On Sunday, 25 June 2017 16.38.00 WEST Steve Grubb wrote: > For example, when I run RStudio, I get: > > R graphics engine version 12 is not supported by this version of RStudio. > The Plots tab will be disabled until a newer version of RStudio is > installed. You need to update Rstudio to a newer version. That fixes this issue.

> Also: > > Warning: Error in library: there is no package called ‘shinyjs’ > Stack trace (innermost first): > > 41: library > > 1: shiny::runApp > > Error : there is no package called ‘shinyjs’ > > So, basically, anyone updating to the new R is dead in the water. It > really needs to be rolled back to 3.3.3 in F24 & F25. Which leads to > another issue...where is R in the bugzilla database? I can't find it. The only packages that needed to be rebuild are those that rebuild packages that register C or Fortran functions:

From https://cran.r-project.org/doc/manuals/r-release/NEWS.html """ PACKAGE INSTALLATION ... Packages which register native routines for .C or .Fortran need to be re- installed for this version (unless installed with R-devel SVN revision r72375 or later). ... """

IOW, F26 can have 3.4 just fine since its not shipped.

On Sun, 2017-06-25 at 17:08 +0100, José Abílio Matos wrote: > On Sunday, 25 June 2017 16.38.00 WEST Steve Grubb wrote: > > > For example, when I run RStudio, I get: > > > > R graphics engine version 12 is not supported by this version of > > RStudio. The Plots tab will be disabled until a newer version of RStudio > > is installed. > > You need to update Rstudio to a newer version. That fixes this issue. > > > Also: > > > > Warning: Error in library: there is no package called ‘shinyjs’ > > Stack trace (innermost first): > > > > 41: library > > > > 1: shiny::runApp > > > > Error : there is no package called ‘shinyjs’ > > > > So, basically, anyone updating to the new R is dead in the water. It > > really needs to be rolled back to 3.3.3 in F24 & F25. Which leads to > > another issue...where is R in the bugzilla database? I can't find it. > > The only packages that needed to be rebuild are those that rebuild > packages that register C or Fortran functions: By policy, the Rstudio update and updates for all those packages should have been included with the update to R itself: https://fedoraproject.org/wiki/Updates_Policy#Updating_inter-dependent_packa ges

"When one updated package requires another (or more than one other), the packages should be submitted together as a single update. For instance, if package A depends on packages B and C, and you want to update to a new version of package A which requires new versions of B and C, you must submit a single update containing the updated versions of all three packages. It is a bad idea to submit three separate updates, because if the update for package A is pushed stable before the updates for packages B and C, it will cause dependency problems." Also, it may be worth considering whether updating R to 3.4 is in line with these parts of the policy: "Releases of the Fedora distribution are like releases of the individual packages that compose it. A major version number reflects a more-or-less stable set of features and functionality. As a result, we should avoid major updates of packages within a stable release. Updates should aim to fix bugs, and not introduce features, particularly when those features would materially affect the user or developer experience. The update rate for any given release should drop off over time, approaching zero near release end-of-life; since updates are primarily bugfixes, fewer and fewer should be needed over time."

"Package maintainers MUST: Avoid Major version updates, ABI breakage or API changes if at all possible. Avoid changing the user experience if at all possible. Avoid updates that are trivial or don't affect any Fedora users." It would be great if the relevant maintainers could take these points into consideration for future updates. Thanks!

On Tuesday, June 27, 2017 3:47:42 PM EDT Adam Williamson wrote: > On Sun, 2017-06-25 at 17:08 +0100, José Abílio Matos wrote: > > On Sunday, 25 June 2017 16.38.00 WEST Steve Grubb wrote: > > > > > For example, when I run RStudio, I get: > > > > > > R graphics engine version 12 is not supported by this version of > > > RStudio. The Plots tab will be disabled until a newer version of > > > RStudio is installed. > > > > You need to update Rstudio to a newer version. That fixes this issue. > > > > > Also: > > > > > > Warning: Error in library: there is no package called ‘shinyjs’ > > > Stack trace (innermost first): > > > 41: library > > > 1: shiny::runApp > > > Error : there is no package called ‘shinyjs’ > > > > > > So, basically, anyone updating to the new R is dead in the water. It > > > really needs to be rolled back to 3.3.3 in F24 & F25. Which leads to > > > another issue...where is R in the bugzilla database? I can't find it. > > > > The only packages that needed to be rebuild are those that rebuild > > packages that register C or Fortran functions: > > By policy, the Rstudio update and updates for all those packages should > have been included with the update to R itself: > > https://fedoraproject.org/wiki/Updates_Policy#Updating_inter-dependent_pac > kages There really is no way to do this. We don't ship RStudio because its so hard to package. I've tried creating a spec file and putting all the pieces where it belongs but it doesn't work unless its installed to /usr/local which is against Fedora policies. The best I could come up with is to create a blog to tell everyone how to build it for themselves.

There just simply needs to be a rule of no update to a major release during a shipped Fedora OS. R 3.3 -> 3.3.1 is fine as that's bug fixes. R 3.3.3 -> 3.4 has potential to wreck things on already shipped OS. > "When one updated package requires another (or more than one other), > the packages should be submitted together as a single update. For > instance, if package A depends on packages B and C, and you want to > update to a new version of package A which requires new versions of B > and C, you must submit a single update containing the updated versions > of all three packages. It is a bad idea to submit three separate > updates, because if the update for package A is pushed stable before > the updates for packages B and C, it will cause dependency problems." > > Also, it may be worth considering whether updating R to 3.4 is in line > with these parts of the policy: > > "Releases of the Fedora distribution are like releases of the > individual packages that compose it. A major version number reflects a > more-or-less stable set of features and functionality. As a result, we > should avoid major updates of packages within a stable release. Updates > should aim to fix bugs, and not introduce features, particularly when > those features would materially affect the user or developer > experience. The update rate for any given release should drop off over > time, approaching zero near release end-of-life; since updates are > primarily bugfixes, fewer and fewer should be needed over time." It sounds like this ^^^ fits best. > "Package maintainers MUST: > > > Avoid Major version updates, ABI breakage or API changes if at all > > possible. > > Avoid changing the user experience if at all possible. > Avoid updates that are trivial or don't affect any Fedora users." > > > It would be great if the relevant maintainers could take these points > into consideration for future updates. Thanks! I'm trying to consider what to do to get a working RStudio again. I suppose I

pulled in a little over 20 dependency packages. Also, what about the missing bz entry? I'd file a bug but I can't find an "R"

Thanks, -Steve _______________________________________________ devel mailing list -- devel(a)lists.fedoraproject.org To unsubscribe send an email to devel-leave(a)lists.fedoraproject.org

First, I appreciate the work that goes into maintaining the R ecosystem. But there is now a problem that I don't think people would allow in other languages. Namely, the recently pushed R-3.4 update breaks the whole world. Something this disruptive usually has to wait for a new Fedora release to get pushed out. I really think the same thing has to apply for R.