Page 20 of 20

Posted: Wed 02 Oct 2013, 04:08
by technosaurus
I posted some light weight xpm tango icons here
over 200 icons each in both 32x32(gz compressed to 49kb) and 16x16(gz compressed to 27kb) versions
I did a some tricks to ensure that they shared an indexed color palette of 64 colors to help with compression and manually edited them to make them look decent. Even without compression they are significantly smaller than the RGBA icons included in the default Tango package.
I like the responsiveness of jwm with only xpm enabled (especially when the file system is already compressed - indexed xpm is smaller and faster)

Posted: Fri 11 Oct 2013, 19:40
by Iguleder
goingnuts, I wrote a udev replacement you could use. It relies on devtmpfs (i.e doesn't create device nodes), but mdev can compensate for that.

It's a small, single-process solution for loading kernel modules, without support for firmware loading or libudev. Is it useful for pupngo?

EDIT: opened a thread.

Posted: Sun 13 Oct 2013, 00:12
by starhawk
goingnuts, I've sent you a PM about The Infernal Dell.

Posted: Mon 24 Feb 2014, 08:28
by greengeek
technosaurus wrote:I typically do my dev work in wary for broader compatibility, dir2sfs is squash version dependent though so for pupngo, I would use a 4.1.x derivative like akita.
I've been trying to remaster a Turbopup variant using pup431 for the remaster but no success. I assumed that 431 would be ok for remastering a 420 based pup.

Now I look back at your comment I realise that I previously thought that both pupngo and turbopup were based on pup 420 (although I see now you pointed out here that pupngo is 412 based...).

Is pupngo definitely based on 412 rather than 420?

Would you expect 431 to have trouble remastering a 412 based pup or 420 based pup?

Should I use Akita to remaster turbopup rather than using 431?
cheers!

Posted: Mon 24 Feb 2014, 14:40
by starhawk
Use Woofy, sc0ttman's remaster tool. But remember with TurboPup -- it's built like a Formula 1 car -- it's *dang* fast, but it's not very rugged, and when it breaks, it breaks spectacularly.

I tried to get TurboPup to remaster cleanly with Woofy about four or five times and I wound up with one badly broken Puppy every single time.

I wish you luck in your endeavors -- and I think you'll need plenty of it!

Posted: Mon 24 Feb 2014, 18:07
by greengeek
Thanks Starhawk. I'm trying a manual remaster first - I want to go real slow, step by step so I can figure out where things go wrong. (Not brave enough to tackle turbopup extreme yet, I'm starting with turbopup alpha3a which is more mild)

My manual method worked fine for pupngo so I was hoping it would also work for Turbopup. I used Akita to do pupngo, as suggested by technosaurus, so I should probably give that a try. I was surprised that I couldnt do it using 431 - I would've thought that was similar enough to 412 and 420 to handle the unsquashfs/makesquashfs but I just ended up with kernel panics.

I'm hoping there is a "mksquash" version reason why this is occurring, rather than just my methodolgical failure...

When I first learned how to remaster I had so many failures and unsatisfactory outcomes using automated and semi-automated remastering tools I've become reluctant to trust them. Did you use woofy for vinnie?

Posted: Mon 24 Feb 2014, 18:33
by starhawk
I indeed used Woofy for Vinnie -- I'm getting better, it only took four tries to get it right :lol: all 3 "failures" (I use the term quite loosely here) were either because I did something I shouldn't have, or didn't do something I should have. In other words, cranial-RAM instability causing PEBCAK errors ;) I have that a lot.

More concisely: if I hadn't forgotten things mid-build, I would've easily gotten Vinnie out on the first try. Maybe next Vinnie (which I still hope will happen -- alas, I've been hit with multiple misfortunes on the computer that's powerful enough to do it, so that I can't use that system right now) which will quite likely be based on X-Slacko 1.1 and have PAE as well as Non-PAE versions.

A side note for PAEophiles (lol) -- as much as PAE is an annoyance when trying to boot on eg most Pentium M and Celeron M laptops... it does perform one helpful function. *If* you have 64bit hardware *and* you want to run a 32bit OS anyways for some reason, you can use PAE and access all your RAM, rather than be restricted to ~3gb of it. I do find it practical in that one single narrow circumstance... but the rest of the time it isn't useful and just kinda wastes performance that I'll probably need -- assuming it doesn't just keep me from booting Puppy.

*ahem*

I've never really messed with TurboPup Alpha3a (or anything other than Xtreme 1.0). BTW Akita is Wary5 based IIRC so you probably won't have too much luck with that -- try Puplite 5.0 -- IIRC it's actually based on TurboPup somehow.

Posted: Mon 24 Feb 2014, 19:45
by Ted Dog
MAY HAVE TO CHALLENGE TURBOPUPS CLAIM....

I will release FASTDOG64 in a week or slow ah or so.

I can't think it could be any faster.... but I'll need to compare myself..



where does this turbopup live and was is its remaster... If we are using same techniques then no need to have two remaster methods..