"Ok, the last -rc obviously wasn't the last one after all, since here's a new one," noted Linus Torvalds, announcing the 2.6.26-rc9 kernel. He continued, "enough changes that we needed another -rc, and the regression list isn't emptying fast enough either (probably because a number of people, including reporters, are vacationing)." He went on to summarize:
"The actual bulk of this all is a new UVC video driver for the standard USB Video Class specification. It's a new driver, so shouldn't cause any regressions, but it's fairly sizable [...] ie 78% is just that one new driver, and almost 92% is driver updates in general (although some of them are reverts, so they show up as diffs against -rc8, but they actually cause the _total_ diff against 2.6.25 to shrink a bit). The fs updates are partly some minor updates to 9p, ecryptfs, proc and udf, but partly some delayed cleanup patches that went through Al. Bad Al. But when Al sends me patches, I apply them. I worry what would happen if I didn't. The rest is mainly small fixes (one-liners and 'few-liners') all over the place, many of them merged from Andrew's -mm queue."
From: Linus Torvalds <torvalds@...> Subject: Linux 2.6.26-rc9 Date: Jul 5, 7:15 pm 2008Ok, the last -rc obviously wasn't the last one after all, since here's a
new one. Enough changes that we needed another -rc, and the regression
list isn't emptying fast enough either (probably because a number of
people, including reporters, are vacationing).But this should in fact close a number of entries, so it's not all doom
and gloom.The actual bulk of this all is a new UVC video driver for the standard USB
Video Class specification. It's a new driver, so shouldn't cause any
regressions, but it's fairly sizable, which explains the dirstat:1.3% drivers/media/video/cx18/
1.4% drivers/media/video/saa7134/
78.0% drivers/media/video/uvc/
83.4% drivers/media/video/
84.4% drivers/media/
1.4% drivers/net/wireless/
2.3% drivers/net/
91.9% drivers/
2.6% fs/
1.7% net/ie 78% is just that one new driver, and almost 92% is driver updates in
general (although some of them are reverts, so they show up as diffs
against -rc8, but they actually cause the _total_ diff against 2.6.25 to
shrink a bit).The fs updates are partly some minor updates to 9p, ecryptfs, proc and
udf, but partly some delayed cleanup patches that went through Al. Bad Al.
But when Al sends me patches, I apply them. I worry what would happen if I
didn't.The rest is mainly small fixes (one-liners and "few-liners") all over the
place, many of them merged from Andrew's -mm queue.Linus
---
Adrian Bunk (2):
pkt_sched: Remove CONFIG_NET_SCH_RR
include/linux/netdevice.h: don't export MAX_HEADER to userspaceAkinobu Mita (1):
add kernel-doc for simple_read_from_buffer and memory_read_from_bufferAl Viro (2):
fix cgroup-inflicted breakage in block_dev.c
[netdrvr] netxen: fix netxen_pci_tbl[] breakageAlan Cox (1):
tty: Fix inverted logic in send_breakAlan Stern (1):
USB: don't lose disconnections during suspendAlex Chiang (1):
PCI: acpiphp: cleanup notify handler on all root bridgesAndi Kleen (1):
[netdrvr] Fix IOMMU overflow checking in s2io.cAndre Haupt (1):
hamradio: remove unused variableAndreas Gruenbacher (1):
vfs: make d_path() consistent across mount operationsAndres Salomon (2):
olpc: sdhci: add quirk for the Marvell CaFe's vdd/powerup issue
olpc: sdhci: add quirk for the Marvell CaFe's interrupt timeoutAndrew G. Morgan (2):
security: filesystem capabilities: fix fragile setuid fixup code
security: filesystem capabilities: fix CAP_SETPCAP handlingAndrew Morton (3):
arch/x86/mm/init_64.c: early_memtest(): fix types
Fix clear_refs_write() use of struct mm_walk
Fix pagemap_read() use of struct mm_walkAndy Gospodarek (1):
e1000: only enable TSO6 via ethtool when using correct hardwareAndy Walls (5):
V4L/DVB (8063): cx18: Fix unintended auto configurations in cx18-av-core
V4L/DVB (8066): cx18: Fix audio mux input definitions for HVR-1600 Line In 2 and FM radio
V4L/DVB (8067): cx18: Fix firmware load for case when digital capture happens first
V4L/DVB (8068): cx18: Add I2C slave reset via GPIO upon initialization
V4L/DVB (8069): cx18: Fix S-Video and Compsite inputs for the Yuan MPC718 and enable card entryAnton Vorontsov (1):
fsl_diu_fb: fix build with CONFIG_PM=y, plus fix some warningsAntti Palosaari (3):
V4L/DVB (8012): gl861: sleep a little to avoid I2C errors
V4L/DVB (8013): gl861: remove useless identify_state
V4L/DVB (8015): gl861: replace non critical msleep(0) with msleep(1) to be on the safe sideArjan van de Ven (2):
V4L/DVB (8108): Fix open/close race in saa7134
softlockup: print a module list on being stuckAtsushi Nemoto (3):
tc35815: Mark carrier-off before starting PHY
tc35815: Fix receiver hangup on Rx FIFO overflow
[MIPS] cevt-txx9: Reset timer counter on initializationAustin Lund (1):
V4L/DVB (8042): DVB-USB UMT-010 channel scan oopsBalbir Singh (1):
delay accounting: maintainer updateBartlomiej Zolnierkiewicz (5):
Revert "BAST: Remove old IDE driver"
ide: fix /proc/ide/ide?/mate reporting
ide: fix hwif->gendev refcounting
ide: ide_unregister() warm-plug bugfix
ide: ide_unregister() locking bugfixBastian Blank (1):
Alpha Linux kernel fails with inconsistent kallsyms dataBastien Nocera (1):
Input: add KEY_MEDIA_REPEAT definitionBen Dooks (5):
I2C: S3C2410: Check ACK on byte transmission
I2C: S3C2410: Fixup error codes returned rom a transfer.
I2C: S3C2410: Add MODULE_ALIAS() for s3c2440 device.
i2c: Documentation: fix device matching description
OHCI: Fix problem if SM501 and another platform driver is selectedBen Hutchings (3):
qla3xxx: Hold RTNL while calling dev_close()
Hold RTNL while calling dev_close()
PCI: Restrict VPD read permission to rootBenjamin Li (1):
PCI: Limit VPD read/write lengths for Broadcom 5706, 5708, 5709 rev.Bryan Wu (1):
Blackfin arch: fix up section mismatch warningCatalin Marinas (1):
[ARM] 5131/1: Annotate platform_secondary_init with trace_hardirqs_offChristoph Lameter (2):
slub: Do not use 192 byte sized cache if minimum alignment is 128 byte
Christoph has movedDaniel J Blueman (1):
x86: section/warning fixesDavid Brownell (4):
USB: ehci - fix timer regression
USB: ohci - record data toggle after unlink
rtc: rtc_read_alarm() handles wraparound
gpio: pca953x (i2c) handles max7310 tooDavid Howells (2):
mn10300: export certain arch symbols required to build allmodconfig
mn10300: provide __ucmpdi2() for MN10300David Rientjes (1):
mempolicy: mask off internal flags for userspace APIDavid S. Miller (1):
[SCSI] esp: Fix OOPS in esp_reset_cleanup().David Woodhouse (1):
kbuild: fix a.out.h export to userspace with O= build.Denis V. Lunev (1):
[patch 4/4] flock: remove unused fields from file_lock_operationsDevin Heitmueller (4):
V4L/DVB (8010): em28xx: Properly register extensions for already attached devices
V4L/DVB (8011): em28xx: enable DVB for HVR-900
V4L/DVB (8017): Ensure em28xx extensions only get run against devs that support them
V4L/DVB (8018): Add em2860 chip IDDivyesh Shah (1):
block: Fix the starving writes bug in the anticipatory IO schedulerDmitri Belimov (1):
V4L/DVB (8020): Fix callbacks functions of saa7134_empressDmitry Adamushko (1):
sched: fix cpu hotplugDmitry Torokhov (1):
Input: fix locking in force-feedback coreDoug Chapman (1):
[IA64] export account_system_vtimeEduard - Gabriel Munteanu (2):
Fix and clean top .gitignore
Move _RET_IP_ and _THIS_IP_ to include/linux/kernel.hEli Cohen (1):
net/inet_lro: remove setting skb->ip_summed when not LRO-ableEmmanuel Grumbach (2):
mac80211: fix an oops in several failure paths in key allocation
mac80211: don't accept WEP keys other than WEP40 and WEP104Eric Van Hensbergen (1):
9p: fix O_APPEND in legacy modeGautham R Shenoy (1):
rcu: fix hotplug vs rcu raceGeert Uytterhoeven (2):
drivers/input/ff-core.c needs
MAINTAINERS: update the email address of Andreas DilgerGreg Kroah-Hartman (2):
USB: another option device id
USB: add a pl2303 device idGuennadi Liakhovetski (3):
V4L/DVB (8039): pxa-camera: fix platform_get_irq() error handling.
V4L/DVB (8040): soc-camera: remove soc_camera_host_class class
serial: fix serial_match_port() for dynamic major tty-device numbersH. Peter Anvin (1):
x86 ACPI: normalize segment descriptor register on resumeHans Verkuil (3):
V4L/DVB (8007): cx18/cx25840: the S-Video LUMA input can use all In1-In8 inputs
V4L/DVB (8008): cx18: remove duplicate audio and video input enums
V4L/DVB (8092): videodev: simplify and fix standard enumerationHarvey Harrison (1):
prism: islpci_eth.c endianness fixHiroshi Shimamoto (1):
Update taskstats-struct document for scaled time accountingHuacai Chen (1):
rtc: fix CMOS time error after writing /proc/acpi/alarmHugh Dickins (1):
x86: fix Intel Mac booting with EFIIngo Molnar (1):
[netdrvr] 3c59x: remove irqs_disabled warning from local_bh_enableIvo van Doorn (2):
rt2x00: Fix unbalanced mutex locking
rt2x00: Fix lock dependency errrorJ. Bruce Fields (1):
svcrpc: fix handling of garbage argsJames Bottomley (2):
[SCSI] esp: tidy up target reference counting
firmware: fix the request_firmware() dummyJan Blunck (1):
[patch 1/4] vfs: path_{get,put}() cleanupsJan Engelhardt (1):
[patch 2/4] fs: make struct file arg to d_path constJan Kara (3):
ext3: add missing unlock to error path in ext3_quota_write()
ext4: add missing unlock to an error path in ext4_quota_write()
reiserfs: add missing unlock to an error path in reiserfs_quota_write()Jarkko Nikula (1):
ARM: OMAP: DMA: Don't mark channel active in omap_enable_channel_irqJaya Kumar (3):
[ARM] 5115/1: pxafb: fix ifdef for command line option handling
[ARM] 5116/1: pxafb: cleanup and fix order of failure handling
[ARM] 5117/1: pxafb: fix __devinit/exit annotationsJeff Kirsher (2):
igb: fix EEH recovery during reset on PPC
e1000e: fix EEH recovery during reset on PPCJens Axboe (1):
Properly notify block layer of sync writesJeremy Fitzhardinge (1):
xen: fix address truncation in pte mfn<->pfn conversionJess Guerrero (1):
ntfs: update help textJesse Barnes (1):
DRM/i915: only use tiled blits on 965+John Blackwood (1):
mm: switch node meminfo Active & Inactive pages to KbytesJohn Linn (2):
powerpc/bootwrapper: update for initrd with simpleImage
powerpc/legacy_serial: Bail if reg-offset/shift properties are presentJon K Hellan (1):
USB: New device ID for ftdi_sio driverJoonwoo Park (1):
textsearch: fix Boyer-Moore text search bugJozsef Kadlecsik (1):
netfilter: nf_conntrack_tcp: fixing to check the lower bound of valid ACKJulia Lawall (1):
drivers/net/r6040.c: Eliminate double sizeofJulius Volz (1):
netlink: Fix some doc comments in net/netlink/attr.cKevin Hao (1):
e100: Do pci_dma_sync after skb_alloc for proper operation on ixp4xxKomuro (1):
pcnet_cs, axnet_cs: clear bogus interrupt before request_irqLaurent Pinchart (1):
V4L/DVB (8145a): USB Video Class driverLen Brown (1):
dock: bay: Don't call acpi_walk_namespace() when ACPI is disabled.Li Zefan (3):
CONNECTOR: add a proc entry to list connectors
cgroups: document the effect of attaching PID 0 to a cgroup
devcgroup: fix odd behaviour when writing 'a' to devices.allowMarcin Slusarz (3):
vfs: fix ERR_PTR abuse in generic_readlink
V4L/DVB (8022): saa7134: fix race between opening and closing the device
V4L/DVB (8100): V4L/vivi: fix possible memory leak in vivi_fillbuffMark Lord (1):
sata_mv: safer logic for limit_warningsMatthew Wilcox (1):
[SCSI] ses: Fix timeoutMatthias Schwarzott (1):
V4L/DVB (8027): saa7134: Avermedia A700: only s-video and composite input are workingMatthieu CASTET (1):
USB: mass storage: new id for US_SC_CYP_ATACBMauro Carvalho Chehab (6):
V4L/DVB (8004): Fix INPUT dependency at budget-ci
V4L/DVB (8005): Fix OOPS if frontend is null
V4L/DVB (8026): Avoids an OOPS if dev struct can't be successfully recovered
V4L/DVB (8028): Improve error messages for tda1004x attach
V4L/DVB (8029): Improve error message at tda1004x_attach
V4L/DVB (8178): uvc: Fix compilation breakage for the other drivers, if uvc is selectedMax Asbock (1):
x86: shift bits the right way in native_read_tscpMel Gorman (1):
Do not overwrite nr_zones on !NUMA when initialising zlcache_ptrMichael Buesch (4):
b43: Do not return TX_BUSY from op_tx
b43legacy: Do not return TX_BUSY from op_tx
b43: Fix possible MMIO access while device is down
b43legacy: Fix possible NULL pointer dereference in DMA codeMichael Halcrow (1):
ecryptfs: remove unnecessary mux from ecryptfs_init_ecryptfs_miscdev()Michael Hamel (1):
rtc-x1205: Fix alarm setMichael Kerrisk (5):
vfs: utimensat(): be consistent with utime() for immutable and append-only files
vfs: utimensat(): ignore tv_sec if tv_nsec == UTIME_OMIT or UTIME_NOW
vfs: utimensat(): fix error checking for {UTIME_NOW,UTIME_OMIT} case
vfs: utimensat(): fix write access check for futimens()
man-pages is supportedMichael Krufky (7):
V4L/DVB (8034): tda18271: fix IF notch frequency handling
V4L/DVB (8035): tda18271: dont touch EB14 if rf_cal lookup is out of range
V4L/DVB (8036): tda18271: toggle rf agc speed mode on TDA18271HD/C2 only
V4L/DVB (8037): tda18271: ensure that the thermometer is off during channel configuration
V4L/DVB (8043): au0828: add support for additional USB device id's
V4L/DVB (8044): au8522: tuning optimizations
V4L/DVB (8061): cx18: only select tuner / frontend modules if !DVB_FE_CUSTOMISEMichal Schmidt (1):
s2io: fix documentation about intr_typeMiguel Ojeda (1):
Miguel Ojeda has movedMike Miller (1):
cciss: read config to obtain max outstanding commands per controllerMiklos Szeredi (2):
vfs: dcache sparse fixes
vfs: dcache cleanupsMilan Broz (1):
dm crypt: use cond_reschedMiquel van Smoorenburg (1):
tcp: calculate tcp_mem based on low memory instead of all memoryNeil Brown (3):
Ensure interrupted recovery completed properly (v1 metadata plus bitmap)
Don't acknowlege that stripe-expand is complete until it really is.
Fix error paths if md_probe fails.Octavian Purdila (1):
tcp: fix for splice receive when used with software LROOleg Nesterov (1):
get_user_pages(): fix possible page leak on oomOliver Endriss (4):
V4L/DVB (8071): tda10023: Fix possible kernel oops during initialisation
V4L/DVB (8073): av7110: Catch another type of ARM crash
V4L/DVB (8074): av7110: OSD transfers should not be interrupted
V4L/DVB (8075): stv0299: Uncorrected block count and bit error rate fixedOliver Neukum (3):
USB: fix cdc-acm resume()
USB: fix Oops on loading ipaq module since 2.6.26
USB: adding comment for ipaq forcing number of portsPatrick McHardy (4):
netfilter: ip6table_mangle: don't reroute in LOCAL_IN
net-sched: change tcf_destroy_chain() to clear start of filter list
net-sched: fix filter destruction in atm/hfsc qdisc destruction
net: fib_rules: fix error code for unsupported familiesPaul Jackson (2):
cpusets: document proc status cpus and mems allowed lists
doc: document the relax_domain_level kernel boot argumentPaul Larson (1):
ixgbe: fix EEH recovery during reset on PPCPaul Mackerras (1):
Update maintainers for powerpcPaul Moore (1):
netlabel: Fix a problem when dumping the default IPv6 static labelsPaul Mundt (1):
lib: taint kernel in common report_bug() WARN path.Pavel Emelyanov (1):
inet fragments: fix race between inet_frag_find and inet_frag_secret_rebuildPavel Machek (1):
Doc*/kernel-parameters.txt: fix stale referencesPavel Roskin (2):
hostap: don't report useless WDS frames by default
hostap: fix sparse warningsPekka Enberg (2):
ipg: fix jumbo frame compilation
ipg: use NULL, not zero, for pointersPeng Haitao (2):
kernel/audit.c: nlh->nlmsg_type is gotten more than once
remove useless argument type in audit_filter_user()Peter Zijlstra (1):
mm: dirty page accounting vs VM_MIXEDMAPPhilipp Zabel (2):
w100fb: do not depend on SHARPSL
w100fb: add 80 MHz modelinePierre Ossman (1):
mmc: don't use DMA on newer ENE controllersRafael J. Wysocki (1):
x86 ACPI: fix resume from suspend to RAM on uniprocessor x86-64Rainer Weikusat (1):
af_unix: fix 'poll for write'/connected DGRAM socketsRaistlin (1):
sched: fix divide error when trying to configure rt_period to zeroRajiv Andrade (1):
Update MAINTAINERS file for the TPM device driverRalf Baechle (1):
[MIPS] Fix bug in atomic_sub_if_positive.Randy Dunlap (2):
audit: fix kernel-doc parameter notation
doc: doc maintainersRene Herman (1):
thermal: Create CONFIG_THERMAL_HWMON=nRichard Sharpe (1):
firewire: fw-sbp2: fix parsing of logical unit directoriesRon Rindjunsky (1):
iwlwifi: improve scanning band selection managementRussell King (1):
[ARM] Export dma_sync_sg_for_device()Samuel Ortiz (1):
MFD maintainerSebastian Siewior (1):
spi: fix the read path in spidevSonic Zhang (1):
Blackfin arch: fix bug - kernel boot fails when Spinlock and rw-lock debugging enabledStefan Becker (1):
USB: fix interrupt disabling for HCDs with shared interrupt handlersStephen Hemminger (2):
tcp: /proc/net/tcp rto,ato values not scaled properly (v2)
ipv4: fix sysctl documentation of time related valuesStephen M. Cameron (1):
cciss: fix regression that no device nodes are created if no logical drives are configured.Stephen Rothwell (2):
Introduce rculist.h
cpumask: introduce new APIsSteven Rostedt (1):
hrtimer: prevent migration for raising softirqSteven Toth (2):
V4L/DVB (8096): au8522: prevent false-positive lock status
V4L/DVB (8097): xc5000: check device hardware state to determine if firmware download is neededTAKADA Yoshihito (1):
ptrace GET/SET FPXREGS brokenTejun Heo (6):
sata_uli: hardreset is broken
sata_sil24: add DID for another adaptec flavor
ahci: always clear all bits in irq_stat
libata-sff: improve HSM violation reporting
pci: VT3336 can't do MSI either
ahci: give another shot at clearing all bits in irq_statThara Gopinath (1):
ARM: OMAP: Correcting the gpmc prefetch control register addressThomas Bogendoerfer (2):
[MIPS] IP32: Fix unexpected irq 71
[MIPS] IP22: Fix crashes due to wrong L1_CACHE_BYTESThomas Gleixner (1):
x86: fix NODES_SHIFT Kconfig rangeTim Farrington (1):
V4L/DVB (8048): saa7134: Fix entries for Avermedia A16d and Avermedia E506Tim Yamin (1):
powerpc/mpc5200: Fix lite5200b suspend/resumeTomas Janousek (1):
udf: Fix regression in UDF anchor block detectionTony Luck (1):
[IA64] Bugfix for system with 32 cpusTony Vroon (1):
mac80211: implement EU regulatory domainUli Luckas (1):
[ARM] 5109/1: Mark rtc sa1100 driver as wakeup source before registering itVegard Nossum (2):
debugobjects: fix lockdep warning
ACPI: don't walk tables if ACPI was disabledWANG Cong (1):
pkt_sched: ERR_PTR() ususally encodes an negative errno, not positive.Wang Chen (4):
netdevice: Fix typo of dev_unicast_add() comment
netlink: Unneeded local variable
net: Tyop of sk_filter() comment
netdevice: Fix wrong string handle in kernel command line parsingWill Newton (1):
sisusbvga: Fix oops on disconnect.Wolfram Sang (1):
i2c: Fix bad hint about irqs in i2c.hYOSHIFUJI Hideaki (1):
ipv6 route: Convert rt6_device_match() to use RT6_LOOKUP_F_xxx flags.Zhang, Yanmin (1):
x86: fix cpu hotplug crashmaximilian attems (1):
hdaps: add support for various newer Lenovo thinkpads--

*sigh*
"and the regression list isn't emptying fast enough either (probably because a number of people, including reporters, are vacationing)."
Shaping up to be another promising Linux kernel release =(
don't worry ...
If this were Microsoft, -rc9 would have been something like the 50th 'patch Tuesday' update. Linus has some degree of integrity and cares more about the quality of the product than just pushing something out the door for the lemmings to squander their money on.
I would go further and say
I would go further and say that Microsoft would have released rc4 as the product and a just month later everyone would have to do 50+ updates to get their system up to current and this update process would require 3 to 5 reboots and 75MB+ of downloads.
ever measured how many megabytes of downloads...
ever measured how many megabytes of downloads one has to download on an average desktop-oriented Linux distro, about a month after its release?
Yeah, comparing a Linux distro which has tons of software with Windows which has a notepad and not much more is not fair ;)
Measure it in hours lost.
I spent the better part of an entire work day a few weeks back doing a fresh install of Windows XP. Add in all the necessary drivers, service packs, reboots and more drivers and reboots and antivirus updates and it came to about 4-5 hours. Doing the same on Ubuntu 7.10 around the same date (dist-upgrading to 8.04) took about 2 and a half hours, and it did it in one shot without requiring me to be there to click a button every 15-20 minutes.
Luckily I didn't have to update the windows PC after that - halfway through installing the patches windows update stopped working completely.
baloney
Please. Read LKML just since rc9. Boatloads of glaring, lowlevel kernel bugs. It is always this way. Linus's dev cycle is a problem with Linux. Perhaps THE problem.
The whole mentality about version numbers gives it away. Add another filesystem, a new memory manager, a new CPU scheduler, and new IRQ balancer, some new drivers, 1,000 other miscellany patches, and what is the version increment? +0.0.1.
Can you imagine what Linus would treat as change from v.2.6 to say, v.3.0? Think here. The goal is not to make something that works and stop changing it. Or sensible QA along a development path. The goal is to keep everyone breathless about what kernel.org will do next. The moment that things "just work" - there'll be no more need for Linus.
Every distro hangs breathlessly on the next kernel release, which will finally fix all the bugs from the previous kernel release, which were introduced because the move from X.Y.Z.3 to X.Y.Z.4 broke everything, because some kernel.org developer wanted to play. And Linus says, in effect, phuck you, this is how we do things. Especially to hardware vendors like nVidia, who mustr rewrite drivers for each kernel bump. And the masses say "Linux is broken" and the distros tell them "wait for the next kernel" and round and round we go.
And everyone wonders if this will finally be the year of the Linux desktop.
Especially to hardware
Especially to hardware vendors like nVidia, who mustr rewrite drivers for each kernel bump
Good, it SHOULD be painful to keep the source closed
And the masses say "Linux is broken"
Ehm, no they don't
And the funny thing?
I do, unfortunately, run the NVidia closed drivers. You want to know the funny thing, though? I've upgraded the kernel several times, and haven't had to download a new driver from them. It's as if, you know, they didn't have to rewrite it with every kernel bump.
I did, however, have to recompile it each time. So, really, it's a reminder to me, every time I upgrade, that their's something goofy here. They've outsourced that pain to the customer. The only time NVidia notices is when there's actually a major interface shift, and they're caught being behind the curve. That doesn't happen very much at all.
I wonder what the impact of this dynamic is on a large scale. For folks like me who know and understand the situation, I'm likely to grumble at NVidia when this happens. But, for the average person out there, who are they going to grumble to? "I upgraded my kernel, and my kernel doesn't work with all my things because I have to recompile one of them." Hmm... The end user touches the kernel and something breaks. I don't think it matters that the thing that broke is fundamentally "broken," I think it's the kernel that gets the "blame" even if it's misassigned.
I guess it's little question where all the "Please make a stable interface for drivers like NVidia" crowd comes from.
--
Program Intellivision and play Space Patrol!
bug fixing?
Dunno, but wouldn't a bugfix release make sense right about now?
You know, do some cleaning up before more features (and yet more bugs) are pushed in... :(
UVC video driver for the standard USB Video Class specification
Thanks!
A small step for Linux, a great step for camsluts! :D
Finally, maybe my webcam will work! :)
<cheek><tongue/></cheek>
Will this be what it takes for Linux to become widely accepted?
The release of the Apple LaserWriter made Mac OS the favorite of desktop publishers and artists, a position Mac OS enjoys to this day, and Mac OS is still seen as synonymous with "artsy".
Will UVC be Linux's LaserWriter?
:^P
Doesn't mean ...
This doesn't mean that all of a sudden, all USB cameras start working. This "just" applies to the newer, UVC compliant ones (a Vista requirement, I might add). Personally I use OpenBSD, which has uvideo(4) for UVC devices. My newer Logitech camera works fine. My old oh-so-crappy camera doesn't. Linux supports it, though (cpia-something, IIRC).
!!A small step for mankind,
!!A small step for mankind, a giant leap for rock'n'roll!!
Oh wait, this isn't the KMFDM forum? :)
A1?
What's A1 here? I'm only familiar with the steak sauce.
AL Viro
AL Viro
Linux steak source
Linux steak source