cloudstack-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Daan Hoogland <daan.hoogl...@gmail.com>
Subject fixPath (was: committer wanted for review)
Date Wed, 19 Jun 2013 19:37:49 GMT
John and others,

I have been looking for the point where the wrong path is instantiated.
After analyses I came to the DownloadAnswer class which contains the
original fixPath method that I c&p'ed to do my thing. I cannot support this
with logging however. Where the DownloadAnswer is created eludes me
however. I got trapped between DownloadManagerImpl and DownloadListener.
The creation of the answer was as far as I could tell in
handleDownloadProgressCmd but again I can not support this with logging.

As the reason I suppect eclipse class-file caching but also this theory
seems to not work.

Anyone's got a good clue for me?

I have burned to much time on this so I will save the patch for if  it is
locally needed by users, but I still want to find the best solution.

As for John's argument of creating technical debt, I am now convinced that
I am not adding but only using the present debt that is in there. The
DownloadAnswer.fixup() method is doing the same on a more obscure place
then my solution.

Also, if we decide to apply my changes anyway I think we should up the
loglevel at least as much as acceptable as to keep pointing to the
technical debt that we have at the moment.

On Tue, Jun 18, 2013 at 5:28 PM, Daan Hoogland <daan.hoogland@gmail.com>wrote:

>  On 2013/6/18 16:49 , John Burwell wrote:
>
> Second, please don't take my feedback as passing judgements such as things being ugly
or
>
>  don't worry, I like the discussion and i don't mind losing an argument if
> the best solution arises from it. Let's see about that.
>
> --
>
> <http://daan.sbpad6.nl/>
>

Mime
  • Unnamed multipart/related (inline, None, 0 bytes)
View raw message