Re: [Rd] Option "installWithVers" seems to impact new.packages() badly?

2005-10-04 Thread A.J. Rossini
Please grow up, Brian. If you want to side track the issue, why don't you wait for Robert, etc, to answer? Are you pissed at that particular change, so that if anyone brings up a problem with it, and who might happen to have had a conversation with them once or twice, that you will jump on them

Re: [Rd] Option "installWithVers" seems to impact new.packages() badly?

2005-10-04 Thread Prof Brian Ripley
Please don't mangle replies like this to make them unreadable. Peter's comment is spot-on. The request for versioned installs and the original code came from the BioC team via Robert Gentleman. I presume it was a deliberate choice that a versioned install of 'ash' is reported as an installati

Re: [Rd] Option "installWithVers" seems to impact new.packages() badly?

2005-10-04 Thread A.J. Rossini
On 04 Oct 2005 16:42:14 +0200, Peter Dalgaard <[EMAIL PROTECTED]> wrote: > > > > > The issue is that installed.packages() has never been taught about > > > > versioned installs, and you might like to take that up with your BioC > > > > colleagues who added the idea. > > > "A.J. Rossini" <[EMAIL PR

Re: [Rd] Option "installWithVers" seems to impact new.packages() badly?

2005-10-04 Thread Peter Dalgaard
"A.J. Rossini" <[EMAIL PROTECTED]> writes: > Also, I don't know what this has to do with BioC. This is core R, so > pertinent to this list. Otherwise, perhaps it should be taken out? ... > > > The issue is that installed.packages() has never been taught about > > > versioned installs, and you

Re: [Rd] Option "installWithVers" seems to impact new.packages() badly?

2005-10-04 Thread A.J. Rossini
Also, I don't know what this has to do with BioC. This is core R, so pertinent to this list. Otherwise, perhaps it should be taken out? On 10/4/05, A.J. Rossini <[EMAIL PROTECTED]> wrote: > I expect "new.packages()" to not report on packages which I've > installed, as per documentation. > > best

Re: [Rd] Option "installWithVers" seems to impact new.packages() badly?

2005-10-04 Thread A.J. Rossini
I expect "new.packages()" to not report on packages which I've installed, as per documentation. best, -tony On 10/4/05, Prof Brian Ripley <[EMAIL PROTECTED]> wrote: > Hard to know what you expect here. There is no recent change that I can > see. > > I removed ash and then installed it with versi

Re: [Rd] Option "installWithVers" seems to impact new.packages() badly?

2005-10-04 Thread Prof Brian Ripley
Hard to know what you expect here. There is no recent change that I can see. I removed ash and then installed it with version. new.packages() showed it as new in both 2.1.1 and 2.2.0-beta. The issue is that installed.packages() has never been taught about versioned installs, and you might li

[Rd] Option "installWithVers" seems to impact new.packages() badly?

2005-09-30 Thread A.J. Rossini
In Rdevel, SVN version built this morning around 10am central european time, it looks like install.packages(new.packages(),installWithVers=TRUE) seem to ignore the version information -- that is, it reinstalls current versions of packages. This did not happen before I used "installWithVers=TR