What steps will reproduce the problem?
1. A student who is over their quota limit saves a PyScripter script.
2. The file shows up in their folder, but with file size zero.
3. PyScripter thinks the file has been successfully saved, and will exit.
What is the expected output? What do you see instead?
I would like it to raise an error "File save failed, or Delayed Write Failed".
Instead, PyScripter thinks the file save has worked.
What version of the product are you using? On what operating system?
PyScripter 2.4.3.1, Python 3.2.2 (32-bit) Windows 7
Please provide any additional information below.
Our lab setup is insanely complex: students have a mapped drive home directory
with Novell back-end servers and file systems. Disk quotas are enforced. This
is complicated by roaming profiles that may need to be synched between the
local machine and the central store. At the campus network layer, the Novell
servers and the Storage Network are VMWare virtualized, replicated,
load-balanced, and can automatically be migrated bewteen physical server racks
and campus locations. So we have enough layers to ensure that nobody actually
seems to know where anything is really running!
When the student is already over or at quota (i.e. their disk is full),
PyScripter appears to save the files fine, but files end up at zero size.
We cannot reproduce the problem on directly connected storage (e.g. a disk or
USB drive), and we know that other tools that we've tested in the lab (Notepad,
Word, etc.) all correctly report that there was a problem saving to the
(over-quota) home drive.
I am an end-user teacher using this lab (so my understanding of the underlying
situation may not be 100% accurate, and I have zero control), but our tech
support guys have investigated, they can reproduce the problem (only on
PyScripter), but they have not come up with any convincing explanation or
solution.
So this ticket asks for a review/walkthough of the code that does the exception
handling for file saves, specifically to ensure that the exceptions that are
caught are as general as possible, just in case an "over quota" exception, a
timeout, or a failure to sync drives, or some slightly unusual exception that
may be thrown in the myriad of layers that we are using is somehow being
allowed to slip through unnoticed.
Thanks
Peter
Original issue reported on code.google.com by cspwc...@gmail.com on 27 Mar 2012 at 6:14
Original issue reported on code.google.com by
cspwc...@gmail.comon 27 Mar 2012 at 6:14