On Fri 2004-03-12 at 10:35:18 -0600, kfogel@collab.net wrote:
> Benjamin Pflugmann <benjamin-svn-dev@pflugmann.de> writes:
> > When reading source code which you are unfamiliar with, functions
> > named dothis(), dothis2(), dothis3() are one of the worst things,
> > IMHO.
>
> I think this "2" (or "N") suffix is actually something of a
> convention. I've seen it before in more than one place, and never
> been confused about what it meant: an API needed variants. Consider
> popen(), popen2(), popen3().
Bad example, I don't know Python :)
Well, I looked it up (that's why I know it's propably Python) and
popen2() and popen3() seem like an exception to me, because there is
some easy "memory hook" indeed (the number is the amount of pipes
returned), not simply an meaningless enumeration. Well, at least for
those two... popen4() seems be one of those abominations that I fear,
and I found no popen() in these Python docs so I don't comment on
that.
> It's hard to say in absolute terms "It's confusing" or "It's not
> confusing". All I can tell you is that I've not been confused by it,
> and other APIs have chosen a similar course.
If you have no problems with enumerations, and it's only me, fine.
When there's no problem, don't fix it. :-)
> I can put a note in HACKING about what '@deprecated' means -- i.e,
> that it's the *API* that's deprecated, not necessarily the name. We
> should simply assume an arbitrary amount of API change when 2.0 comes
> out. It's not like it'll be that hard for anyone to fix their apps.
> You try to compile your code, it doesn't work, you go do the obvious
> thing (removing the "N" suffixes from functions, basically).
Jupp, that wasn't my concern.
HAND,
Benjamin.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Fri Mar 12 20:18:31 2004