>>>>> "Vincent" == Vincent J Carey, <stvjc@channing.harvard.edu> writes: ............ Vincent> robert's email is generated from Vincent> bioconductor-admin, but i made the reply to Vincent> bioconductor to reach the core. please advise if Vincent> changes to my mailing strategy are desired. The mails from both mailman (like this) or majordomo (like the usual R lists) come with a From: <you, the="" poster=""> Sender: <listname>-admin@.... {"-admin" for mailman / "-owner" for majordomo} To: <listname> @.... (and several other headers, in particular, mailman has half a dozen of "List-*: <url>" headers.) A reply or followup of any decent MUA (mail user agent) should produce a To: <him, the="" poster=""> @ .. CC: <listname> @ .. where it may depend which is To and which (or if there) is CC. Yes, there are bad pieces of software who do reply to "admin"/"owner". Yes, they are a real pain, particularly when I am the admin/owner :-) Martin

>>>>> "Vincent" == Vincent J Carey, <stvjc@channing.harvard.edu> writes: >> My proposal is to set up a discussion mailing list (that >> we all subscribe to) for Bioconductor. I don't think that >> we have anything to Vincent> i take it that mailing to Vincent> bioconductor@stat.math.ethz.ch constitutes even at Vincent> this point a discussion mailing list Vincent> an open question is how to structure the Vincent> communications system. bioconductor-core is Vincent> necessary for continuing threads like this one. Vincent> bioconductor-help, bioconductor-devel also seem to Vincent> make sense. I would propose to change this (current group) to "bioconductor-core" -- which I would propose to call "biocore" (the the other being too long for my taste). and make "bioconductor" a new (currently only) public list. >> discuss yet -- which is a problem. We have had four >> requests to join Vincent> i think we have a fair amount to discuss as core, Vincent> and those who are non-core may want to describe Vincent> their concerns. probably not too early to start, Vincent> but a bit too early to publicize very widely >> all with .de suffixes but there will be more. Any other >> ideas? >> Next, libraries. I am hoping to have our new programmer ^^^^^^^^^ the word is really "packages" A library is either a *.so, *.dll, *.a (etc) file or one of the [lib.loc] directories where packages are located >> (starting in Jan) learn how to test libraries and to >> handle all submissions (by doing some simple testing and >> then add them to a site like CRAN). Do we want to >> develop and start to populate the site now with the >> proviso that we will put in more stringent tests later? >> The reason for this is to gain some momentum and >> hopefully to find some more people who might be willing >> to do some of the organizational work that will be >> needed. Vincent> i am in favor of deployment soon. i will be Vincent> working on getting test-functionality equivalent to Vincent> Biobase. pointers on how to get there would be Vincent> worth writing down. >> Fritz/Kurt: is it feasible to set up a web page for >> bioconductor off of CRAN using mainly the existing >> structure and to populate it with unvetted software. I >> expect that the new programmer will be able to take this >> over in late January. So it would be the initial setup >> and population that would be required. Vincent> i hope this is possible, otherwise we need to get Vincent> sourceforge or anoncvs running locally. i would Vincent> put some effort into that if i knew that ad hoc Vincent> security considerations would not lay that effort Vincent> to waste at some arbitrary future time >> Organization: - we need someone (volunteers please take >> a giant step forward) to organize some mechanism for >> keeping track of what we are doing and what other people >> are doing regarding building packages Vincent> http://biowww.dfci.harvard.edu/~bioconductor/ Vincent> already has a little bit in this regard. we need Vincent> to distinguish between internal core-level tracking Vincent> that can be used for strategizing and evaluation Vincent> and public tracking that is a public service and Vincent> agenda publicity Vincent> i am happy to put core-generated changes into the Vincent> aforementioned web page on short notice. thank you. However, this should eventually move to www.bioconductor.org, right? <....> Martin