Re: OrchidWiz CD
- From: "Ted Byers" <r.ted.byers@xxxxxxxxxx>
- Date: Sun, 4 Dec 2005 09:30:50 -0500
"?" <pakrat@xxxxxxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message
news:slrndp521q.an8.pakrat@xxxxxxxxxxxxxxxxxxxxxxxxxxxx
> On Sat, 3 Dec 2005 19:56:50 -0500 in
> <qbqdnSFSNshOow_enZ2dnUVZ_v6dnZ2d@xxxxxxxxxx> Ted Byers
> <r.ted.byers@xxxxxxxxxx> wrote:
>>>
>> For a program like this, this would be the smartest option. A relatively
>> new PC or laptop would certainly be capable of giving decent performance
>> with this kind of application.
>
> And a bit too bulky to carry around in someone else's greenhouse
> complex.
True, but that is a context where one could obtain good performance from a
client/server application in which most of the application resides on a SOHO
wireless LAN. In this conext, the LAN will have plenty of unused bandwidth,
so communication between the PDA and the desktop computer running the server
application would be almost instantaneous. The program running on the PDA
would be what we call a thin client with almost all processing being done by
the server application.
>>
>> As always, the trick is to design your application for a specific target
>> platform, and recognize that some platforms are totally inadequate for
>> some
>> applications at a given time. A PDA may be totally inappropriate for an
>> application with the amount of data managed by OrchidWiz, right now, but
>> the
>> PDA of 2015 may be well suited to such an application.
>
> IMHO, OrchidWiz currently has the shortcoming of the application
> cannot be broken from the collection of data.
I am not sure I understand you here. A capable software engineer could
break a given application into any collection of components that he sees as
appropriate. One of the things I can see PDAs of today as being able to
handle well is that of data collection and storage. Coupled with a GPS
device, such as system would be priceless for field biologists who'd prefer
to do medium to long term studies on a selection of specimens using
non-destructive techniques and technologies. All it would take is a
software engineer to develop a little application to tie it all together
(probably about a man-year for a commercial quality data collection
application).
> Given the data, a geek can kludge up an adequate interface to access
> the relevant subsets of the data from a PDA.
> I know that I could have used some remote access to genus description
> and renaming tables Saturday and one or two pictures.
> The bandwidth of the various PDAs with long distance wireless data
> access would have been adequate to retrieve that information, even
> though current generation PDAs would have been illsuited to
> running the entire application.
>
Here it seems like you're visualizing the kind of client/server application
I was coemplating about.
> [Various Snippage concerning power of PDAs]
>
> One thing I'll never understand is why tasks
> that could be done on the pocket computers of 20 years ago
> now require almost the same resources as tasks that couldn't
> be done on the pocket computers of 20 years ago...
>
This is a complex issue, but I can offer suggestions on a couple factors.
First, many software houses are too cheap to hire experienced software
engineers, engineers who know how to use available hardware well. After
all, one can get a kid who still doesn't know he's still wet behind the ears
for half the visible cost of an experienced software engineer. Some of
these kids will be a whiz for certain tasks, but most will flounder for a
while, taking several times as long to complete a task as an experienced
software engineer would take. They just don't have the experience. IMHO,
an ideal software development team will include both a number of these kids
and a few old guys like me. ;-) Then you'd get the enthusiasm and energy
of the young guided by the experience of the old. Second, many 'software
engineers' lack the discipline of delivering only what is needed. I have
seen plenty of applications in which the developer added features just
because he could, and largely for the purpose of impressing his peers.
Third, most applications, and operating systems, suffer bloating because of
feature creep and gold plating. Fourth, many of these programs are built to
run on Windows or the standard distributions of Linux. MS has provided a
stripped down version of Windows for portable devices, and one can strip
down a distribution of Linux to a similar degree, but I doubt that either OS
could be stripped down enough to run on an 8086 PC. Even a supercomputer of
20 years ago would have been hard pressed to run either OS as they exist
now. That said I have seen special, relatively modern versions of unix that
have been developed specifically to run on anemic processors such as the
8086, but this is for embedded applications that no consumer will ever use
directly. Some of the smart applicances use it, but more often the
applications control equipment in industrial settings. The companies that
specialize on this market have been developing such tiny versions of unix
for years specifically for embedded applications. I have talked with
hardware engineers who do this, and sometimes they will even forgo the OS
and embed those features of the OS that they need right in the application
they're developing, using either C or assembler. But you and I will never
start, or try to use, such applications. They are designed to run the
moment the smart device is powered up, and then run continuously, without
user input, until it is powered down.
Yes, you're right in the trend you're seeing. It is not inevitable, but it
is understandable in terms of other trends in the industry, including the
fact that, in a sense, new applications carry the baggage of twenty years of
IT history to one extent or another. For consumer applications, that
baggage is embedded through the OS and would be outrageously expensive to
remove.
Does this help?
Cheers,
Ted
--
R.E. (Ted) Byers, Ph.D., Ed.D.
R & D Decision Support Solutions
http://www.randddecisionsupportsolutions.com/
Healthy Living Through Informed Decision Making
.
- References:
- Re: OrchidWiz CD
- From: Lady Blacksword
- Re: OrchidWiz CD
- From: K Barrett
- Re: OrchidWiz CD
- From: wendy7
- Re: OrchidWiz CD
- From: Ted Byers
- Re: OrchidWiz CD
- From: ?
- Re: OrchidWiz CD
- Prev by Date: Re: to let spike or not to let spike?
- Next by Date: Orchids at Harbour Bay Plaza, Sewall's Pt. FL
- Previous by thread: Re: OrchidWiz CD
- Next by thread: Re: OrchidWiz CD
- Index(es):
Relevant Pages
|