Re: [fossil-dev] Unversioned file size calculation bugs in /stat

2017-10-02 Thread Warren Young 
On Oct 1, 2017, at 8:31 PM, Richard Hipp  wrote:
> 
> Fixed now, I think.

Installed and verified, thank you!

> The problem was that fsize would sometimes get
> changed during the computation of the compression ratio, which in turn
> was throwing off the unversioned file size percentage.

Ah.  In case it wasn’t obvious, my misdiagnosis was because I was reading the 
label of the third line as “Unversioned Artifact Size”.  *facepalm*
___
fossil-dev mailing list
fossil-dev@mailinglists.sqlite.org
http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/fossil-dev

Re: [fossil-dev] Unversioned file size calculation bugs in /stat

2017-10-01 Thread Richard Hipp 
On 10/1/17, Warren Young  wrote:
> The “unversioned” lines on this page make no sense:
>
> https://tangentsoft.com/pidp8i/stat
>
> Look first at the fifth line: how can the unversioned files occupy 484% of
> the repository space?

Fixed now, I think.  The problem was that fsize would sometimes get
changed during the computation of the compression ratio, which in turn
was throwing off the unversioned file size percentage.


>
> Also, is line 137 in stat.c right?
>
> pct = (iStored*100 + fsize - 1)/fsize;

That causes the size to be rounded up rather than rounded down.  I
have now changed this to round to the nearest.


-- 
D. Richard Hipp
d...@sqlite.org
___
fossil-dev mailing list
fossil-dev@mailinglists.sqlite.org
http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/fossil-dev

[fossil-dev] Unversioned file size calculation bugs in /stat

2017-10-01 Thread Warren Young 
The “unversioned” lines on this page make no sense:

https://tangentsoft.com/pidp8i/stat

Look first at the fifth line: how can the unversioned files occupy 484% of the 
repository space?  

If you take the “4 files” link from that line, it totals to 15.3 MB, but on 
line 3 of the /stat output, it says 101.0 MB, which is clearly where the 
calculation gets thrown.

I looked into fixing it, but it’s not clear what the simple fix is.  Over in 
uvlist_page(), it’s doing the total in C code, whereas in stat_page() it’s 
doing it in SQL.

Also, is line 137 in stat.c right?

pct = (iStored*100 + fsize - 1)/fsize;

Shouldn’t that be

pct = iStored*100/fsize;

?
___
fossil-dev mailing list
fossil-dev@mailinglists.sqlite.org
http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/fossil-dev