De-tarballing
Pages: 1 2
Dave Higton (281) 668 posts |
Sorry, I hadn’t realised that there are three links – two of them are to the same zip, the other is to a self-extracting archive. I can’t remember which one I used. I’ll have to collect both and try each. |
Trevor Johnson (329) 1645 posts |
Thanks. I can do the same – although you’ll probably beat me to it! If they’re different, presumably they’ll have different version numbers… or at least different timestamps. |
Andrew Conroy (370) 740 posts |
I downloaded the src-disc.5.17.tar.bz this morning and set it decompressing with UnTarBZ2. I’ve just logged on to the BB from work and it appears to have fully decompressed it. I had alignment exceptions off. |
Dave Higton (281) 668 posts |
The RISC OS source code page really doesn’t show the MD5 hash of the files, does it? I’m not dreaming? |
Trevor Johnson (329) 1645 posts |
You’re not. That certainly looks like the time after the date! |
Dave Higton (281) 668 posts |
OK… so how am I supposed to check that the MD5 hash of the archive is the same as that computed by ROOL and therefore that the tarball I’ve downloaded is correct? |
Dave Higton (281) 668 posts |
Compare with (most of?) the other download pages, which seem to have an MD5 hash along with each file. |
Dave Higton (281) 668 posts |
Wow. It’s a good job I was sitting down… I just noticed that UnTarBZ2 has completed! That’s UnTarBZ2 as created from the self-extracting archive, working on the OMAP source tarball from this morning, with alignment exceptions off and the system heap set to its maximum value of 4080K (and why, for the benefit of my curiosity, is that not 4096K?) I’ll have to try again with UnTarBZ2 from the zipped version. Anyway, now I’m in a position of strength. I have a fresh set of source files to work with! |
Dave Higton (281) 668 posts |
I’ve e-mailed info@riscosopen.org to ask if the fact that only some files and archives have an MD5 hash is known. I’ve also set a version of UnTarBZ2, unzipped from the zipfile on the ROOL site, working on the same tarball. We’ll see if the result is any different. |
Dave Higton (281) 668 posts |
And, yes, indeed, it is different… it cr@ps out a few minutes in. So the versions of UnTarBZ2 from the self-extracting utility and the zip will probably turn out to be different. That’s a test for tomorrow. |
Trevor Johnson (329) 1645 posts |
Beware that some components will fail (with newer sources than those you’ve extracted) due to the recent renaming of macros. |
Trevor Johnson (329) 1645 posts |
Perhaps some pages don’t run/work properly with make-md5.rb. [Edit: Or maybe it’s here instead.] It might be worth logging a bug report (even if it just points here), so it’s recorded for easy reference. |
Trevor Johnson (329) 1645 posts |
I’m running the 1 2011-03-15 03:45:02 |
Trevor Johnson (329) 1645 posts |
The extraction completed in both cases. Will have to check the contents later. |
James Lampard (51) 120 posts |
I think there is 16K of space for the SVC stack permanently allocated in the same dynamic area. |
Trevor Johnson (329) 1645 posts |
The contents certainly appear to be the same in each case, with files correctly filetyped and suffixes removed :-) |
Trevor Johnson (329) 1645 posts |
The process seems to be working for me now. I have a suggestion for possible inclusion in a future update to !UnTarBZ2:
|
Trevor Johnson (329) 1645 posts |
Is the current arrangement a revision? (I didn’t take a screengrab.) |
Steve Revill (20) 1361 posts |
People should certainly be aware of this topic. |
Pages: 1 2