, they go off and write their own
>What happened, happende, let's not let it happen again. The problems we
>had we had were had for specific reasons - most of them had nothing to
>do with the technology.
They had everything to do with the technology, and whether the
technology made it possible for them to do what it was they wanted to
do in the way that they wanted to do it, or whether some things
'just couldn't be done' or 'were easier to do manually' and so on
and so forth. These are the very working conditions that make people
decide that one technology does not suit their needs and go get, or
write another one.
But given our diverse community, it is not surprising that we cannot
find a fit which suits everybody. Consider CherryPy, to pick
something that we aren't using, and as far as I know are not
considering using, so should not unduly stress people out. I have
heard both of these comments about CherryPy's likeness to PHP. It is
'a major strength that allows people to work in ways they prefer and
enjoy', or 'an abomination, like the PHP it resembles, that makes it
working with it an intolerable experience'. There is no hope in
getting these two reviewers to meet in some happy middle. You might,
if you worked hard at it, create a work experience that both of them
dislike, but pleasing them both is impossible.
There is a reason why we have so many webframeworks, and that does
not reflect badly on us. People really, really, really do care about
how they work, and really prefer to do things in ways that other people
hate. Indeed, the same feature _often_ works that way.
>Bottom line: is the software ready or not?
>We need NOT to integrate the website with INFORMATION on the conference
>with any other conference reg.system, automated or manual. I've managed
>larger conferences that EPC without any automated systems. ust a website
>with info (edicted with a texteditor).
What is relevant is whether the people who plan to manage EP2006 want
integration, most definitely do not want integration, or do not care.
>So, how much time do folks that have interests in EPC using their
>software need to set up working environments so that the rest of us can
I think that, should there be any evaluation to be done, only the
people who have already planned to do a lot of work in organising the
conference should do it. Otherwise we risk getting a system based on
the needs and desires of irrelevant people.
So if CERN is full of people who are already familiar with the
CERN system and prefer to use it because it is what they know and are
familiar with, then that would be reason enough for me to conclude
they should use it. My problem is that so far I haven't heard from
any knowledgable CERN system people. The people who were pushing
it, were doing so, despite being unfamiliar with it. This is
Proof by 'it must work, lots of people have used it'.
But if there is one thing we have figured out, in enormous detail, is
that 'working for lots of people' does not imply 'working for you' or
even 'having what you would consider the most elementary and
necessary features implemented'. People implement features in
accordance to what they consider 'essential' and the variation
among different tools is large.
>Perhaps there are more people than just Strakt and CERN interested in
>having their software evaluated?
>We need to have a decision on this soon. And please - let's keep the
>requirements at a sensible and practical level.
This is the problem. What one person considers 'an unreasonable
requirement' somebody else considers 'an essential requirement'.
>Is 10 days enough time for the interested parties to set their systems up
What I would like to see is a discussion, or a report, or something
from the CERN Europython organising team as to what it is that they
want, in freatures, usability, whatever-. Now it may be that they are
feeling inhibited in discussing it, because they would in general
discuss this in French. I'd say, discuss or post it in French. I'd
just like to make sure that you know what you want, ahead of time,
and, should you prefer to use one system or another, that you have
some reason for using it beyond 'it was developed here and we feel we
have to use it' or 'why not?', or 'somebody told me it was cool'.
Because I have already gone through the 'we should use some system
because somebody else likes it' route. And the 'I never thought about
it much, but surely any system would work in some reasonable way, they
way I consider reasonable ... it ought to have the appropriate
capabilities, even though I have never used it, just because' route.
This is an invitation to an unbelievable amount of stress. You can
make yourself sick over this. I want to spare you all this agony.
I really and truly believe that the conference organisers should use a
system they like and enjoy using. If you don't have a candidate,
possibly because you have never used a system and you want to go try
the Strakt system, to see if you like it, then we can set something
up. If you already have some other system you know and love, this is
fine with me too.
The other thing this report would be useful for is to locate: 'who
besides Benedikt is organising this thing'? Even with track chairs,
Europython is way to much work for one person to handle the 'onsite
preparation details'. If Benedikt needs help, then we need to hear
about this as soon as possible.
take care all,
More information about the EuroPython