From kde-devel Thu Oct 18 20:33:17 2007 From: Michael Pyne Date: Thu, 18 Oct 2007 20:33:17 +0000 To: kde-devel Subject: Re: ktempfile Message-Id: <200710181633.21179.michael.pyne () kdemail ! net> X-MARC-Message: https://marc.info/?l=kde-devel&m=119273967428580 MIME-Version: 1 Content-Type: multipart/mixed; boundary="--===============0578178024==" --===============0578178024== Content-Type: multipart/signed; boundary="nextPart2381911.cyVky6mRa4"; protocol="application/pgp-signature"; micalg=pgp-sha1 Content-Transfer-Encoding: 7bit --nextPart2381911.cyVky6mRa4 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Content-Disposition: inline On Thursday 18 October 2007, Lubos Lunak wrote: > BTW, an opinion of two other kernel developers (that don't have any > specific fs knowledge) is that our response to users complaining about > losing data this way should be telling them to kiss their data good bye a= nd > buy UPS or use a filesystem where it works if they don't like it. Instead of changing KSaveFile or KConfig couldn't we just change=20 KBookmarkManager and the KonqHistory code to not use KSaveFile and instead= =20 just use QFile and rename() (or add a flag to KSaveFile to allow not syncin= g=20 before the rename)? I don't think the problem is that KSaveFile calls sync() since I'd actually= =20 almost expect that behavior from its API documentation, but instead that co= de=20 uses KSaveFile where the atomicity/sync guarantee isn't essential. Regards, - Michael Pyne --nextPart2381911.cyVky6mRa4 Content-Type: application/pgp-signature; name=signature.asc Content-Description: This is a digitally signed message part. -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.6 (GNU/Linux) iD8DBQBHF8MQqjQYp5Omm0oRAtiiAKCCuqRDFSo2PgWlmArcX+5bAxZfvQCfZ61V CTwidPoVtfl038Cv2lTFl3U= =kxu4 -----END PGP SIGNATURE----- --nextPart2381911.cyVky6mRa4-- --===============0578178024== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline >> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe << --===============0578178024==--