Skip to content
Snippets Groups Projects
Commit 0897a2d0 authored by edgrif's avatar edgrif
Browse files

final version of initial description.

parent be58e651
No related branches found
No related tags found
No related merge requests found
......@@ -6,7 +6,6 @@ Cycle 1: Feb 5th
Cycle 2: June 5th
Cycle 3: Oct 5th
but I don't know what the cycles are.....
......@@ -17,8 +16,8 @@ On page called PHS 398 Research Plan there are the following sections:
Specific Aims
-------------
To produce a genome viewer with a display speed comparable to current high speed
computer games such as car racing programs.
To produce a genome viewer with a display speed comparable to current high
speed computer games such as car racing programs.
To show that such a viewer will significantly increase the effectiveness of
curation and annotation of genomes with particular reference to the display of:
......@@ -37,9 +36,9 @@ Background and Significance
Current genome viewers are either quite low performance because they are
web-based or even when coded in efficient languages such as C or C++ are some
way behind gaming software because they do not make use of the graphics hardware
to be found on even modest desktop machines. This is a problem becuase of two
different but related developments:
way behind gaming software because they do not make use of the graphics
hardware to be found on even modest desktop machines. This is a problem becuase
of two different but related developments:
- the amount of alignment data in the form of nuclotide and peptide
matches is increasing extremely rapidly meaning that current browsers
......@@ -51,12 +50,12 @@ different but related developments:
possible to compare variation between and within species but this
requires the display of much larger amounts of data.
Exploiting the graphics hardware traditionally requires use of esoteric software
toolkits mostly only available to gaming enthusiasts, recently however a number
of Open Source toolkits have emerged (e.g. Clutter) which provide more straight
forward access to the hardware. They provide well supported and portable
toolkits which largely shield the programmer from the intricacies of the
hardware.
Exploiting the graphics hardware traditionally requires use of esoteric
software toolkits mostly only available to gaming enthusiasts, recently however
a number of Open Source toolkits have emerged (e.g. Clutter) which provide more
straight forward access to the hardware. They provide well supported and
portable toolkits which largely shield the programmer from the intricacies of
the hardware.
......@@ -67,16 +66,17 @@ An analysis of three common open source conventional toolkits for writing
graphical applications shows that their performance while good cannot
realistically be significantly improved (USE ROYS ASSESSMENT OF FOOCANVAS,
GOOCANVAS AND THE OTHER ONE...). Generally to get good reactivity means drawing
a large number of features up front so that subsequent scrolling/panning/zooming
works well. This means a sometimes substantial delay because the drawing calls
cause a bottleneck and also uses substantial amounts of memory. Features simply
cannot be drawn fast enough on the fly to avoid this.
a large number of features up front so that subsequent
scrolling/panning/zooming works well. This means a sometimes substantial delay
because the drawing calls cause a bottleneck and also uses substantial amounts
of memory. Features simply cannot be drawn fast enough on the fly to avoid
this.
Gaming software takes a different approach and objects must be drawn in real
time _as_ the user scrolls/pans/zooms. This is possible because the game
software makes good use of the graphics hardware. Where such hardware assistance
is either disabled or is not available gaming programs are not much faster than
normal graphics programs.
software makes good use of the graphics hardware. Where such hardware
assistance is either disabled or is not available gaming programs are not much
faster than normal graphics programs.
......@@ -95,8 +95,8 @@ software is stable and its performance dependable.
2) Prototype viewers
Different viewing modes wil be implemented as prototypes to assess with the help
of annotators their use and speed.
Different viewing modes wil be implemented as prototypes to assess with the
help of annotators their use and speed.
......
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment