Lots of configs

David Helkowski dhelkowski at sbgnet.com
Tue Mar 8 14:03:47 CET 2011

To write this sort of message, and to the list no doubt, is nothing
short of immature. In so much as what I said caused such a response,
I apologize for those having bothered to read this.

That said, I am going to response to the points made. I would appreciate 
a 3rd
party ( well a 4th at this point ), who has more experience and maturity,
would chip in and provide some order to this discussion.

On 3/8/2011 2:39 AM, Poul-Henning Kamp wrote:
> In message<4D74D763.706 at sbgnet.com>, David Helkowski writes:
>> The best way would be to use a jump table.
>> By that, I mean to make multiple subroutines in C, and then to jump to
>> the different subroutines by looking
>> up pointers to the subroutines using a string hashing/lookup system.
> The sheer insanity of this proposal had me wondering which vending
> machine gave you a CS degree instead of the cola you ordered.
They don't teach jump tables in any college I know of. I believe I first 
learned about
them in my own readings of 'Peter Norton's Assembly Language'; a book I 
first read
perhaps about 15 years ago. I still have the book on the shelf. I don't 
think Peter Norton
would ever call an ingenious solution to a challenging problem 'sheer 
insanity'. He would
very likely laugh at the simplicity of what I am suggesting.
> But upon reading:
>> I attempted to do this myself when I first started using
>> varnish, but I was having problems with varnish crashing
>> when attempting to use the code I wrote in C. There may be
>> limitations to the C code that can be used.
> I realized that you're probably just some troll trying to have
> a bit of a party here on our mailing list, or possibly some teenager
> in his mothers basement, from where you "rulez teh w0rld" because
> he is quite clearly Gods Gift To Computers.
This is called an Ad hominen attack. Belittling those you interact with 
in no way
betters your opinion. I am also not sure why this is a response to what 
you quoted
me on. I wrote what I did because I am actually curious if someone has 
time and
effort to get hash tables working in VCL. I would to see a working 
rendition of it.
I didn't really spend much time attempting to make it work, because my 
own usage
of VCL didn't end up requiring it.

That is, my statement here is an admission of my own lack of knowledge 
of the
limitations of inline C in VCL. I am not trolling and would seriously 
like to see working
hash tables.
> Or quite likely both.
> The fact that you have to turn to Wikipedia to find out how many
> instructions a contemporary CPU can execute per second, and then
> get the answer wrong by about an order of magnitude makes me almost
> sad for you.
I will test your code and write a subroutine demonstrating the reality 
of the
numbers I have quoted. Once I have done that I will respond to this 
> But you may have a future in you still, but there are a lot of good
> books you will have read to unlock it.
> I would recommend you start out with "The Mythical Man Month", and
> continue with pretty much anything Kernighan has written on the
> subject of programming.
I have read many discussions on the book in question, and am quite 
familiar with
the writing of Kernighan and Ritchie. They are well written authors on 
the C language.
Their methodologies are also outdated. Their book a on C is over 20 
years old at this
point. Obviously good information doesn't expire, but a lot of good 
things have been
learned since then.

I am not interested in playing knowledge based games. Programming is not 
a trivia
game; it is about applying workable solutions to real world problems in 
an efficient
> At some point, you will understand what Dijkstra is talking about here:
>      http://www.cs.utexas.edu/users/EWD/transcriptions/EWD01xx/EWD117.html
No doubt this is a well written piece that bears a response of its own. 
I am not going to
respond to this link with any detail at the moment, because you haven't 
bothered to
explain the purpose of putting it here; other than to link to something 
more well written
than your own childish attack.
> Until then, you should not attempt to do anything with a computer
> that could harm other people.
I hardly see how answering a request for the right way to do something 
with the
appropriate correct way is something that will harm. It is up to the 
reader to decide
what method they which to use.

Also, I am concerned with your lack of confidence in other users of 
Varnish. I think
that there are many learned users of it, and a good number of them are 
quite capable
of taking my hash table suggestion and making it a usable reality. Once 
it is a reality
it could easily be used by other less experienced users of Varnish.

How is having an open discussion about an efficient solution to a 
recurring problem
> And now: Please shut up before I mock you.
If you wish to mock; feel free. I would prefer if you send me a direct 
email and do not
send such nonsense to the list, nor to other uninvolved parties.
> Poul-Henning

More information about the varnish-misc mailing list