Page 7 of 7

Re: Blake's 7 feedback and review thread

Posted: Wed Nov 29, 2017 8:29 pm
by Chema
I am not sure how this will be organized, and I also have little experience with repositories, but you are updating the main trunk with your on going modifications.

I don't know if that is the way to go, but be sure the code still compiles properly. I may update things soon and would like to avoid ending up with bogus sources in my laptop. It is my only backup!

Also tou are not using any accents or special French symbols. If you're going to use them, remember I have to tweak my compiler to handle them! Please read my post about all this.

Re: Blake's 7 feedback and review thread

Posted: Thu Nov 30, 2017 5:49 am
by coco.oric
I don't know if that is the way to go, but be sure the code still compiles properly. I may update things soon and would like to avoid ending up with bogus sources in my laptop. It is my only backup!
And i'm not experienced with repos. I use the icon from the svn application.
I've a backup too.

All that i do is (up to now) is put a #ifdef FRENCH ... #endif in some .os files (the first ones in the directory by alpha order).
So it should'nt modify others compilations.

I'll stop until you'll update the game then i'll fork it repo (branch-tag). If i understood correctly, this will keep the original repo, letting people changing french sentences (unless i'm alone to traduce it and in this case, i can keep all the files on my hd and expecting the translation within one year, for christmas 2018)

Re: Blake's 7 feedback and review thread

Posted: Wed Dec 06, 2017 11:14 pm
by Chema
Hi. Not sure if we missed this reference (for the first post: http://mag.mo5.com/actu/127023/blakes-7 ... -sur-oric/)

Dbug, any chance of a new FloppyBuilder which corrects the CRC thingy? Oh, and I think Fabrice Frances noticed the generated interleave was not the one selected in the config file, and it was not even constant (going off by 1 or 2 places somewhat randomly - I can't find the email to check the values now).

I have a version which runs on Jasmin, Microdisc, and Telestrat with better disk routines. The only thing it cannot do is booting on a Telestrat (it is impossible with the same disk), but Fabrice sent me a small program that asks for a Microdisc/Jasmin disk and boots it.

Re: Blake's 7 feedback and review thread

Posted: Thu Dec 07, 2017 8:36 am
by Dbug
Chema wrote:
Wed Dec 06, 2017 11:14 pm
Hi. Not sure if we missed this reference (for the first post: http://mag.mo5.com/actu/127023/blakes-7 ... -sur-oric/)
We did not have it, just added it.
Chema wrote:
Wed Dec 06, 2017 11:14 pm
Dbug, any chance of a new FloppyBuilder which corrects the CRC thingy? Oh, and I think Fabrice Frances noticed the generated interleave was not the one selected in the config file, and it was not even constant (going off by 1 or 2 places somewhat randomly - I can't find the email to check the values now).
I can try to see for a fix this weekend.
Which version are you using?

Re: Blake's 7 feedback and review thread

Posted: Thu Dec 07, 2017 9:37 am
by Chema
I think it is the last one; when I run it with no arguments, it displays:

Code: Select all

FloppyBuilder - Version 1.001 - This program is a part of the OSDK

Re: Blake's 7 feedback and review thread

Posted: Thu Dec 07, 2017 11:27 am
by iss
Attached is 'Floppy.cpp' from the SVN with fixed CRC. I added missing 'compute_crc' calls where they are needed.
Floppy.cpp.zip
(8.83 KiB) Downloaded 7 times

Re: Blake's 7 feedback and review thread

Posted: Sat Dec 09, 2017 12:03 pm
by Dbug
Could you check with this version, it should have the CRC correctly fixed.
FloppyBuilder.zip
Floppy builder 1.2
(104.41 KiB) Downloaded 5 times
Regarding the interleave, if you could give me more details about the problems :D

Re: Blake's 7 feedback and review thread

Posted: Mon Dec 11, 2017 8:58 am
by Chema
Thanks for the update! I will check as soon as possible. I have been too busy lately :oops:

About the interleaving factor, while helping me, Fabrice Frances created a physical disk with writedsk and examined it with Nibble. Writedsk creates the correct sector order, but he told me that:
> the order of sectors I can see with this tool :
> 5,9,12,15,2,6,10,13,16,3,7,11,14,17,4,8,1
> ... which I find a bit strange though, because the interleaving is not
> regular... Starting from sector 1, the intervals are
> +4,+4,+3,+3,+4,+4,+4,+3,+3,+4,+4,+4,+3,+3,+4,+4
> Is there a purpose for this ?
when I asked for an interleave of 5. I guess it is simply a different interpretation about how interleaving works, because in the order produced by FloppyBuilder sector i+1 is 5 sectors ahead sector i, which is what I supposed should be. I will ask FF again.

EDIT: I have asked FF and he confirms the order seems to be correct, so no need to change anything :mrgreen:

Re: Blake's 7 feedback and review thread

Posted: Mon Dec 11, 2017 7:18 pm
by Dbug
The main reason for the interleave to not be regular, is that when you have reached the end of the track, you need to loop back to start, and of course you can't reuse the sectors that have already been set, so you may have to jump over some.

So basically the ideal interleave pattern is also pretty much dependent of the number of sectors per track :)