Re: Use of pseudo-tags to mark translation issues
On Fri, Sep 19, 2003 at 08:23:17AM +0200, Philippe Faes wrote:
> On Fri, 2003-09-19 at 02:39, Adam Heath wrote:
> > On Thu, 18 Sep 2003, Philippe Faes wrote:
> > > I can see that is is useful for us to cathegorise bugs, and to be able
> > > to find i18n bugs automatically. However, the BTS already has a system
> > > of tags AND priorities.
> > > Why aren't the existing tags system expanded to include the pseudo-tags
> > > that Adam suggests? (eg. doc,arch,intl,assert,...) I don't see a good
> > > reason to have tags AND pseudo-tags. They serve the same purpose don't
> > > they?
> > Because the number of tags would get unwieldy. And the bts doesn't have
> > priorities, it has severities, which are different. And even if it did have
> > priorities, they wouldn't be enough to classify a bug.
> I believe my argument still stands.
> Our options are: a zillion standardised pseudo-tags (which people would
> have to look up when they use them) plus a few real tags which are
> supported by 'reportbug'
> OR: a zillion + a few real tags, all supported by the BTS and by
> I am not convinced that the first option is better. In fact it looks
> more like a hack than a permanent solution. I still think we should
> consider moving the pseudo-tags to the BTS tag system.
I consider the pseudo tags as an experimentation of which usefull real patch
would have to be added to the BTS.
I'm not sure that the mechanism proposed on d-dpkg is enough for i18n bugs.
For example, there is no good solution yet to say that a l10n bug is solved
by translator, and just need a commit from maintainer.
But I'm sure that such good idea will come from experimentations with those
A related argument for the use of pseudo tags and not real tag is that this
is really easy to add a pseudo tag (just retitle the bug) while adding a
real tag needs to offer support in the BTS and all related tools. If one
week after, we discover that it would be more pleasant to call the pseudo
tag another way, too bad.
The good solution in my mind would be another dozen of standardized tag,
once we discover which from the zillion of pseudo tags are really usefull.
Source is provided to this software because we believe users have the right
to know exactly what a program is going to do before they run it.