October 2002 - Cosmo-torun

Re: interesting cosmo CMB software?
by Michal Frackowiak 29 Oct '02

29 Oct '02

On Mon, 2002-10-28 at 19:42, Boud Roukema wrote: > What do people think of: > > http://arcturus.mit.edu/cosmics/ > It was the first free code for calculating density peturbations with an emphasis on CMB fluctuations. CMBFast is based on it. But it calculates perturbations in linear regime using full set (not full, but cut anyway) of Boltzmann equations - as opposed to some radical approximation by CMBFast. The inconsistency however is only about 1-2%, but required comp. power is MUCH bigger. In my oppinion there is no point in dealing with this soft - CMBFast is much faster and gives you practically the same accuracy. It also has more features e.g. grav. lensing, reionization. You would need a supercomputer to run COSMICS. And just a PC to run CMBFast. regards - michal

1 0

phenomenological
by Boud Roukema 23 Oct '02

23 Oct '02

Hi I'm having an interesting (random) discussion with Sebastian, and pointed out some nice galaxy formation pages at the IAP: http://brunello.iap.fr/galics/GalICS/main_frames.html There's a useful word related to these: "phenomenological" Its meaning is somewhere in between "theoretical" and "observational", i.e. finding formulae which are reasonable from a theoretical point of view, but not directly justified, and which can be "fit" to observations. My guess is the constraints we have on Omega_m and Omega_Lambda could probably be described as phenomenological, since although they are very directly linked to GR, which definitely counts as theory, they're not (yet) linked into the rest of physics except by a range of "reasonable" hypotheses. The challenge is to get these solid observational and/or (preferably "and") theoretical support... pozd boud

1 0

PROPOSALS su/web/TWiki/backup/quotas DEADLINE TUE 22 OCT 00:30
by Boud Roukema 14 Oct '02

14 Oct '02

Witam, The deadline passed and there were several useful comments, onlist and offlist - thanks everyone. Here are some updates and specific proposals. ########## PROPOSALS su/web/TWiki/backup/quotas V0.1 ################## ### deadline for comments/veto Tue 22 Oct 00:30 ### (1) superusers - There is now one staff member and one student who have been suggested and who have accepted, in principle, becoming superusers. I would be happy for both of them to be su, and think 3 people is sufficient. (Other people can become ordinary users, of course, and since there is plenty of scratch space, that's enough to develop experience in GNU/Linux, IMHO.) Anyone who wishes to know their names please contact me offlist. (2) web server - I've (tried to) install the default apache server of Deb3.0, and if noone objects, would like to put it online once the decision deadline passes. Are there any recommendations for firewalls etc, or good Deb /other linux links/hints for this? (3) (T)Wiki server for * cosmic topology page - complete references to all published articles (last time I counted, there were only 36 authors during the 20th century + 1 in the last year of the 19th century) * standard ruler page * meta-page for other cosmology stuff * galaxy/AGN formation/evolution pages If there are no objections, I'd like to try installing this quickly, since I think it's a key piece of software for working together efficiently. (4) backup strategy? Initial backup strategy: - daily hard disk backups (incremental) of /home - weekly hard disk backup (full) of /home + backup of same on to CDROM(s) (max 3 CDROMs since 2GB partition) (5) initial quotas: - 200MB per user on /home - 2GB per user on a /scratch disk (6) - mail server? For the moment seems safer *not* to run adjani as a mail server, and there's no obvious need anyway. So I propose to *not* run a mail server, at least for the moment. Any comments (or veto, if for a fundamental reason of what we are here for) on (1)-(6) welcome by the deadline. ##################################################################### ########## PROPOSAL for machine installation V0.02 ################## ### deadline passed ### > - Deb 3.0r0 or RedHat ? > I would like to install Debian 3.0r0, with the 2.4.18bf kernel > flavour. This includes software like SnapPea which is useful for ... There was one strong recommendation in favour of Deb 3.0r0 :) from Michał Frąckowiak. > - superusers - I think that Andrzej M and I should be superusers, but For security reasons, it's probably better to discuss this more offlist in terms of individual names. There was one objection to having a student or two as su, but the same person agreed that if it's someone we know and trust, who knows GNU/linux, it could be OK. Another opinion was in favour of having a student or two as su. See above for a specific proposal. > - name of machine (silk, sahni, liddle, peebles, novikov, levin?) Andrzej pointed out that the "dead astronomer" tradition is really just for the solaris machines. Michał F had a bunch of suggestions which show that we have different personal tastes ;), and was not enthusiastic for using the name of a dead (or live) cosmologist. But since this is a question of taste anyway, I finally chose "Adjani" after Isabelle Adjani :). > - partition suggestions (there are 2 disks each of 60Gb) Geniu and Michał H made several useful comments - thanks. > My idea of 650MB /home is that CDROM backups should be relatively easy. However, given that disks are so big today, should we create much more backed up home space? E.g. using hard disk backups on a daily basis + tape backups once a week? Or daily hard disk + up to 3 CDROMs once a week (I don't think I would have more patience than copying 3 CDROMS, even that would be pretty boring, done every week)? 3 CDROM => approx 2GB => 200MB for 10 users. Probably plenty for the next 12 months. See proposal above. In any case, the only decision required immediately is how big the partitions should be. And of course, there could be parts of a scratch disk which are linked with ln -s from the real /home in order to look like home disks so that users know/expect that they are backed up. ######################################################################

1 0

PROPOSAL for longer term policy on new computer V0.01
by Boud Roukema 08 Oct '02

08 Oct '02

########## PROPOSAL for longer term policy V0.01 ######### ### no deadline for the moment ### - susers - boud + amr? + a student? - web server? - home space per user - h.m. MB? With 650MB on /home, and potentially a dozen people, that's only about 50MB each. This brings up the question of... - backup strategy? - scratch space? Probably not much to say here - except the question of whether the observers can use this for radio data while waiting for us to find other cosmo uses for this disk space. What sort of policy could be reasonable? - mail server? - (T)Wiki server for * cosmic topology page - complete references to all published articles (last time I counted, there were only 36 authors during the 20th century + 1 in the last year of the 19th century) * standard ruler page * meta-page for other cosmology stuff * galaxy/AGN formation/evolution pages - sympa vs mailman? Sympa running at the moment on "astro" is fine, but I'm just wondering... - other software? CMB stuff? specific packages? N-body stuff for galaxy formation? - standard CPU rules - more than 5 min => nice 19 - quotas for accounts and/or scratch space, decision-making on acc and/or scratch space? ######################################################################

1 0

PROPOSAL for machine installation V0.01
by Boud Roukema 08 Oct '02

08 Oct '02

########## PROPOSAL for machine installation V0.01 ################## ### deadline Thu 10/10/02 19:00 (decisions a matter of taste, IMHO) ### - Deb 3.0r0 or RedHat ? I would like to install Debian 3.0r0, with the 2.4.18bf kernel flavour. This includes software like SnapPea which is useful for cosmic topology, and also gives the users (cosmologists) maximum control and freedom to manage the software. I don't imagine we could get easy consensus on Deb vs Redhat vs Mandrake (or other) ;), but please make any comments you feel like anyway. - superusers - I think that Andrzej M and I should be superusers, but at least one or two other people than me should also be su, e.g. a student or two. This is really a more serious and longer term decision, but if anyone wants to help install, now is the time to speak up, since I'm eager to get started... :) - name of machine (silk, sahni, liddle, peebles, novikov, levin?) In the tradition of TCfA, I guess we should pick the name of a famous astronomer. So for this machine, it should be a famous cosmologist. IMHO, it would be fun to name it after a *living* cosmologist, and preferably someone who's also a cool dude ;). Since this is a question of taste, I'm not sure how we should decide, and probably if I make an authoritarian decision, it would be reasonable since it won't really affect other people that much. But suggestions welcome :). At the moment the names I've thought of include: * liddle * silk * peebles * sahni * levin * padmanabhan * novikov - partition suggestions (there are 2 disks each of 60Gb) / 185MB /boot 15MB /tmp 150MB /home 650MB /usr 3GB /usr/local 2GB /var 1GB /var/mail 1GB /home2 2GB /scratch1 10GB /scratch[2-9] 10GB -> 80Gb /scratch10 20GB My idea of 650MB /home is that CDROM backups should be relatively easy. Any good ideas for partition? ######################################################################

1 0

PROPOSAL for a decision-making method for new computer V0.01
by Boud Roukema 08 Oct '02

08 Oct '02

Witam wszystkim, A computer has been bought for the TCfA cosmology group! :) This is a big investment and I would like to try to apply modern management methods - specifically, consensus decision-making techniques - for deciding on how to use and run it, rather than just one or two people deciding. Please comment, criticise this proposal. (If you think it's fundamentally against what we are here for, you could argue to veto the proposal itself!) Next message: concrete proposals for the machine. boud ########## PROPOSAL for a decision-making method V0.01 ################## First, a reminder on the AIM of our group: * research and education-leading-to-research in observational cosmology * Second, background reading on what consensus is and why it's effective: flow chart: http://www.consensus.net/flowchart.html word description: http://www.consensus.net/ocac2.html Third, a simplified consensus method: DEFINITION: "anyone" is defined here as undergrad students, masters/PhD students, postdocs or faculty studying or teaching at TCfA/CAMK-Toru� in cosmology related fields. (I will put a proposal for "cafe+" type meetings on http://www.astro.uni.torun.pl/sympa/cosmo-spotka/2002-10/ soon...) PROPOSALS: WHO - anyone (the proposer) can make a proposal for use of the computer, but hopefully it should be someone willing to do the work once the decision is made! HOW - the proposal is made on the most appropriate mailing list, e.g. cosmo-torun HOW - anyone can make comments, constructive criticisms, to try to improve the proposal HOW - the proposer writes successive versions of the proposal based on the comments DECISION - after discussion seems to have converged, the proposer (or other moderator) sends an email to the same list with Subject: PROPOSAL <name of proposal> DEADLINE FOR COMMENTS/VETO <deadline>, where the deadline is X days after the date of the email. X = 2 if the decision is essentially a matter of taste X = 7 if the decision is more serious X = 1 if there is some unexpected really urgent decision to be made STEPPING ASIDE - anyone who is against the proposal, but accepts that his/her criticisms have been listened to and are not important enough for a veto, can "step aside", and these criticisms are added to the proposal. This way, others are aware of the (claimed) weak points. VETO - anyone can veto the decision, but only based on the aim of our group (see AIM: above), i.e. not for trivial reasons. He/she should explain the reason for the veto. #######################################################################

1 0

Is DE related to CDM?
by Boud Roukema 01 Oct '02

01 Oct '02

Witam What if DE (dark energy) is not something from early universe physics but instead is directly related to the properties of the CDM (cold dark matter) particles as some sort of negative pressure? Two articles on this idea: http://de.arXiv.org/abs/astro-ph/0209584 Schwartz http://de.arXiv.org/abs/hep-ph/0209322 Gondolo \& Freese (Varun Sahni et al. had some ideas earlier - last 12-18 months - on how DE and non-baryonic DM could be the same thing, but I think it was just phenomenological. Though I'm not sure.) pozd boud

1 0