On Monday 27 July 2009 18:42:11 Ian Clarke wrote:
> On Mon, Jul 27, 2009 at 6:27 AM, Matthew
> Toseland<t...@amphibian.dyndns.org> wrote:
> >> > I disagree, php or java reduces performance and increases costs, for all 
> >> > hosting options.
> >>
> >> Are you serious?  I can imagine someone making that kind of argument
> >> against dynamic page generation in 1994, but not now.
> >
> > It happens to be true. On grounds of security, performance and cost. We get 
> > a lot of hits, and any cheap hosting option, including the Google app 
> > platform, will be much slower with .php's than with static HTML.
> 
> Well, firstly you can't use PHP on the Google App platform, it has to
> be Python or Java, and I suspect in either case it will be as fast as
> we could ever need it to be.
> 
> > And we don't want an expensive hosting option: we want to save money.
> 
> Crippling our website to save a few dollars per month (and that is, at
> most, all it would be) is extremely short-sighted.  Google App Engine
> is free up to a pretty high traffic volume AFAIK.  I think we'd
> probably have a hard time finding hosting that *didn't* support some
> form of dynamic page generation.
> 
> >> I don't like using Javascript for this, and I certainly don't like the
> >> assumption that we are no-longer permitted to do any dynamic page
> >> generation server side.  That is crazy.
> >
> > We don't use it now. We haven't used it for years, all the php stuff is is 
> > SSIs, which
> > can be compiled in advance. But please step back a bit: we are talking 
> > about what
> > happens *when javascript is turned off*. Only paranoid geeks turn off 
> > javascript. If
> > they have to choose their OS then that's no great hardship for them.
> 
> Except that it appears to be unreliable even with Javascript turned
> on.  I experienced problems with it in Safari 4 (with Javascript most
> assuredly switched on).

Right, so we need to determine whether we can reliably detect the operating 
system in the 3 major browsers using javascript only. If we can't then clearly 
we need to do it on the server. Acceptable?
...
On Tuesday 28 July 2009 19:59:43 Florent Daignière wrote:
> * Ian Clarke <i...@locut.us> [2009-07-27 12:42:11]:
> > Crippling our website to save a few dollars per month (and that is, at
> > most, all it would be) is extremely short-sighted.  Google App Engine
> > is free up to a pretty high traffic volume AFAIK.  I think we'd
> > probably have a hard time finding hosting that *didn't* support some
> > form of dynamic page generation.
> > 
> 
> Hmmm, Freenet?
> 
> To do what is proposed (change the download button/text depending on the
> user-agent/OS), you need redirects: not dynamic page generation.
> 
Okay so there is a cheap solution as long as we have access to the server 
config. Great.

Attachment: signature.asc
Description: This is a digitally signed message part.

_______________________________________________
Devl mailing list
Devl@freenetproject.org
http://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl

Reply via email to