On Aug 14, 2007, at 8:16 AM, Adam R. Maxwell wrote:

> On Aug 14, 2007, at 04:42, Christiaan Hofman wrote:
>
>> That explains it. We check for %%EOF to find out if the file has been
>> fully written. Otherwise we have no way to know if a file is complete
>> (and tex writes incrementally, so loading too earl;y can lead to a
>> crash).
>
> Sounds like PLplot writes the same way as TeX (and it's even more
> important to check the PS file, since conversion can be slow).  The
> Apple-generated eps/ps files seem to have a %%EOF, as do ones from
> dvips.  A bunch of the .ps files in the GhostScript distribution don't
> have a %%EOF, though.  If it's within the PS spec, it sounds like the
> easiest fix would be for PLplot to add a %%EOF, or do it yourself from
> the shell or in C.
>
> --  
> Adam
>

Problem solved. Thanks, everyone! The PLplot guys added the %%EOF to  
one of their three non-Apple PS engine's output--the other two were OK.

Jerry


>>
>> On 14 Aug 2007, at 1:36 PM, Jerry wrote:
>>
>>> No, the PS files do not end with $$EOF, but with these few lines (I
>>> checked two files). The last character in the file is a line
>>> terminator.
>>>
>>> ------- Last few lines --------
>>>  S
>>> 1 W S
>>> eop
>>>
>>> %%Trailer
>>> %%Pages: 4
>>> @end
>>>
>>> ------- Last few lines --------
>>>
>>> Jerry
>>>
>>>
>>> On Aug 14, 2007, at 2:27 AM, Christiaan Hofman wrote:
>>>
>>>> It shouldn't really matter how data is written to the file, it only
>>>> matters that the file is not deleted (we're using kqueue to track
>>>> changes to the file). So with what you're saying it should work.
>>>> Does
>>>> the PS file end with %%EOF ?
>>>>
>>>> Christiaan
>>>>
>>>> On 14 Aug 2007, at 9:12 AM, Jerry wrote:
>>>>
>>>>>
>>>>> On Aug 9, 2007, at 4:46 AM, Christiaan Hofman wrote:
>>>>>
>>>>>> In fact PS files are followed for file changes. So the problem is
>>>>>> somewhere else. Probably the PS file is deleted before it is
>>>>>> replaced, in which case it's lost track of.
>>>>>>
>>>>>> Christiaan
>>>>>>
>>>>>
>>>>> Sorry for the slow follow-up.
>>>>>
>>>>> FWIW, the creation date is left the same (after a run of my  
>>>>> PLplot-
>>>>> using program) but the modification date is updated.
>>>>>
>>>>> Also (not sure if these are relevant), the Node ID, Inode's  
>>>>> Device,
>>>>> and the Spotlight Item ID are left unchanged (as reported by Path
>>>>> Finder).
>>>>>
>>>>> Here is a comment that I got from the PLplot list:
>>>>>
>>>>>  When plplot writes any file it uses fopen to open the file first,
>>>>> with
>>>>>  the "wb+" options to make the file writeable and to truncate the
>>>>> file if
>>>>>  it already exists. This is the standard C way to deal with  
>>>>> opening
>>>>>  files so I'm surprised it doesn't work. What do the Skim people
>>>>>  "recommend" as a way of opening files?
>>>>>
>>>>>  Are you using the ps driver or the psttf driver? The psttf driver
>>>>> has to
>>>>>  do some more complicated manouvers to get a C++ stream. It first
>>>>> opens
>>>>>  and truncates the file as above. It the closes the C stream and
>>>>> opens
>>>>>  the file again as a C++ stream.
>>>>>
>>>>> Jerry
>>>>>
>>>>>
>>>>>
>>>>>> On 9 Aug 2007, at 1:06 PM, Jerry wrote:
>>>>>>
>>>>>>>
>>>>>>> On Aug 9, 2007, at 2:43 AM, Christiaan Hofman wrote:
>>>>>>>
>>>>>>>>
>>>>>>>> On 9 Aug 2007, at 6:37 AM, Adam R. Maxwell wrote:
>>>>>>>>
>>>>>>>>>
>>>>>>>>> On Aug 8, 2007, at 21:17, Jerry wrote:
>>>>>>>>>
>>>>>>>>>> I'd file this as a bug report but I'm not sure it's a bug.
>>>>>>>>>>
>>>>>>>>>> I'm not using Skim for TeX stuff but as a viewer for plots
>>>>>>>>>> which
>>>>>>>>>> happen to be generated by the excellent PLplot. PLplot writes
>>>>>>>>>> Postscript files. I've enabled "Check for file changes" in  
>>>>>>>>>> the
>>>>>>>>>> prefs
>>>>>>>>>> but Skim does not re-load the file after a run of my plot-
>>>>>>>>>> generating
>>>>>>>>>> program finishes.
>>>>>>>>>
>>>>>>>>> I believe that only files which are loaded as PDF can be
>>>>>>>>> monitored; PS
>>>>>>>>> goes through an intermediate step internally whereby it's
>>>>>>>>> converted to
>>>>>>>>> PDF for display, just like Apple's Preview.
>>>>>>>>>
>>>>>>>>>> I have to make an edit to the (old) version then
>>>>>>>>>> select Revert in order to get the newly-make PS file. Is the
>>>>>>>>>> "Check
>>>>>>>>>> for file changes" related only to TeX stuff or can it monitor
>>>>>>>>>> any
>>>>>>>>>> file for changes. A simple "Reload" command might be a good
>>>>>>>>>> compromise between ease of use and not accidentally
>>>>>>>>>> reloading a
>>>>>>>>>> file
>>>>>>>>>> (and loosing notes). Also, a periodic re-checking of the
>>>>>>>>>> file on
>>>>>>>>>> disk
>>>>>>>>>> might generate too much work if the file (due to me  
>>>>>>>>>> running my
>>>>>>>>>> plotting program) takes several seconds or minutes to be re-
>>>>>>>>>> written.
>>>>>>>>>> The more I think about it, a manual Reload function might be
>>>>>>>>>> sweet.
>>>>>>>>>
>>>>>>>>> Revert serves this purpose, and you can also use it from
>>>>>>>>> AppleScript.
>>>>>>>>> If there's a script involved in using PLplot, you could use
>>>>>>>>> osascript
>>>>>>>>> to reload.
>>>>>>>>>
>>>>>>>>> --  
>>>>>>>>> adam
>>>>>>>>
>>>>>>>> But Revert would never work for a PS file, just as Save is
>>>>>>>> disabled.
>>>>>>>> PS is not a native type for Skim, it's only a viewer for that
>>>>>>>> type.
>>>>>>>> This is standard document based app behavior, and for good
>>>>>>>> reasons.
>>>>>>>>
>>>>>>>> Christiaan
>>>>>>>>
>>>>>>>>
>>>>>>> I'm not sure what all the issues are--I'm just saying that for
>>>>>>> the
>>>>>>> particular situation where I have to repeatedly open the same
>>>>>>> (Postscript) file, having to select the Note tool, make a  
>>>>>>> mark on
>>>>>>> the
>>>>>>> file (making the loaded file dirty, I suppose), then selecting
>>>>>>> Revert
>>>>>>> and answering a dialog boxe is a fair number of steps especially
>>>>>>> when
>>>>>>> I have to do it a whole lot of times in a day. A simple Reload,
>>>>>>> as
>>>>>>> opposed to a Revert (which assumes that the loaded file has been
>>>>>>> modified in memory), would certainly be nice to have.
>>>>>>>
>>>>>>> Jerry
>>>>>>>
>>>>>>>
>>>>>>>
>>>
>>> -------------------------------------------------------------------- 
>>> --
>>> ---
>>> This SF.net email is sponsored by: Splunk Inc.
>>> Still grepping through log files to find problems?  Stop.
>>> Now Search log events and configuration files using AJAX and a
>>> browser.
>>> Download your FREE copy of Splunk now >>  http://get.splunk.com/
>>> _______________________________________________
>>> Skim-app-users mailing list
>>> [email protected]
>>> https://lists.sourceforge.net/lists/listinfo/skim-app-users
>>
>>
>> --------------------------------------------------------------------- 
>> ----
>> This SF.net email is sponsored by: Splunk Inc.
>> Still grepping through log files to find problems?  Stop.
>> Now Search log events and configuration files using AJAX and a
>> browser.
>> Download your FREE copy of Splunk now >>  http://get.splunk.com/
>> _______________________________________________
>> Skim-app-users mailing list
>> [email protected]
>> https://lists.sourceforge.net/lists/listinfo/skim-app-users
>
>
> ---------------------------------------------------------------------- 
> ---
> This SF.net email is sponsored by: Splunk Inc.
> Still grepping through log files to find problems?  Stop.
> Now Search log events and configuration files using AJAX and a  
> browser.
> Download your FREE copy of Splunk now >>  http://get.splunk.com/
> _______________________________________________
> Skim-app-users mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/skim-app-users
>


-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/
_______________________________________________
Skim-app-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/skim-app-users

Reply via email to