[R] Packages sometimes don't update, but no error or warning is thrown

Duncan Murdoch murdoch@dunc@n @end|ng |rom gm@||@com
Tue Feb 13 11:36:04 CET 2024 

I think you should ask more from the BioC folks, or do more digging 
yourself.  At a minimum we would need to know what arguments were used 
in the call that failed before we could help you with this.  So if this 
ever happens again, here's what you can do:

  Run debug(install.packages), then the call to BiocManager::install. 
It will break to the debugger when install.packages is called.  It will 
print the whole source of install.packages which is quite long, but at 
the very top it will show something like

   debugging in: install.packages(pkgs = x)

The "install.packages(pkgs = x)" will be the call from whatever function 
calls install.packages.  The arguments shown may be constants or 
variables like x; the argument names may or may not be shown, depending 
on the Bioconductor code.  If they are variables, they are in the scope 
of the caller, not install.packages, whereas the argument name "pkgs" 
would be in scope and you could print it.  Record all the arguments that 
were passed, then quit using Q.

Now run undebug(install.packages), and re-run the install.packages() 
call from the command line, substituting the variables you recorded.  If 
that shows the same problems as when called from BiocManager::install, 
show us the log of what happened.  If not, it's a Bioconductor issue.

Duncan Murdoch

On 13/02/2024 5:07 a.m., gernophil using gmx.net wrote:
> Yes you're right that it started as an Bioconductor issue. The reason I 
> am writing it here, if because of one of the core members of the 
> Bioconductor project stated this:
> "(...) But anyway, it's almost certainly an install.packages issue 
> rather than BiocManager::install, so you might check over at 
> r-help using r-project.org."
> It's hard to give an explicit example here since it's always a bit time 
> dependent, when and if this happens. And "unfortunately", right now all 
> packages can be updated, but I'll come back to this as soon as I 
> experience it again. It's more of a general feature request then a help 
> request, but as you say "In my experience update.packages() asks whether 
> you want the source or binary install and respects your choice." 
> However, at least for me, it does not do this anymore, but simply 
> silently fails to update those packages.
> The Bioconductor thread is the most informative for this 
> (https://support.bioconductor.org/p/9156283 
> <https://support.bioconductor.org/p/9156283/#9156308>) in case anyone 
> wants to check the situation in more detail.
> Philipp
> *Gesendet:* Dienstag, 13. Februar 2024 um 10:53 Uhr
> *Von:* "Duncan Murdoch" <murdoch.duncan using gmail.com>
> *An:* "R Project Help" <R-help using r-project.org>
> *Betreff:* [R] Packages sometimes don't update, but no error or warning 
> is thrown
> 
> Those three references are about RStudio and Bioconductor (or at least
> they start out that way, I didn't read through the long threads).
> Neither of those is relevant here, but from my quick scan it appears the
> issue is that those systems detect a package in source form is
> available, then install an older binary instead.
> 
> If you have an example where update.packages() does this, that would be
> on topic on this list. Please post just that here: your calls, and all
> the output you received. In my experience update.packages() asks
> whether you want the source or binary install and respects your choice.
> 
> Duncan Murdoch
> 
> On 13/02/2024 3:59 a.m., gernophil--- via R-help wrote:
>  > Hey everyone,
>  >
>  > this question is related to this 
> (https://community.rstudio.com/t/packages-are-not-updating/166214/3 
> <https://community.rstudio.com/t/packages-are-not-updating/166214/3>), 
> this (https://www.biostars.org/p/9586316/#9586323 
> <https://www.biostars.org/p/9586316/#9586323>) and this 
> (https://support.bioconductor.org/p/9156283/#9156308 
> <https://support.bioconductor.org/p/9156283/#9156308>). The two latter 
> ones are pots from myself.
>  >
>  > To sum it up: If I am updating packages (be it via Bioconductor or 
> CRAN) some packages simply don’t update, but they also don’t throw any 
> warning or error that they have not been updated. The reason behind this 
> is most likely that there is an update, but it’s not yet available as a 
> binary – in my case as a "mac.binary.big-sur-arm64", since I am on Apple 
> Silicon. My pkgType is set to "both" (default), but my .Platform$pkgType 
> is "mac.binary.big-sur-arm64" (also default).
>  >
>  > If the system shows an update is available and I click update (or 
> update via install.packages()), but the package will not be updated, I 
> would expect any kind of message that the package will not be updated, 
> since no newer binary is available or a prompt, if I want to compile 
> from source. However, the only message I get is:
>  > ```
>  > trying URL '<url_to_package>'
>  > Content type 'application/x-gzip' length XXXXXXX bytes (X.X MB)
>  > ==================================================
>  > downloaded X.X MB
>  >
>  >
>  > The downloaded binary packages are in
>  >                 /var/folders/ws/XXX/T//YYY/downloaded_packages
>  > ```
>  > To see, if the package was actually updated, I have to check for 
> updates again (or check the version of the installed package). Is this 
> the expected behavior and if so, why? (Also, why is there a double slash 
> in the path to the downloaded packages?)
>  >
>  > Thanks for your help.
>  >
>  > Best,
>  > Philipp
>  >
>  > ______________________________________________
>  > R-help using r-project.org mailing list -- To UNSUBSCRIBE and more, see
>  > https://stat.ethz.ch/mailman/listinfo/r-help 
> <https://stat.ethz.ch/mailman/listinfo/r-help>
>  > PLEASE do read the posting guide 
> http://www.R-project.org/posting-guide.html 
> <http://www.R-project.org/posting-guide.html>
>  > and provide commented, minimal, self-contained, reproducible code.
> 
> ______________________________________________
> R-help using r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help 
> <https://stat.ethz.ch/mailman/listinfo/r-help>
> PLEASE do read the posting guide 
> http://www.R-project.org/posting-guide.html 
> <http://www.R-project.org/posting-guide.html>
> and provide commented, minimal, self-contained, reproducible code.

More information about the R-help mailing list