Language = Clarion 5.5EE, platform is Win98 and Win2000
Server.
My Blowfish code passes all the test vectors provided by
the original creator of Blowfish, located here,
http://www.counterpane.com/blowfish.html
But as I stated, it was through the help of others who
have successfully authored their own client code that I
found out that the de-facto blowfish standard does not
work with Tucows. In fact one person told me that they had
CounterPane help them with getting the Blowfish alogorithm
working and CounterPane told them the library you refer to
is slightly different than the Blowfish spec. Thus without
Tucows documenting this descrepency I'm SOL.
We could easily digress into a spitting match about
Blowfish and Tocows implementation, but my point is that
the interface should not be so difficult to implement
across different platforms as to allow such a spitting
match to be a possibility .....
Again, I am stating my request for platform independent
support. Something as simple as enom's interface would be
nice but not necessary. A free well documented Windows OS
DLL would be more than adiquit and would support virtually
all languages. Even better would be incremental test
vectors for each Tucows API layer, but last time I asked
for that it was suggested that I was a hacker attempting
to reverse engineer other resellers private keys
........... (sigh)
Thanks!
On Tue, 11 Mar 2003 08:35:12 -0500
Roger Ward <roger.ward@natnet.com> wrote:
>
>Can you be more specific? You vent and yet don't say what
>language you wrote
>yours in.... if you've written it in C++/C/PHP/Perl, yet
>need a blowfish
>implementation... USE libmcrypt! It works!
>
>-Roger
>
>
>
>Quoting inbox@fpgatools.com:
>
>>
>> Fortunately many poeple came to my assistance in trying
>>to
>> write my own interface ..... Then I was told that even
>> though my Blowfish algorithm satifies all the test
>>vectors
>> it will not work for API since the Perl implementation
>> they use is slightly different than "the standard"
>>......
>> I gave up at that point ....
>>
>>
>> So I guess I have to continue to use enom's nice API for
>> registration and then manually transfer to Tucows? Or is
>> Tucows going to simplify their API? Why can't Tucows
>>make
>> my life so easy?
>>
>>
>> To be clear --
>>
>> I really *DISLIKE* using enom and am in no why
>>attempting
>> to promote them. But I am attempting to shame Tucows
>>into
>> filling in the only hole I've found in their otherwise
>> stellar service: an API that is *REASONABLE* to
>>implement
>> in the Resellers language of choice!
>>
>>
>> Thank you for this opertunity to vent. ;)
>>
>>
>> Joe Rhett <jrhett@isite.net> wrote:
>> >Did you get any takers on this? I wrote a C++ client
>> >back in late 2000,
>> >but it is far far out of date now.
>> >
>> >On Fri, Jan 24, 2003 at 01:33:25PM -0500, Frank
>>Michlick
>> >wrote:
>> >> Hi,
>> >>
>> >> If anyone has implemented a client code to connect to
>> >>our API in C++,
>> >> please email me (fmichlick@tucows.com), I'd be
>> >>interested in talking
>> >> to you.
>> >>
>> >> Thanks + best regards,
>> >> Frank Michlick
>> >
>> >--
>> >Joe Rhett
>> >
>> Chief
>> >Geek
>> >JRhett@ISite.Net
>> > ISite Services,
>>Inc.
>>
>>
>
This archive was generated by hypermail 2.1.3 : Tue Oct 19 2004 - 23:37:39 EDT