Velocity Reviews - Computer Hardware Reviews

Velocity Reviews > Newsgroups > Computing > Digital Photography > Re: Possible new feature for next Photoshop

Reply
Thread Tools

Re: Possible new feature for next Photoshop

 
 
Ryan McGinnis
Guest
Posts: n/a
 
      10-26-2011
On Mon, 24 Oct 2011, nospam wrote:

> In article <alpine.DEB.2.02.1110242130410.2055@ryan-BigStorm>, Ryan
> McGinnis <(E-Mail Removed)> wrote:
>
>> Adobe does not "limit compatibility". Camera makers continually change
>> the RAW file format over and over again, requiring Adobe to task a team of
>> coders to recode their RAW engine to deal with them. These employees cost
>> money.

>
> that's because every new sensor is a little different. they don't sit
> around and think gee how can we change something.


Every sensor is different, but with standardization, this could be
overcome. Every camera is different, but so far all programs that I use
to view JPEGs don't care (or even know) what camera or software the JPEG
came from. There is no reason this could not be the case with RAW as well
-- indeed, DNG (Adobe's format) supports all cameras. As you put it,
there is little motivation for cameramakers to standardize. This is not
Adobe's fault.


-Ryan McGinnis
The BIG Storm Picture: http://bigstormpicture.com PGP Key 0x65115E4C
Follow my storm chasing adventures at http://bigstormpicture.blogspot.com
Images@Getty: http://bit.ly/oDW1pT Images@Alamy:<a href="http://bit.ly"...p://bit.ly</a>/aMH6Qd
 
Reply With Quote
 
 
 
 
Wolfgang Weisselberg
Guest
Posts: n/a
 
      10-27-2011
Pete A <(E-Mail Removed)> wrote:

> The 386 and later CPUs have 4 privilege rings, but most OS designers
> use only ring 0 (kernel) and ring 3 (user). This may have been
> acceptable for NT 4, but is totally unacceptable for the complexity of
> more recent OSs including modern UNIX and Linux systems. Even OS/2 used
> 3 of the rings (ring 2 was used for operations such as user-mode I/O).


Feel free to write a patch for Linux. While doing that remember
that there are many other CPU vendors and types outside the 386
anchestry, so please keep your code as generic as possible.


> As a practical example, Sony BMG published CDs with copy protection and
> DRM, which installed a rootkit onto the machine without the user being
> aware of it (a rootkit hijacks part of the OS kernel and is extremely
> difficult to detect). If the OS had been designed properly with driver
> code running at ring 1 instead of 0, this would have been impossible,
> as would much other malware.


Are you absolutely sure?
Are you really sure that drivers wouldn't be able to contain
the functions of a rootkit as wanted by Sony?
Are you really sure that you'll allow a ring 1 driver to write
to the same partition the boot code lies on, and if not, how do
you think a ring 0 driver wouldn't be a driver?

> The only thing that should be allowed to change kernel code is a vendor
> supplied patch or update.


This is too stupid for words.

You probably suffer from "Patches fall from the sky and are
all distributed by Microsoft --- and there are no operating
systems outside Windows, anyway" syndrome.

> User installed software should never be able
> to modify kernel code, not even if the user is a member of the system
> administrator group of users.


So do a
chown root: $kernel_code
chmod go-w $kernel_code
chmod go-w dirs($kernel_code)
and only the root user (note: user, not group!) can change the
kernel code. Easy as apple pie and doesn't even need ACLs.

Oh, users cannot install software to anywhere but public
directories or their own home directory, anyway.


> The third biggest problem with modern OSs is the user experience. Many
> users get so frustrated when logged in correctly as a non-privileged
> user that they use the machine logged into an admin account. There is
> simply no excuse for this incredibly poor user experience in MS, UNIX,
> and Linux.


Do you have any recent user experience with Unix or Linux?

Probably not.

Unless you are doing sysadminy things (i.e. not behaving as a
user), you don't need privileges.


> Until the OS designers use all 4 privilege levels provided by the CPU
> and MS creates a robust shell _for_ Explorer and the user apps to all
> share, we are stuck with increasingly slow, unstable and insecure
> systems.


Hey, are you sure you don't need 5 rings? Or 6? Or 4096?
Just because intel-86 doesn't support any more rings doesn't
mean we don't need them --- and just because intel-86 offers them
doesn't mean we need them.

And tell me ... increasingly slow? I haven't noticed in the last
20 years. Unstable? Sorry, faults in user software (ring 3)
and faults in the hardware (independent of rings) isn't helped by
4 rings. As far as I can see, the linux kernel is rather stable
and isn't getting worse. Insecure? Sure, with the number of code
lines growing fast in applications and the number of applications
that interact with the intenet in some way--- but 4 privilege
levels won't help a bit against compromised SSL host keys, adobe
flash bugs, BEAST attacks against SSL, XSS holes, etc. etc. etc.

In other words: I think you're quite completely wrong and see
only a very small part of reality.


> And my final observation is that while OSs continue to be written in C
> and applications written in C++, we are doomed. This is insanity: at
> the very least, core OS modules and all code that parses user input
> should be written in Pascal. Pascal has inherent run-time range and
> boundary checking and works in harmony with the x86 series (almost zero
> overhead) instruction to implement these checks - buffer overflow (aka
> buffer overrun) injection attacks would be impossible yet they are
> still causing vulnerabilities in modern software. C was, and still is,
> a jack of all trades and master of none.


Pascal doesn't even allow you to copy a 19 character 'string'
(actually a packed array of char) variable into a 20 character
'string' variable. (Look it up!)
Any 'Pascal' that allows that is using some extension.

And remember, everyone would be much safer if everyone wore
a straightjacket.

Much more important than some runtime checks is the errors per
functional unit for the relevant code type.

> I read that the nuclear power
> station up the coast from me has two and a half million lines of C code
> controlling it - it doesn't take a rocket scientist to figure out that
> there are probably quite a few bugs in that amount of C code


Sure, and the same ones would be in Pascal. Up to the "range
check failed, code aborted, not inserting the control rods ..."

-Wolfgang
 
Reply With Quote
 
 
 
 
Pete A
Guest
Posts: n/a
 
      10-28-2011
On 2011-10-28 00:17:30 +0100, Wolfgang Weisselberg said:

> Pete A <(E-Mail Removed)> wrote:
>
>> [...]

>
>> The only thing that should be allowed to change kernel code is a vendor
>> supplied patch or update.

>
> This is too stupid for words.
>
> You probably suffer from "Patches fall from the sky and are
> all distributed by Microsoft --- and there are no operating
> systems outside Windows, anyway" syndrome.


Yeah. What's your excuse for writing total *******s in your reply?

 
Reply With Quote
 
Wolfgang Weisselberg
Guest
Posts: n/a
 
      10-28-2011
Pete A <(E-Mail Removed)> wrote:
> On 2011-10-16 02:24:17 +0100, Charles E. Hardwidge said:


>> broader customer empowerment and cost issues. Perhaps both management and
>> customer education is helpful?


> Only to those who sincerely want to learn something. I have a funny
> feeling the number is way below 10%. I don't think the modern customer
> wants to be empowered; I think they prefer moaning about the state of
> everything. Empower them and they will have nothing to moan about.
> Being responsible and accountable for one's knowledge and actions has
> become a dirty phrase, normally spoken only by "the enemy of the
> people."


Hmmm, and how does *that* match to the health reform in the US,
which AFAIK didn't happen? As in "We don't want basic health
insurance for everyone, we want to be empowered to have *no*
health insurance at all"?

(Which from my POV is more than utterly stupid. I cannot
comprehend a society that actually *wants* gaping holes that it
then tries to close with voluntary actions of charity. Not only
is that self-serving and self-aggrandizing (first hurt them,
then heal them a bit for making oneself look better), it's also
inherently unfair (charity isn't very equally distributed, so some
will get lots and others naught) and puts the recipient into an
uncomfortable position of dependency (can't speak out against X
or loose the charity) and make him half a beggar (is having power
(whether you misuse or use it or not) over other people one put
into that position christian? Or is that the slave holder gene?)

But then I live in a country where the workers parties demanded
such things (back when they were only third class voters!) and got
a very weak response in 1883 (that's eighteen-hundred-eighty-three)
in form of a law that introduced health insurance for (nearly)
everyone.

As I cannot comprehend them I cannot tell them what to do, they
will have to find their own way.)

-Wolfgang
 
Reply With Quote
 
Wolfgang Weisselberg
Guest
Posts: n/a
 
      10-28-2011
Charles E. Hardwidge <(E-Mail Removed)> wrote:

> The MS firewall is completely functional but most of its capabilities aren't
> exposed via a GUI, including outbound protection. MS deemed two way firewall
> protection "wasn't necessary".


The MS 'firewall' is a desktop machine packet filter, not
a firewall. (A firewall is a security concept, which usually
includes filters (e.g. packet filters, deep inspection, blocked
ports), but also e.g. rules of behaviour (e.g. "do not download from
dubious web sites", "run AV software and update it", "no external
USB sticks on the premises" and so on.)

The functionality of a desktop 'firewall' can be described on
topic:
http://www.e-2005.de/potw/unischranke.jpg


> I've written on two way permission based circles of protection before (which
> can reach down to the bit level and scale across the whole internet) and
> this gets pretty boring after a while.


So where's the paper, and where is the proof of concept code for
your idea? And why has noone implemented that yet? It seems
something like that would be needed in a lot of places ...


> The rolling out of smart meters in the UK was halted not because of cost or
> user reasons but the fact the government **** itself silly that a teenage
> hacker in a basement flat in Nairobi could switch off the UK.


And you know that because mothman told you so, or do you have
some more reliable source?


-Wolfgang
 
Reply With Quote
 
Wolfgang Weisselberg
Guest
Posts: n/a
 
      10-28-2011
Ryan McGinnis <(E-Mail Removed)> wrote:
> On Sun, 16 Oct 2011, Wolfgang Weisselberg wrote:
>> Eric Stevens <(E-Mail Removed)> wrote:


>>> There is also the very considerable testing required to ensure that
>>> your software doesn't conflict with commonly available software from
>>> others such as McAfee, Firefox etc.


>> And who does that testing? Nearly noone. If it conflicts with a
>> virus scanner, fix the scanner. If it conflicts with Firefox ---
>> and why should it, unless it uses it a lot --- claim that only
>> IE is supported. Done.


> As someone who beta tested CS4 and CS5, yes, there are people doing
> testing. Quite a few. Have you ever actually worked for or with a large
> software company, or is the extent of your coding all Open Source / GNU?


I've worked for companies who *live* *and* *die* by the software
they produce[1]. I'm the one who's clamouring for more testing and
and more automated testing. Sometimes I don't have to actually
convert people --- I only have to get them to take the resources
to test more/better.

I have quite a good idea how much, or rather, how little testing is
done and how "conflict[s] with commonly available [but completely
unrelated] software" is handled.

As to GNU --- tests have shown long ago that GNU software is
more stable and has less bugs than vendor supplied software.
Especially on an 1:1 comparison, since the programs do the same
stuff and should have identical outputs. (Not to mention that GNU
software usually has lots more features and much less annoyances
and shortcomings.)

-Wolfgang

[1] literally. Microsoft, for example, can survive something
like Vista, Microsoft Bob, and many more. Where I worked,
such a misstep would have been fatal.
 
Reply With Quote
 
Wolfgang Weisselberg
Guest
Posts: n/a
 
      10-28-2011
Ryan McGinnis <(E-Mail Removed)> wrote:
> On Mon, 17 Oct 2011, Wolfgang Weisselberg wrote:
>> nospam <(E-Mail Removed)> wrote:


>> [DNG converter]


>>> it's not a work around. if someone gets a new camera they can get a
>>> newer version of camera raw and use dng converter to batch convert all
>>> of them to dng and keep using their existing old version of photoshop.
>>> it works fine and is fully supported. it's also totally free.


>> Totally free? You get the source code? You get the rights
>> to use, read, improve, share and share the improved versions?


>> Or is that just "free beer, but certainly *no* free speech"?


> Free means "it costs nothing" in common language.


So "it costs nothing speech" is what the US upholds?
Fits.


> If I give you a
> photograph of mine "for free", that means it costs nothing.


But nospam didn't write 'it's for free'. Nospam wrote "it's
also totally free." TOTALLY free.


> Most people
> would not then assume that they're allowed to make 20,000 copies of the
> photo and resell them.


If it's TOTALLY free, I would assume exactly that. So would
enough other people.

And if it's about software, I'd assume that twice as much.

-Wolfgang
 
Reply With Quote
 
Wolfgang Weisselberg
Guest
Posts: n/a
 
      10-28-2011
Pete A <(E-Mail Removed)> wrote:
> On 2011-10-28 00:17:30 +0100, Wolfgang Weisselberg said:
>> Pete A <(E-Mail Removed)> wrote:


>>> [...]


>>> The only thing that should be allowed to change kernel code is a vendor
>>> supplied patch or update.


>> This is too stupid for words.


>> You probably suffer from "Patches fall from the sky and are
>> all distributed by Microsoft --- and there are no operating
>> systems outside Windows, anyway" syndrome.


> Yeah. What's your excuse for writing total *******s in your reply?


*******s? Because e.g. Linux kernel developers *have* to compile
and change the kernel code, without going through a 'vendor'
every 10 minutes?

Not to mention that quite a few people want the freedom to change,
compile, add patches to their own kernel.

But in Pete's world users are not enabled and have to use Windows
(thank you, secure boot). Or Macs. A few may be allowed to use
Red Hat or SuSE (and nothing else) (but will be unable to boot
Windows) and that's that. Welcome to the new freedom.


Next year: "The only camera that should be allowed to take
photos is a pinhole phone camera."

-Wolfgang
 
Reply With Quote
 
Wolfgang Weisselberg
Guest
Posts: n/a
 
      10-29-2011
Ryan McGinnis <(E-Mail Removed)> wrote:
> On Mon, 24 Oct 2011, nospam wrote:
>> McGinnis <(E-Mail Removed)> wrote:


>>> Adobe does not "limit compatibility". Camera makers continually change
>>> the RAW file format over and over again, requiring Adobe to task a team of
>>> coders to recode their RAW engine to deal with them. These employees cost
>>> money.


>> that's because every new sensor is a little different. they don't sit
>> around and think gee how can we change something.


> Every sensor is different, but with standardization, this could be
> overcome.


True.

That would however mean that all sensors would have to conform to
whatever the standard allows. And we all know how far-looking
standards can be. Like ... being able to address only 1 MByte
(640k should be enough for everyone), FAT and large disks
(that's why we have FAT32 and co. now ... for now), disks >
~2 GB and booting, the various SD, SDHC, SDXC variants, and a
million other examples.

So that would hinder experimentation and improvements.


> Every camera is different, but so far all programs that I use
> to view JPEGs don't care (or even know) what camera or software the JPEG
> came from.


Many programs cannot even deal properly with aRGB JPEGs, never mind
other colourspaces. Or anything outside the basic JPEG structure.

And basic JPEG is an end product. Made for an unchanging target.

> There is no reason this could not be the case with RAW as well


RAW is a source format, which changes as technology improves
or changes.

> -- indeed, DNG (Adobe's format) supports all cameras.


No, it doesn't. Try Lytro.
It's support at best allows most data to be written. It doesn't
mean the data is understood. Rectangular pixels, for example,
or hexagonal ones --- whoever reads the DNG must *know* not only
these facts, but also how to *deal* with them.

So at the very best DNG is somewhat like unicode: you can write
lots of characters --- but you also must display the glyphs of the
characters in some font (which means you need to have the font)
and even with all that you'll likely not make any sense out of
most languages.

> As you put it,
> there is little motivation for cameramakers to standardize. This is not
> Adobe's fault.


Why should camera makers help Adobe, anyway, and punish Adobe's
competition? After all, most camera makers have their own
specialized RAW converter that does exactly what the camera
needs ...

-Wolfgang
 
Reply With Quote
 
Pete A
Guest
Posts: n/a
 
      10-30-2011
On 2011-10-29 00:23:31 +0100, Wolfgang Weisselberg said:

> Pete A <(E-Mail Removed)> wrote:
>> On 2011-10-28 00:17:30 +0100, Wolfgang Weisselberg said:
>>> Pete A <(E-Mail Removed)> wrote:

>
>>>> [...]

>
>>>> The only thing that should be allowed to change kernel code is a vendor
>>>> supplied patch or update.

>
>>> This is too stupid for words.

>
>>> You probably suffer from "Patches fall from the sky and are
>>> all distributed by Microsoft --- and there are no operating
>>> systems outside Windows, anyway" syndrome.

>
>> Yeah. What's your excuse for writing total *******s in your reply?

>
> *******s? Because e.g. Linux kernel developers *have* to compile
> and change the kernel code, without going through a 'vendor'
> every 10 minutes?


Kernel developers write kernel code and make it available to end users
via a vending mechanism. Strangely enough, these are called patches or
updates as appropriate. RPMs are one example of packaging and delivery
methods. Those who release Linux kernels are the vendors. So, what was
"too stupid for words?"

> Not to mention that quite a few people want the freedom to change,
> compile, add patches to their own kernel.


Now you have digressed from secure OSs. For experimental systems or
systems for which security is not a high priority, of course anyone can
modify their own OS.

> But in Pete's world users are not enabled and have to use Windows
> (thank you, secure boot). Or Macs. A few may be allowed to use
> Red Hat or SuSE (and nothing else) (but will be unable to boot
> Windows) and that's that. Welcome to the new freedom.


Quite a few corporations would be a little upset if an employee changed
kernel code!

> Next year: "The only camera that should be allowed to take
> photos is a pinhole phone camera."


Nah, phone cameras are less secure than stand-alone models.


By the way, Pascal strings are NOT packed arrays of char - you would
know this if you had ever used the language. Furthermore, any Pascal
that allows the copy in your example is not using an extension, it is
using a broken compiler

 
Reply With Quote
 
 
 
Reply

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Re: Possible new feature for next Photoshop PeterN Digital Photography 10 10-23-2011 07:33 PM
Re: Possible new feature for next Photoshop Trevor Digital Photography 9 10-17-2011 05:08 PM
Best Photoshop Feature You've Never Heard Of? Annika1980 Digital Photography 2 12-12-2005 04:50 PM
CurrentElement->next = CurrentElement->next->next (UNDEFINED?) Deniz Bahar C Programming 2 03-09-2005 12:45 AM



Advertisments