[Qt-interest] QAbstractFileEngine problems

[Thiago Macieira]

On Wednesday 7. July 2010 20.42.02 Jason H wrote:
> Why not just keep what they have now... If Qt is moving to open governance,
> can't we just elect to keep it? It is fine if you deprecate it, as long as
> *something* replaces it, hopefully for the better. But to drop something
> because it is "slow" and replace it with nothing isn't really a good
> reason to 1) do it then and 2) do it ever.

Oh, yes it is. Performance is very important and if the current model imposes 
bottlenecks, the current model has to change.

> My current thinking is a VFS layer is cool as sh!t because it abstracts the
> structured storage of binary?) data from the implementation. Who doesn't
> want that?! KIOSlaves are the bomb!

Yeah, and most VFS functionality works on top of a non-virtual functionality. 
Qt doesn't offer a non-virtual method of accessing the filesystem and THAT is 
the problem.

> Also Scott, since the code is already LGPL'd why not keep it?

Scott replied to the wrong thread. He was asking about the VS addin.

> So far, Qt and I seem to be on a path to parting ways. QDom and
> QAbstractFileEngine. Why do my projects always collide with the future of
> Qt? What am I doing wrong? I am hoping that things like open governance
> means we get to say, no, "we want to keep this" a little while longer...
> At least with XML there are other APIs that can be adapted to XML
> processing as a whole is not dropped.

Like any other Qt 4 technology, the file engine support cannot be removed until 
Qt 4. So it's staying exactly where it is. Apps using that will continue to 
work. The same applies to QDom. The fact that I recommend against those 
technologies doesn't mean you must not use them. It's just a recommendation 
that you have to weigh in your decision.

Qt engineers will not further optimise those technologies. We think we can 
spend our time in things that give developers more value. At one point, like 
the QDom, we may even stop fixing anything but critical bugs. And for Qt 5, 
whenever that happens, we will remove any unmaintained technologies.

Open Governance means that you can step up and do the work. It doesn't mean 
you get to tell us where to work, but it allows you to put your money where 
your mouth is and step up for maintaining and improving the technology. 
(Alternatively, you can also use real money and pay commercial support or 
consulting)

It also means that if you feel strongly about a VFS layer, you can work on it. 
If you manage to make the abstract file engine that VFS layer that really works 
and is fast, then all the better. If you need to create new classes or even a 
new module, that's just as great.

And personally, I'd love to see a VFS layer. But it's not in our current plans 
to do one.

-- 
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
  Senior Product Manager - Nokia, Qt Development Frameworks
      PGP/GPG: 0x6EF45358; fingerprint:
      E067 918B B660 DBD1 105C  966C 33F5 F005 6EF4 5358
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 190 bytes
Desc: This is a digitally signed message part.
Url : http://lists.qt-project.org/pipermail/qt-interest-old/attachments/20100707/1f129036/attachment.bin