New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We鈥檒l occasionally send you account related emails.

Already on GitHub? Sign in to your account

Rationalise is ready i think to be merged #1011

Merged
merged 65 commits into from May 21, 2017

Conversation

Projects
None yet
4 participants
@01micko
Contributor

01micko commented May 14, 2017

It does need a full review so I won't commit myself.

Changes from testing:

  • a, y and f drive full support
  • copying of packages to rootfs-complete (and a y z drives) now a function
  • fdrive support with various mechanisms including kernel-kit, ability to include packages and downloading and packaging of non-free firmware (b43, nouveau, dvb to name a few)
  • full integration into __0build.conf as not to impede 'headless' (well it's not quite headless 馃槢 ) build
  • others I've forgotten (run git log .. its all recorded)

Happy reviewing!

@01micko

This comment has been minimized.

Show comment
Hide comment
@01micko

01micko Apr 15, 2017

Contributor

@wdlkmpx this seems to work quite well as a function (oh yes fixed boo boo in 5056268 copied over wrong file - oops!) and facilitates adding a ydrv (or more) in the future if the need/desire is there.

It's a bit ugly appending the dummy value but because the read delim is ' ' (space) the very last entry wasn't getting picked up - both in rootfs-complete and the adrv. I'm not sure how well it will work if a ZHACK exists though so it's not ready for prime time yet.

Contributor

01micko commented on e347652 Apr 15, 2017

@wdlkmpx this seems to work quite well as a function (oh yes fixed boo boo in 5056268 copied over wrong file - oops!) and facilitates adding a ydrv (or more) in the future if the need/desire is there.

It's a bit ugly appending the dummy value but because the read delim is ' ' (space) the very last entry wasn't getting picked up - both in rootfs-complete and the adrv. I'm not sure how well it will work if a ZHACK exists though so it's not ready for prime time yet.

01micko and others added some commits Apr 15, 2017

fix weird adrive bug concerning libreoffice and rox.
This could be a bug of either but I suspect the latter.
When trying to open a doc or whatever with 'open with'
a 'permission denied' error occurs. If the .desktop extension
is present all is fixed, regardles of the prefix... test with
'a.desktop' :/
some changes/fixes for the adrv stuff
i think it works better this way
i had a corrupted kernel tarball, so
a more complete test may happen tomorrow
it seems to work fine

@01micko
@01micko

This comment has been minimized.

Show comment
Hide comment
@01micko

01micko Apr 18, 2017

Contributor

Looks sound, I didn't even think of using tr .... and i admit appending the dummy was ugly!

What do you think? add support for ydrv/fdrv? (with a note 'experts only')? I'm not sure zdrv is appropriate at this stage, might think about it.

Contributor

01micko commented on 221e073 Apr 18, 2017

Looks sound, I didn't even think of using tr .... and i admit appending the dummy was ugly!

What do you think? add support for ydrv/fdrv? (with a note 'experts only')? I'm not sure zdrv is appropriate at this stage, might think about it.

This comment has been minimized.

Show comment
Hide comment
@ninaholic

ninaholic Apr 19, 2017

Contributor

Nice idea. I have a large list of programs that can be sent to adrv 馃檲

Contributor

ninaholic replied Apr 19, 2017

Nice idea. I have a large list of programs that can be sent to adrv 馃檲

This comment has been minimized.

Show comment
Hide comment
@wdlkmpx

wdlkmpx Apr 20, 2017

Contributor

I just built a stretch iso, I didn't see any issues related to the adrv, so i guess it's working.. might be time to merge rationalise back to testing. but i'll keep using the current install for some time, while i'm trapped in the fido account to fix some puppy scripts and my own scripts.. and because i have other things to do atm..

Yeah i think adding support for ydrv/fdrv is a good idea. I'd certainly take a look and review..

I was also thinking about using some "prebuilt" drvs, including testdisk, flashrom, ddrescue, etc.. interesting stuff compiled almost statically in older puppies or by the initrd-progs build system (musl static) to make them portable..

Contributor

wdlkmpx replied Apr 20, 2017

I just built a stretch iso, I didn't see any issues related to the adrv, so i guess it's working.. might be time to merge rationalise back to testing. but i'll keep using the current install for some time, while i'm trapped in the fido account to fix some puppy scripts and my own scripts.. and because i have other things to do atm..

Yeah i think adding support for ydrv/fdrv is a good idea. I'd certainly take a look and review..

I was also thinking about using some "prebuilt" drvs, including testdisk, flashrom, ddrescue, etc.. interesting stuff compiled almost statically in older puppies or by the initrd-progs build system (musl static) to make them portable..

This comment has been minimized.

Show comment
Hide comment
@01micko

01micko Apr 20, 2017

Contributor

@ninaholic nice catch. I suppose we can allow this but with restrictions. A broken system may result if things like sfs_load go to the {a,y,f,z}drv.

Contributor

01micko replied Apr 20, 2017

@ninaholic nice catch. I suppose we can allow this but with restrictions. A broken system may result if things like sfs_load go to the {a,y,f,z}drv.

This comment has been minimized.

Show comment
Hide comment
@01micko

01micko Apr 20, 2017

Contributor

Actually, sfs_load is a bad example as it is already filtered but if a user includes ptheme and wallpaper then they will be broken if the adrv gets changed, so in a case like pmusic then sure, it can be moved.

Anyway, I'll filter what apps i reckon and we can adjust the list later.

Contributor

01micko replied Apr 20, 2017

Actually, sfs_load is a bad example as it is already filtered but if a user includes ptheme and wallpaper then they will be broken if the adrv gets changed, so in a case like pmusic then sure, it can be moved.

Anyway, I'll filter what apps i reckon and we can adjust the list later.

This comment has been minimized.

Show comment
Hide comment
@01micko

01micko Apr 21, 2017

Contributor

@ninaholic pmusic was copied to the main sfs if you chose it because that happens before everything else.

Maybe it's best not to include pmusic (or whatever) in the rootfs-packages choose dialog or list and add it o the adrive later. I'll see what I can do..

Contributor

01micko replied Apr 21, 2017

@ninaholic pmusic was copied to the main sfs if you chose it because that happens before everything else.

Maybe it's best not to include pmusic (or whatever) in the rootfs-packages choose dialog or list and add it o the adrive later. I'll see what I can do..

This comment has been minimized.

Show comment
Hide comment
@01micko

01micko Apr 21, 2017

Contributor

Rightio.. 103ec9e adds full support for a,y,f drives. You can have any, all or none and should work ok. rootfs-packages is supported by an extra block in __0func but the package you wish to send to a or y must not be included in the checkbox field or config file. So this is somewhat of a work around but bear in mind that is only to facilitate the few. 馃槢

@wdlkmpx I would say this is still 'beta' so review and we'll wait on real tests before merging back to testing.

Contributor

01micko replied Apr 21, 2017

Rightio.. 103ec9e adds full support for a,y,f drives. You can have any, all or none and should work ok. rootfs-packages is supported by an extra block in __0func but the package you wish to send to a or y must not be included in the checkbox field or config file. So this is somewhat of a work around but bear in mind that is only to facilitate the few. 馃槢

@wdlkmpx I would say this is still 'beta' so review and we'll wait on real tests before merging back to testing.

01micko added some commits May 13, 2017

add nouveau firmware to get_fw script.
also enable cross checking DOTconfig if built a kernel with kernel-kit
now optionally saves downloads
make the new firmware routine compatible with generated fdrive
also fixed a path change from when kernel-kit was modified
remove the old firmwares, now pets up at ibiblio.
Newer versions are (optionally) downloaded in 3builddistro-Z
add to 3.44.67 DOT config; CONFIG_FW_LOADER_USER_HELPER_FALLBACK=y
this might help some pedantic intel wireless chipsets to load their firmware
@01micko

This comment has been minimized.

Show comment
Hide comment
@01micko

01micko May 14, 2017

Contributor

doh! should have been 4.4.67!

Contributor

01micko commented on c095f29 May 14, 2017

doh! should have been 4.4.67!

fix bug extracting fresh kernel package from kernel-kit
the intention going forward is to support just a huge_kernel**.tar archive
as no compression is need, so just a waste of cpu.
@01micko

This comment has been minimized.

Show comment
Hide comment
@01micko

01micko May 14, 2017

Contributor

For the curios.. isos, devx, kernel-source (32 + 64) are uploading to ibiblio

Look for 700/testing ... ETA 08.28.35 from right now. Approx iso sizes are 259 MB and 366 MB... respectively 馃

Contributor

01micko commented May 14, 2017

For the curios.. isos, devx, kernel-source (32 + 64) are uploading to ibiblio

Look for 700/testing ... ETA 08.28.35 from right now. Approx iso sizes are 259 MB and 366 MB... respectively 馃

@ninaholic

This comment has been minimized.

Show comment
Hide comment
@ninaholic

ninaholic May 14, 2017

Contributor

Hello @01micko ! I tried your Slacko 6.9.6.7 32-bit, it looks nice. I connect to the internet no problems, but the browser button did not respond. Maybe my Pentium M laptop is too old?

Here is screenshot of terminal output: https://ibb.co/ciOVMQ

Contributor

ninaholic commented May 14, 2017

Hello @01micko ! I tried your Slacko 6.9.6.7 32-bit, it looks nice. I connect to the internet no problems, but the browser button did not respond. Maybe my Pentium M laptop is too old?

Here is screenshot of terminal output: https://ibb.co/ciOVMQ

@01micko

This comment has been minimized.

Show comment
Hide comment
@01micko

01micko May 14, 2017

Contributor
Contributor

01micko commented May 14, 2017

@01micko

This comment has been minimized.

Show comment
Hide comment
@01micko

01micko May 14, 2017

Contributor

@ninaholic palemoon is rebuilt and uploaded to ibiblio

http://distro.ibiblio.org/puppylinux/pet_packages-slacko14/palemoon-27.3.0-i686_701s_sse.pet
~31.5 MB (~700 kB smaller than the original)

Here is the full optimization line:
--enable-optimize=\"-O2 -march=pentium3 -msse -mfpmath=sse -pipe -fomit-frame-pointer\"

I suggest to unload the adrv (if you have an install - just mv it) , reboot, then install the pet. You can grab it with wget. That's how I tested it (posting from it now), but my cpu has all the instructions necessary and all my old computers have died. 馃槥 so I can't do a full test. I did try with qemu but I don't think that qemu's virtual cpu is accurate.

Contributor

01micko commented May 14, 2017

@ninaholic palemoon is rebuilt and uploaded to ibiblio

http://distro.ibiblio.org/puppylinux/pet_packages-slacko14/palemoon-27.3.0-i686_701s_sse.pet
~31.5 MB (~700 kB smaller than the original)

Here is the full optimization line:
--enable-optimize=\"-O2 -march=pentium3 -msse -mfpmath=sse -pipe -fomit-frame-pointer\"

I suggest to unload the adrv (if you have an install - just mv it) , reboot, then install the pet. You can grab it with wget. That's how I tested it (posting from it now), but my cpu has all the instructions necessary and all my old computers have died. 馃槥 so I can't do a full test. I did try with qemu but I don't think that qemu's virtual cpu is accurate.

@01micko

This comment has been minimized.

Show comment
Hide comment
Contributor

01micko commented May 15, 2017

@ninaholic

This comment has been minimized.

Show comment
Hide comment
@ninaholic

ninaholic May 15, 2017

Contributor

Thanks @01micko the pet works 馃憤 Congrats on releasing new Slacko version! :)

Contributor

ninaholic commented May 15, 2017

Thanks @01micko the pet works 馃憤 Congrats on releasing new Slacko version! :)

@01micko

This comment has been minimized.

Show comment
Hide comment
@01micko

01micko May 15, 2017

Contributor

OK thanks for testing @ninaholic. I'll do a rebuild with the new pet and upload a delta. It will only be a few meg.

Edit: wow! 24 MB but still better than grabbing an iso full size.

Contributor

01micko commented May 15, 2017

OK thanks for testing @ninaholic. I'll do a rebuild with the new pet and upload a delta. It will only be a few meg.

Edit: wow! 24 MB but still better than grabbing an iso full size.

@01micko

This comment has been minimized.

Show comment
Hide comment
@01micko

01micko May 19, 2017

Contributor

@wdlkmpx @mavrothal @woodenshoe-wi @ninaholic @peabee @zigbert @gyrog : if nobody objects I am going to merge this in 24 hours or so.

Contributor

01micko commented May 19, 2017

@wdlkmpx @mavrothal @woodenshoe-wi @ninaholic @peabee @zigbert @gyrog : if nobody objects I am going to merge this in 24 hours or so.

@wdlkmpx wdlkmpx merged commit 9859ea8 into testing May 21, 2017

@wdlkmpx

This comment has been minimized.

Show comment
Hide comment
@wdlkmpx

wdlkmpx May 21, 2017

Contributor

We'll find bugs eventually..

Contributor

wdlkmpx commented May 21, 2017

We'll find bugs eventually..

@01micko

This comment has been minimized.

Show comment
Hide comment
@01micko

01micko May 22, 2017

Contributor

I suppose this branch has served it's purpose and should be deleted. I'll wait until everyone knows that I'm deleting it and then do the deed; or anyone else with the right privileges can delete if they wish.

If we need a new branch for wild experiments then we can create one at any time.

Contributor

01micko commented May 22, 2017

I suppose this branch has served it's purpose and should be deleted. I'll wait until everyone knows that I'm deleting it and then do the deed; or anyone else with the right privileges can delete if they wish.

If we need a new branch for wild experiments then we can create one at any time.

@peabee

This comment has been minimized.

Show comment
Hide comment
@peabee

peabee May 22, 2017

Contributor

LxPupSc built with "new" woof-ce testing:

BUILD_FROM_WOOF='testing;deca49a;2017-05-21 20:38:32 +0800'

Didn't use the new adrv/fdrv facilities....
No obvious issues in build or in built system....
pupMd5sum needs enhancing to produce sha256 checksums (and SourceForge does not provide them....) so md5 will be needed for the foreseeable....

Contributor

peabee commented May 22, 2017

LxPupSc built with "new" woof-ce testing:

BUILD_FROM_WOOF='testing;deca49a;2017-05-21 20:38:32 +0800'

Didn't use the new adrv/fdrv facilities....
No obvious issues in build or in built system....
pupMd5sum needs enhancing to produce sha256 checksums (and SourceForge does not provide them....) so md5 will be needed for the foreseeable....

@01micko

This comment has been minimized.

Show comment
Hide comment
@01micko

01micko May 23, 2017

Contributor

so md5 will be needed for the foreseeable....

There's no plans to deprecate, just add sha256 as an alternative.

Contributor

01micko commented May 23, 2017

so md5 will be needed for the foreseeable....

There's no plans to deprecate, just add sha256 as an alternative.

@01micko 01micko deleted the rationalise branch May 23, 2017

@peabee

This comment has been minimized.

Show comment
Hide comment
Contributor

peabee commented May 23, 2017

@01micko 01micko restored the rationalise branch May 25, 2017

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment