Hi, what is the reason to keep supporting K24 versions? Given the compatibility table, all supported routers run the K26 version as well. Is K24 still worth the effort of our brave developers?
Being less demanding for hardware resources, the kernel 2.4-based builds are still more appropriate for routers with 16MB RAM / 4MB flash.
However, I'm seriously considering to stop releasing new the K24 updates. Not right now, but in the near future…
K24 is faster to boot, and subjectively faster in response - not a scientific benchmark but "pixelserv" can send 500 blank gifs in 3secs on K24 and 4secs on K26 25% difference?
True enough, but the sheer number of (understandable) K24 regressions lately combined with the stability of K26 (it's all I run on _any_ hardware now, including an *OLD* 54G), suggests that K24 might better be left behind sooner rather than later. Imagine if Microsoft continued to target the i386 as a base platform, with all the requisite constraints, just because there might be a few people out there still crazy enough to run on such old hardware. Don't get me wrong, I've a *deep* soft spot in my heart for the old 54G series where it all started, but at some point technology and life have to move on!
My Tomato utilities site: http://multics.minidns.net/blog/articles/tomato_utilities
I want to install this on a WRT160N v1, which is 4MB flash. Is K24 still "more appropriate" for this router? Or is the stability of K26 vs. the K24 regressions lately favor a K26 install on WRT160N v1 instead?
IIRC, K26 is pretty big. And on a 4mb flash router will pretty much consume all the space leaving little ot none jffs.
Don't forget that most routers are cheap and don't have as much memory (ram & flash) as more expensive ones like RT-N16.
I for one am very glad to see K24 still supported. My GL is a flawless performer here at home on a modest setup where I have absolutely no desire or need to waste $$$ on a new router. I also prefer to keep it as lean and mean as possible. TB's NoUSB Std K24 does the trick, provides me with a solid QoS, and most of the enhancements of the K26 build.
The day development ceases on the K24 branch will be a sad day for me and my GL.
Dosn't have to be, I also have a wrt54GL and have it running a k26 build, and no problems at the moment.
The only difference is 500kb less of availible ram (maybe 1mb dont remember), as the kernel occupies more, but runs grate.
Rodney's comment about i386 support is spot on. The IBM PC derivatives were held back for many, many years because of backwards compatibility - to cater mainly for a few banking corporations.
Anyway, why not drop it. There's no point in keeping on upgrading old routers with new firmware when the old version was stable and clearly faster. Enthusiasts should move on - other less demanding users can use 2.6 - I see no problem with that.
Hopefully, maybe as 2.6 progresses it might even get faster too (what is the reason for the slowdown anyway?).
Download site http://www.4shared.com/dir/v1BuINP3/Toastman_Builds.html (inc. torrent downloads)
Make a Paypal Donation http://toastmanfirmware.yolasite.com/donations.php
I dont think theres any slowdown issues on the 2.6 builds (particularly I dont feel my old WRT54GL slower in any way, maybe other hardcore users find it diferent), just that the kernel is bigger, so… it takes 1MB more memory.
Maybe to some users thats a problem with old routers just with 16MB (like mine, but I doubt it), id say the 2.6 builds manages memory better than 2.4 builds, so Im with the ditching of the 2.4 builds, no reason to still support them.