[blivet:master+? 0/9] More work on blivet.size

David Lehman dlehman at redhat.com
Thu Oct 23 20:43:33 UTC 2014


On 10/23/2014 03:35 PM, Anne Mulhern wrote:
>
>
>
>
> ----- Original Message -----
>> From: "David Lehman" <dlehman at redhat.com>
>> To: anaconda-patches at lists.fedorahosted.org
>> Sent: Thursday, October 23, 2014 4:05:43 PM
>> Subject: Re: [blivet:master+? 0/9] More work on blivet.size
>>
>> On 10/23/2014 12:57 PM, mulhern wrote:
>>> Fixes bugs of varying importance and simplifies a bit.
>>>
>>> Removes a change that was intended to make sort of ambiguous specs, like
>>> "8 M", which sort of could be "8 MiB" or "8 MB" be interpreted as "8 MiB".
>>> This is sort of a good idea, because we like to prefer binary, but
>>> problematic
>>> because the change broke a lot of stuff and was inconsistent within size,
>>> convertTo() interpreting "M" as "MB". This is why I went back to previous,
>>> where "M" is interpreted as "MB". To do it properly would require a bit of
>>> thought, as _makeSpecs() believe that "M" is a specifier for "MB" and not
>>> for
>>> "MiB".
>>
>> We need to find a way to retain the current behavior ("X" -> "XiB").
>> Assuming decimal makes no sense at all, really. It sounds like we also
>> need to fix convertTo to do the same.
>>
>> I haven't read the patches yet, just wanted to speak up about the above.
>>
>> David
>>
>
> Things are easy if we choose not to interpret anything as decimal units.

We should interpret only explicit decimal units as decimal units. We 
can't get away with not being able to understand/handle decimal units at 
all, though.

>
> I would be just fine w/ that...if it would work for all.
>
> We already don't even pretend to consider attempting to display values
> (via humanReadable()) with decimal units.

Right. We only want to use them (for anything) if explicitly requested 
by the user.

>
>>>
>>> Made convertTo() handle translated specs. The reasoning behind this is that
>>> it was easy and convertTo() was treating some values as if they were
>>> translated and others as if they were not. So adding a switch for
>>> translation,
>>> and making it consistent within that seems fine. Making it only handle
>>> English specs would be very easy, too, however.
>>>
>>> mulhern (9):
>>>     Use b"", not u"", for _EMPTY_PREFIX.
>>>     Translate less eagerly in Size.humanReadable()
>>>     Make _lowerASCII() python 3 compatible and add a method header.
>>>     Change _makeSpecs() so that it translates for the empty string.
>>>     Add new method getSpecifiers().
>>>     Rewrite _parseSpec()
>>>     Rewrite convertTo().
>>>     Remove now useless imports, constants, functions.
>>>     Add a number of new tests.
>>>
>>>    blivet/size.py     | 195
>>>    ++++++++++++++++++++++++++++++-----------------------
>>>    tests/size_test.py | 127 ++++++++++++++++++++++++++++++----
>>>    2 files changed, 222 insertions(+), 100 deletions(-)
>>>
>>
>> _______________________________________________
>> anaconda-patches mailing list
>> anaconda-patches at lists.fedorahosted.org
>> https://lists.fedorahosted.org/mailman/listinfo/anaconda-patches
>>
> _______________________________________________
> anaconda-patches mailing list
> anaconda-patches at lists.fedorahosted.org
> https://lists.fedorahosted.org/mailman/listinfo/anaconda-patches
>



More information about the anaconda-patches mailing list