> On Apr 26, 2016, at 10:23 AM, Gauravjeet Singh <gaurav.ishwer...@gmail.com> 
> wrote:
> 
> This is my first year as GSoC with BRL-CAD. I hope to get most of it.
> 
> I would like to know if everything's good with my proposal. I
> submitted my proposal at the very end, so couldn't get a good amount
> of feedback on it.

It’s pretty solid, but there are some changes that should be discussed.  
There’s also a lot of past work that needs to be consolidated with a few ideas 
integrated.

For starters, make sure you’re familiar with these in detail:

http://brlcad.org/wiki/OGV
http://brlcad.org/wiki/Online_Geometry
http://brlcad.org/wiki/Online_Geometry/TODO

It’s a lot to read, but critical to for having a coherent conversation.  You’ll 
notice a lot of duplication of ideas and information.  This all needs to be 
consolidated together.  Perhaps you can take a stab at combining the 
“future/todo/next/ideas” information from those three pages all onto the latter 
http://brlcad.org/wiki/Online_Geometry/TODO page?  Include any todo items from 
your your proposal that aren’t listed.

Obviously be sure to meet with and take input from Harmanpreet Singh, 
Inderpreet Singh, Deepak Sharma, Shubham Chauhan, and others that have gotten 
us to where OGV is today.  Our meritocracy rules only kick in when there is 
unreconcilable disagreement, which just means we should strive to reach 
consensus.

I have lots of comments provided below, but discuss them with your mentors as 
my comments are mere feedback.

> I listed all the changes I wish to do with OGV in my proposal, I would
> like to know if there's any amount of addition/deletion required.

Once the info above is consolidated, that should give a better picture on how 
to move forward, turning your proposal into a project plan.  The overarching 
goal should be bringing things to a production state — as your proposal already 
states — introducing a few or no new features if possible.

Towards that, I think you should drop advanced searching, color editing, 
customized access, and phantomjs screenshot thumbnails (use rt instead).

> Here's the link to my OGV proposal
> http://brlcad.org/wiki/User:Gauravjeet/OGV_Proposal

The overall structure in your proposal is great.  After consolidating/updating 
the wiki docs, code cleanup is next priority and the first work you identified.

I’m ambivalent about adding a schema to Mongo — we still need to confirm that 
Mongo can be set up on FreeBSD, which was a problem last year and the main 
reason why OGV isn’t live on brlcad.org (Mongo needed some kernel mod 
enabled(!)).

The single biggest limitation of OGV right now is probably the OBJ backend — 
the need to convert all models into triangles in order to display them.  This 
is not something robustly done in most any CAD system.  So you can make it 
faster, but getting to OBJ in the first place for any random model can be 
really hard.  Also, note that most real models have thousands or tens of 
thousands of objects, not hundreds.  We don’t have any in our repository 
currently because they’re simply too big.

That said, do not worry about the size of files on the server, consider it 
unlimited space.

Working on handling the geometry more efficiently is good, but focusing on OBJ 
might not be the best way.  One possible idea might be to directly display the 
.g file as NURBS instead of converting by using http://verbnurbs.com  

Also, a minor nitpick on the GUI — user should not know that the model was 
converted to OBJ.  That is purely a backend method for getting triangles.  Any 
references or display of a “.obj” suffix in the GUI should be removed to only 
present the original object names as they are in the .g file.  Instead of 
“converting” on upload, it should probably more generically say “processing”.

There is already color information in the .g file, so displaying that properly 
would be more important than letting users edit colors.

I love your live and weekly deployment objective.

One point not covered yet that I think will be important is the ability to 
instantiate OGV over a folder of existing .g files, and have them be queued up 
for processing by OBJ when we point it at a directory of pre-uplaoded files on 
the server.  For example, there is a db directory in the BRL-CAD distribution 
on the server.  It would be great if we could simply point OGV at that 
directory to catalog everything in there.

Cheers!
Sean


------------------------------------------------------------------------------
Find and fix application performance issues faster with Applications Manager
Applications Manager provides deep performance insights into multiple tiers of
your business applications. It resolves application problems quickly and
reduces your MTTR. Get your free trial!
https://ad.doubleclick.net/ddm/clk/302982198;130105516;z
_______________________________________________
BRL-CAD Developer mailing list
brlcad-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/brlcad-devel

Reply via email to