I'm looking for objective of my master thesis

[Daniel Hans]

Hello,

I am doing my final year of computer science studies at University of
Warsaw, Poland. Recently I have been looking for the objective of my
master thesis. I am attending at 'Computer System Security and
Cryptography' seminar at my university and, therefore, the requirement
is that it has to be in this field. For some time I have participated
in some open source projects development and came up with a concept of
working for open source community. I asked my friend from Google if he
knew any projects hosted by Google which could have some security
concerns. In his answer, he told me about the Native Client project:-)
I read some documentation, then played with it for a while and think
it is a nice idea:-)

I have seen the Security Contest you ran, so you certainly have to
deal with security problems (as probably most browser apps), but I am
afraid that such issues would not be enough for a master thesis…
Maybe there is a chance, that you have in mind a security related
feature that could be introduced to Native Client, on which I could
work? It could require doing some research, tests, everything… If you
graduated, you must know better:-)

I would appreciate for your response. If you need some information
from me, I will gladly answer to your questions. If there is nothing I
can do, maybe you know some open source projects which would require
some more complex work in security area?

Regards,
Daniel

[Bennet Yee (余仕斌)]

hi daniel,

do you have a research advisor? i don't know how polish universities
work, but in many US universities, masters students can choose from
course-based or research-based tracks, and research track students
will need to have a professor who will guide the research. several
google NaCl team members have been professors in our past lives, but a
professor at your institution is needed, and generally it helps to
work with her/him to finalize the research focus.

that said, we can try to come up with security-related projects that
would be appropriate for an MS thesis project. since you hadn't
mentioned crypto except as part of the seminar title, is it safe to
assume that you're more interested in the OS/Web security side of
things?

should we take this off list?

-bsy

>
> Regards,
> Daniel
>
> --
>
> You received this message because you are subscribed to the Google Groups "Native Client Discuss" group.
> To post to this group, send email to native-cli...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/native-client-discuss?hl=.
>
>
>



--
bennet s yee
i usually don't capitalize due to mild tendonitis

[Daniel Hans]

Hello,

thank you for your response!

Ok. So I will try to explain how it works in Poland. We do not have
any choice between course-based and research based tracks. Each
student has to take a number of courses, but also write a master
thesis which describes student's more or less comprehensive work on a
given subject.
I have a tutor who may be called a research advisor, but his main role
is to keep track on progress of my work. In practice, I have to meet
with him like once a month and give a short update on how the things
go. He also has to approve the subject of work, but it should not be a
problem, because I have initially talked to him and he agreed on work
on some open source project.

> google NaCl team members have been professors in our past lives, but a
> professor at your institution is needed, and generally it helps to
> work with her/him to finalize the research focus.

It is great to know that you have academic experience:) If it is
needed, I will give my tutor a contact to you.

> that said, we can try to come up with security-related projects that
> would be appropriate for an MS thesis project.  since you hadn't
> mentioned crypto except as part of the seminar title, is it safe to
> assume that you're more interested in the OS/Web security side of
> things?

Yes, it is. I am more interested in that area. The title of the
seminary is general, so that we can focus on many various (theoretical
or practical) topics.

Regards,
Daniel

> should we take this off list?
>
> -bsy
>
>
>
> > Regards,
> > Daniel
>
> > --
>
> > You received this message because you are subscribed to the Google Groups "Native Client Discuss" group.
> > To post to this group, send email to native-cli...@googlegroups.com.

> > For more options, visit this group athttp://groups.google.com/group/native-client-discuss?hl=.

[Bennet Yee (余仕斌)]

[ native-client-discuss moved to bcc, since there's some general
advice here, but i think future replies will not be of general
interest; feel free to cc n-c-d if you feel otherwise. ]

hi,

sounds good.

>
> > google NaCl team members have been professors in our past lives, but a
> > professor at your institution is needed, and generally it helps to
> > work with her/him to finalize the research focus.
>
> It is great to know that you have academic experience:) If it is
> needed, I will give my tutor a contact to you.
>
> > that said, we can try to come up with security-related projects that
> > would be appropriate for an MS thesis project.  since you hadn't
> > mentioned crypto except as part of the seminar title, is it safe to
> > assume that you're more interested in the OS/Web security side of
> > things?
>
> Yes, it is. I am more interested in that area. The title of the
> seminary is general, so that we can focus on many various (theoretical
> or practical) topics.

one issue that i'd advise you on is that NaCl's main code base is
pretty fast moving -- and unless you're willing to have to deal with
frequent synchronization w/ the subversion repository and the
consequent merge conflicts, i'd recommend that you look at designing
and implementing applications of NaCl rather than features that would
require changes to the underlying core system. personally i find that
i lose significant time to dealing with merges, and if you're not
working on NaCl stuff full time i think you'll find it very annoying.
the alternative is to work from one particular revision number /
snapshot, but it makes it just makes it harder to integrate later --
if you want to at all -- research projects don't necessarily have to
contribute to the main repo, but if your project is to be on an open
source project, it seems silly if there's little hope of checking your
code in. furthermore, we are sticklers (or try to be) when it comes
to code review -- both for style and for security -- and w/o being
able to get your code reviewed for check in periodically, you'd end up
with a huge change list and nobody will want to review a large CL.

anyway, does that sound reasonable as a first step towards narrowing
down to an MS-sized project?

-bsy

>
> Regards,
> Daniel
>
> > should we take this off list?
> >
> > -bsy
> >
> >
> >
> > > Regards,
> > > Daniel
> >
> > > --
> >
> > > You received this message because you are subscribed to the Google Groups "Native Client Discuss" group.
> > > To post to this group, send email to native-cli...@googlegroups.com.
> > > For more options, visit this group athttp://groups.google.com/group/native-client-discuss?hl=.
> >
> > --
> > bennet s yee
> > i usually don't capitalize due to mild tendonitis
>
> --
>
> You received this message because you are subscribed to the Google Groups "Native Client Discuss" group.
> To post to this group, send email to native-cli...@googlegroups.com.

> For more options, visit this group at http://groups.google.com/group/native-client-discuss?hl=.