On Tuesday 11 March 2008 12:41:58 Wookey wrote:
> On 2008-03-10 20:37 +0100, Dirk Behme wrote:
> > Charles Manning wrote:
> > > Hi All
> > >
> > > I have entered YAFFS for Google Summer of Code. We'll know in a few
> > > days if we're accepted.
> > >
> > > We have put a preliminary ideas list up at 
> > > http://www.yaffs.net/ideas-list
> > >
> > > Any more ideas? If so, please send them to (on or off list) so that I
> > > can include them.
> >
> > One additional: I'm not so familiar with past discussion, but what's
> > about doing some work to get YAFFS itself closer to mainline
> > integration? Don't know what will be necessary for this though.
>
> Whilst this is highly worthwile, it is not really the sort of project
> that normally gets SoC approval. SoC is intended to let students learn
> to write new code on discrete projects small enough for them to have a
> reasonable chance of succeeding. Tarting up existing code to get it
> into kernel shape doesn't really count.

This is indeed true and I don't think fixing mkyaffs/unyaffs are enough on 
their own. Perhaps when combined with other work this could be reasonable.

The main reason I'm asking the list is to see if we can form a big enough 
project that encompasses this stuff,

The performance measurement/analysis is sufficient in its own right as a 
project I think.

>
> On the subject of actual integration, one of the main things that is
> necessary is to get the inband tags code into CVS, because this makes
> YAFFS much better-matched to the way MTD in the kernel is headed, and
> thus should prevent resistance to it going in.

I shall be away next week. The week after I shall by trying to consolidate all 
the inband tags stuff etc.

>
> Just to note another recent request for kernel inclusion I had OPen
> Embedded asking what the chances of YAFFS getting into mainlines were,
> as they'd really like it to be much easier for people to just pick
> 'YAFFS' when building an OE image.
>
> Wookey