Finally I downloaded the latest I7 for Windows. Unluckily this version does not work anymore with “Italian” by Massimo Stella (whose latest version is available here sourceforge.net/projects/milleun … p/download) while the previous 6L38 worked fine.
Since one of the most interesting features of the latest I7 was the advanced support to non-english languages, I wonder if anyone has experienced similar problems with other translations (Spanish and French in particular).
The error I get compiling a project just after the adding of “(in Italian)” to the the title is the following:
An internal error has occurred: Parse tree broken. The error was
detected at line 1656 of “inform7/Chapter 13/Parse Tree.w”. This should
never happen, and I am now halting in abject failure.
I didn’t have the time to look at the new version of Inform because of the French Comp, which takes me all my time.
It ends today, so after that I’ll try to see if I can make the French extension work.
There’s certainly a problem. You can demonstrate this with a very cut-down test:
/c/ essere ...
-) in the Preform grammar.
In Italian essere nel is a verb meaning to be in.
In Italian essere nello is a verb meaning to be in.
In Italian essere nella is a verb meaning to be in.
In Italian essere nei is a verb meaning to be in.
In Italian essere negli is a verb meaning to be in.
In Italian essere nelle is a verb meaning to be in.
The Kitchen is a room.
I don’t know if there’s a workaround.
To be clear, the Preform system still counts as “experimental and undocumented”.
I hope to find one, since all my framework (to write interactive fiction in italian) is based on that extension.
For what is worth, you can get the same error with the English language.[code]“Test”
There is a room.
[To retrofit is a verb. [That one works fine]]
In English to retrofit is a verb. [That one makes Inform crash.][/code]
So apparently the problem is linked with the sentence “In [language] [a verb] is a verb”.
I hope it will be fixed soon, too.
Oh, thanks. I’ll add that to the bug report.
Spanish extension uses the same preform grammar so, it is in wait-till-bug-dies mode ON.
According to you, is there any workaround?
well, this is not a workaround per se, but…
You can dispense with some language feature by commenting the “is a verb” lines and fixes those related :
- [verb] in language-adaptative-responses sets.
- code that use the specifics verb in the source.
(Spanish extension does not use extensively spanish-verbs neither in Responses and source code.)
Mainly in order to have a 6M62 compilable code; seeking other problems.
Well, I don’t think it’s worth changing all the responses and all my code which uses the verbs, to revert when the bug will be fixed.
I prefer to wait.
Me too: the Sarganar’s idea is not very practicable in the italian case, because the broken feature is extensively used.
However the Sarganar’s idea is the only way to go (other bugs apart) if an italian author wants to use 6M62.
I saw from Peter’s Inform issues summary that this is still (unsurprisingly) causing some grief. Here is the latest word from Graham on what is going on with the language issue:
“This is a part of Inform currently undergoing rewriting, with the aim of improving Inform’s adaptability to non-English languages (for example, to support noun cases in German). The specific issue in 0001817 is gone, but Preform-hacking code will need to be adjusted on the next release because the internal grammar is changing. I hope to have a preview release at some point for translators to experiment with.”
Thanks for the reply! (and of cours, we shouldn’t complain that much, a lot of work is being done.)
And thanks Peter for all the work.
Very good news! I can’t wait to play with the new system.