Hi, Actually, I already submitted a patch to qmailadmin for this (against 1.0.12) about 2 weeks ago. The patch depends on vpopmail-5.3.19.
qmailadmin was changed to not have to understand anything, but instead call an API that returns the value as an int. The parsing belongs inside the API routine. Currently, qmailadmin doesn't read or understand *anything* from the user except a field expected to be in megabytes from the POST. It uses that field, or the default from the vlimits API if the field isn't supplied. vmoduser needs to be updated to write a properly formatted data to the field. Currently, it just writes the information verbatim from the command line. Thanks, Brian > On Monday, March 17, 2003, at 05:56 AM, Brian Kolaci wrote: > > The abbreviations for Megabytes & Kilobytes "M", "m", "K" and "k" > > have all been deprecated in the latest release of vpopmail > > and qmailadmin. You should use all numbers now. The limits API > > doesn't support this (and probably won't due to the database > > holding numbers and not strings for these values). > > > > Read the vpopmail README.quotas file and you'll see the string format > > for quotas and an explanation. Its along the lines ######S,####C > > (size & count only, M & K for meg & kilo are gone now). > > > > Just use a number. > > What do you guys think of the following: > > 1) qmailadmin should read and understand the old-style format of xxM > (quota = xx*1024*1024) or xxK (quota = xx*1024). > > 2) qmailadmin should read and understand the new-style format of xxS > (and xxS,yyC) (quota = xx) and even the format of plain xx (without the > S). > > 3) qmailadmin should always write a modified quota out as #####S. > > 4) vmoduser should be updated to convert xxM and xxK formats into > straight bytes (or, even better, xxS format). > > I'm willing to make these changes and submit a patch, but I want to > make sure we all agree that it's the best way to go. > > FYI, I need to update my last patch anyway, since I checked my C > reference on sprintf formatting strings and learned that %lf is the > correct format for a double. It appears that at least under Linux, %lf > and %f are treated as a double, but I don't know if other platforms are > that way. > > -- > Tom Collins > [EMAIL PROTECTED] >
