[MERGE] Split test_readv_with_adjust_for_latency to achieve better defect localization :)

[John Arbash Meinel]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Vincent Ladeuil wrote:
| Well, the subject is not a lie, but...
|
| The true reason is that test_readv_with_adjust_for_latency was:
|
| - creating a file via local file system (lfs for short),
| - querying that file via http,
| - creating the same file via lfs with a totally different content
| - querying that file via http for the new content.
|
| Now, in the context of the local_test_server plugin, lighttpd (a
| web server) get tricked and served the second request with the
| content of the first file.
|
| There is little interest in debugging lighttpd further so I chose
| the easy path and split the test (which was too big anyway ;).
|
|     Vincent
|
|
| P.S.: John ! Look ! No typo in lighttpd ! :)
|
|

BB:approve

I don't see a specific reason why the test couldn't be split into two.

John
=:->

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (Cygwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iEYEARECAAYFAkhGy4gACgkQJdeBCYSNAAOo6wCgoVSKEmXdvPu6s3M7eNoS5/uK
cgAAn0xhP/Anydm/vPB7QAC4vt6sXpzE
=FskO
-----END PGP SIGNATURE-----