[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: 199d6 feature creep...



Quoting Greg Pettyjohn:
> Forgive me, the former software tester never fails to criticise
> even a good feature...
> :-)

No, no! Please! Thanks for the great feedback!

> 1. If you compress an s-expr and then save, as you promised, the file
> format is now binary instead of text. If you expand all s-exprs and
> then save again, the file will be text again. This could be a
> nuisance if someone needed to look at the file using some editor
> besides DrScheme.
>
> 2. I noticed that you have a "save definitions as text". When I save
> a file with collapsed s-exprs this is what happens:
>
> collapsed version:
> (define test
>   ...)
> 
> text version:
> (define test
>   .)
> 
> (The same phenomenon is observed with cut and paste). My hope was
> that "save definitions as text" would expand all collapsed s-exprs in
> the text version. This would help to avoid the nuisance described
> above.

This is a bug. I have committed and tagged a fix so that when you copy
and paste between drscheme's you preserve the collapsed nature, but
when you paste to another program or save as text, you get the original
text.

> 3. I collapsed an s-expr and then saved. Then I did a: (load
> "test.scm") where test.scm is the name of the file with collapsed
> s-exprs. It worked just fine when I load from DrScheme, but fails
> when I load from MzScheme. This could be a real nuisance if you are
> working on a MzScheme program, but using DrScheme as the editor.

Sadly, there is nothing to be done, if you must use mzscheme. There are
plans, however, so that you will be able to load the file in mred,
which hopefully would be good enough for most purposes. I expect this
support to be in mred by the time we release 200 (if not sooner;
Matthew Flatt might be able say more precisely).

> 4. I collapsed an s-expr and then searched for a string that I knew
> was in the s-expr... no dice. This could be a nuisance if you did a
> "replace all" only to discover that it overlooked some hidden
> instances.

Yes, I'm not sure what should be done about that. Ideally, the
collapsed sexpression would pop open when the search string was in it
(or partially in it). Sadly, the text% class doesn't support this very
well, and it would be a fairly major undertaking to fix it. I'm afraid
we are going to have to live with this shortcoming for a while.

Robby