Windows: Over 36 Thousand Registry Modifications ... Just to Install a Printer
Homer
[quote]
"there are *ZERO* registry entries added for the driver, because it's
entirely done by windows during hardware detection." ~ Fuddie
[quote]
http://groups.google.com/group/comp.os.linux.advocacy/msg/ab2ccc26ef48d4a8
Here's the Procmon (update of Regmon) logs of what happens to the
Registry when one installs a HP 6940 Printer under windows.
During this process there were actually over 2 *Million* system events,
but I've filtered them to only include the following operations:
RegCreateKey
RegDeleteKey
RegSetValue
RegDeleteValue
I've also filtered any personally identifiable information, and as much
irrelevant crud as I could identify, but you'll have to appreciate that
with over 36 thousand lines - I'm bound to have missed something.
Here's the full log, containing all events of the above "operation" types:
http://slated.org/files/Logfile.full.CSV.txt
I then filtered out anything not specific to HP's software:
msiexec
Explorer.EXE
svchost.exe
services.exe
Setup.exe
winlogon.exe
attrib.exe
Wrapper.exe
MsiExec.exe
spoolsv.exe
\"System\"
rundll32.exe
regsvr32.exe
setup.exe
verclsid.exe
_INS5576._MP
That last one is technically part of the setup process, but is actually
InstallShield data rather than an actual HP component, so I left it out.
This then leaves the following:
hpzwup01.exe
hpzshl01.exe
hpzrcv01.exe
hpzprl01.exe
hpzarp01.exe
hpzdui01.exe
hpzmsi01.exe
hpzsetup.exe
hpzcdl01.exe
HpqUnApl.exe
hpqtra08.exe
HPBOID.EXE
HPBPRO.EXE
hpqSTE08.exe
HPZipm12.exe
hpqpprop.exe
hprblog.exe
hposid01.exe
hpqWRG.exe
hpqdirec.exe
hpqtbx01.exe
The above HP components altered the Registry a total of 4,869 times
during the install process:
http://slated.org/files/Logfile.filtered.CSV.txt
Note in particular the entries which specifically deal with hardware
enumeration and drivers, such as this:
"18:47:37.1542790","HPBOID.EXE","2876","RegSetValue","HKCR\CLSID\{D713F357-7920-4B91-9EB6-49054709EC7A}\LocalServer32\(Default)","SUCCESS","Type:
REG_SZ, Length: 108, Data:
C:\WINDOWS\system32\spool\drivers\w32x86\3\HPBOID.EXE"
And this:
"18:48:30.9991618","hpqtra08.exe","3704","RegSetValue","HKLM\System\CurrentControlSet\Enum\USB\Vid_03f0&Pid_8904\MY822CS0DM04Q9\Device
Parameters\SoftwareCUEContextID","SUCCESS","Type: REG_SZ, Length: 102,
Data: #Hewlett-Packard#HP Deskjet 6940 series#1227552509"
And this:
"18:48:31.0186330","hpqtra08.exe","3704","RegSetValue","HKLM\System\CurrentControlSet\Enum\USB\Vid_03f0&Pid_8904&MI_00\7&20e5f2ab&0&0000\Device
Parameters\DDDevNodeColor","SUCCESS","Type: REG_DWORD, Length: 4, Data: 1
Indeed there are 81 entries for "Enum" alone, another 13 for "drivers",
and a total of 1,062 entries for "HKCU" (Current User). All searches
were done with "grep -i" (case insensitive).
Here's some more stats:
RegCreateKey: 3,429 times
RegDeleteKey: 405 times
RegSetValue: 972 times
RegDeleteValue: 62 times
And some people wonder why Windows is so slow; wastes so much resources;
and becomes so unstable and filled with orphaned crud over time. With
this much overcomplexity, and such an incompetent; chaotic; amateurish
design ... is it any wonder? How anything manages to work under Windows
at /all/ is a miracle of good fortune, especially hardware drivers.
Just a reminder:
[quote]
"there are *ZERO* registry entries added for the driver, because it's
entirely done by windows during hardware detection." ~ Fuddie
[quote]
--
K.
http://slated.org
.----
| "At the time, I thought C was the most elegant language and Java
| the most practical one. That point of view lasted for maybe two
| weeks after initial exposure to Lisp." ~ Constantine Vetoshev
`----
Fedora release 8 (Werewolf) on sky, running kernel 2.6.25.11-60.fc8
22:35:00 up 19 days, 6:17, 5 users, load average: 0.09, 0.14, 0.16
Fingel
unecessarily so, but I had no idea of the magnitude. That is just
wierd...
--
3A:29
Chris Ahlstrom
this bit o' wisdom:
> On 2008-11-24, Homer <use...@slated.org> wrote:
>> First the context:
>>
>> [quote]
>> "there are *ZERO* registry entries added for the driver, because it's
>> entirely done by windows during hardware detection." ~ Fuddie
>> [quote]
>>
>> http://groups.google.com/group/comp.os.linux.advocacy/msg/ab2ccc26ef48d4a8
>>
>> msiexec
>> Explorer.EXE
>> svchost.exe
>> services.exe
>> Setup.exe
>> winlogon.exe
>> attrib.exe
>> Wrapper.exe
>> MsiExec.exe
>> spoolsv.exe
>> \"System\"
>> rundll32.exe
>> regsvr32.exe
>> setup.exe
>> verclsid.exe
>> _INS5576._MP
>>
>> That last one is technically part of the setup process, but is actually
>> InstallShield data rather than an actual HP component, so I left it out.
>>
>> This then leaves the following:
>>
>> hpzwup01.exe
>> hpzshl01.exe
>> . . .
>> hpqtbx01.exe
>>
>> The above HP components altered the Registry a total of 4,869 times
>> during the install process:
>>
>> http://slated.org/files/Logfile.filtered.CSV.txt
>>
>> [quote]
>> "there are *ZERO* registry entries added for the driver, because it's
>> entirely done by windows during hardware detection." ~ Fuddie
>> [quote]
>>
> That is incredible. I knew installing hardware in windows was slow and
> unecessarily so, but I had no idea of the magnitude. That is just
> wierd...
Looks like the Explorer team doesn't communicate with the Setup team and the
WinLogon team, the Wrapper team, the rundll32 team, the regsvr32 team, the
verclsid team; and the Hewlett-Packard Windows printer driver team doesn't
talk to any of them.
--
In case it's not obvious, any solution to this problem that introduces a
dependency on Java is profoundly uninteresting to me. In fact, my
indifference to that could only be described as "sexual" in intensity.
-- Jamie Zawinski
Cork Soaker
> First the context:
>
<snip>
Yes, a failure during an Installshield setup leaves a bloody awful mess
to clean up.
However, I don't see how this advocates Linux?
Sandeep Kumar
Try fixing CUPS when it takes a crap.
Not easy at all.
Erik Funkenbusch
> First the context:
>
> [quote]
> "there are *ZERO* registry entries added for the driver, because it's
> entirely done by windows during hardware detection." ~ Fuddie
> [quote]
>
> http://groups.google.com/group/comp.os.linux.advocacy/msg/ab2ccc26ef48d4a8
No, that's only part of the context. The rest of the context is that we're
talking about a Digital Camera driver, not a printer driver. Further, the
context is specifically for the driver, not the additional software
included (you can install the driver without the software).
Finally, this is pre-hardware detection. Obviously registry entries are
created by the hardware wizard once the hardware is detected and installed,
but the *REAL* context of the thread is what happens before that.
> Just a reminder:
>
> [quote]
> "there are *ZERO* registry entries added for the driver, because it's
> entirely done by windows during hardware detection." ~ Fuddie
> [quote]
Stop lying, Homer. You have entirely changed the context of the
discussion.
LusoTec
> Try fixing CUPS when it takes a crap.
> Not easy at all.
Depending on the problem, you can just delete the config files and
reconfigure. If needed, also reinstall the packages. This has solved almost
all my CUPS problem in minutes. In the few occasions I had to debug the
printing problems that reconfiguring/reinstalling did not solve was related
to buggy drivers that caused problems (e.g. incorrectly detecting
out-of-paper state and not continuing after paper was supplied).
Regards.
Peter Köhlmann
Let my think when I had the last time problems with CUPS.
Oh, now I remember. Never
--
Any idiot can run XP. And usually does.
Gregory Shearman
Yeah. It. Just. Works.
Looks like the flatfish troll has been uncovered. Mentioning problems
with CUPS is a troll dead giveaway.
--
Regards,
Gregory.
Gentoo Linux - Penguin Power
Ezekiel
"Peter Köhlmann" <peter.k...@arcor.de> wrote in message
news:492fe48d$0$31333$9b4e...@newsspool4.arcor-online.net...
Sounds like it was the same time that I had my last printing problem with
Windows.
Hadron
I'm beginning to think Peter doesn't even use Linux. He never had a
problem with his video card, never with CUPs etc etc.
http://www.google.com/search?q=CUPS+problems
We know he's a professional Windows programmer, so the chances are he
just has a Linux box at home to practice his l337 "ls" skills.
--
"It explains a lot. I've not heard of anyone I know, anywhere, buying XP,
and I've not seen it sold whilst I've been in any shops."
comp.os.linux.advocacy - where they put the lunacy in advocacy
Ezekiel
"Hadron" <hadro...@gmail.com> wrote in message
news:ggosmq$i63$1...@reader.motzarella.org...
> "Ezekiel" <Ze...@z.com> writes:
>
>> "Peter Köhlmann" <peter.k...@arcor.de> wrote in message
>> news:492fe48d$0$31333$9b4e...@newsspool4.arcor-online.net...
>>> flatfish (Sandeep Kumar) nymshifted:
>>>
>>>> On Fri, 28 Nov 2008 01:30:15 +0000, Cork Soaker wrote:
>>>>
>>>>> Homer wrote:
>>>>>> First the context:
>>>>>>
>>>>>
>>>>> <snip>
>>>>>
>>>>> Yes, a failure during an Installshield setup leaves a bloody awful
>>>>> mess
>>>>> to clean up.
>>>>>
>>>>> However, I don't see how this advocates Linux?
>>>>
>>>> Try fixing CUPS when it takes a crap.
>>>> Not easy at all.
>>>
>>
>>> Let my think when I had the last time problems with CUPS.
>>> Oh, now I remember. Never
>>
>> Sounds like it was the same time that I had my last printing problem
>> with
>> Windows.
>
> I'm beginning to think Peter doesn't even use Linux. He never had a
> problem with his video card, never with CUPs etc etc.
>
> http://www.google.com/search?q=CUPS+problems
>
> We know he's a professional Windows programmer, so the chances are he
> just has a Linux box at home to practice his l337 "ls" skills.
>
I don't know. I just think that "problems" are over exaggerated by both
sides here in COLA. Generally speaking... the computers simply work
provided that the user setting them up has some idea of what they're doing.
Hadron
Well, if that "some idea" means not using certain things because they
simply do not work with that HW then fair enough. Example : I gave up on
ATI Video cards and moved to evil NVidia. Why? Because the ATI OpenGL
driver installation was a joke.
But Koehlmann and co are a waste of space in this "share and share
alike" OSS/Linux world. Why? Because they are no help to anyone since
anyone with an issue is called a "liar" and told that it "works for me".
--
"Don't like Linux.. don't use it. Simple."
-- Rick <no...@nomail.com> in comp.os.linux.advocacy, alt.os.windows-xp
Sandeep Kumar
> Stop lying, Homer. You have entirely changed the context of the
> discussion.
So what else is new?
DFS
> Let my think when I had the last time problems with CUPS.
> Oh, now I remember. Never
You kidding, dumbkopf? Day after day you come up with these stupid samples
of one.
Peter Köhlmann
You mean that using linux for more than 10 years and *never* having had any
problems with CUPS, across a wide variety of distros, is a sample of one?
So the other linux boxen I installed, but don't run myself, don't count?
Idiot
--
I say you need to visit Clues 'R' Us. They are having a special on
slightly used clues.
Cork Soaker
> "Ezekiel" <Ze...@z.com> writes:
>
>> "Peter Köhlmann" <peter.k...@arcor.de> wrote in message
>> news:492fe48d$0$31333$9b4e...@newsspool4.arcor-online.net...
>>> flatfish (Sandeep Kumar) nymshifted:
>>>
>>>> On Fri, 28 Nov 2008 01:30:15 +0000, Cork Soaker wrote:
>>>>
>>>>> Homer wrote:
>>>>>> First the context:
>>>>>>
>>>>> <snip>
>>>>>
>>>>> Yes, a failure during an Installshield setup leaves a bloody awful mess
>>>>> to clean up.
>>>>>
>>>>> However, I don't see how this advocates Linux?
>>>> Try fixing CUPS when it takes a crap.
>>>> Not easy at all.
>>> Let my think when I had the last time problems with CUPS.
>>> Oh, now I remember. Never
>> Sounds like it was the same time that I had my last printing problem with
>> Windows.
>
> I'm beginning to think Peter doesn't even use Linux. He never had a
> problem with his video card, never with CUPs etc etc.
>
> http://www.google.com/search?q=CUPS+problems
Well that's proved it, Hardon liar!
Idiot.
Cork Soaker
Never used it.
A paperless office is perfectly reasonable.
Hadron
Yes. Peter is an idiot. And sometimes a liar.
He either does not use Linux or he is incredible lucky OR he wrote his
own distro or he is telling lies. You take your pick.
--
"His asshole is so reamed out he has room for an oxygen
tank, too."
-- Tattoo Vampire loooking for new accomodation in comp.os.linux.advocacy
Cork Soaker
Saying things make it true Hardon. Stupid lying k00ktroll.
Looks, Windows has never had a problem:
http://www.google.co.uk/search?q=sex+with+donkeys
You're a Google-spastic.
Gregory Shearman
>
>> Peter Köhlmann wrote:
>>
>>> Let my think when I had the last time problems with CUPS.
>>> Oh, now I remember. Never
>>
>> You kidding, dumbkopf? Day after day you come up with these stupid
>> samples of one.
>
> You mean that using linux for more than 10 years and *never* having had any
> problems with CUPS, across a wide variety of distros, is a sample of one?
> So the other linux boxen I installed, but don't run myself, don't count?
CUPS was the saviour of printing on Linux. I remember the shitfight
printing was on linux before CUPS and I thank those developers from the
bottom of my heart.
> Idiot
Sandeep Kumar
> On 2008-11-28, Peter Köhlmann <peter.k...@arcor.de> wrote:
>> DFS wrote:
>>
>>> Peter K��ann wrote:
>>>
>>>> Let my think when I had the last time problems with CUPS.
>>>> Oh, now I remember. Never
>>>
>>> You kidding, dumbkopf? Day after day you come up with these stupid
>>> samples of one.
>>
>> You mean that using linux for more than 10 years and *never* having had any
>> problems with CUPS, across a wide variety of distros, is a sample of one?
>> So the other linux boxen I installed, but don't run myself, don't count?
>
> CUPS was the saviour of printing on Linux. I remember the shitfight
> printing was on linux before CUPS and I thank those developers from the
> bottom of my heart.
Yea...
So did Eric Raymond...
http://www.catb.org/~esr/writings/cups-horror.html
Too bad you weren't around to give him lessons in how CUPS works and how to
configure it.
The Ghost In The Machine
I've had minor problems with CUPS, mostly because at
home my printer is actually attached to another machine
with an older version thereof. I think part of it is
authorization-related as well; my laptop in particular
I've not authorized to use my LAN yet (DHCP and all that),
and I've not set it all up.
These sorts of things I generally expect; the actual
printing part works very well -- when it works.
CUPS has the additional property of talking to Microsoft
print servers, at least here at $EMPLOYER; the main
drawback is that I have to put my username-password in
as part of the print specification -- and change it every
3 months.
Of course, I might print once every month, so as concerns
go, it's not exactly high priority. ;-)
I'm not all that thrilled with the HTML interface (or
with HTML in general, really), but as interfaces go it's
generally easy to use and does what's needed.
It's a lot simpler than "lprng" -- which presumably
is deprecated by now -- or Microsoft's print interfaces,
presumably.
Gregory Shearman
> On Nov 28, 4:31 am, Peter Köhlmann <peter.koehlm...@arcor.de> wrote:
>> flatfish (Sandeep Kumar) nymshifted:
>>
>> > On Fri, 28 Nov 2008 01:30:15 +0000, Cork Soaker wrote:
>>
>> >> Homer wrote:
>> >>> First the context:
>>
>> >> <snip>
>>
>> >> Yes, a failure during an Installshield setup leaves a bloody awful mess
>> >> to clean up.
>>
>> >> However, I don't see how this advocates Linux?
>>
>> > Try fixing CUPS when it takes a crap.
>> > Not easy at all.
>>
>> Let my think when I had the last time problems with CUPS.
>> Oh, now I remember. Never
>> --
>> Any idiot can run XP. And usually does.
>
> I've had minor problems with CUPS, mostly because at
> home my printer is actually attached to another machine
> with an older version thereof. I think part of it is
> authorization-related as well; my laptop in particular
> I've not authorized to use my LAN yet (DHCP and all that),
> and I've not set it all up.
I've had none. I've also had no problems since I dumped Canon and went
for HP.
>
> I'm not all that thrilled with the HTML interface (or with HTML in
> general, really), but as interfaces go it's generally easy to use and
I think the web interface is ugly, clunky and not so easy to navigate.
It hasn't really interfered with controlling the printer.
> does what's needed.
>
God, I hope so. What a fucking monstrosity.
> now -- or Microsoft's print interfaces, presumably.