[PATCH 1/2] Add a help button to every Anaconda screen

Anne Mulhern amulhern at redhat.com
Thu Aug 28 12:44:33 UTC 2014





----- Original Message -----
> From: "Vratislav Podzimek" <vpodzime at redhat.com>
> To: "anaconda patch review" <anaconda-patches at lists.fedorahosted.org>
> Sent: Thursday, August 28, 2014 3:27:46 AM
> Subject: Re: [PATCH 1/2] Add a help button to every Anaconda screen
> 
> On Wed, 2014-08-27 at 11:38 -0400, Anne Mulhern wrote:
> > 
> > 
> > 
> > ----- Original Message -----
> > > From: "Vratislav Podzimek" <vpodzime at redhat.com>
> > > To: "anaconda patch review" <anaconda-patches at lists.fedorahosted.org>
> > > Sent: Wednesday, August 27, 2014 10:43:58 AM
> > > Subject: Re: [PATCH 1/2] Add a help button to every Anaconda screen
> > > 
> > > On Wed, 2014-08-27 at 10:36 -0400, Anne Mulhern wrote:
> > > > > > Actually,
> > > > > > 
> > > > > > I think that, ideally, we would store all this data in an sqlite
> > > > > > database, rather than in separate files in separate directories.
> > > > > > We would do the same for the CLI help text, now that it
> > > > > > has gotten so mature, and we would be able to use _exactly_ the
> > > > > > same database schema for both, since, at least in theory, CLI
> > > > > > help text should be available in any language.
> > > > > > 
> > > > > > However, we would have to handle version control of the structure
> > > > > > of the database and also its contents.
> > > > > > 
> > > > > > It might be that adding an sqlite dependency to anaconda
> > > > > > is not so attractive to some.
> > > > > > 
> > > > > > Also, I don't know how translation is expected to work with the
> > > > > > contents of databases.
> > > > > > 
> > > > > > So maybe this idea is a bit too ambitious. But now is as good
> > > > > > a time to think about it as any.
> > > > > I don't think we need such a big-hammer approach. Using directories
> > > > > for
> > > > > translations is really common and not so hard.
> > > > > 
> > > > 
> > > > "too ambitious" is my way of saying using a database would be
> > > > harder but also the kind of thing you might see in a more mature
> > > > application.
> > > > I don't think there's any question that using directories for this
> > > > problem is the path of least resistance. A lot of the hardness
> > > > of using the database might however be due to unfamiliarity and
> > > > novelty,
> > > > rather than anything really fundamental. IMHO, we should consider
> > > > not taking the path of least resistance occasionally, i.e., doing
> > > > something unfamiliar and novel and hence a bit difficult for us at the
> > > > time,
> > > > especially as technologies mature and the path of least resistance
> > > > turns
> > > > inexorably into the stuff that everybody used to do before the
> > > > clever ones learned a better way. On the other hand, I'm certainly
> > > > willing to be persuaded that right now is not a good time to choose
> > > > to try to break out of this particular rut.
> > > > 
> > > > Is it the case that the CLI help text should, at least in theory,
> > > > be translatable?
> > > I don't think the CLI options' (--ks=, --keyboard=, etc.) help text will
> > > ever be translated. But the text mode and GUI help texts (for earch
> > > "screen") should be. The same applies to ransom notes and license
> > > agreement.
> > 
> > Well, given that and given that dcantrell has opened the door to at least
> > some form of TUI help, I'ld like to ask mkolman to hold onto his current
> > GUI
> > help patches a bit longer, while it is considered how they'ld relate to the
> > rest
> > of the multi-language stuff and help text.
> Why? The proposed code uses the same approach we use in all the other
> places. And publican (the tool for creating documentation) uses
> directories such as en_US/, cs_CZ/, etc. too.
> 

In order to avoid having avoidable reversals of coding decisions
baked forever into the git history. Unnecessary commits, especially
ones that mostly, but not quite, reverse previous commits, make
cherry-picking, rebasing, etc., that much more painful. Due diligence
in planning for something we believe is going to happen anyway
ought to reduce the occurrence of these somewhat.

- mulhern

> --
> Vratislav Podzimek
> 
> Anaconda Rider | RHCE | Red Hat, Inc. | Brno - Czech Republic
> 
> 
> _______________________________________________
> anaconda-patches mailing list
> anaconda-patches at lists.fedorahosted.org
> https://lists.fedorahosted.org/mailman/listinfo/anaconda-patches
> 


More information about the anaconda-patches mailing list