ReiserFS IFSD for Windows

ReiserFS IFSD for Windows

[Mark Piper]

This summer, I will be starting a 3.5-month project to create an 
Installable File System Driver (IFSD) to read ReiserFS under Microsoft 
Windows, similar to ext2fsd.   (The project is a practicum to complete 
my degree in software engineering from  Carnegie Mellon.)

I have three questions:

* Is anyone out there interested in using the tool as it is developed?

* Is anyone out there interested in mentoring the development of the 
tool?  (This basically consists of answering specific technical 
questions when they arise.)

* Other than compliance to the GPL, are there any other liscensing 
factors I should be aware of?

For anyone interested, my initial plan is to port the read-only ReiserFS 
3 code from the GRUB bootloader, and turn it into an IFSD using the same 
techniques as in ext2fsd -- although I'd love to hear other ideas, and 
opinions of whether the project might be feasible with ReiserFS 4.

  Thanks,
     Mark Piper

Re: ReiserFS IFSD for Windows

[Bedros Hanounik]

[-- Attachment #1: Type: text/plain, Size: 1659 bytes --]

I'll be interested in testing your code. I have an old 20GB HD which I can 
play with. 

Currently, I have a USB drive formatted in NTFS, because, I want it to be 
accessible by friend's XP machine in case of emergencies. and I would 
convert it to ReiserFS once your code is stable enough.

having a read only mode is fine with me; write mode would be a nice option, 
but it's not important in my case.

My impression is Hans doesn't like the Idea, but a stripped down version of 
ReiserFS driver for windows could be a good thing, and encourages more 
people to use ReiserFS.


-Bedros


On 5/12/05, Mark Piper <mwp@cs.cmu.edu> wrote:
> 
> This summer, I will be starting a 3.5-month project to create an
> Installable File System Driver (IFSD) to read ReiserFS under Microsoft
> Windows, similar to ext2fsd. (The project is a practicum to complete
> my degree in software engineering from Carnegie Mellon.)
> 
> I have three questions:
> 
> * Is anyone out there interested in using the tool as it is developed?
> 
> * Is anyone out there interested in mentoring the development of the
> tool? (This basically consists of answering specific technical
> questions when they arise.)
> 
> * Other than compliance to the GPL, are there any other liscensing
> factors I should be aware of?
> 
> For anyone interested, my initial plan is to port the read-only ReiserFS
> 3 code from the GRUB bootloader, and turn it into an IFSD using the same
> techniques as in ext2fsd -- although I'd love to hear other ideas, and
> opinions of whether the project might be feasible with ReiserFS 4.
> 
> Thanks,
> Mark Piper
> 
>

[-- Attachment #2: Type: text/html, Size: 2031 bytes --]

Re: ReiserFS IFSD for Windows

[Christian Iversen]

[-- Attachment #1: Type: text/plain, Size: 733 bytes --]

On Thursday 12 May 2005 18:59, Mark Piper wrote:
> This summer, I will be starting a 3.5-month project to create an
> Installable File System Driver (IFSD) to read ReiserFS under Microsoft
> Windows, similar to ext2fsd.   (The project is a practicum to complete
> my degree in software engineering from  Carnegie Mellon.)
>
> I have three questions:
>
> * Is anyone out there interested in using the tool as it is developed?

I would happily use it, as I have a hard time reading my Linux-partition on my 
dual-boot notebook. (if you meant commercially interested, then no :)

I'd also be happy to test it. 

Is there any chance the read-write code will eventually be integrated? 

-- 
Regards,
Christian Iversen

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: ReiserFS IFSD for Windows

[Hans Reiser]

Mark Piper wrote:

> This summer, I will be starting a 3.5-month project to create an
> Installable File System Driver (IFSD) to read ReiserFS under Microsoft
> Windows, similar to ext2fsd.   (The project is a practicum to complete
> my degree in software engineering from  Carnegie Mellon.)

I would prefer that you not do it.  Namesys believes that people who use
proprietary operating systems should pay to use reiserfs.

I can suggest quite a few other tasks that would make a more impressive
practicum (copy on write links, various other new and needed plugins).

>
> I have three questions:
>
> * Is anyone out there interested in using the tool as it is developed?
>
> * Is anyone out there interested in mentoring the development of the
> tool?  (This basically consists of answering specific technical
> questions when they arise.)
>
> * Other than compliance to the GPL, are there any other liscensing
> factors I should be aware of?
>
> For anyone interested, my initial plan is to port the read-only
> ReiserFS 3 code from the GRUB bootloader, and turn it into an IFSD
> using the same techniques as in ext2fsd -- although I'd love to hear
> other ideas, and opinions of whether the project might be feasible
> with ReiserFS 4.
>
>  Thanks,
>     Mark Piper
>
>
>
>
>

Re: ReiserFS IFSD for Windows

[David Masover]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hans Reiser wrote:
> Mark Piper wrote:
> 
> 
>>This summer, I will be starting a 3.5-month project to create an
>>Installable File System Driver (IFSD) to read ReiserFS under Microsoft
>>Windows, similar to ext2fsd.   (The project is a practicum to complete
>>my degree in software engineering from  Carnegie Mellon.)
> 
> 
> I would prefer that you not do it.  Namesys believes that people who use
> proprietary operating systems should pay to use reiserfs.
> 
> I can suggest quite a few other tasks that would make a more impressive
> practicum (copy on write links, various other new and needed plugins).

Could COW really be a plugin?

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org
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=gdCE
-----END PGP SIGNATURE-----

Re: ReiserFS IFSD for Windows

[Hans Reiser]

David Masover wrote:

> Hans Reiser wrote:
>
> >Mark Piper wrote:
>
>
> >>This summer, I will be starting a 3.5-month project to create an
> >>Installable File System Driver (IFSD) to read ReiserFS under Microsoft
> >>Windows, similar to ext2fsd.   (The project is a practicum to complete
> >>my degree in software engineering from  Carnegie Mellon.)
>
>
> >I would prefer that you not do it.  Namesys believes that people who use
> >proprietary operating systems should pay to use reiserfs.
>
> >I can suggest quite a few other tasks that would make a more impressive
> >practicum (copy on write links, various other new and needed plugins).
>
>
> Could COW really be a plugin?
>
Sure, or inheritance or quite a few others.

Re: ReiserFS IFSD for Windows

[Leif W]

> Subject:    Re: ReiserFS IFSD for Windows
> From:       Hans Reiser <reiser () namesys ! com>
> Date:       2005-05-13 18:14:32
> Message-ID: 4284EE88.2020309 () namesys ! com
>
>> Mark Piper wrote:
>>
>> This summer, I will be starting a 3.5-month project to create an
>> Installable File System Driver (IFSD) to read ReiserFS under 
>> Microsoft
>> Windows, similar to ext2fsd.   (The project is a practicum to 
>> complete
>> my degree in software engineering from  Carnegie Mellon.)
>
> I would prefer that you not do it.  Namesys believes that people who 
> use
> proprietary operating systems should pay to use reiserfs.

I've recently wondered about such functionality, if it existed or if 
there were plans to develop.  I would be very interested to make use of 
that software.  Now I see the response and have some strong rebuttals.

Hans, you say that proprietary users should pay to use ReiserFS.  But I 
wonder what you mean or if you're sincere.  There is no evidence of any 
way to "buy" such software on the Namesys website.  So in what manner do 
people normally "pay" for this functionality?  You want people to pay 
$25 to ask if it exists, and then what?  Tell it's for sale and charge 
them again?  As far as any of the world knows, an IFSD does not exist, 
so you could be telling people that the Namesys team will write it from 
scratch at $5k/mo.  Then what of subsequent requests?  Resell the code? 
Or do you erase it and rewrite from scratch upon the next request?  :p 
If you have code and it's for sale, then you really need to attend a 
class on basic business and marketing.  Or at the very least put up an 
itemized price list and description on a web page, with a means of 
contact.  Hopefully it'd be as simple to use as some shopping cart 
solution.  :p

I've talked to some actual Russians (where the website states Namesys 
custom/contract coders are located), and they survive on much less than 
$500/month, perhaps $300/month.  So where do you get off charging ten 
times that?  Only about 1 month's worth of work to survive a year.  I 
only wish I could find a job that paid so well.  Even in the US, I wish 
I could find a job to pay $5k/month.  Most here manage to survive on 
less than half of that.  There is great competition for such well paying 
jobs.

But anyways, I digress.  So I wonder, perhaps you mean that proprietary 
users must "pay" in a metaphorical sense.  Pay for their decision to use 
proprietary software.  The cost is interoperability.  So, you eschew the 
tenets of proprietary software by adopting the same closed source 
mentality to prevent interoperability.  Interesting contradiction in 
your flawed logic.  I suppose you believe anyone who ever paid for 
anything that you don't have should pay you money, just because you hold 
a grudge about the use of money in society which from time to time 
excludes some people from some items, experiences, benefits, and so on.

In laymen's terms (and analogy), you don't want to share your ball with 
anyone else who bought a ball so you'll act like a child and go take it 
home and pout over it and tell yourself what a great ball player you 
are, in the imaginary games you'd play in your head while everyone else 
is out in the real world getting dirty and interacting with each other.

And what do you know of my attitudes from proprietary or free software, 
or my economic or educational background?  I grew up very poor, with no 
guidance in computers.  I didn't discover my computer interest until 
college, which I had entered early and for free, due solely to early 
merits of my scholastic performance.  My interest in computers took 
over, but I didn't have money to finish my education, let alone build a 
computer.  So I worked and saved, bought some text books, studied on my 
own with no computer, and finally built my own.  What I've learned of 
source code, what use I've made of hardware over the years has largely 
if not solely been inspired and influenced by free software projects 
such as ReiserFS.  Whenever I have worked on any job involving software, 
I have requested that if it would be generally useful, that it be 
released under a free license (I prefer GNU's GPL or in some cases 
LGPL).  I've donated what meager amounts of money I can to some projects 
here and there.

It's not always by my choice that I end up in Windows, but when I do, I 
like to have access to at least some similar functionality.

Apache.
Perl.
PHP.
MySQL.
Gaim.
GIMP.
GTK+.
wxWidgets.
Python.
MinGW/MSYS (collections of GNU tools).
Mozilla.

The list goes on...

Imagine where any of these software packages would be if they had not 
opted for interoperability and compatibility.  They have inspired me to 
think of interoperability.  If I need to work to live, then I need to 
work when I can, and as such I may not be in a position to dictate to 
the company or the industry that Linux or FreeBSD must be used.  Or else 
what?  Or else I am requested to resign or fired outright, or merely I 
am just never hired, and I starve and die in the street.  The end?  I'd 
like it not to be.  So I must survive another day.  I do favor those 
companies that predominantly use Linux or FreeBSD, and in the past I 
have taken a serious pay cut to work at such a company (I made $14k in a 
YEAR, no health or dental, no 401k).  Simply because in comparison to 
the myriad inexplicable and truly idiotic Windows errors that can't be 
fixed or worked around, FreeBSD and Linux are a sheer pleasure to work 
with.  Blissful by comparison.  I'd rather make less money and enjoy 
what I do, and the tools I work with are an important factor.  As such, 
portable programs make my life easier no matter where I happen to be at 
any given moment.

In either work or in personal environment, I find it helpful to have two 
or more systems, each with different OSes and versions, especially if 
doing cross-platform development or testing for free software projects 
or not.  All in the interest of interoperability.  But I may not always 
have the resources available to run multiple machines, hence the need 
for a dual, tri, quad, or N-boot computer.  It is in this case when 
interoperability is a key factor for filesystems.  And it grieves me 
that Namesys has in it's folly taken such a narrow minded view of the 
software world in this modern day.  I start to wonder if that prejudiced 
mentality permeates other aspects of life, such as race, religion, 
politics, language, cultures and geographical locations.  I certainly 
hope not.

The first software company I worked for was doing a lot of web based 
stuff.  That is what really grounded my appreciation for software that 
can talk to each other.  So, while my co-workers were concerned only 
with IE4 or IE5, I was also interested in Netscape and Mozilla being 
able to view a page, and even Lynx at least able to view the 
information.  When I was at home with a Linux-only system, I also became 
much more aware of the effects of being summarily excluded by website 
makers who were only interested in proprietary code and felt no need to 
provide Mozilla support.  It forces you to have a much more open mind 
and discipline towards software design.

Alone, I am neither the quickest, most learned nor most skilled 
programmer.  Also, I am not "made of money", as they say, so can not 
underwrite the full costs of development of software such as a Windows 
IFSD.  Yet, here you have someone in society, who is possessed of enough 
skill and money to be in a position to write such software, who is 
freely offering to undertake the project, who has a vital stake in it's 
completion and therefore likely committed to follow through, with the 
benefit to them being a degree and a good career builder, and the 
benefit to everyone else being interoperability.  This is what I always 
found to be in the spirit of free software.  You made free software such 
that others could learn from and extend it, to make life better for 
everyone else.  Yet you shoot it down.  The payoff to you would be 
people buying the software.  That in itself indicates a conflict of 
interest.

The motives I question.  The logic I question.  The ethics I question.


Leif


> I can suggest quite a few other tasks that would make a more 
> impressive
> practicum (copy on write links, various other new and needed plugins).
>
>>
>> I have three questions:
>>
>> * Is anyone out there interested in using the tool as it is 
>> developed?
>>
>> * Is anyone out there interested in mentoring the development of the
>> tool?  (This basically consists of answering specific technical
>> questions when they arise.)
>>
>> * Other than compliance to the GPL, are there any other licensing
>> factors I should be aware of?
>>
>> For anyone interested, my initial plan is to port the read-only
>> ReiserFS 3 code from the GRUB bootloader, and turn it into an IFSD
>> using the same techniques as in ext2fsd -- although I'd love to hear
>> other ideas, and opinions of whether the project might be feasible
>> with ReiserFS 4.
>>
>>  Thanks,
>>     Mark Piper
>>
>>
>>
>>
>> 

Re: ReiserFS IFSD for Windows

[Hans Reiser]

I believe that once someone produces a piece of software that is the
best in the world, they are more than entitled to earn what the average
american programmer earns even if they were born in Russia.  My $30/hr.
guys are underpaid.

Namesys will eventually produce a windows driver, once we deal with a
whole bunch of other tasks before us (like getting into the kernel).  We
will charge money for that driver.  Thanks for your enthusiasm for that
driver.

Hans

Leif W wrote:

>> Subject:    Re: ReiserFS IFSD for Windows
>> From:       Hans Reiser <reiser () namesys ! com>
>> Date:       2005-05-13 18:14:32
>> Message-ID: 4284EE88.2020309 () namesys ! com
>>
>>> Mark Piper wrote:
>>>
>>> This summer, I will be starting a 3.5-month project to create an
>>> Installable File System Driver (IFSD) to read ReiserFS under Microsoft
>>> Windows, similar to ext2fsd.   (The project is a practicum to complete
>>> my degree in software engineering from  Carnegie Mellon.)
>>
>>
>> I would prefer that you not do it.  Namesys believes that people who use
>> proprietary operating systems should pay to use reiserfs.
>
>
> I've recently wondered about such functionality, if it existed or if
> there were plans to develop.  I would be very interested to make use
> of that software.  Now I see the response and have some strong rebuttals.
>
> Hans, you say that proprietary users should pay to use ReiserFS.  But
> I wonder what you mean or if you're sincere.  There is no evidence of
> any way to "buy" such software on the Namesys website.  So in what
> manner do people normally "pay" for this functionality?  You want
> people to pay $25 to ask if it exists, and then what?  Tell it's for
> sale and charge them again?  As far as any of the world knows, an IFSD
> does not exist, so you could be telling people that the Namesys team
> will write it from scratch at $5k/mo.  Then what of subsequent
> requests?  Resell the code? Or do you erase it and rewrite from
> scratch upon the next request?  :p If you have code and it's for sale,
> then you really need to attend a class on basic business and
> marketing.  Or at the very least put up an itemized price list and
> description on a web page, with a means of contact.  Hopefully it'd be
> as simple to use as some shopping cart solution.  :p
>
> I've talked to some actual Russians (where the website states Namesys
> custom/contract coders are located), and they survive on much less
> than $500/month, perhaps $300/month.  So where do you get off charging
> ten times that?  Only about 1 month's worth of work to survive a
> year.  I only wish I could find a job that paid so well.  Even in the
> US, I wish I could find a job to pay $5k/month.  Most here manage to
> survive on less than half of that.  There is great competition for
> such well paying jobs.
>
> But anyways, I digress.  So I wonder, perhaps you mean that
> proprietary users must "pay" in a metaphorical sense.  Pay for their
> decision to use proprietary software.  The cost is interoperability. 
> So, you eschew the tenets of proprietary software by adopting the same
> closed source mentality to prevent interoperability.  Interesting
> contradiction in your flawed logic.  I suppose you believe anyone who
> ever paid for anything that you don't have should pay you money, just
> because you hold a grudge about the use of money in society which from
> time to time excludes some people from some items, experiences,
> benefits, and so on.
>
> In laymen's terms (and analogy), you don't want to share your ball
> with anyone else who bought a ball so you'll act like a child and go
> take it home and pout over it and tell yourself what a great ball
> player you are, in the imaginary games you'd play in your head while
> everyone else is out in the real world getting dirty and interacting
> with each other.
>
> And what do you know of my attitudes from proprietary or free
> software, or my economic or educational background?  I grew up very
> poor, with no guidance in computers.  I didn't discover my computer
> interest until college, which I had entered early and for free, due
> solely to early merits of my scholastic performance.  My interest in
> computers took over, but I didn't have money to finish my education,
> let alone build a computer.  So I worked and saved, bought some text
> books, studied on my own with no computer, and finally built my own. 
> What I've learned of source code, what use I've made of hardware over
> the years has largely if not solely been inspired and influenced by
> free software projects such as ReiserFS.  Whenever I have worked on
> any job involving software, I have requested that if it would be
> generally useful, that it be released under a free license (I prefer
> GNU's GPL or in some cases LGPL).  I've donated what meager amounts of
> money I can to some projects here and there.
>
> It's not always by my choice that I end up in Windows, but when I do,
> I like to have access to at least some similar functionality.
>
> Apache.
> Perl.
> PHP.
> MySQL.
> Gaim.
> GIMP.
> GTK+.
> wxWidgets.
> Python.
> MinGW/MSYS (collections of GNU tools).
> Mozilla.
>
> The list goes on...
>
> Imagine where any of these software packages would be if they had not
> opted for interoperability and compatibility.  They have inspired me
> to think of interoperability.  If I need to work to live, then I need
> to work when I can, and as such I may not be in a position to dictate
> to the company or the industry that Linux or FreeBSD must be used.  Or
> else what?  Or else I am requested to resign or fired outright, or
> merely I am just never hired, and I starve and die in the street.  The
> end?  I'd like it not to be.  So I must survive another day.  I do
> favor those companies that predominantly use Linux or FreeBSD, and in
> the past I have taken a serious pay cut to work at such a company (I
> made $14k in a YEAR, no health or dental, no 401k).  Simply because in
> comparison to the myriad inexplicable and truly idiotic Windows errors
> that can't be fixed or worked around, FreeBSD and Linux are a sheer
> pleasure to work with.  Blissful by comparison.  I'd rather make less
> money and enjoy what I do, and the tools I work with are an important
> factor.  As such, portable programs make my life easier no matter
> where I happen to be at any given moment.
>
> In either work or in personal environment, I find it helpful to have
> two or more systems, each with different OSes and versions, especially
> if doing cross-platform development or testing for free software
> projects or not.  All in the interest of interoperability.  But I may
> not always have the resources available to run multiple machines,
> hence the need for a dual, tri, quad, or N-boot computer.  It is in
> this case when interoperability is a key factor for filesystems.  And
> it grieves me that Namesys has in it's folly taken such a narrow
> minded view of the software world in this modern day.  I start to
> wonder if that prejudiced mentality permeates other aspects of life,
> such as race, religion, politics, language, cultures and geographical
> locations.  I certainly hope not.
>
> The first software company I worked for was doing a lot of web based
> stuff.  That is what really grounded my appreciation for software that
> can talk to each other.  So, while my co-workers were concerned only
> with IE4 or IE5, I was also interested in Netscape and Mozilla being
> able to view a page, and even Lynx at least able to view the
> information.  When I was at home with a Linux-only system, I also
> became much more aware of the effects of being summarily excluded by
> website makers who were only interested in proprietary code and felt
> no need to provide Mozilla support.  It forces you to have a much more
> open mind and discipline towards software design.
>
> Alone, I am neither the quickest, most learned nor most skilled
> programmer.  Also, I am not "made of money", as they say, so can not
> underwrite the full costs of development of software such as a Windows
> IFSD.  Yet, here you have someone in society, who is possessed of
> enough skill and money to be in a position to write such software, who
> is freely offering to undertake the project, who has a vital stake in
> it's completion and therefore likely committed to follow through, with
> the benefit to them being a degree and a good career builder, and the
> benefit to everyone else being interoperability.  This is what I
> always found to be in the spirit of free software.  You made free
> software such that others could learn from and extend it, to make life
> better for everyone else.  Yet you shoot it down.  The payoff to you
> would be people buying the software.  That in itself indicates a
> conflict of interest.
>
> The motives I question.  The logic I question.  The ethics I question.
>
>
> Leif
>
>
>> I can suggest quite a few other tasks that would make a more impressive
>> practicum (copy on write links, various other new and needed plugins).
>>
>>>
>>> I have three questions:
>>>
>>> * Is anyone out there interested in using the tool as it is developed?
>>>
>>> * Is anyone out there interested in mentoring the development of the
>>> tool?  (This basically consists of answering specific technical
>>> questions when they arise.)
>>>
>>> * Other than compliance to the GPL, are there any other licensing
>>> factors I should be aware of?
>>>
>>> For anyone interested, my initial plan is to port the read-only
>>> ReiserFS 3 code from the GRUB bootloader, and turn it into an IFSD
>>> using the same techniques as in ext2fsd -- although I'd love to hear
>>> other ideas, and opinions of whether the project might be feasible
>>> with ReiserFS 4.
>>>
>>>  Thanks,
>>>     Mark Piper
>>>
>>>
>>>
>>>
>>>
>
>
>
>
>