[BioC] Getting LimmaGUI to run

stefano iacus stefano.iacus at unimi.it
Fri Jul 15 18:27:07 CEST 2005


On 15/lug/05, at 17:47, chris umbricht wrote:


> Stefano,
>
> re: "if R quits "unexpectedly" the finder will tell you, just press  
> the button concerning the crash report"
>
> R quits immediately after starting up, and the finder does not  
> generate a crash report.
>
umh, probably the report is generated but not shown by the Finder.

>
> As I understood from both your and Rob's comments, it appears  
> simply that my .RData file contains X11 dependent processes, eg tkltk.
> So I can avoid the problem by prestarting the X11 window, which is  
> fine.
>
> My question to you is, how do I undo that dependecy in the R.app  
> (1.12, build 1622) short of trashing the RData file from within the  
> terminal window?
>
>

try the following with X11 running
launch R.app (should not quit)
and in R Console

 > (.packages())

does tcltk appear in this list? If so type

 > detach(package:tcltk)

and check that it is no longer there with
 > (.packages())

now
 > save.image()

quit R, shut down the X11 window server and try to restart R.app. Let  
me know
stefano



> Am I wrong in assuming the workspace menu items are designed to get  
> back to the original, non-X11 dependent environment? If so, is  
> there another way?
>
> thanks
>
> chris
>
>
> On Jul 15, 2005, at 5:44 AM, stefano iacus wrote:
>
>
>
>>
>> On 11/lug/05, at 22:26, chris umbricht wrote:
>>
>>
>>
>>> Rob,
>>>
>>> thanks for your feedback, and I do understand your directions,  
>>> i.e. I
>>> think I will be able to navigate this X11 dependency with your
>>> suggestions.
>>>
>>> I was trying to give Stefano some feedback, since he is heavily
>>> involved in developing the OS X version of R.App, and I'm trying to
>>> distinguish between my misunderstanding some elements of the GUI vs
>>> having a corrupted installation. It seems to me that having to  
>>> delete
>>> the .RData file from the terminal should not be necessary as routine
>>> measure...
>>>
>>> I'm hoping he will chime in, since he asked me for more details
>>> earlier, but I'm not sure what he needs.
>>>
>>>
>>
>> it R quits "unexpectedly" the finder will tell you, just press the  
>> button concerning the crash report (I think is something like  
>> "send report")
>> stefano
>>
>>
>>
>>
>>>
>>>
>>> On Jul 11, 2005, at 4:18 PM, Rob J Goedman wrote:
>>>
>>>
>>>
>>>
>>>> Chris,
>>>>
>>>> Remove .RData (as in previous emails).
>>>> Start R and close R. When it asks if you want to save
>>>> the session, choose Save to create a new .RData file.
>>>>
>>>> I nearly always select Don't Save, which will not create a new
>>>> .Rdata file.
>>>>
>>>> If you don't load tcltk, jpeg and/or png, e.g. used in marray,  
>>>> (maybe
>>>> there are a few other packages that will require X11), you don't  
>>>> need
>>>> X11 and it will not end up as a dependency in .RData.
>>>>
>>>> Or you can automatically start X11 when you login (Drag it to Login
>>>> items of your Account is System Preferences).
>>>>
>>>> Rob
>>>>
>>>>
>>>> On Jul 11, 2005, at 12:44 PM, chris umbricht wrote:
>>>>
>>>>
>>>>
>>>>
>>>>> Right, so how do I change the .RData back to not containing tcltk
>>>>> calls?
>>>>>
>>>>> I tried the (to me) obvious steps of unchecking anything tcltk- 
>>>>> related
>>>>> in the manager, and trying to clear the workspace etc. without
>>>>> apparent
>>>>> effect...
>>>>>
>>>>> c.
>>>>>
>>>>>
>>>>> On Jul 11, 2005, at 12:39 PM, stefano iacus wrote:
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>> Hi Chris,
>>>>>> no it has always been the case. And the problem is that  
>>>>>> probably your
>>>>>> .RData contains calls to tcltk.
>>>>>> Any call to tcltk, X11 etc requires a X11 server to be running.
>>>>>> btw, I need more detailed bug report that this.
>>>>>> stefano
>>>>>>
>
>



More information about the Bioconductor mailing list