[R-pkg-devel] mistake in submission of MODIStsp

Dear all,

     I recon I did a mistake in my recent submission of v 1.3.3 of package
MODIStsp.

As you can see from the current check results at

https://cran.rstudio.com/web/checks/check_results_MODIStsp.html

, I am obviously missing an import call for package "gWidgetsRGtk2"
? in NAMESPACE?
. I'm sorry about that.

On this topic, I have two questions:

1) How should I proceed to submit a patched version ? Do I need to wait for
all checks to be completed or
?should
I start a
?re-
submission already ?

2) Although this was obviously a mistake on my side, to avoid similar
problems in the future I was wondering why the checks I did on windows
locally (with as--cran option active) and with winbuilder (on r-devel), and
on linux with travis-ci (also with as--cran option active) didn't catch the
problem. Are there different check settings active on the different builds
?

Thanks in advance
? for any advice?
,

Lorenzo

	[[alternative HTML version deleted]]

On 13.08.2017 15:19, Lorenzo Busetto wrote:

Dear all,

     I recon I did a mistake in my recent submission of v 1.3.3 of package
MODIStsp.

As you can see from the current check results at

https://cran.rstudio.com/web/checks/check_results_MODIStsp.html

, I am obviously missing an import call for package "gWidgetsRGtk2"
? in NAMESPACE?
. I'm sorry about that.

On this topic, I have two questions:

1) How should I proceed to submit a patched version ? Do I need to wait
for
all checks to be completed or
?should
I start a
?re-
submission already ?

Ideally wait one or two days to see whether this is the only problem, then
resubmit.


2) Although this was obviously a mistake on my side, to avoid similar

problems in the future I was wondering why the checks I did on windows
locally (with as--cran option active) and with winbuilder (on r-devel),
and
on linux with travis-ci (also with as--cran option active) didn't catch
the
problem. Are there different check settings active on the different builds
?

Yes, but in this case all shluld have got the same problem unless you are
using the function only under non Windows OS?

Best,
Uwe Ligges



Thanks in advance

? for any advice?
,

Lorenzo

        [[alternative HTML version deleted]]

Thanks for your answer. I'll proceed as suggested.

Concerning the checks, besides my own check on windows 10, both the
submission to winbuilder:

(https://win-builder.r-project.org/T7H7Nd2vH7Rh/00check.log)

and a travis-ci run on linux:

(https://travis-ci.org/lbusett/MODIStsp/builds/263772681)

didn't catch the issue.

Lorenzo


On 13 August 2017 at 22:27, Uwe Ligges <ligges at statistik.tu-dortmund.de>
wrote:

On 13.08.2017 15:19, Lorenzo Busetto wrote:

Dear all,

     I recon I did a mistake in my recent submission of v 1.3.3 of
package
MODIStsp.

As you can see from the current check results at

https://cran.rstudio.com/web/checks/check_results_MODIStsp.html

, I am obviously missing an import call for package "gWidgetsRGtk2"
? in NAMESPACE?
. I'm sorry about that.

On this topic, I have two questions:

1) How should I proceed to submit a patched version ? Do I need to wait
for
all checks to be completed or
?should
I start a
?re-
submission already ?

Ideally wait one or two days to see whether this is the only problem,
then resubmit.


2) Although this was obviously a mistake on my side, to avoid similar

problems in the future I was wondering why the checks I did on windows
locally (with as--cran option active) and with winbuilder (on r-devel),
and
on linux with travis-ci (also with as--cran option active) didn't catch
the
problem. Are there different check settings active on the different
builds
?

Yes, but in this case all shluld have got the same problem unless you are
using the function only under non Windows OS?

Best,
Uwe Ligges



Thanks in advance

? for any advice?
,

Lorenzo

        [[alternative HTML version deleted]]