|
|
Posted by Ben on 03/30/07 18:41
"shimmyshack" <matt.farey@gmail.com> wrote in message
news:1175275132.374121.229580@p77g2000hsh.googlegroups.com...
> On Mar 30, 5:43 pm, "Ben" <nos...@thankyou.com> wrote:
>> General question for the group. I have a friend who claims that PHP/Java
>> grids can't touch .NET with respect to dev time, reliability and speed.
>> I'm
>> not in a position to debate but would like to know the thoughts given the
>> following requirements:
>>
>> - Grid will handle 1,000 row, 25 columns record sets on a regular basis
>>
>> - Multiple users hitting same data set so automatic refreshes will need
>> to
>> occur frequently, 10 second intervals for example, and a clean/fast
>> repaint
>> is key.
>>
>> - Maybe of less significance but drag n drop is required.
>>
>> Don't want to debate the requirements themselves, they are what they are.
>> But if you had to choose PHP/Java vs .NET (ASP/VB) to develop said grid,
>> which is "better" and why? The back end DB for both scenarios will be
>> either SQL, MySQL or Oracle, client choice and don't want to debate
>> offering
>> a choice of the DB solutions either please.
>>
>> TYIA for any thoughts.
>>
>> =B
>
> What you are /really/ asking is whether php can provide fast access to
> the database. The rest of it, the grid (html table markup) repainting
> (rendering) drag and drop (javascript) mtuliple users (server) doesn't
> have much to do with php.
> But you should know IIRC php offloads its maths and various other
> things onto native C libraries, so I can't see that .NET will have any
> improvement whatsoever it the speed of execution. (.NET surely
> compiles to byte code which is then executed in M$'s runtime env)
> As to the rest, the only thing which you have control over are the
> server and javascript, chances are the opensource libraries for drag
> drop and other functionality will be WAY better than anything M$ can
> possibly offer. It, as a company has a woeful record of client jscript
> - even the postback was broken for ages!) it is slow to respond to
> obvious advances, whereas opencourse can be tweaked and improved daily
> and frequently is, so you can pair your library right down and if all
> you need is drag and drop you can optimise it for that.
> As for server setup, I guess you are talking php on IIS, so fast_cgi
> benchmarks are what you want, I read somewhere that this compares very
> well.
> But this will always come down to opinion, he has never used a well
> setup Linux version probably, and doesnt understand the underlying
> technologies well enough to be the judge of any of this. He is just
> religious about .NET because he knows he can "drag and drop" forms and
> so on, and "ajaxify" his code with a few clicks, great - his focus is
> on speedy coding so he can spend his big bucks on a new gadget, and
> let someone else worry about matters of security and professionalism,
> and tight code. Not my persaonl idea of what a professional programmer
> should be.
>
This is exactly the type of response I was looking for. My *current*
ignorance prevents a rebuttal based on experience so I just listen. He
actually said, "Well in ten years no one knows if PHP and Java will still be
around but we know M$ will be there." and. "Look at the added development
cost of PHP/Java, I could knock this grid out in two hours with M$ and spent
a day with this and have nothing." Problem is he's got influence and I
totally disagree with the direction....so...plan is to build a working
proto-type in PHP/Java/MySQL. So I need to get busy, lack of experience
doesn't scare me and I can always fall back on groups.
As previously posted I'm a noob to web apps but do have 17 years experience
coding, from PAL, Clipper/Xbase and VFP with some mapping sprinkled in for
good measure. My primary VFP app is basically a html string generator from
very diverse data sources ranging from Unix/Ingress, DbaseIII to SQL. VFP
is the data engine and parses data and writes strings at supersonic speeds.
So for a "no interaction" web pages generator it was smokin. All output is
to *tight* html files (generated by optimized code not report writer) strung
together to create a web-report archive and (dont laugh) all from funcs I
developed that are fast fast fast.....but the day has come for a true
interactive web app which requires me to go outside my little box and dive
in. I like PHP, the PHP community, have followed for years but no time to
jump in....and from a business perspective, M$ has no cost benefit that I
can see. Maybe faster dev of half-assed apps but when you look at total
cost....no brainer I think.
Anyway, thanks again.
=B
Navigation:
[Reply to this message]
|