[EuroPython] List of things that need doing
Dario Lopez-Kästen
dario at ita.chalmers.se
Mon Apr 4 23:18:01 CEST 2005
Jacob Hallén wrote:
> Coverage was especially bad in VR, though I think there were some complaints
> about parts of VM as well. We may expect coverage to be bad in both VA and
> VB, since they are further away from the access point, with several walls in
> between.
Right, this is expeced and fits the current level of service. There is
no warranty that Radio may reach down there due to the positioning of
the access points. There is cable based net access in two of the rooms
but they are for the speakers, not the audience. There is a planned
increase in coverage to the upper floors as well, and this MAY affect us
positively, if it gets done in time for the conference.
And like I said, we may be able to borrow an extra acess point by the
time of the conference that we may be able to use. Maybe.
I am not sure how this may increase the coverage in the rooms.
I'll ask about possible options and report back.
> If there is a switch close to the existing access point, it would be rather
> simple to set up a couple of temporary extra access points. We can buy 2
> access points and 2 100-meter ethernet cables, so we can distribute the
> access out to the places where it is most needed.
I am sorry to sound grumpy, but this now constitutes about the Fifth
time I have said that this is a) not supported, and b) will not likely
work. Everything there is NOMAD based (NOMAD is the system we use for
the publicly available wireless and cabled network infrastructure). The
NOMAD network is not designed, on purpose, to allow using NAT routers
and other devices such as hubs or switches. By the dates the conference
is being held, this policy is being more enforced than last year
(because of technical upgrades to the service).
It will not work to bring equipment to set up temporary access points.
The infrastructure is designed on purpose to prevent the kind of
temporary solution you suggest above.
You are welcome to try, it should not cause any too serious network
trouble for the rest of the NOMAD network (if it does and it is serious,
I'll probably get a nasty call to unplug the offending equipment so the
network can be brought back up).
> Also, there were quite a few people who complained about there being no place
> to go to if you needed wired access. This is probably going to be less of a
> problem this year, but after seeing that there were still qute a few people
> who used wired access at Pycon, I can see that many people are still
> dependent on wired connections.
There are several NOMAD ethernet wall sockets availabel in the
Arktitekture Light Garden (the big area to the right of the entrance).
They will lileky be spread around the walls buy we are welcome to use
them if we need them.
> In any case, we should have a few wireless cards avalable for people to borrow
> this year. We have some left from last year, and we can buy a few new ones.
oki.
>> > - We also need to get accounts created for all attendees, including
>> > accounts for on-site registrants.
>>
>>did this work allright last year? (all accounts where prefabricated by
>>the network group). If not, what needs to be improved?
>
>
> From my perspective, it worked excellently. The one thing we need to make sure
> is that we have a surplus of accounts, for the people who manage to lose
> theirs and the ones who register on-site. We should print separate slips for
> these before the conference, so that we have them ready when they are needed.
> Last year we had to cannibalise the accounts of people who never showed up.
> It worked, but we ran the risk of people showing up late and having their
> accounts gone.
Right, I need an estimate about how many people are attendind some 2-3
weeks in advance, and I will add a suitable extra amount of "emergency"
accounts (20-30 should do it, right?).
> There were only a few people who complained about having authenticated login
> to access the network. I think we can just ignore those complaints, since I
> think it is unreasonable to expect an unauthenticated service.
Well, since there is no alternative to authenticated net access, they
are welcome to complain all they want, but all we can do is ignore them,
unfortunately.
/dario
--
-- -------------------------------------------------------------------
Dario Lopez-Kästen, IT Systems & Services Chalmers University of Tech.
"...and click? damn, I need to kill -9 Word again..." - b using macosx
More information about the EuroPython
mailing list