[Kst] Kst Crash

Thu Jan 30 01:06:51 UTC 2014

On 30/01/2014 10:29 AM, Ben Lewis wrote:
> On 29/01/2014 9:13 PM, Peter Kümmel wrote:
>> On 18.01.2014 11:34, Ben Lewis wrote:
>>> Hi Peter / Kst,
>>>
>>> The changes you've made work great! I can now load, large, live data files. Thank you.
>>>
>>> Also, there are now no delays in the data wizard. Thank you.
>>>
>>> It seems that the new progress bar only gets used on files over a certain size. I like that.
>>>
>>> There are still some minor things that need to be fixed before the next release:
>>>
>>>   * Progress bar gets stuck at 50%
>>
>> Is fixed.
> Great! Works well.
>>
>>>   * I sometimes get snippets of data missing in a live plot. If I restart Kst and reload the 
>>> data there are no missing
>>>     bits. This seems to happen when there is a large amount of network traffic.
>>
>> The file is not local?
> That's right, the data file is not local. It is generated on a remote Beckhoff PLC (WindowsXP 
> Embedded) and stored on a USB memory stick. The memory stick, on the PLC, is mapped to a local 
> drive on my laptop. Kst is run on my laptop and reads the data file from the mapped location. When 
> I configure the ASCII data file settings I set the update type to "time interval". For all intents 
> and purposes this works exactly the same as if the data file were local. Is there a problem with 
> the data file not being local?
>>
>>>   * Every so often Kst seems to read an entire vector from the start after the initial read is 
>>> done. I noticed this when
>>>     the temporary console log file was enabled. I can still see one of the symptoms, which is an 
>>> increased CPU load on
>>>     the remote PC every time Kst reads the entire vector. This is not very noticeable on small 
>>> files but the file size
>>>     gets larger it becomes much more noticeable.
>>
>> When your data file grows while the initial read, it would be read again as soon as the 
>> monitoring begins.
> This problem is not related to the initial read, it occurs at semi regular intervals while ever 
> Kst is updating. It can be easily observed in the temporary console log.
> It makes Kst unusable for very large, live, data files.
>
>
>>
>>>   * "Out of Memory" error http://kde.6490.n7.nabble.com/Out-of-memory-error-td1555215.html
>>
>> I could not reproduce, but  I fixed the miss leading error description.
> The new description is much more useful. I do not get the error when "use threads" is disabled.
> What I find interesting though, is that when I get the error, after I click OK, it continues to 
> work regardless.
>>
>>>   * "FFT Length" is editable when "Interleaved Average" is deselected. 
>>> https://bugs.kde.org/show_bug.cgi?id=328371
>>>   * Time axis does not update after changing sample frequency. 
>>> https://bugs.kde.org/show_bug.cgi?id=328481 It looks like
>>>     the behaviour of this bug has changed but it is still not right yet. I can provide more 
>>> details if required.
>>>   * Temporary memory is not freed after reading data from an ASCII file. As it stands at the 
>>> moment approximately twice
>>>     as much RAM is required than is really necessary.
>>
>> This is by design, but yes, for large files it could be a problem.
>> Could you create a bug report?
> OK.
>>
>>>
>>> If you need clarification or further testing on any of these please let me know. I'm happy to 
>>> help out in any way I can.
> I have found another bug in the ASCII data wizard. The time taken to get through the first screen 
> is directly proportional to the size of the previous file opened. Unless the file no longer 
> exists. In which case there is no delay. I will submit a new bug report.
Forget about this. I can't reproduce it now.
>> Many thanks,
>> Peter
>>
>>>
>>> Regards, Ben
>>>
>>>
>>> On 17/01/2014 10:09 PM, Peter Kümmel wrote:
>>>> On 06.01.2014 11:38, Peter Kümmel wrote:
>>>>>> It seems that regular updates should be disabled during the use of the ASCII Wizard and
>>>>>> during the first read. Once loaded, regular updates can resume. What do you think?
>>>>>
>>>>> Good catch Ben! Yes, this should be fixed.
>>>>
>>>> Fixed, monitoring the file now starts when initial reading is done.
>>>>
>>>>>
>>>>>>
>>>>>> Ben
>> _______________________________________________
>> Kst mailing list
>> Kst at kde.org
>> https://mail.kde.org/mailman/listinfo/kst
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/kst/attachments/20140130/92ed4f19/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/png
Size: 244465 bytes
Desc: not available
URL: <http://mail.kde.org/pipermail/kst/attachments/20140130/92ed4f19/attachment-0001.png>