Re: [Fwd: Re: Wikipedia criticism about root]

From: Andy Buckley <andy.buckley_at_durham.ac.uk>
Date: Fri, 30 Jun 2006 10:22:19 +0100


Valeri Onuchin wrote:
> Hi Andy, {sorry, team. I do not want to go into discussions.
> I just want to "extract constcructive things" from this "holly war").
>
> First af all, thanks for the critics (!).
> My personal experience convinced me that even unfair and FUD critics
> can help to improve "the product".

Before continuing, I'd like to address the accusations of "holy war" and "FUD"! What I've done is supply some opinions based on my own observations while using ROOT and observing its development over a roughly 5 year period. I've produced a whole raft of accusations about the stability or sanity of CINT, the usability of ACLIC, and persistency format problems (which have been greatly expanded upon by others) and I've yet to be accurately contradicted on any of them. I freely admit that I know less about ROOT's current status than the developers, which is why I'm surprised that none have actually produced corrections on technical points.

Since I've produced examples and I've mainly received replies of the form "Authority A says that ROOT is good and therefore it must be", I'd be less hasty in applying the FUD label to *my* postings!

And lest we forget, the original point of this thread was not to engage in polemics about what's good and what's bad. It was to work out if the criticisms on the Wikipedia page are to be considered accurate or not. I don't think anyone's shown that they are not --- in fact the single edit added to the wiki (I guess as a result of this discussion) has added further criticism which I was unaware of.

> You wrote:
> "class structure or GUI are my primary concerns. I find it
> hard to use ROOT as a reliable component in a well-engineered
> application, because many of the design idioms in ROOT data classes seem
> designed for the convenience of the ROOT GUI rather than a user who only
> wants to use a select few classes."
>
> I didn't find in your text the critics on "ROOT GUI" (only mentioning about
> "ROOT guibuilding is added to the ROOT core" - which is not true).
> What do you mean by the words "... GUI are my primary concerns"?

The GUI may not be in the Core library, but objects which are not strictly GUI do seem to be designed for its convenience. The best example I can think of is that when you pass a histogram to a canvas (which, despite being displayable, is not necessarily displayed, therefore not strictly a GUI object to my mind), then delete the canvas. The histogram will be deleted. What if you were just "lending" it to the canvas? But this approach makes it easy to avoid memory leaks in the GUI, which is the only reason I can imagine for it working that way. Using shared pointers would avoid the memory leaks and work as expected, but CINT didn't understand them when I tried. Perhaps this has changed?

But you're right that the actual user interaction with the Qt GUI is not my concern. Do you have any comments on what I/we've been saying about data formats, OO design or the CINT interface?

Andy Received on Fri Jun 30 2006 - 11:22:36 MEST

This archive was generated by hypermail 2.2.0 : Mon Jan 01 2007 - 16:31:59 MET