[Info-vax] Potential loss of data problem in sftp client in TCP/IP Services ?

[Simon Clubley]

On 2012-03-13, Simon Clubley <clubley at remove_me.eisner.decus.org-Earth.UFP> wrote:
> On 2012-03-13, VAXman-  @SendSpamHere.ORG <VAXman- at SendSpamHere.ORG> wrote:
>> In article <jjnc5c$1l5$1 at dont-email.me>, Simon Clubley <clubley at remove_me.eisner.decus.org-Earth.UFP> writes:
>>>
>>>Thanks for doing this. Just for confirmation, was this V5.7, ECO3 ?
>>
>> Yes.
>>
>
> [snip]
>
>>
>> $ CREATE SIMON.TEST
>> AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
>> BBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBB                               
>> CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC
>> DDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDD
>> EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE
>> FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF
>> GGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGG
>> HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH
>> IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
>> JJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJ
>> KKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKK
>> LLLLLLLLLLLLLLLLLLLLLLLL
>> MMMMM[ Exit ]
>> $ sftp -oPort=###### vaxman at 192.168.#.###
>> sftp> put simon.test
>>
>> I see everything when I open the file with 'gedit'.
>>
>
> Thank you for running this test Brian.
>
> This is consistant with the reports I have received from support over
> the last few hours that they cannot reproduce it in V5.7 ECO 3 but can
> only reproduce it in the Engineering images they sent me.
>
> On the plus side, this means the problem does not appear to be in
> any public ECO kit. On the down side, it looks like the Engineering
> images have a nice new file truncation bug in them...
>
> (I'm assuming here that once they find a bug in one supported version,
> they still check for the same bug in all supported versions.)
>

I should expand on that last statement since it's clearly open to
misunderstanding; it was just something which occurred to me while posting.

Although, as mentioned, I am running TCP/IP V5.6, and hence the image id
in the engineering image HP supplied has a V5.6 id, the image HP supplied
has been very clearly built from the latest code base because of all the
enhanced functionality in it.

I realised I've been implicitly assuming that HP are fixing common mode
problems in all supported versions at the same time just like DEC have
always done. However it also occured to me, that due to the types of
procedural problems we have seen recently, this may not be the case.

Hence it suddenly occured to me it was possible this was a bug fixed on
the V5.7 stream which didn't get moved into the V5.6 stream even though
the V5.6 stream is clearly using code from the latest code base; therefore
this might be a known bug which didn't get fixed when it should have been.

However, thinking about it, it's far more likely that this is a new
problem which would have also affected the next V5.7 ECO as well if
it wasn't fixed; I have not seen any evidence common mode problems are
not getting fixed across all versions once they are identified.

Sorry for the confusion,

Simon.

-- 
Simon Clubley, clubley at remove_me.eisner.decus.org-Earth.UFP
Microsoft: Bringing you 1980s technology to a 21st century world