Data.Binary.Get.runGetState at position 296: demandInput: not enough bytes

[Henk te Sligte]

Hi Joachim,


On Wed, Oct 30, 2013 at 6:46 PM, Joachim Breitner
<mail at joachim-breitner.de>wrote:

> Dear Henk,
>
>
> Am Mittwoch, den 30.10.2013, 12:13 +0100 schrieb Henk te Sligte:
> > I am trying to get arbtt working, but whatever I do, I keep getting
> > the error:
> > arbtt-stats: Data.Binary.Get.runGetState at position 296: demandInput:
> not enough bytes
> >
> >
> >
> > I have removed the log file, I have tried to run arbtt-recover, I have
> > installed it with and without cabal, but I keep getting this error.
> >
> > The only way I can find out what windows have been running on my
> > system is by running arbtt-dump | more and trying to read the active
> > window manually. Is there something I am doing wrong?
>
> you are now the second user reporting this error, and you even tried
> various things that I would recommend... but unfortunately, I don’t have
> a good idea what might be causing this.
>
> What version of arbtt are you using?
>

I am using version 0.7:
henk at linux-lu3t:~> arbtt-stats --version
arbtt-stats 0.7


>
> When you run arbtt-dump, do you also get the error message?
>

Yes, I do get the TimeLogEntries, but the error message is also printed (I
guess to stderr, because when I dump the contents to a file, the error
message is printed:)
henk at linux-lu3t:~> arbtt-dump > test1.txt
arbtt-dump: Data.Binary.Get.runGetState at position 396: demandInput: not
enough bytes



> Does arbtt-recover go through? What does it print?
>

arbtt-recover goes through but I remember having errors as well when
working with a larger log file. I might be wrong though, I'm not sure about
this. In this case, I don't get errors:
henk at linux-lu3t:~> arbtt-recover
Found header, continuing... (38513 bytes to go)

In this case I don't get errors after renaming capture.log.recovered to
capture.log. I'm definitely sure I did get the demandInput: not enough
bytes error before with a larger log file.
henk at linux-lu3t:~/.arbtt> mv capture.log capture.old
henk at linux-lu3t:~/.arbtt> mv capture.log.recovered capture.log
henk at linux-lu3t:~/.arbtt> arbtt-stats
Total time per
tag
==================
Tag_|_Time_|_Percentage_

(in case you were wondering why I didn't use --logfile:)
henk at linux-lu3t:~> arbtt-stats --logfile=~/.arbtt/capture.log.recovered
arbtt-stats: ~/.arbtt/capture.log.recovered: openBinaryFile: does not exist
(No such file or directory)



> When you remove the log file and start with a fresh file, does
> arbtt-stats immediately give the error message, or only after it has
> been running for a while?
>

Yes, it starts immediately with the same error:
henk at linux-lu3t:~/.arbtt> rm capture.log
henk at linux-lu3t:~/.arbtt> (arbtt-capture &)
henk at linux-lu3t:~/.arbtt> arbtt-stats
arbtt-stats: Data.Binary.Get.runGetState at position 429: demandInput: not
enough bytes


> Did you modify the sample rate?
>

No, I have just the minimal categorize.cfg:
henk at linux-lu3t:~/.arbtt> cat categorize.cfg
{$idle > 60 ==> tag inactive}

If you are interested, I'd be happy to send you a log file so you can
analyze this better.
I don't have any haskell knowlegde, but I'm willing to install experimental
versions to help resolve this issue.

Greetings,
Henk te Sligte
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.nomeata.de/pipermail/arbtt/attachments/20131031/7c381546/attachment.htm>