Closed if-kenn closed 7 years ago
Sorry for the inconvenience, the copy of pandoc-1.17.2.zip we host has actually changed a couple of times now for not-very-interesting reasons. We should be good now.
I'd recommend updating to Gabor's universal pandoc 1.19 binaries. We're using them in RStudio as is CRAN on their build servers On Mon, Jul 3, 2017 at 6:25 PM Joe Cheng notifications@github.com wrote:
Closed #303 https://github.com/rstudio/shiny-server/issues/303.
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/rstudio/shiny-server/issues/303#event-1148752451, or mute the thread https://github.com/notifications/unsubscribe-auth/AAfafUT7DVMjynMARHim49Tsq3CKJ_h1ks5sKWpygaJpZM4OLYZJ .
@jjallaire I thought @gaborcsardi's 1.19 binaries don't support RHEL 5, is that wrong?
FWIW I tried the binaries quickly on a CentOS 5.11 docker machine and they appear to work there. The fact that they're completely static executables (not even dynamically linking to libc
) should protect us from potential incompatibilities with older C libraries as well.
They don't support RHEL 5. Note that the v1.17 binaries you are using also don't support RHEL 5 (they bundle 1.15 in place of it I believe). RStudio Server also no longer supports RHEL 5 (no one has yet complained). Also RedHat doesn't support RHEL 5! (it has hit EOL).
On Mon, Jul 3, 2017 at 9:37 PM, Joe Cheng notifications@github.com wrote:
@jjallaire https://github.com/jjallaire I thought Gabor's 1.19 binaries don't support RHEL 5, is that wrong?
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/rstudio/shiny-server/issues/303#issuecomment-312760189, or mute the thread https://github.com/notifications/unsubscribe-auth/AAGXxx4izjC6n36qfQmAgt0EQSlSziFBks5sKZdZgaJpZM4OLYZJ .
@fereshtehRS @ajmcoqui @tareefk @alandipert FYI, sounds like SSP 1.5.4 will need to drop support for RedHat 5/CentOS 5, to take a new pandoc binary.
@jjallaire Are you sure that the new, static pandoc binary does support RHEL 5? Unless the system is 32bit, it should, I think.
I cannot try it with RHEL 5, but I did try it with CentOS 5, and it was fine, and Kevin tried it as well.
Dropping RHEL 5 support makes a lot of sense, I just want to know about any potential issues.
No I'm not sure at all, I was just assuming because the static binary I built on Ubuntu 10.04 didn't work on RHEL 5.
So @jcheng5 maybe Gabor's binary would work for your RHEL 5 distribution after all!
On Fri, Jul 7, 2017 at 5:10 PM, Gábor Csárdi notifications@github.com wrote:
@jjallaire https://github.com/jjallaire Are you sure that the new, static pandoc binary does support RHEL 5? Unless the system is 32bit, it should, I think.
I cannot try it with RHEL 5, but I did try it with CentOS 5, and it was fine, and Kevin tried it as well.
Dropping RHEL 5 support makes a lot of sense, I just want to know about any potential issues.
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/rstudio/shiny-server/issues/303#issuecomment-313793765, or mute the thread https://github.com/notifications/unsubscribe-auth/AAGXx7VDO_hHS7fZ6M0Hfut_hPoogZKkks5sLp6vgaJpZM4OLYZJ .
It looks like the checksum hash for pandoc is not correct, I am getting:
Looking at https://github.com/rstudio/shiny-server/blob/master/external/pandoc/CMakeLists.txt#L15 I see 2bf1d83ce3bf3ae856d2af9453832952.
When I manually download http://s3.amazonaws.com/rstudio-buildtools/pandoc-1.17.2.zip to get the MD5 I get 9d53413f3d69fbf0bd369f5cebb7dc8b.